rdelaage
/
postgresql
mirror of https://git.postgresql.org/git/postgresql.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
postgresql/src/backend/executor/nodeModifyTable.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

4479 lines
140 KiB
C
Raw Normal View History

Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*-------------------------------------------------------------------------
*
* nodeModifyTable.c
* routines to handle ModifyTable nodes.
*
Update copyright for 2023 Backpatch-through: 11
2023-01-02 21:00:37 +01:00
* Portions Copyright (c) 1996-2023, PostgreSQL Global Development Group
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
Remove cvs keywords from all files.
2010-09-20 22:08:53 +02:00
* src/backend/executor/nodeModifyTable.c
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*
*-------------------------------------------------------------------------
*/
/* INTERFACE ROUTINES
* ExecInitModifyTable - initialize the ModifyTable node
* ExecModifyTable - retrieve the next tuple from the node
* ExecEndModifyTable - shut down the ModifyTable node
* ExecReScanModifyTable - rescan the ModifyTable node
*
* NOTES
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* The ModifyTable node receives input from its outerPlan, which is
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
* the data to insert for INSERT cases, the changed columns' new
* values plus row-locating info for UPDATE and MERGE cases, or just the
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* row-locating info for DELETE cases.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* MERGE runs a join between the source relation and the target
* table; if any WHEN NOT MATCHED clauses are present, then the
* join is an outer join. In this case, any unmatched tuples will
* have NULL row-locating info, and only INSERT can be run. But for
* matched tuples, then row-locating info is used to determine the
* tuple to UPDATE or DELETE. When all clauses are WHEN MATCHED,
* then an inner join is used, so all tuples contain row-locating info.
*
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* If the query specifies RETURNING, then the ModifyTable returns a
* RETURNING tuple after completing each row insert, update, or delete.
* It must be called again to continue the operation. Without RETURNING,
* we just loop within the node until all the work is done, then
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* return NULL. This avoids useless call/return overhead. (MERGE does
* not support RETURNING.)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*/
#include "postgres.h"
Don't include heapam.h from others headers. heapam.h previously was included in a number of widely used headers (e.g. execnodes.h, indirectly in executor.h, ...). That's problematic on its own, as heapam.h contains a lot of low-level details that don't need to be exposed that widely, but becomes more problematic with the upcoming introduction of pluggable table storage - it seems inappropriate for heapam.h to be included that widely afterwards. heapam.h was largely only included in other headers to get the HeapScanDesc typedef (which was defined in heapam.h, even though HeapScanDescData is defined in relscan.h). The better solution here seems to be to just use the underlying struct (forward declared where necessary). Similar for BulkInsertState. Another problem was that LockTupleMode was used in executor.h - parts of the file tried to cope without heapam.h, but due to the fact that it indirectly included it, several subsequent violations of that goal were not not noticed. We could just reuse the approach of declaring parameters as int, but it seems nicer to move LockTupleMode to lockoptions.h - that's not a perfect location, but also doesn't seem bad. As a number of files relied on implicitly included heapam.h, a significant number of files grew an explicit include. It's quite probably that a few external projects will need to do the same. Author: Andres Freund Reviewed-By: Alvaro Herrera Discussion: https://postgr.es/m/20190114000701.y4ttcb74jpskkcfb@alap3.anarazel.de
2019-01-15 00:54:18 +01:00
#include "access/heapam.h"
Split tuple struct defs from htup.h to htup_details.h This reduces unnecessary exposure of other headers through htup.h, which is very widely included by many files. I have chosen to move the function prototypes to the new file as well, because that means htup.h no longer needs to include tupdesc.h. In itself this doesn't have much effect in indirect inclusion of tupdesc.h throughout the tree, because it's also required by execnodes.h; but it's something to explore in the future, and it seemed best to do the htup.h change now while I'm busy with it.
2012-08-30 22:15:44 +02:00
#include "access/htup_details.h"
tableam: Add and use scan APIs. Too allow table accesses to be not directly dependent on heap, several new abstractions are needed. Specifically: 1) Heap scans need to be generalized into table scans. Do this by introducing TableScanDesc, which will be the "base class" for individual AMs. This contains the AM independent fields from HeapScanDesc. The previous heap_{beginscan,rescan,endscan} et al. have been replaced with a table_ version. There's no direct replacement for heap_getnext(), as that returned a HeapTuple, which is undesirable for a other AMs. Instead there's table_scan_getnextslot(). But note that heap_getnext() lives on, it's still used widely to access catalog tables. This is achieved by new scan_begin, scan_end, scan_rescan, scan_getnextslot callbacks. 2) The portion of parallel scans that's shared between backends need to be able to do so without the user doing per-AM work. To achieve that new parallelscan_{estimate, initialize, reinitialize} callbacks are introduced, which operate on a new ParallelTableScanDesc, which again can be subclassed by AMs. As it is likely that several AMs are going to be block oriented, block oriented callbacks that can be shared between such AMs are provided and used by heap. table_block_parallelscan_{estimate, intiialize, reinitialize} as callbacks, and table_block_parallelscan_{nextpage, init} for use in AMs. These operate on a ParallelBlockTableScanDesc. 3) Index scans need to be able to access tables to return a tuple, and there needs to be state across individual accesses to the heap to store state like buffers. That's now handled by introducing a sort-of-scan IndexFetchTable, which again is intended to be subclassed by individual AMs (for heap IndexFetchHeap). The relevant callbacks for an AM are index_fetch_{end, begin, reset} to create the necessary state, and index_fetch_tuple to retrieve an indexed tuple. Note that index_fetch_tuple implementations need to be smarter than just blindly fetching the tuples for AMs that have optimizations similar to heap's HOT - the currently alive tuple in the update chain needs to be fetched if appropriate. Similar to table_scan_getnextslot(), it's undesirable to continue to return HeapTuples. Thus index_fetch_heap (might want to rename that later) now accepts a slot as an argument. Core code doesn't have a lot of call sites performing index scans without going through the systable_* API (in contrast to loads of heap_getnext calls and working directly with HeapTuples). Index scans now store the result of a search in IndexScanDesc->xs_heaptid, rather than xs_ctup->t_self. As the target is not generally a HeapTuple anymore that seems cleaner. To be able to sensible adapt code to use the above, two further callbacks have been introduced: a) slot_callbacks returns a TupleTableSlotOps* suitable for creating slots capable of holding a tuple of the AMs type. table_slot_callbacks() and table_slot_create() are based upon that, but have additional logic to deal with views, foreign tables, etc. While this change could have been done separately, nearly all the call sites that needed to be adapted for the rest of this commit also would have been needed to be adapted for table_slot_callbacks(), making separation not worthwhile. b) tuple_satisfies_snapshot checks whether the tuple in a slot is currently visible according to a snapshot. That's required as a few places now don't have a buffer + HeapTuple around, but a slot (which in heap's case internally has that information). Additionally a few infrastructure changes were needed: I) SysScanDesc, as used by systable_{beginscan, getnext} et al. now internally uses a slot to keep track of tuples. While systable_getnext() still returns HeapTuples, and will so for the foreseeable future, the index API (see 1) above) now only deals with slots. The remainder, and largest part, of this commit is then adjusting all scans in postgres to use the new APIs. Author: Andres Freund, Haribabu Kommi, Alvaro Herrera Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-11 20:46:41 +01:00
#include "access/tableam.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "access/xact.h"
Remove WITH OIDS support, change oid catalog column visibility. Previously tables declared WITH OIDS, including a significant fraction of the catalog tables, stored the oid column not as a normal column, but as part of the tuple header. This special column was not shown by default, which was somewhat odd, as it's often (consider e.g. pg_class.oid) one of the more important parts of a row. Neither pg_dump nor COPY included the contents of the oid column by default. The fact that the oid column was not an ordinary column necessitated a significant amount of special case code to support oid columns. That already was painful for the existing, but upcoming work aiming to make table storage pluggable, would have required expanding and duplicating that "specialness" significantly. WITH OIDS has been deprecated since 2005 (commit ff02d0a05280e0). Remove it. Removing includes: - CREATE TABLE and ALTER TABLE syntax for declaring the table to be WITH OIDS has been removed (WITH (oids[ = true]) will error out) - pg_dump does not support dumping tables declared WITH OIDS and will issue a warning when dumping one (and ignore the oid column). - restoring an pg_dump archive with pg_restore will warn when restoring a table with oid contents (and ignore the oid column) - COPY will refuse to load binary dump that includes oids. - pg_upgrade will error out when encountering tables declared WITH OIDS, they have to be altered to remove the oid column first. - Functionality to access the oid of the last inserted row (like plpgsql's RESULT_OID, spi's SPI_lastoid, ...) has been removed. The syntax for declaring a table WITHOUT OIDS (or WITH (oids = false) for CREATE TABLE) is still supported. While that requires a bit of support code, it seems unnecessary to break applications / dumps that do not use oids, and are explicit about not using them. The biggest user of WITH OID columns was postgres' catalog. This commit changes all 'magic' oid columns to be columns that are normally declared and stored. To reduce unnecessary query breakage all the newly added columns are still named 'oid', even if a table's column naming scheme would indicate 'reloid' or such. This obviously requires adapting a lot code, mostly replacing oid access via HeapTupleGetOid() with access to the underlying Form_pg_*->oid column. The bootstrap process now assigns oids for all oid columns in genbki.pl that do not have an explicit value (starting at the largest oid previously used), only oids assigned later by oids will be above FirstBootstrapObjectId. As the oid column now is a normal column the special bootstrap syntax for oids has been removed. Oids are not automatically assigned during insertion anymore, all backend code explicitly assigns oids with GetNewOidWithIndex(). For the rare case that insertions into the catalog via SQL are called for the new pg_nextoid() function can be used (which only works on catalog tables). The fact that oid columns on system tables are now normal columns means that they will be included in the set of columns expanded by * (i.e. SELECT * FROM pg_class will now include the table's oid, previously it did not). It'd not technically be hard to hide oid column by default, but that'd mean confusing behavior would either have to be carried forward forever, or it'd cause breakage down the line. While it's not unlikely that further adjustments are needed, the scope/invasiveness of the patch makes it worthwhile to get merge this now. It's painful to maintain externally, too complicated to commit after the code code freeze, and a dependency of a number of other patches. Catversion bump, for obvious reasons. Author: Andres Freund, with contributions by John Naylor Discussion: https://postgr.es/m/20180930034810.ywp2c7awz7opzcfr@alap3.anarazel.de
2018-11-21 00:36:57 +01:00
#include "catalog/catalog.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "commands/trigger.h"
Centralize executor-related partitioning code. Some code is moved from partition.c, which has grown very quickly lately; splitting the executor parts out might help to keep it from getting totally out of control. Other code is moved from execMain.c. All is moved to a new file execPartition.c. get_partition_for_tuple now has a new interface that more clearly separates executor concerns from generic concerns. Amit Langote. A slight comment tweak by me. Discussion: http://postgr.es/m/1f0985f8-3b61-8bc4-4350-baa6d804cb6d@lab.ntt.co.jp
2017-11-15 16:23:28 +01:00
#include "executor/execPartition.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "executor/executor.h"
#include "executor/nodeModifyTable.h"
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
#include "foreign/fdwapi.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "miscadmin.h"
#include "nodes/nodeFuncs.h"
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
#include "optimizer/optimizer.h"
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
#include "rewrite/rewriteHandler.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "storage/bufmgr.h"
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
#include "storage/lmgr.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "utils/builtins.h"
Convert ExecComputeStoredGenerated to use tuple slots This code was still using the old style of forming a heap tuple rather than using tuple slots. This would be less efficient if a non-heap access method was used. And using tuple slots is actually quite a bit faster when using heap as well. Also add some test cases for generated columns with null values and with varlena values. This lack of coverage was discovered while working on this patch. Discussion: https://www.postgresql.org/message-id/flat/20190331025744.ugbsyks7czfcoksd%40alap3.anarazel.de
2019-05-15 19:37:52 +02:00
#include "utils/datum.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
#include "utils/memutils.h"
Move Trigger and TriggerDesc structs out of rel.h into a new reltrigger.h This lets us stop including rel.h into execnodes.h, which is a widely used header.
2011-02-23 18:18:09 +01:00
#include "utils/rel.h"
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
typedef struct MTTargetRelLookup
{
Oid relationOid; /* hash key, must be first */
int relationIndex; /* rel's index in resultRelInfo[] array */
} MTTargetRelLookup;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* Context struct for a ModifyTable operation, containing basic execution
* state and some output variables populated by ExecUpdateAct() and
* ExecDeleteAct() to report the result of their actions to callers.
*/
typedef struct ModifyTableContext
{
/* Operation state */
ModifyTableState *mtstate;
EPQState *epqstate;
EState *estate;
/*
* Slot containing tuple obtained from ModifyTable's subplan. Used to
* access "junk" columns that are not going to be stored.
*/
TupleTableSlot *planSlot;
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/* MERGE specific */
MergeActionState *relaction; /* MERGE action in progress */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* Information about the changes that were made concurrently to a tuple
* being updated or deleted
*/
TM_FailureData tmfd;
/*
* The tuple projected by the INSERT's RETURNING clause, when doing a
* cross-partition UPDATE
*/
TupleTableSlot *cpUpdateReturningSlot;
} ModifyTableContext;
/*
* Context struct containing output data specific to UPDATE operations.
*/
typedef struct UpdateContext
{
bool updated; /* did UPDATE actually occur? */
bool crossPartUpdate; /* was it a cross-partition update? */
Pre-beta mechanical code beautification. Run pgindent, pgperltidy, and reformat-dat-files. This set of diffs is a bit larger than typical. We've updated to pg_bsd_indent 2.1.2, which properly indents variable declarations that have multi-line initialization expressions (the continuation lines are now indented one tab stop). We've also updated to perltidy version 20230309 and changed some of its settings, which reduces its desire to add whitespace to lines to make assignments etc. line up. Going forward, that should make for fewer random-seeming changes to existing code. Discussion: https://postgr.es/m/20230428092545.qfb3y5wcu4cm75ur@alvherre.pgsql
2023-05-19 23:24:48 +02:00
TU_UpdateIndexes updateIndexes; /* Which index updates are required? */
Move ModifyTableContext->lockmode to UpdateContext Should have been done this way to start with, but I failed to notice This way we avoid some pointless initialization, and better contains the variable to exist in the scope where it is really used. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 11:18:04 +02:00
/*
* Lock mode to acquire on the latest tuple version before performing
* EvalPlanQual on it
*/
LockTupleMode lockmode;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
} UpdateContext;
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
static void ExecBatchInsert(ModifyTableState *mtstate,
ResultRelInfo *resultRelInfo,
TupleTableSlot **slots,
TupleTableSlot **planSlots,
int numSlots,
EState *estate,
bool canSetTag);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
static void ExecPendingInserts(EState *estate);
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
static void ExecCrossPartitionUpdateForeignKey(ModifyTableContext *context,
ResultRelInfo *sourcePartInfo,
ResultRelInfo *destPartInfo,
ItemPointer tupleid,
TupleTableSlot *oldslot,
TupleTableSlot *newslot);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
static bool ExecOnConflictUpdate(ModifyTableContext *context,
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ResultRelInfo *resultRelInfo,
ItemPointer conflictTid,
TupleTableSlot *excludedSlot,
bool canSetTag,
TupleTableSlot **returning);
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
static TupleTableSlot *ExecPrepareTupleRouting(ModifyTableState *mtstate,
EState *estate,
PartitionTupleRouting *proute,
ResultRelInfo *targetRelInfo,
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
TupleTableSlot *slot,
ResultRelInfo **partRelInfo);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
static TupleTableSlot *ExecMerge(ModifyTableContext *context,
ResultRelInfo *resultRelInfo,
ItemPointer tupleid,
bool canSetTag);
static void ExecInitMerge(ModifyTableState *mtstate, EState *estate);
static bool ExecMergeMatched(ModifyTableContext *context,
ResultRelInfo *resultRelInfo,
ItemPointer tupleid,
bool canSetTag);
static void ExecMergeNotMatched(ModifyTableContext *context,
ResultRelInfo *resultRelInfo,
bool canSetTag);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* Verify that the tuples to be produced by INSERT match the
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* target relation's rowtype
*
* We do this to guard against stale plans. If plan invalidation is
* functioning properly then we should never get a failure here, but better
* safe than sorry. Note that this is called after we have obtained lock
* on the target rel, so the rowtype can't change underneath us.
*
* The plan output is represented by its targetlist, because that makes
* handling the dropped-column case easier.
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
*
* We used to use this for UPDATE as well, but now the equivalent checks
* are done in ExecBuildUpdateProjection.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*/
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
static void
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
ExecCheckPlanOutput(Relation resultRel, List *targetList)
{
TupleDesc resultDesc = RelationGetDescr(resultRel);
int attno = 0;
ListCell *lc;
foreach(lc, targetList)
{
TargetEntry *tle = (TargetEntry *) lfirst(lc);
Form_pg_attribute attr;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
Assert(!tle->resjunk); /* caller removed junk items already */
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
if (attno >= resultDesc->natts)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("table row type and query-specified row type do not match"),
errdetail("Query has too many columns.")));
Change tupledesc->attrs[n] to TupleDescAttr(tupledesc, n). This is a mechanical change in preparation for a later commit that will change the layout of TupleDesc. Introducing a macro to abstract the details of where attributes are stored will allow us to change that in separate step and revise it in future. Author: Thomas Munro, editorialized by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CAEepm=0ZtQ-SpsgCyzzYpsXS6e=kZWqk3g5Ygn3MDV7A8dabUA@mail.gmail.com
2017-08-20 20:19:07 +02:00
attr = TupleDescAttr(resultDesc, attno);
attno++;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
if (!attr->attisdropped)
{
/* Normal case: demand type match */
if (exprType((Node *) tle->expr) != attr->atttypid)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("table row type and query-specified row type do not match"),
errdetail("Table has type %s at ordinal position %d, but query expects %s.",
format_type_be(attr->atttypid),
attno,
format_type_be(exprType((Node *) tle->expr)))));
}
else
{
/*
* For a dropped column, we can't check atttypid (it's likely 0).
* In any case the planner has most likely inserted an INT4 null.
* What we insist on is just *some* NULL constant.
*/
if (!IsA(tle->expr, Const) ||
!((Const *) tle->expr)->constisnull)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("table row type and query-specified row type do not match"),
errdetail("Query provides a value for a dropped column at ordinal position %d.",
attno)));
}
}
if (attno != resultDesc->natts)
ereport(ERROR,
(errcode(ERRCODE_DATATYPE_MISMATCH),
errmsg("table row type and query-specified row type do not match"),
errdetail("Query has too few columns.")));
}
/*
* ExecProcessReturning --- evaluate a RETURNING list
*
Fix comment for ExecProcessReturning resultRelInfo is the argument for the function, not projectReturning. Author: Etsuro Fujita Discussion: https://postgr.es/m/5AA8E11E.1040609@lab.ntt.co.jp
2018-03-14 14:28:08 +01:00
* resultRelInfo: current result rel
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* tupleSlot: slot holding tuple actually inserted/updated/deleted
* planSlot: slot holding tuple returned by top subplan node
*
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
* Note: If tupleSlot is NULL, the FDW should have already provided econtext's
* scan tuple.
*
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* Returns a slot holding the result tuple
*/
static TupleTableSlot *
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
ExecProcessReturning(ResultRelInfo *resultRelInfo,
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
TupleTableSlot *tupleSlot,
TupleTableSlot *planSlot)
{
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
ProjectionInfo *projectReturning = resultRelInfo->ri_projectReturning;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
ExprContext *econtext = projectReturning->pi_exprContext;
/* Make tuple and any needed join variables available to ExecProject */
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
if (tupleSlot)
econtext->ecxt_scantuple = tupleSlot;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
econtext->ecxt_outertuple = planSlot;
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
/*
* RETURNING expressions might reference the tableoid column, so
* reinitialize tts_tableOid before evaluating them.
*/
econtext->ecxt_scantuple->tts_tableOid =
RelationGetRelid(resultRelInfo->ri_RelationDesc);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* Compute the RETURNING expressions */
Remove obsoleted code relating to targetlist SRF evaluation. Since 69f4b9c plain expression evaluation (and thus normal projection) can't return sets of tuples anymore. Thus remove code dealing with that possibility. This will require adjustments in external code using ExecEvalExpr()/ExecProject() - that should neither be hard nor very common. Author: Andres Freund and Tom Lane Discussion: https://postgr.es/m/20160822214023.aaxz5l4igypowyri@alap3.anarazel.de
2017-01-19 23:12:38 +01:00
return ExecProject(projectReturning);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
* ExecCheckTupleVisible -- verify tuple is visible
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
*
* It would not be consistent with guarantees of the higher isolation levels to
* proceed with avoiding insertion (taking speculative insertion's alternative
* path) on the basis of another tuple that is not visible to MVCC snapshot.
* Check for the need to raise a serialization failure, and do so as necessary.
*/
static void
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
ExecCheckTupleVisible(EState *estate,
Relation rel,
TupleTableSlot *slot)
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
{
if (!IsolationUsesXactSnapshot())
return;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
if (!table_tuple_satisfies_snapshot(rel, slot, estate->es_snapshot))
Don't throw serialization errors for self-conflicts in INSERT ON CONFLICT. A transaction that conflicts against itself, for example INSERT INTO t(pk) VALUES (1),(1) ON CONFLICT DO NOTHING; should behave the same regardless of isolation level. It certainly shouldn't throw a serialization error, as retrying will not help. We got this wrong due to the ON CONFLICT logic not considering the case, as reported by Jason Dusek. Core of this patch is by Peter Geoghegan (based on an earlier patch by Thomas Munro), though I didn't take his proposed code refactoring for fear that it might have unexpected side-effects. Test cases by Thomas Munro and myself. Report: <CAO3NbwOycQjt2Oqy2VW-eLTq2M5uGMyHnGm=RNga4mjqcYD7gQ@mail.gmail.com> Related-Discussion: <57EE93C8.8080504@postgrespro.ru>
2016-10-24 00:36:13 +02:00
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
Datum xminDatum;
TransactionId xmin;
bool isnull;
xminDatum = slot_getsysattr(slot, MinTransactionIdAttributeNumber, &isnull);
Assert(!isnull);
xmin = DatumGetTransactionId(xminDatum);
Don't throw serialization errors for self-conflicts in INSERT ON CONFLICT. A transaction that conflicts against itself, for example INSERT INTO t(pk) VALUES (1),(1) ON CONFLICT DO NOTHING; should behave the same regardless of isolation level. It certainly shouldn't throw a serialization error, as retrying will not help. We got this wrong due to the ON CONFLICT logic not considering the case, as reported by Jason Dusek. Core of this patch is by Peter Geoghegan (based on an earlier patch by Thomas Munro), though I didn't take his proposed code refactoring for fear that it might have unexpected side-effects. Test cases by Thomas Munro and myself. Report: <CAO3NbwOycQjt2Oqy2VW-eLTq2M5uGMyHnGm=RNga4mjqcYD7gQ@mail.gmail.com> Related-Discussion: <57EE93C8.8080504@postgrespro.ru>
2016-10-24 00:36:13 +02:00
/*
* We should not raise a serialization failure if the conflict is
* against a tuple inserted by our own transaction, even if it's not
* visible to our snapshot. (This would happen, for example, if
* conflicting keys are proposed for insertion in a single command.)
*/
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
if (!TransactionIdIsCurrentTransactionId(xmin))
Don't throw serialization errors for self-conflicts in INSERT ON CONFLICT. A transaction that conflicts against itself, for example INSERT INTO t(pk) VALUES (1),(1) ON CONFLICT DO NOTHING; should behave the same regardless of isolation level. It certainly shouldn't throw a serialization error, as retrying will not help. We got this wrong due to the ON CONFLICT logic not considering the case, as reported by Jason Dusek. Core of this patch is by Peter Geoghegan (based on an earlier patch by Thomas Munro), though I didn't take his proposed code refactoring for fear that it might have unexpected side-effects. Test cases by Thomas Munro and myself. Report: <CAO3NbwOycQjt2Oqy2VW-eLTq2M5uGMyHnGm=RNga4mjqcYD7gQ@mail.gmail.com> Related-Discussion: <57EE93C8.8080504@postgrespro.ru>
2016-10-24 00:36:13 +02:00
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
errmsg("could not serialize access due to concurrent update")));
Don't throw serialization errors for self-conflicts in INSERT ON CONFLICT. A transaction that conflicts against itself, for example INSERT INTO t(pk) VALUES (1),(1) ON CONFLICT DO NOTHING; should behave the same regardless of isolation level. It certainly shouldn't throw a serialization error, as retrying will not help. We got this wrong due to the ON CONFLICT logic not considering the case, as reported by Jason Dusek. Core of this patch is by Peter Geoghegan (based on an earlier patch by Thomas Munro), though I didn't take his proposed code refactoring for fear that it might have unexpected side-effects. Test cases by Thomas Munro and myself. Report: <CAO3NbwOycQjt2Oqy2VW-eLTq2M5uGMyHnGm=RNga4mjqcYD7gQ@mail.gmail.com> Related-Discussion: <57EE93C8.8080504@postgrespro.ru>
2016-10-24 00:36:13 +02:00
}
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
}
/*
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
* ExecCheckTIDVisible -- convenience variant of ExecCheckTupleVisible()
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
*/
static void
ExecCheckTIDVisible(EState *estate,
ResultRelInfo *relinfo,
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
ItemPointer tid,
TupleTableSlot *tempSlot)
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
{
Relation rel = relinfo->ri_RelationDesc;
/* Redundantly check isolation level */
if (!IsolationUsesXactSnapshot())
return;
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
if (!table_tuple_fetch_row_version(rel, tid, SnapshotAny, tempSlot))
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
elog(ERROR, "failed to fetch conflicting tuple for ON CONFLICT");
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
ExecCheckTupleVisible(estate, rel, tempSlot);
ExecClearTuple(tempSlot);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
}
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
/*
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
* Initialize to compute stored generated columns for a tuple
*
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
* This fills the resultRelInfo's ri_GeneratedExprsI/ri_NumGeneratedNeededI
* or ri_GeneratedExprsU/ri_NumGeneratedNeededU fields, depending on cmdtype.
* If cmdType == CMD_UPDATE, the ri_extraUpdatedCols field is filled too.
*
* Note: usually, a given query would need only one of ri_GeneratedExprsI and
* ri_GeneratedExprsU per result rel; but MERGE can need both, and so can
* cross-partition UPDATEs, since a partition might be the target of both
* UPDATE and INSERT actions.
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
*/
Make new GENERATED-expressions code more bulletproof. In commit 8bf6ec3ba I assumed that no code path could reach ExecGetExtraUpdatedCols without having gone through ExecInitStoredGenerated. That turns out not to be the case in logical replication: if there's an ON UPDATE trigger on the target table, trigger.c will call this code before anybody has set up its generated columns. Having seen that, I don't have a lot of faith in there not being other such paths. ExecGetExtraUpdatedCols can call ExecInitStoredGenerated for itself, as long as we are willing to assume that it is only called in CMD_UPDATE operations, which on the whole seems like a safer leap of faith. Per report from Vitaly Davydov. Discussion: https://postgr.es/m/d259d69652b8c2ff50e14cda3c236c7f@postgrespro.ru
2023-01-15 19:14:52 +01:00
void
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
ExecInitStoredGenerated(ResultRelInfo *resultRelInfo,
EState *estate,
CmdType cmdtype)
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
{
Relation rel = resultRelInfo->ri_RelationDesc;
TupleDesc tupdesc = RelationGetDescr(rel);
int natts = tupdesc->natts;
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
ExprState **ri_GeneratedExprs;
int ri_NumGeneratedNeeded;
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
Bitmapset *updatedCols;
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
MemoryContext oldContext;
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
/* Nothing to do if no generated columns */
if (!(tupdesc->constr && tupdesc->constr->has_generated_stored))
return;
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
/*
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
* In an UPDATE, we can skip computing any generated columns that do not
* depend on any UPDATE target column. But if there is a BEFORE ROW
* UPDATE trigger, we cannot skip because the trigger might change more
* columns.
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
*/
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
if (cmdtype == CMD_UPDATE &&
!(rel->trigdesc && rel->trigdesc->trig_update_before_row))
updatedCols = ExecGetUpdatedCols(resultRelInfo, estate);
else
updatedCols = NULL;
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
/*
* Make sure these data structures are built in the per-query memory
* context so they'll survive throughout the query.
*/
oldContext = MemoryContextSwitchTo(estate->es_query_cxt);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
ri_GeneratedExprs = (ExprState **) palloc0(natts * sizeof(ExprState *));
ri_NumGeneratedNeeded = 0;
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
for (int i = 0; i < natts; i++)
{
if (TupleDescAttr(tupdesc, i)->attgenerated == ATTRIBUTE_GENERATED_STORED)
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
{
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
Expr *expr;
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
/* Fetch the GENERATED AS expression tree */
expr = (Expr *) build_column_default(rel, i + 1);
if (expr == NULL)
elog(ERROR, "no generation expression found for column number %d of table \"%s\"",
i + 1, RelationGetRelationName(rel));
Optimize update of tables with generated columns When updating a table row with generated columns, only recompute those generated columns whose base columns have changed in this update and keep the rest unchanged. This can result in a significant performance benefit. The required information was already kept in RangeTblEntry.extraUpdatedCols; we just have to make use of it. Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b05e781a-fa16-6b52-6738-761181204567@2ndquadrant.com
2020-02-17 15:19:58 +01:00
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
/*
* If it's an update with a known set of update target columns,
* see if we can skip the computation.
*/
if (updatedCols)
{
Bitmapset *attrs_used = NULL;
pull_varattnos((Node *) expr, 1, &attrs_used);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
if (!bms_overlap(updatedCols, attrs_used))
continue; /* need not update this column */
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
}
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
/* No luck, so prepare the expression for execution */
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
ri_GeneratedExprs[i] = ExecPrepareExpr(expr, estate);
ri_NumGeneratedNeeded++;
/* If UPDATE, mark column in resultRelInfo->ri_extraUpdatedCols */
if (cmdtype == CMD_UPDATE)
resultRelInfo->ri_extraUpdatedCols =
bms_add_member(resultRelInfo->ri_extraUpdatedCols,
i + 1 - FirstLowInvalidHeapAttributeNumber);
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
}
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
}
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
/* Save in appropriate set of fields */
if (cmdtype == CMD_UPDATE)
{
/* Don't call twice */
Assert(resultRelInfo->ri_GeneratedExprsU == NULL);
resultRelInfo->ri_GeneratedExprsU = ri_GeneratedExprs;
resultRelInfo->ri_NumGeneratedNeededU = ri_NumGeneratedNeeded;
}
else
{
/* Don't call twice */
Assert(resultRelInfo->ri_GeneratedExprsI == NULL);
resultRelInfo->ri_GeneratedExprsI = ri_GeneratedExprs;
resultRelInfo->ri_NumGeneratedNeededI = ri_NumGeneratedNeeded;
}
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
MemoryContextSwitchTo(oldContext);
}
/*
* Compute stored generated columns for a tuple
*/
void
ExecComputeStoredGenerated(ResultRelInfo *resultRelInfo,
EState *estate, TupleTableSlot *slot,
CmdType cmdtype)
{
Relation rel = resultRelInfo->ri_RelationDesc;
TupleDesc tupdesc = RelationGetDescr(rel);
int natts = tupdesc->natts;
ExprContext *econtext = GetPerTupleExprContext(estate);
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
ExprState **ri_GeneratedExprs;
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
MemoryContext oldContext;
Datum *values;
bool *nulls;
/* We should not be called unless this is true */
Assert(tupdesc->constr && tupdesc->constr->has_generated_stored);
/*
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
* Initialize the expressions if we didn't already, and check whether we
* can exit early because nothing needs to be computed.
Optimize update of tables with generated columns When updating a table row with generated columns, only recompute those generated columns whose base columns have changed in this update and keep the rest unchanged. This can result in a significant performance benefit. The required information was already kept in RangeTblEntry.extraUpdatedCols; we just have to make use of it. Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b05e781a-fa16-6b52-6738-761181204567@2ndquadrant.com
2020-02-17 15:19:58 +01:00
*/
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
if (cmdtype == CMD_UPDATE)
{
if (resultRelInfo->ri_GeneratedExprsU == NULL)
ExecInitStoredGenerated(resultRelInfo, estate, cmdtype);
if (resultRelInfo->ri_NumGeneratedNeededU == 0)
return;
ri_GeneratedExprs = resultRelInfo->ri_GeneratedExprsU;
}
else
{
if (resultRelInfo->ri_GeneratedExprsI == NULL)
ExecInitStoredGenerated(resultRelInfo, estate, cmdtype);
/* Early exit is impossible given the prior Assert */
Assert(resultRelInfo->ri_NumGeneratedNeededI > 0);
ri_GeneratedExprs = resultRelInfo->ri_GeneratedExprsI;
}
Optimize update of tables with generated columns When updating a table row with generated columns, only recompute those generated columns whose base columns have changed in this update and keep the rest unchanged. This can result in a significant performance benefit. The required information was already kept in RangeTblEntry.extraUpdatedCols; we just have to make use of it. Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b05e781a-fa16-6b52-6738-761181204567@2ndquadrant.com
2020-02-17 15:19:58 +01:00
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
oldContext = MemoryContextSwitchTo(GetPerTupleMemoryContext(estate));
values = palloc(sizeof(*values) * natts);
nulls = palloc(sizeof(*nulls) * natts);
Convert ExecComputeStoredGenerated to use tuple slots This code was still using the old style of forming a heap tuple rather than using tuple slots. This would be less efficient if a non-heap access method was used. And using tuple slots is actually quite a bit faster when using heap as well. Also add some test cases for generated columns with null values and with varlena values. This lack of coverage was discovered while working on this patch. Discussion: https://www.postgresql.org/message-id/flat/20190331025744.ugbsyks7czfcoksd%40alap3.anarazel.de
2019-05-15 19:37:52 +02:00
slot_getallattrs(slot);
memcpy(nulls, slot->tts_isnull, sizeof(*nulls) * natts);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
for (int i = 0; i < natts; i++)
{
Convert ExecComputeStoredGenerated to use tuple slots This code was still using the old style of forming a heap tuple rather than using tuple slots. This would be less efficient if a non-heap access method was used. And using tuple slots is actually quite a bit faster when using heap as well. Also add some test cases for generated columns with null values and with varlena values. This lack of coverage was discovered while working on this patch. Discussion: https://www.postgresql.org/message-id/flat/20190331025744.ugbsyks7czfcoksd%40alap3.anarazel.de
2019-05-15 19:37:52 +02:00
Form_pg_attribute attr = TupleDescAttr(tupdesc, i);
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
if (ri_GeneratedExprs[i])
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
{
Datum val;
bool isnull;
Fix calculation of which GENERATED columns need to be updated. We were identifying the updatable generated columns of inheritance children by transposing the calculation made for their parent. However, there's nothing that says a traditional-inheritance child can't have generated columns that aren't there in its parent, or that have different dependencies than are in the parent's expression. (At present it seems that we don't enforce that for partitioning either, which is likely wrong to some degree or other; but the case clearly needs to be handled with traditional inheritance.) Hence, drop the very-klugy-anyway "extraUpdatedCols" RTE field in favor of identifying which generated columns depend on updated columns during executor startup. In HEAD we can remove extraUpdatedCols altogether; in back branches, it's still there but always empty. Another difference between the HEAD and back-branch versions of this patch is that in HEAD we can add the new bitmap field to ResultRelInfo, but that would cause an ABI break in back branches. Like 4b3e37993, add a List field at the end of struct EState instead. Back-patch to v13. The bogus calculation is also being made in v12, but it doesn't have the same visible effect because we don't use it to decide which generated columns to recalculate; as a consequence of which the patch doesn't apply easily. I think that there might still be a demonstrable bug associated with trigger firing conditions, but that's such a weird corner-case usage that I'm content to leave it unfixed in v12. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqFshLKNvQUd1DgwJ-7tsTp=dwv7KZqXC4j2wYBV1aCDUA@mail.gmail.com Discussion: https://postgr.es/m/2793383.1672944799@sss.pgh.pa.us
2023-01-05 20:12:17 +01:00
Assert(attr->attgenerated == ATTRIBUTE_GENERATED_STORED);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
econtext->ecxt_scantuple = slot;
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
val = ExecEvalExpr(ri_GeneratedExprs[i], econtext, &isnull);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Fix possible crash with GENERATED ALWAYS columns In some corner cases, this could also lead to corrupted values being included in the tuple. Users who are concerned that they are affected by this should first upgrade and then perform a base backup of their database and restore onto an off-line server. They should then query each table with generated columns to ensure there are no rows where the generated expression does not match a newly calculated version of the GENERATED ALWAYS expression. If no crashes occur and no rows are returned then you're not affected. Fixes bug #16369. Reported-by: Cameron Ezell Discussion: https://postgr.es/m/16369-5845a6f1bef59884@postgresql.org Backpatch-through: 12 (where GENERATED ALWAYS columns were added.)
2020-04-18 04:10:37 +02:00
/*
* We must make a copy of val as we have no guarantees about where
* memory for a pass-by-reference Datum is located.
*/
if (!isnull)
val = datumCopy(val, attr->attbyval, attr->attlen);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
values[i] = val;
nulls[i] = isnull;
Convert ExecComputeStoredGenerated to use tuple slots This code was still using the old style of forming a heap tuple rather than using tuple slots. This would be less efficient if a non-heap access method was used. And using tuple slots is actually quite a bit faster when using heap as well. Also add some test cases for generated columns with null values and with varlena values. This lack of coverage was discovered while working on this patch. Discussion: https://www.postgresql.org/message-id/flat/20190331025744.ugbsyks7czfcoksd%40alap3.anarazel.de
2019-05-15 19:37:52 +02:00
}
else
{
if (!nulls[i])
values[i] = datumCopy(slot->tts_values[i], attr->attbyval, attr->attlen);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
}
}
Convert ExecComputeStoredGenerated to use tuple slots This code was still using the old style of forming a heap tuple rather than using tuple slots. This would be less efficient if a non-heap access method was used. And using tuple slots is actually quite a bit faster when using heap as well. Also add some test cases for generated columns with null values and with varlena values. This lack of coverage was discovered while working on this patch. Discussion: https://www.postgresql.org/message-id/flat/20190331025744.ugbsyks7czfcoksd%40alap3.anarazel.de
2019-05-15 19:37:52 +02:00
ExecClearTuple(slot);
memcpy(slot->tts_values, values, sizeof(*values) * natts);
memcpy(slot->tts_isnull, nulls, sizeof(*nulls) * natts);
ExecStoreVirtualTuple(slot);
ExecMaterializeSlot(slot);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
MemoryContextSwitchTo(oldContext);
}
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/*
* ExecInitInsertProjection
* Do one-time initialization of projection data for INSERT tuples.
*
* INSERT queries may need a projection to filter out junk attrs in the tlist.
*
* This is also a convenient place to verify that the
* output of an INSERT matches the target table.
*/
static void
ExecInitInsertProjection(ModifyTableState *mtstate,
ResultRelInfo *resultRelInfo)
{
ModifyTable *node = (ModifyTable *) mtstate->ps.plan;
Plan *subplan = outerPlan(node);
EState *estate = mtstate->ps.state;
List *insertTargetList = NIL;
bool need_projection = false;
ListCell *l;
/* Extract non-junk columns of the subplan's result tlist. */
foreach(l, subplan->targetlist)
{
TargetEntry *tle = (TargetEntry *) lfirst(l);
if (!tle->resjunk)
insertTargetList = lappend(insertTargetList, tle);
else
need_projection = true;
}
/*
* The junk-free list must produce a tuple suitable for the result
* relation.
*/
ExecCheckPlanOutput(resultRelInfo->ri_RelationDesc, insertTargetList);
/* We'll need a slot matching the table's format. */
resultRelInfo->ri_newTupleSlot =
table_slot_create(resultRelInfo->ri_RelationDesc,
&estate->es_tupleTable);
/* Build ProjectionInfo if needed (it probably isn't). */
if (need_projection)
{
TupleDesc relDesc = RelationGetDescr(resultRelInfo->ri_RelationDesc);
/* need an expression context to do the projection */
if (mtstate->ps.ps_ExprContext == NULL)
ExecAssignExprContext(estate, &mtstate->ps);
resultRelInfo->ri_projectNew =
ExecBuildProjectionInfo(insertTargetList,
mtstate->ps.ps_ExprContext,
resultRelInfo->ri_newTupleSlot,
&mtstate->ps,
relDesc);
}
resultRelInfo->ri_projectNewInfoValid = true;
}
/*
* ExecInitUpdateProjection
* Do one-time initialization of projection data for UPDATE tuples.
*
* UPDATE always needs a projection, because (1) there's always some junk
* attrs, and (2) we may need to merge values of not-updated columns from
* the old tuple into the final tuple. In UPDATE, the tuple arriving from
* the subplan contains only new values for the changed columns, plus row
* identity info in the junk attrs.
*
* This is "one-time" for any given result rel, but we might touch more than
Comment cleanup for a1115fa07. Amit Langote Discussion: https://postgr.es/m/CA+HiwqEcawatEaUh1uTbZMEZTJeLzbroRTz9_X9Z5CFjTWJkhw@mail.gmail.com
2021-04-07 18:21:54 +02:00
* one result rel in the course of an inherited UPDATE, and each one needs
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
* its own projection due to possible column order variation.
*
* This is also a convenient place to verify that the output of an UPDATE
* matches the target table (ExecBuildUpdateProjection does that).
*/
static void
ExecInitUpdateProjection(ModifyTableState *mtstate,
ResultRelInfo *resultRelInfo)
{
ModifyTable *node = (ModifyTable *) mtstate->ps.plan;
Plan *subplan = outerPlan(node);
EState *estate = mtstate->ps.state;
TupleDesc relDesc = RelationGetDescr(resultRelInfo->ri_RelationDesc);
int whichrel;
List *updateColnos;
/*
* Usually, mt_lastResultIndex matches the target rel. If it happens not
* to, we can get the index the hard way with an integer division.
*/
whichrel = mtstate->mt_lastResultIndex;
if (resultRelInfo != mtstate->resultRelInfo + whichrel)
{
whichrel = resultRelInfo - mtstate->resultRelInfo;
Assert(whichrel >= 0 && whichrel < mtstate->mt_nrels);
}
updateColnos = (List *) list_nth(node->updateColnosLists, whichrel);
/*
* For UPDATE, we use the old tuple to fill up missing values in the tuple
* produced by the subplan to get the new tuple. We need two slots, both
* matching the table's desired format.
*/
resultRelInfo->ri_oldTupleSlot =
table_slot_create(resultRelInfo->ri_RelationDesc,
&estate->es_tupleTable);
resultRelInfo->ri_newTupleSlot =
table_slot_create(resultRelInfo->ri_RelationDesc,
&estate->es_tupleTable);
/* need an expression context to do the projection */
if (mtstate->ps.ps_ExprContext == NULL)
ExecAssignExprContext(estate, &mtstate->ps);
resultRelInfo->ri_projectNew =
ExecBuildUpdateProjection(subplan->targetlist,
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
false, /* subplan did the evaluation */
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
updateColnos,
relDesc,
mtstate->ps.ps_ExprContext,
resultRelInfo->ri_newTupleSlot,
&mtstate->ps);
resultRelInfo->ri_projectNewInfoValid = true;
}
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/*
* ExecGetInsertNewTuple
* This prepares a "new" tuple ready to be inserted into given result
* relation, by removing any junk columns of the plan's output tuple
* and (if necessary) coercing the tuple to the right tuple format.
*/
static TupleTableSlot *
ExecGetInsertNewTuple(ResultRelInfo *relinfo,
TupleTableSlot *planSlot)
{
ProjectionInfo *newProj = relinfo->ri_projectNew;
ExprContext *econtext;
/*
* If there's no projection to be done, just make sure the slot is of the
* right type for the target rel. If the planSlot is the right type we
* can use it as-is, else copy the data into ri_newTupleSlot.
*/
if (newProj == NULL)
{
if (relinfo->ri_newTupleSlot->tts_ops != planSlot->tts_ops)
{
ExecCopySlot(relinfo->ri_newTupleSlot, planSlot);
return relinfo->ri_newTupleSlot;
}
else
return planSlot;
}
/*
* Else project; since the projection output slot is ri_newTupleSlot, this
* will also fix any slot-type problem.
*
* Note: currently, this is dead code, because INSERT cases don't receive
* any junk columns so there's never a projection to be done.
*/
econtext = newProj->pi_exprContext;
econtext->ecxt_outertuple = planSlot;
return ExecProject(newProj);
}
/*
* ExecGetUpdateNewTuple
* This prepares a "new" tuple by combining an UPDATE subplan's output
* tuple (which contains values of changed columns) with unchanged
* columns taken from the old tuple.
*
* The subplan tuple might also contain junk columns, which are ignored.
* Note that the projection also ensures we have a slot of the right type.
*/
TupleTableSlot *
ExecGetUpdateNewTuple(ResultRelInfo *relinfo,
TupleTableSlot *planSlot,
TupleTableSlot *oldSlot)
{
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
ProjectionInfo *newProj = relinfo->ri_projectNew;
ExprContext *econtext;
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/* Use a few extra Asserts to protect against outside callers */
Assert(relinfo->ri_projectNewInfoValid);
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
Assert(planSlot != NULL && !TTS_EMPTY(planSlot));
Assert(oldSlot != NULL && !TTS_EMPTY(oldSlot));
econtext = newProj->pi_exprContext;
econtext->ecxt_outertuple = planSlot;
econtext->ecxt_scantuple = oldSlot;
return ExecProject(newProj);
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* ----------------------------------------------------------------
* ExecInsert
*
* For INSERT, we have to insert the tuple into the target relation
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
* (or partition thereof) and insert appropriate tuples into the index
* relations.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* slot contains the new tuple value to be stored.
*
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* Returns RETURNING result if any, otherwise NULL.
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
* *inserted_tuple is the tuple that's effectively inserted;
Fix various typos and incorrect/outdated name references Author: Alexander Lakhin Discussion: https://postgr.es/m/699beab4-a6ca-92c9-f152-f559caf6dc25@gmail.com
2023-04-19 03:50:33 +02:00
* *insert_destrel is the relation where it was inserted.
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
* These are only set on success.
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
*
* This may change the currently active tuple conversion map in
* mtstate->mt_transition_capture, so the callers must take care to
* save the previous value to avoid losing track of it.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* ----------------------------------------------------------------
*/
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
static TupleTableSlot *
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecInsert(ModifyTableContext *context,
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
ResultRelInfo *resultRelInfo,
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
TupleTableSlot *slot,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
bool canSetTag,
TupleTableSlot **inserted_tuple,
ResultRelInfo **insert_destrel)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ModifyTableState *mtstate = context->mtstate;
EState *estate = context->estate;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Relation resultRelationDesc;
List *recheckIndexes = NIL;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
TupleTableSlot *planSlot = context->planSlot;
Fix reporting of violations in ExecConstraints, again. We decided in f1b4c771ea74f42447dccaed42ffcdcccf3aa694 to pass the original slot to ExecConstraints(), but that breaks when there are BEFORE ROW triggers involved. So we need to do reverse-map the tuples back to the original descriptor instead, as Amit originally proposed. Amit Langote, reviewed by Ashutosh Bapat. One overlooked comment fixed by me. Discussion: http://postgr.es/m/b3a17254-6849-e542-2353-bde4e880b6a4@lab.ntt.co.jp
2017-04-10 18:20:08 +02:00
TupleTableSlot *result = NULL;
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
TransitionCaptureState *ar_insert_trig_tcs;
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
ModifyTable *node = (ModifyTable *) mtstate->ps.plan;
OnConflictAction onconflict = node->onConflictAction;
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
PartitionTupleRouting *proute = mtstate->mt_partition_tuple_routing;
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
MemoryContext oldContext;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
* If the input result relation is a partitioned table, find the leaf
* partition to insert the tuple into.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*/
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
if (proute)
{
ResultRelInfo *partRelInfo;
slot = ExecPrepareTupleRouting(mtstate, estate, proute,
resultRelInfo, slot,
&partRelInfo);
resultRelInfo = partRelInfo;
}
ExecMaterializeSlot(slot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
resultRelationDesc = resultRelInfo->ri_RelationDesc;
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/*
* Open the table's indexes, if we have not done so already, so that we
* can add new index entries for the inserted tuple.
*/
if (resultRelationDesc->rd_rel->relhasindex &&
resultRelInfo->ri_IndexRelationDescs == NULL)
ExecOpenIndices(resultRelInfo, onconflict != ONCONFLICT_NONE);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* BEFORE ROW INSERT Triggers.
*
* Note: We fire BEFORE ROW TRIGGERS for every attempted insertion in an
* INSERT ... ON CONFLICT statement. We cannot check for constraint
* violations before firing these triggers, because they can change the
* values to insert. Also, they can run arbitrary user-defined code with
* side-effects that we can't cancel by just not inserting the tuple.
*/
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
if (resultRelInfo->ri_TrigDesc &&
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
resultRelInfo->ri_TrigDesc->trig_insert_before_row)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
/* Flush any pending inserts, so rows are visible to the triggers */
if (estate->es_insert_pending_result_relations != NIL)
ExecPendingInserts(estate);
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
if (!ExecBRInsertTriggers(estate, resultRelInfo, slot))
return NULL; /* "do nothing" */
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
/* INSTEAD OF ROW INSERT Triggers */
if (resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_insert_instead_row)
{
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
if (!ExecIRInsertTriggers(estate, resultRelInfo, slot))
return NULL; /* "do nothing" */
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
}
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
else if (resultRelInfo->ri_FdwRoutine)
{
Fix usage of "tableoid" in GENERATED expressions. We consider this supported (though I've got my doubts that it's a good idea, because tableoid is not immutable). However, several code paths failed to fill the field in soon enough, causing such a GENERATED expression to see zero or the wrong value. This occurred when ALTER TABLE adds a new GENERATED column to a table with existing rows, and during regular INSERT or UPDATE on a foreign table with GENERATED columns. Noted during investigation of a report from Vitaly Ustinov. Back-patch to v12 where GENERATED came in. Discussion: https://postgr.es/m/CAM_DEiWR2DPT6U4xb-Ehigozzd3n3G37ZB1+867zbsEVtYoJww@mail.gmail.com
2021-05-21 21:02:06 +02:00
/*
* GENERATED expressions might reference the tableoid column, so
* (re-)initialize tts_tableOid before evaluating them.
*/
slot->tts_tableOid = RelationGetRelid(resultRelInfo->ri_RelationDesc);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
/*
* Compute stored generated columns
*/
if (resultRelationDesc->rd_att->constr &&
resultRelationDesc->rd_att->constr->has_generated_stored)
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
ExecComputeStoredGenerated(resultRelInfo, estate, slot,
CMD_INSERT);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
/*
* If the FDW supports batching, and batching is requested, accumulate
* rows and insert them in batches. Otherwise use the per-row inserts.
*/
if (resultRelInfo->ri_BatchSize > 1)
{
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
bool flushed = false;
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
/*
Update comment in ExecInsert() regarding batch insertion. Remove the stale text that is a leftover from an earlier version of the patch to add support for batch insertion, and adjust the wording in the remaining text. Back-patch to v14 where batch insertion came in. Review and wording adjustment by Tom Lane. Discussion: https://postgr.es/m/CAPmGK14goatHPHQv2Aeu_UTKqZ%2BBO%2BP%2Bzd3HKv5D%2BdyyfWKDSw%40mail.gmail.com
2022-09-29 09:55:00 +02:00
* When we've reached the desired batch size, perform the
* insertion.
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
*/
if (resultRelInfo->ri_NumSlots == resultRelInfo->ri_BatchSize)
{
ExecBatchInsert(mtstate, resultRelInfo,
resultRelInfo->ri_Slots,
resultRelInfo->ri_PlanSlots,
resultRelInfo->ri_NumSlots,
estate, canSetTag);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
flushed = true;
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
}
oldContext = MemoryContextSwitchTo(estate->es_query_cxt);
if (resultRelInfo->ri_Slots == NULL)
{
resultRelInfo->ri_Slots = palloc(sizeof(TupleTableSlot *) *
resultRelInfo->ri_BatchSize);
resultRelInfo->ri_PlanSlots = palloc(sizeof(TupleTableSlot *) *
resultRelInfo->ri_BatchSize);
}
Optimize creation of slots for FDW bulk inserts Commit b663a41363 introduced bulk inserts for FDW, but the handling of tuple slots turned out to be problematic for two reasons. Firstly, the slots were re-created for each individual batch. Secondly, all slots referenced the same tuple descriptor - with reasonably small batches this is not an issue, but with large batches this triggers O(N^2) behavior in the resource owner code. These two issues work against each other - to reduce the number of times a slot has to be created/dropped, larger batches are needed. However, the larger the batch, the more expensive the resource owner gets. For practical batch sizes (100 - 1000) this would not be a big problem, as the benefits (latency savings) greatly exceed the resource owner costs. But for extremely large batches it might be much worse, possibly even losing with non-batching mode. Fixed by initializing tuple slots only once (and reusing them across batches) and by using a new tuple descriptor copy for each slot. Discussion: https://postgr.es/m/ebbbcc7d-4286-8c28-0272-61b4753af761%40enterprisedb.com
2021-06-11 20:19:48 +02:00
/*
* Initialize the batch slots. We don't know how many slots will
* be needed, so we initialize them as the batch grows, and we
* keep them across batches. To mitigate an inefficiency in how
* resource owner handles objects with many references (as with
* many slots all referencing the same tuple descriptor) we copy
Use appropriate tuple descriptor in FDW batching The FDW batching code was using the same tuple descriptor both for all slots (regular and plan slots), but that's incorrect - the subplan may use a different descriptor. Currently this is benign, because batching is used only for INSERTs, and in that case the descriptors always match. But that would change if we allow batching UPDATEs. Fix by copying the appropriate tuple descriptor. Backpatch to 14, where the FDW batching was implemented. Author: Amit Langote Backpatch-through: 14, where FDW batching was added Discussion: https://postgr.es/m/CA%2BHiwqEWd5B0-e-RvixGGUrNvGkjH2s4m95%3DJcwUnyV%3Df0rAKQ%40mail.gmail.com
2021-08-12 21:32:53 +02:00
* the appropriate tuple descriptor for each slot.
Optimize creation of slots for FDW bulk inserts Commit b663a41363 introduced bulk inserts for FDW, but the handling of tuple slots turned out to be problematic for two reasons. Firstly, the slots were re-created for each individual batch. Secondly, all slots referenced the same tuple descriptor - with reasonably small batches this is not an issue, but with large batches this triggers O(N^2) behavior in the resource owner code. These two issues work against each other - to reduce the number of times a slot has to be created/dropped, larger batches are needed. However, the larger the batch, the more expensive the resource owner gets. For practical batch sizes (100 - 1000) this would not be a big problem, as the benefits (latency savings) greatly exceed the resource owner costs. But for extremely large batches it might be much worse, possibly even losing with non-batching mode. Fixed by initializing tuple slots only once (and reusing them across batches) and by using a new tuple descriptor copy for each slot. Discussion: https://postgr.es/m/ebbbcc7d-4286-8c28-0272-61b4753af761%40enterprisedb.com
2021-06-11 20:19:48 +02:00
*/
if (resultRelInfo->ri_NumSlots >= resultRelInfo->ri_NumSlotsInitialized)
{
TupleDesc tdesc = CreateTupleDescCopy(slot->tts_tupleDescriptor);
Use appropriate tuple descriptor in FDW batching The FDW batching code was using the same tuple descriptor both for all slots (regular and plan slots), but that's incorrect - the subplan may use a different descriptor. Currently this is benign, because batching is used only for INSERTs, and in that case the descriptors always match. But that would change if we allow batching UPDATEs. Fix by copying the appropriate tuple descriptor. Backpatch to 14, where the FDW batching was implemented. Author: Amit Langote Backpatch-through: 14, where FDW batching was added Discussion: https://postgr.es/m/CA%2BHiwqEWd5B0-e-RvixGGUrNvGkjH2s4m95%3DJcwUnyV%3Df0rAKQ%40mail.gmail.com
2021-08-12 21:32:53 +02:00
TupleDesc plan_tdesc =
Pre-beta mechanical code beautification. Run pgindent, pgperltidy, and reformat-dat-files. This set of diffs is a bit larger than typical. We've updated to pg_bsd_indent 2.1.2, which properly indents variable declarations that have multi-line initialization expressions (the continuation lines are now indented one tab stop). We've also updated to perltidy version 20230309 and changed some of its settings, which reduces its desire to add whitespace to lines to make assignments etc. line up. Going forward, that should make for fewer random-seeming changes to existing code. Discussion: https://postgr.es/m/20230428092545.qfb3y5wcu4cm75ur@alvherre.pgsql
2023-05-19 23:24:48 +02:00
CreateTupleDescCopy(planSlot->tts_tupleDescriptor);
Optimize creation of slots for FDW bulk inserts Commit b663a41363 introduced bulk inserts for FDW, but the handling of tuple slots turned out to be problematic for two reasons. Firstly, the slots were re-created for each individual batch. Secondly, all slots referenced the same tuple descriptor - with reasonably small batches this is not an issue, but with large batches this triggers O(N^2) behavior in the resource owner code. These two issues work against each other - to reduce the number of times a slot has to be created/dropped, larger batches are needed. However, the larger the batch, the more expensive the resource owner gets. For practical batch sizes (100 - 1000) this would not be a big problem, as the benefits (latency savings) greatly exceed the resource owner costs. But for extremely large batches it might be much worse, possibly even losing with non-batching mode. Fixed by initializing tuple slots only once (and reusing them across batches) and by using a new tuple descriptor copy for each slot. Discussion: https://postgr.es/m/ebbbcc7d-4286-8c28-0272-61b4753af761%40enterprisedb.com
2021-06-11 20:19:48 +02:00
resultRelInfo->ri_Slots[resultRelInfo->ri_NumSlots] =
MakeSingleTupleTableSlot(tdesc, slot->tts_ops);
resultRelInfo->ri_PlanSlots[resultRelInfo->ri_NumSlots] =
Use appropriate tuple descriptor in FDW batching The FDW batching code was using the same tuple descriptor both for all slots (regular and plan slots), but that's incorrect - the subplan may use a different descriptor. Currently this is benign, because batching is used only for INSERTs, and in that case the descriptors always match. But that would change if we allow batching UPDATEs. Fix by copying the appropriate tuple descriptor. Backpatch to 14, where the FDW batching was implemented. Author: Amit Langote Backpatch-through: 14, where FDW batching was added Discussion: https://postgr.es/m/CA%2BHiwqEWd5B0-e-RvixGGUrNvGkjH2s4m95%3DJcwUnyV%3Df0rAKQ%40mail.gmail.com
2021-08-12 21:32:53 +02:00
MakeSingleTupleTableSlot(plan_tdesc, planSlot->tts_ops);
Optimize creation of slots for FDW bulk inserts Commit b663a41363 introduced bulk inserts for FDW, but the handling of tuple slots turned out to be problematic for two reasons. Firstly, the slots were re-created for each individual batch. Secondly, all slots referenced the same tuple descriptor - with reasonably small batches this is not an issue, but with large batches this triggers O(N^2) behavior in the resource owner code. These two issues work against each other - to reduce the number of times a slot has to be created/dropped, larger batches are needed. However, the larger the batch, the more expensive the resource owner gets. For practical batch sizes (100 - 1000) this would not be a big problem, as the benefits (latency savings) greatly exceed the resource owner costs. But for extremely large batches it might be much worse, possibly even losing with non-batching mode. Fixed by initializing tuple slots only once (and reusing them across batches) and by using a new tuple descriptor copy for each slot. Discussion: https://postgr.es/m/ebbbcc7d-4286-8c28-0272-61b4753af761%40enterprisedb.com
2021-06-11 20:19:48 +02:00
/* remember how many batch slots we initialized */
resultRelInfo->ri_NumSlotsInitialized++;
}
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
Fix copying data into slots with FDW batching Commit b676ac443b optimized handling of tuple slots with bulk inserts into foreign tables, so that the slots are initialized only once and reused for all batches. The data was however copied into the slots only after the initialization, inserting duplicate values when the slot gets reused. Fixed by moving the ExecCopySlot outside the init branch. The existing postgres_fdw tests failed to catch this due to inserting data into foreign tables without unique indexes, and then checking only the number of inserted rows. This adds a new test with both a unique index and a check of inserted values. Reported-by: Alexander Pyhalov Discussion: https://postgr.es/m/7a8cf8d56b3d18e5c0bccd6cd42d04ac%40postgrespro.ru
2021-06-16 22:53:31 +02:00
ExecCopySlot(resultRelInfo->ri_Slots[resultRelInfo->ri_NumSlots],
slot);
ExecCopySlot(resultRelInfo->ri_PlanSlots[resultRelInfo->ri_NumSlots],
planSlot);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
/*
* If these are the first tuples stored in the buffers, add the
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
* target rel and the mtstate to the
* es_insert_pending_result_relations and
Fix various typos This fixes many spelling mistakes in comments, but a few references to invalid parameter names, function names and option names too in comments and also some in string constants Also, fix an #undef that was undefining the incorrect definition Author: Alexander Lakhin Reviewed-by: Justin Pryzby Discussion: https://postgr.es/m/d5f68d19-c0fc-91a9-118d-7c6a5a3f5fad@gmail.com
2023-04-18 03:23:23 +02:00
* es_insert_pending_modifytables lists respectively, except in
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
* the case where flushing was done above, in which case they
* would already have been added to the lists, so no need to do
* this.
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
*/
if (resultRelInfo->ri_NumSlots == 0 && !flushed)
{
Assert(!list_member_ptr(estate->es_insert_pending_result_relations,
resultRelInfo));
estate->es_insert_pending_result_relations =
lappend(estate->es_insert_pending_result_relations,
resultRelInfo);
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
estate->es_insert_pending_modifytables =
lappend(estate->es_insert_pending_modifytables, mtstate);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
}
Assert(list_member_ptr(estate->es_insert_pending_result_relations,
resultRelInfo));
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
resultRelInfo->ri_NumSlots++;
MemoryContextSwitchTo(oldContext);
return NULL;
}
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
/*
* insert into foreign table: let the FDW do it
*/
slot = resultRelInfo->ri_FdwRoutine->ExecForeignInsert(estate,
resultRelInfo,
slot,
planSlot);
if (slot == NULL) /* "do nothing" */
return NULL;
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
/*
* AFTER ROW Triggers or RETURNING expressions might reference the
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
* tableoid column, so (re-)initialize tts_tableOid before evaluating
Fix usage of "tableoid" in GENERATED expressions. We consider this supported (though I've got my doubts that it's a good idea, because tableoid is not immutable). However, several code paths failed to fill the field in soon enough, causing such a GENERATED expression to see zero or the wrong value. This occurred when ALTER TABLE adds a new GENERATED column to a table with existing rows, and during regular INSERT or UPDATE on a foreign table with GENERATED columns. Noted during investigation of a report from Vitaly Ustinov. Back-patch to v12 where GENERATED came in. Discussion: https://postgr.es/m/CAM_DEiWR2DPT6U4xb-Ehigozzd3n3G37ZB1+867zbsEVtYoJww@mail.gmail.com
2021-05-21 21:02:06 +02:00
* them. (This covers the case where the FDW replaced the slot.)
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
*/
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
slot->tts_tableOid = RelationGetRelid(resultRelInfo->ri_RelationDesc);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
}
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
else
{
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
WCOKind wco_kind;
Prevent BEFORE triggers from violating partitioning constraints. Since tuple-routing implicitly checks the partitioning constraints at least for the levels of the partitioning hierarchy it traverses, there's normally no need to revalidate the partitioning constraint after performing tuple routing. However, if there's a BEFORE trigger on the target partition, it could modify the tuple, causing the partitioning constraint to be violated. Catch that case. Also, instead of checking the root table's partition constraint after tuple-routing, check it beforehand. Otherwise, the rules for when the partitioning constraint gets checked get too complicated, because you sometimes have to check part of the constraint but not all of it. This effectively reverts commit 39162b2030fb0a35a6bb28dc636b5a71b8df8d1c in favor of a different approach altogether. Report by me. Initial debugging by Jeevan Ladhe. Patch by Amit Langote, reviewed by me. Discussion: http://postgr.es/m/CA+Tgmoa9DTgeVOqopieV8d1QRpddmP65aCdxyjdYDoEO5pS5KA@mail.gmail.com
2017-06-07 18:45:32 +02:00
Don't allow system columns in CHECK constraints, except tableoid. Previously, arbitray system columns could be mentioned in table constraints, but they were not correctly checked at runtime, because the values weren't actually set correctly in the tuple. Since it seems easy enough to initialize the table OID properly, do that, and continue allowing that column, but disallow the rest unless and until someone figures out a way to make them work properly. No back-patch, because this doesn't seem important enough to take the risk of destabilizing the back branches. In fact, this will pose a dump-and-reload hazard for those upgrading from previous versions: constraints that were accepted before but were not correctly enforced will now either be enforced correctly or not accepted at all. Either could result in restore failures, but in practice I think very few users will notice the difference, since the use case is pretty marginal anyway and few users will be relying on features that have not historically worked. Amit Kapila, reviewed by Rushabh Lathia, with doc changes by me.
2013-09-23 19:31:22 +02:00
/*
Fix usage of "tableoid" in GENERATED expressions. We consider this supported (though I've got my doubts that it's a good idea, because tableoid is not immutable). However, several code paths failed to fill the field in soon enough, causing such a GENERATED expression to see zero or the wrong value. This occurred when ALTER TABLE adds a new GENERATED column to a table with existing rows, and during regular INSERT or UPDATE on a foreign table with GENERATED columns. Noted during investigation of a report from Vitaly Ustinov. Back-patch to v12 where GENERATED came in. Discussion: https://postgr.es/m/CAM_DEiWR2DPT6U4xb-Ehigozzd3n3G37ZB1+867zbsEVtYoJww@mail.gmail.com
2021-05-21 21:02:06 +02:00
* Constraints and GENERATED expressions might reference the tableoid
* column, so (re-)initialize tts_tableOid before evaluating them.
Don't allow system columns in CHECK constraints, except tableoid. Previously, arbitray system columns could be mentioned in table constraints, but they were not correctly checked at runtime, because the values weren't actually set correctly in the tuple. Since it seems easy enough to initialize the table OID properly, do that, and continue allowing that column, but disallow the rest unless and until someone figures out a way to make them work properly. No back-patch, because this doesn't seem important enough to take the risk of destabilizing the back branches. In fact, this will pose a dump-and-reload hazard for those upgrading from previous versions: constraints that were accepted before but were not correctly enforced will now either be enforced correctly or not accepted at all. Either could result in restore failures, but in practice I think very few users will notice the difference, since the use case is pretty marginal anyway and few users will be relying on features that have not historically worked. Amit Kapila, reviewed by Rushabh Lathia, with doc changes by me.
2013-09-23 19:31:22 +02:00
*/
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
slot->tts_tableOid = RelationGetRelid(resultRelationDesc);
Don't allow system columns in CHECK constraints, except tableoid. Previously, arbitray system columns could be mentioned in table constraints, but they were not correctly checked at runtime, because the values weren't actually set correctly in the tuple. Since it seems easy enough to initialize the table OID properly, do that, and continue allowing that column, but disallow the rest unless and until someone figures out a way to make them work properly. No back-patch, because this doesn't seem important enough to take the risk of destabilizing the back branches. In fact, this will pose a dump-and-reload hazard for those upgrading from previous versions: constraints that were accepted before but were not correctly enforced will now either be enforced correctly or not accepted at all. Either could result in restore failures, but in practice I think very few users will notice the difference, since the use case is pretty marginal anyway and few users will be relying on features that have not historically worked. Amit Kapila, reviewed by Rushabh Lathia, with doc changes by me.
2013-09-23 19:31:22 +02:00
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
/*
* Compute stored generated columns
*/
if (resultRelationDesc->rd_att->constr &&
resultRelationDesc->rd_att->constr->has_generated_stored)
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
ExecComputeStoredGenerated(resultRelInfo, estate, slot,
CMD_INSERT);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
/*
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
* Check any RLS WITH CHECK policies.
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
*
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
* Normally we should check INSERT policies. But if the insert is the
* result of a partition key update that moved the tuple to a new
* partition, we should instead check UPDATE policies, because we are
* executing policies defined on the target table, and not those
* defined on the child partitions.
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
*
* If we're running MERGE, we refer to the action that we're executing
* to know if we're doing an INSERT or UPDATE to a partition table.
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
*/
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
if (mtstate->operation == CMD_UPDATE)
wco_kind = WCO_RLS_UPDATE_CHECK;
else if (mtstate->operation == CMD_MERGE)
wco_kind = (context->relaction->mas_action->commandType == CMD_UPDATE) ?
WCO_RLS_UPDATE_CHECK : WCO_RLS_INSERT_CHECK;
else
wco_kind = WCO_RLS_INSERT_CHECK;
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
/*
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
* ExecWithCheckOptions() will skip any WCOs which are not of the kind
* we are looking for at this point.
*/
if (resultRelInfo->ri_WithCheckOptions != NIL)
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
ExecWithCheckOptions(wco_kind, resultRelInfo, slot, estate);
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
/*
Don't needlessly check the partition contraint twice Starting with commit f0e44751d717, ExecConstraints was in charge of running the partition constraint; commit 19c47e7c8202 modified that so that caller could request to skip that checking depending on some conditions, but that commit and 15ce775faa42 together introduced a small bug there which caused ExecInsert to request skipping the constraint check but have this not be honored -- in effect doing the check twice. This could have been fixed in a very small patch, but on further analysis of the involved function and its callsites, it turns out to be simpler to give the responsibility of checking the partition constraint fully to the caller, and return ExecConstraints to its original (pre-partitioning) shape where it only checked tuple descriptor-related constraints. Each caller must do partition constraint checking on its own schedule, which is more convenient after commit 2f178441044 anyway. Reported-by: David Rowley Author: David Rowley, Álvaro Herrera Reviewed-by: Amit Langote, Amit Khandekar, Simon Riggs Discussion: https://postgr.es/m/CAKJS1f8w8+awsxgea8wt7_UX8qzOQ=Tm1LD+U1fHqBAkXxkW2w@mail.gmail.com
2018-06-11 22:53:33 +02:00
* Check the constraints of the tuple.
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
*/
Don't needlessly check the partition contraint twice Starting with commit f0e44751d717, ExecConstraints was in charge of running the partition constraint; commit 19c47e7c8202 modified that so that caller could request to skip that checking depending on some conditions, but that commit and 15ce775faa42 together introduced a small bug there which caused ExecInsert to request skipping the constraint check but have this not be honored -- in effect doing the check twice. This could have been fixed in a very small patch, but on further analysis of the involved function and its callsites, it turns out to be simpler to give the responsibility of checking the partition constraint fully to the caller, and return ExecConstraints to its original (pre-partitioning) shape where it only checked tuple descriptor-related constraints. Each caller must do partition constraint checking on its own schedule, which is more convenient after commit 2f178441044 anyway. Reported-by: David Rowley Author: David Rowley, Álvaro Herrera Reviewed-by: Amit Langote, Amit Khandekar, Simon Riggs Discussion: https://postgr.es/m/CAKJS1f8w8+awsxgea8wt7_UX8qzOQ=Tm1LD+U1fHqBAkXxkW2w@mail.gmail.com
2018-06-11 22:53:33 +02:00
if (resultRelationDesc->rd_att->constr)
ExecConstraints(resultRelInfo, slot, estate);
Prevent BEFORE triggers from violating partitioning constraints. Since tuple-routing implicitly checks the partitioning constraints at least for the levels of the partitioning hierarchy it traverses, there's normally no need to revalidate the partitioning constraint after performing tuple routing. However, if there's a BEFORE trigger on the target partition, it could modify the tuple, causing the partitioning constraint to be violated. Catch that case. Also, instead of checking the root table's partition constraint after tuple-routing, check it beforehand. Otherwise, the rules for when the partitioning constraint gets checked get too complicated, because you sometimes have to check part of the constraint but not all of it. This effectively reverts commit 39162b2030fb0a35a6bb28dc636b5a71b8df8d1c in favor of a different approach altogether. Report by me. Initial debugging by Jeevan Ladhe. Patch by Amit Langote, reviewed by me. Discussion: http://postgr.es/m/CA+Tgmoa9DTgeVOqopieV8d1QRpddmP65aCdxyjdYDoEO5pS5KA@mail.gmail.com
2017-06-07 18:45:32 +02:00
Don't needlessly check the partition contraint twice Starting with commit f0e44751d717, ExecConstraints was in charge of running the partition constraint; commit 19c47e7c8202 modified that so that caller could request to skip that checking depending on some conditions, but that commit and 15ce775faa42 together introduced a small bug there which caused ExecInsert to request skipping the constraint check but have this not be honored -- in effect doing the check twice. This could have been fixed in a very small patch, but on further analysis of the involved function and its callsites, it turns out to be simpler to give the responsibility of checking the partition constraint fully to the caller, and return ExecConstraints to its original (pre-partitioning) shape where it only checked tuple descriptor-related constraints. Each caller must do partition constraint checking on its own schedule, which is more convenient after commit 2f178441044 anyway. Reported-by: David Rowley Author: David Rowley, Álvaro Herrera Reviewed-by: Amit Langote, Amit Khandekar, Simon Riggs Discussion: https://postgr.es/m/CAKJS1f8w8+awsxgea8wt7_UX8qzOQ=Tm1LD+U1fHqBAkXxkW2w@mail.gmail.com
2018-06-11 22:53:33 +02:00
/*
* Also check the tuple against the partition constraint, if there is
* one; except that if we got here via tuple-routing, we don't need to
* if there's no BR trigger defined on the partition.
*/
Don't fetch partition check expression during InitResultRelInfo. Since there is only one place that actually needs the partition check expression, namely ExecPartitionCheck, it's better to fetch it from the relcache there. In this way we will never fetch it at all if the query never has use for it, and we still fetch it just once when we do need it. The reason for taking an interest in this is that if the relcache doesn't already have the check expression cached, fetching it requires obtaining AccessShareLock on the partition root. That means that operations that look like they should only touch the partition itself will also take a lock on the root. In particular we observed that TRUNCATE on a partition may take a lock on the partition's root, contributing to a deadlock situation in parallel pg_restore. As written, this patch does have a small cost, which is that we are microscopically reducing efficiency for the case where a partition has an empty check expression. ExecPartitionCheck will be called, and will go through the motions of setting up and checking an empty qual, where before it would not have been called at all. We could avoid that by adding a separate boolean flag to track whether there is a partition expression to test. However, this case only arises for a default partition with no siblings, which surely is not an interesting case in practice. Hence adding complexity for it does not seem like a good trade-off. Amit Langote, per a suggestion by me Discussion: https://postgr.es/m/VI1PR03MB31670CA1BD9625C3A8C5DD05EB230@VI1PR03MB3167.eurprd03.prod.outlook.com
2020-09-16 20:28:11 +02:00
if (resultRelationDesc->rd_rel->relispartition &&
Fix permission checks on constraint violation errors on partitions. If a cross-partition UPDATE violates a constraint on the target partition, and the columns in the new partition are in different physical order than in the parent, the error message can reveal columns that the user does not have SELECT permission on. A similar bug was fixed earlier in commit 804b6b6db4. The cause of the bug is that the callers of the ExecBuildSlotValueDescription() function got confused when constructing the list of modified columns. If the tuple was routed from a parent, we converted the tuple to the parent's format, but the list of modified columns was grabbed directly from the child's RTE entry. ExecUpdateLockMode() had a similar issue. That lead to confusion on which columns are key columns, leading to wrong tuple lock being taken on tables referenced by foreign keys, when a row is updated with INSERT ON CONFLICT UPDATE. A new isolation test is added for that corner case. With this patch, the ri_RangeTableIndex field is no longer set for partitions that don't have an entry in the range table. Previously, it was set to the RTE entry of the parent relation, but that was confusing. NOTE: This modifies the ResultRelInfo struct, replacing the ri_PartitionRoot field with ri_RootResultRelInfo. That's a bit risky to backpatch, because it breaks any extensions accessing the field. The change that ri_RangeTableIndex is not set for partitions could potentially break extensions, too. The ResultRelInfos are visible to FDWs at least, and this patch required small changes to postgres_fdw. Nevertheless, this seem like the least bad option. I don't think these fields widely used in extensions; I don't think there are FDWs out there that uses the FDW "direct update" API, other than postgres_fdw. If there is, you will get a compilation error, so hopefully it is caught quickly. Backpatch to 11, where support for both cross-partition UPDATEs, and unique indexes on partitioned tables, were added. Reviewed-by: Amit Langote Security: CVE-2021-3393
2021-02-08 10:01:51 +01:00
(resultRelInfo->ri_RootResultRelInfo == NULL ||
Don't needlessly check the partition contraint twice Starting with commit f0e44751d717, ExecConstraints was in charge of running the partition constraint; commit 19c47e7c8202 modified that so that caller could request to skip that checking depending on some conditions, but that commit and 15ce775faa42 together introduced a small bug there which caused ExecInsert to request skipping the constraint check but have this not be honored -- in effect doing the check twice. This could have been fixed in a very small patch, but on further analysis of the involved function and its callsites, it turns out to be simpler to give the responsibility of checking the partition constraint fully to the caller, and return ExecConstraints to its original (pre-partitioning) shape where it only checked tuple descriptor-related constraints. Each caller must do partition constraint checking on its own schedule, which is more convenient after commit 2f178441044 anyway. Reported-by: David Rowley Author: David Rowley, Álvaro Herrera Reviewed-by: Amit Langote, Amit Khandekar, Simon Riggs Discussion: https://postgr.es/m/CAKJS1f8w8+awsxgea8wt7_UX8qzOQ=Tm1LD+U1fHqBAkXxkW2w@mail.gmail.com
2018-06-11 22:53:33 +02:00
(resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_insert_before_row)))
ExecPartitionCheck(resultRelInfo, slot, estate, true);
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
if (onconflict != ONCONFLICT_NONE && resultRelInfo->ri_NumIndices > 0)
{
/* Perform a speculative insertion. */
uint32 specToken;
ItemPointerData conflictTid;
bool specConflict;
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
List *arbiterIndexes;
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
arbiterIndexes = resultRelInfo->ri_onConflictArbiterIndexes;
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* Do a non-conclusive check for conflicts first.
*
* We're not holding any locks yet, so this doesn't guarantee that
* the later insert won't conflict. But it avoids leaving behind
* a lot of canceled speculative insertions, if you run a lot of
* INSERT ON CONFLICT statements that do conflict.
*
* We loop back here if we find a conflict below, either during
* the pre-check, or when we re-check after inserting the tuple
Add CHECK_FOR_INTERRUPTS in ExecInsert's speculative insertion loop. Ordinarily the functions called in this loop ought to have plenty of CFIs themselves; but we've now seen a case where no such CFI is reached, making the loop uninterruptible. Even though that's from a recently-introduced bug, it seems prudent to install a CFI at the loop level in all branches. Per discussion of bug #17558 from Andrew Kesper (an actual fix for that bug will follow). Discussion: https://postgr.es/m/17558-3f6599ffcf52fd4a@postgresql.org
2022-08-04 20:10:06 +02:00
* speculatively. Better allow interrupts in case some bug makes
* this an infinite loop.
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
*/
vlock:
Add CHECK_FOR_INTERRUPTS in ExecInsert's speculative insertion loop. Ordinarily the functions called in this loop ought to have plenty of CFIs themselves; but we've now seen a case where no such CFI is reached, making the loop uninterruptible. Even though that's from a recently-introduced bug, it seems prudent to install a CFI at the loop level in all branches. Per discussion of bug #17558 from Andrew Kesper (an actual fix for that bug will follow). Discussion: https://postgr.es/m/17558-3f6599ffcf52fd4a@postgresql.org
2022-08-04 20:10:06 +02:00
CHECK_FOR_INTERRUPTS();
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
specConflict = false;
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
if (!ExecCheckIndexConstraints(resultRelInfo, slot, estate,
&conflictTid, arbiterIndexes))
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
{
/* committed conflict tuple found */
if (onconflict == ONCONFLICT_UPDATE)
{
/*
* In case of ON CONFLICT DO UPDATE, execute the UPDATE
* part. Be prepared to retry if the UPDATE fails because
* of another concurrent UPDATE/DELETE to the conflict
* tuple.
*/
TupleTableSlot *returning = NULL;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
if (ExecOnConflictUpdate(context, resultRelInfo,
&conflictTid, slot, canSetTag,
&returning))
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
{
Fix IndexOnlyScan counter for heap fetches in parallel mode The HeapFetches counter was using a simple value in IndexOnlyScanState, which fails to propagate values from parallel workers; so the counts are wrong when IndexOnlyScan runs in parallel. Move it to Instrumentation, like all the other counters. While at it, change INSERT ON CONFLICT conflicting tuple counter to use the new ntuples2 instead of nfiltered2, which is a blatant misuse. Discussion: https://postgr.es/m/20180409215851.idwc75ct2bzi6tea@alvherre.pgsql
2018-04-10 20:56:15 +02:00
InstrCountTuples2(&mtstate->ps, 1);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
return returning;
}
else
goto vlock;
}
else
{
/*
* In case of ON CONFLICT DO NOTHING, do nothing. However,
* verify that the tuple is visible to the executor's MVCC
* snapshot at higher isolation levels.
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
*
* Using ExecGetReturningSlot() to store the tuple for the
* recheck isn't that pretty, but we can't trivially use
* the input slot, because it might not be of a compatible
* type. As there's no conflicting usage of
* ExecGetReturningSlot() in the DO NOTHING case...
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
*/
Assert(onconflict == ONCONFLICT_NOTHING);
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
ExecCheckTIDVisible(estate, resultRelInfo, &conflictTid,
ExecGetReturningSlot(estate, resultRelInfo));
Fix IndexOnlyScan counter for heap fetches in parallel mode The HeapFetches counter was using a simple value in IndexOnlyScanState, which fails to propagate values from parallel workers; so the counts are wrong when IndexOnlyScan runs in parallel. Move it to Instrumentation, like all the other counters. While at it, change INSERT ON CONFLICT conflicting tuple counter to use the new ntuples2 instead of nfiltered2, which is a blatant misuse. Discussion: https://postgr.es/m/20180409215851.idwc75ct2bzi6tea@alvherre.pgsql
2018-04-10 20:56:15 +02:00
InstrCountTuples2(&mtstate->ps, 1);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
return NULL;
}
}
/*
* Before we start insertion proper, acquire our "speculative
* insertion lock". Others can use that to wait for us to decide
* if we're going to go ahead with the insertion, instead of
* waiting for the whole transaction to complete.
*/
specToken = SpeculativeInsertionLockAcquire(GetCurrentTransactionId());
/* insert the tuple, with the speculative token */
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
table_tuple_insert_speculative(resultRelationDesc, slot,
estate->es_output_cid,
0,
NULL,
specToken);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* insert index entries for tuple */
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
recheckIndexes = ExecInsertIndexTuples(resultRelInfo,
Pass down "logically unchanged index" hint. Add an executor aminsert() hint mechanism that informs index AMs that the incoming index tuple (the tuple that accompanies the hint) is not being inserted by execution of an SQL statement that logically modifies any of the index's key columns. The hint is received by indexes when an UPDATE takes place that does not apply an optimization like heapam's HOT (though only for indexes where all key columns are logically unchanged). Any index tuple that receives the hint on insert is expected to be a duplicate of at least one existing older version that is needed for the same logical row. Related versions will typically be stored on the same index page, at least within index AMs that apply the hint. Recognizing the difference between MVCC version churn duplicates and true logical row duplicates at the index AM level can help with cleanup of garbage index tuples. Cleanup can intelligently target tuples that are likely to be garbage, without wasting too many cycles on less promising tuples/pages (index pages with little or no version churn). This is infrastructure for an upcoming commit that will teach nbtree to perform bottom-up index deletion. No index AM actually applies the hint just yet. Author: Peter Geoghegan <pg@bowt.ie> Reviewed-By: Victor Yegorov <vyegorov@gmail.com> Discussion: https://postgr.es/m/CAH2-Wz=CEKFa74EScx_hFVshCOn6AA5T-ajFASTdzipdkLTNQQ@mail.gmail.com
2021-01-13 17:11:00 +01:00
slot, estate, false, true,
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
&specConflict,
Ignore BRIN indexes when checking for HOT updates When determining whether an index update may be skipped by using HOT, we can ignore attributes indexed by block summarizing indexes without references to individual tuples that need to be cleaned up. A new type TU_UpdateIndexes provides a signal to the executor to determine which indexes to update - no indexes, all indexes, or only the summarizing indexes. This also removes rd_indexattr list, and replaces it with rd_attrsvalid flag. The list was not used anywhere, and a simple flag is sufficient. This was originally committed as 5753d4ee32, but then got reverted by e3fcca0d0d because of correctness issues. Original patch by Josef Simanek, various fixes and improvements by Tomas Vondra and me. Authors: Matthias van de Meent, Josef Simanek, Tomas Vondra Reviewed-by: Tomas Vondra, Alvaro Herrera Discussion: https://postgr.es/m/05ebcb44-f383-86e3-4f31-0a97a55634cf@enterprisedb.com Discussion: https://postgr.es/m/CAFp7QwpMRGcDAQumN7onN9HjrJ3u4X3ZRXdGFT0K5G2JWvnbWg%40mail.gmail.com
2023-03-20 10:34:07 +01:00
arbiterIndexes,
false);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* adjust the tuple's state accordingly */
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
table_tuple_complete_speculative(resultRelationDesc, slot,
specToken, !specConflict);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* Wake up anyone waiting for our decision. They will re-check
* the tuple, see that it's no longer speculative, and wait on our
* XID as if this was a regularly inserted tuple all along. Or if
* we killed the tuple, they will see it's dead, and proceed as if
* the tuple never existed.
*/
SpeculativeInsertionLockRelease(GetCurrentTransactionId());
/*
* If there was a conflict, start from the beginning. We'll do
* the pre-check again, which will now find the conflicting tuple
* (unless it aborts before we get there).
*/
if (specConflict)
{
list_free(recheckIndexes);
goto vlock;
}
/* Since there was no insertion conflict, we're done */
}
else
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
/* insert the tuple normally */
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
table_tuple_insert(resultRelationDesc, slot,
estate->es_output_cid,
0, NULL);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* insert index entries for tuple */
if (resultRelInfo->ri_NumIndices > 0)
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
recheckIndexes = ExecInsertIndexTuples(resultRelInfo,
slot, estate, false,
Ignore BRIN indexes when checking for HOT updates When determining whether an index update may be skipped by using HOT, we can ignore attributes indexed by block summarizing indexes without references to individual tuples that need to be cleaned up. A new type TU_UpdateIndexes provides a signal to the executor to determine which indexes to update - no indexes, all indexes, or only the summarizing indexes. This also removes rd_indexattr list, and replaces it with rd_attrsvalid flag. The list was not used anywhere, and a simple flag is sufficient. This was originally committed as 5753d4ee32, but then got reverted by e3fcca0d0d because of correctness issues. Original patch by Josef Simanek, various fixes and improvements by Tomas Vondra and me. Authors: Matthias van de Meent, Josef Simanek, Tomas Vondra Reviewed-by: Tomas Vondra, Alvaro Herrera Discussion: https://postgr.es/m/05ebcb44-f383-86e3-4f31-0a97a55634cf@enterprisedb.com Discussion: https://postgr.es/m/CAFp7QwpMRGcDAQumN7onN9HjrJ3u4X3ZRXdGFT0K5G2JWvnbWg%40mail.gmail.com
2023-03-20 10:34:07 +01:00
false, NULL, NIL,
false);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
}
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
if (canSetTag)
(estate->es_processed)++;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
/*
* If this insert is the result of a partition key update that moved the
* tuple to a new partition, put this row into the transition NEW TABLE,
* if there is one. We need to do this separately for DELETE and INSERT
* because they happen on different tables.
*/
ar_insert_trig_tcs = mtstate->mt_transition_capture;
if (mtstate->operation == CMD_UPDATE && mtstate->mt_transition_capture
&& mtstate->mt_transition_capture->tcs_update_new_table)
{
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
ExecARUpdateTriggers(estate, resultRelInfo,
NULL, NULL,
NULL,
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
NULL,
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
slot,
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
NULL,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
mtstate->mt_transition_capture,
false);
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
/*
* We've already captured the NEW TABLE row, so make sure any AR
* INSERT trigger fired below doesn't capture it again.
*/
ar_insert_trig_tcs = NULL;
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* AFTER ROW INSERT Triggers */
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
ExecARInsertTriggers(estate, resultRelInfo, slot, recheckIndexes,
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
ar_insert_trig_tcs);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Fix memory leak created by deferrable-index-constraints patches. We need to free the OID list returned by ExecInsertIndexTuples to avoid a query-lifespan memory leak. When many rows require rechecking, this can be a significant leak --- it's even more than the space used for the queued trigger events. Dean Rasheed
2010-01-31 19:15:39 +01:00
list_free(recheckIndexes);
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
/*
* Check any WITH CHECK OPTION constraints from parent views. We are
* required to do this after testing all constraints and uniqueness
* violations per the SQL spec, so we do it after actually inserting the
* record into the heap and all indexes.
*
* ExecWithCheckOptions will elog(ERROR) if a violation is found, so the
Fix various typos and grammar errors in comments. Author: Dmitriy Olshevskiy Discussion: 553D00A6.4090205@bk.ru
2015-04-26 18:42:31 +02:00
* tuple will never be seen, if it violates the WITH CHECK OPTION.
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
*
* ExecWithCheckOptions() will skip any WCOs which are not of the kind we
* are looking for at this point.
*/
WITH CHECK OPTION support for auto-updatable VIEWs For simple views which are automatically updatable, this patch allows the user to specify what level of checking should be done on records being inserted or updated. For 'LOCAL CHECK', new tuples are validated against the conditionals of the view they are being inserted into, while for 'CASCADED CHECK' the new tuples are validated against the conditionals for all views involved (from the top down). This option is part of the SQL specification. Dean Rasheed, reviewed by Pavel Stehule
2013-07-18 23:10:16 +02:00
if (resultRelInfo->ri_WithCheckOptions != NIL)
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
ExecWithCheckOptions(WCO_VIEW_CHECK, resultRelInfo, slot, estate);
WITH CHECK OPTION support for auto-updatable VIEWs For simple views which are automatically updatable, this patch allows the user to specify what level of checking should be done on records being inserted or updated. For 'LOCAL CHECK', new tuples are validated against the conditionals of the view they are being inserted into, while for 'CASCADED CHECK' the new tuples are validated against the conditionals for all views involved (from the top down). This option is part of the SQL specification. Dean Rasheed, reviewed by Pavel Stehule
2013-07-18 23:10:16 +02:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* Process RETURNING if present */
if (resultRelInfo->ri_projectReturning)
Fix RETURNING to work correctly with partition tuple routing. In ExecInsert(), do not switch back to the root partitioned table ResultRelInfo until after we finish ExecProcessReturning(), so that RETURNING projection is done using the partition's descriptor. For the projection to work correctly, we must initialize the same for each leaf partition during ModifyTableState initialization. Amit Langote
2017-01-19 19:20:11 +01:00
result = ExecProcessReturning(resultRelInfo, slot, planSlot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
if (inserted_tuple)
*inserted_tuple = slot;
if (insert_destrel)
*insert_destrel = resultRelInfo;
Fix RETURNING to work correctly with partition tuple routing. In ExecInsert(), do not switch back to the root partitioned table ResultRelInfo until after we finish ExecProcessReturning(), so that RETURNING projection is done using the partition's descriptor. For the projection to work correctly, we must initialize the same for each leaf partition during ModifyTableState initialization. Amit Langote
2017-01-19 19:20:11 +01:00
return result;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
/* ----------------------------------------------------------------
* ExecBatchInsert
*
* Insert multiple tuples in an efficient way.
* Currently, this handles inserting into a foreign table without
* RETURNING clause.
* ----------------------------------------------------------------
*/
static void
ExecBatchInsert(ModifyTableState *mtstate,
ResultRelInfo *resultRelInfo,
TupleTableSlot **slots,
TupleTableSlot **planSlots,
int numSlots,
EState *estate,
bool canSetTag)
{
int i;
int numInserted = numSlots;
TupleTableSlot *slot = NULL;
TupleTableSlot **rslots;
/*
* insert into foreign table: let the FDW do it
*/
rslots = resultRelInfo->ri_FdwRoutine->ExecForeignBatchInsert(estate,
resultRelInfo,
slots,
planSlots,
&numInserted);
for (i = 0; i < numInserted; i++)
{
slot = rslots[i];
/*
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
* AFTER ROW Triggers might reference the tableoid column, so
* (re-)initialize tts_tableOid before evaluating them.
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
*/
slot->tts_tableOid = RelationGetRelid(resultRelInfo->ri_RelationDesc);
/* AFTER ROW INSERT Triggers */
ExecARInsertTriggers(estate, resultRelInfo, slot, NIL,
mtstate->mt_transition_capture);
/*
* Check any WITH CHECK OPTION constraints from parent views. See the
* comment in ExecInsert.
*/
if (resultRelInfo->ri_WithCheckOptions != NIL)
ExecWithCheckOptions(WCO_VIEW_CHECK, resultRelInfo, slot, estate);
}
if (canSetTag && numInserted > 0)
estate->es_processed += numInserted;
Fix buffer refcount leak with FDW bulk inserts The leak would show up when using batch inserts with foreign tables included in a partition tree, as the slots used in the batch were not reset once processed. In order to fix this problem, some ExecClearTuple() are added to clean up the slots used once a batch is filled and processed, mapping with the number of slots currently in use as tracked by the counter ri_NumSlots. This buffer refcount leak has been introduced in b676ac4 with the addition of the executor facility to improve bulk inserts for FDWs, so backpatch down to 14. Alexander has provided the patch (slightly modified by me). The test for postgres_fdw comes from me, based on the test case that the author has sent in the report. Author: Alexander Pyhalov Discussion: https://postgr.es/m/b035780a740efd38dc30790c76927255@postgrespro.ru Backpatch-through: 14
2023-04-25 02:42:19 +02:00
/* Clean up all the slots, ready for the next batch */
for (i = 0; i < numSlots; i++)
{
ExecClearTuple(slots[i]);
ExecClearTuple(planSlots[i]);
}
resultRelInfo->ri_NumSlots = 0;
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
}
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
/*
* ExecPendingInserts -- flushes all pending inserts to the foreign tables
*/
static void
ExecPendingInserts(EState *estate)
{
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
ListCell *l1,
*l2;
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
forboth(l1, estate->es_insert_pending_result_relations,
l2, estate->es_insert_pending_modifytables)
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
{
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
ResultRelInfo *resultRelInfo = (ResultRelInfo *) lfirst(l1);
ModifyTableState *mtstate = (ModifyTableState *) lfirst(l2);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
Assert(mtstate);
ExecBatchInsert(mtstate, resultRelInfo,
resultRelInfo->ri_Slots,
resultRelInfo->ri_PlanSlots,
resultRelInfo->ri_NumSlots,
estate, mtstate->canSetTag);
}
list_free(estate->es_insert_pending_result_relations);
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
list_free(estate->es_insert_pending_modifytables);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
estate->es_insert_pending_result_relations = NIL;
Remove new structure member from ResultRelInfo. In commit ffbb7e65a, I added a ModifyTableState member to ResultRelInfo to save the owning ModifyTableState for use by nodeModifyTable.c when performing batch inserts, but as pointed out by Tom Lane, that changed the array stride of es_result_relations, and that would break any previously-compiled extension code that accesses that array. Fix by removing that member from ResultRelInfo and instead adding a List member at the end of EState to save such ModifyTableStates. Per report from Tom Lane. Back-patch to v14, like the previous commit; I chose to apply the patch to HEAD as well, to make back-patching easy. Discussion: http://postgr.es/m/4065383.1669395453%40sss.pgh.pa.us
2022-12-08 08:15:00 +01:00
estate->es_insert_pending_modifytables = NIL;
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
}
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* ExecDeletePrologue -- subroutine for ExecDelete
*
* Prepare executor state for DELETE. Actually, the only thing we have to do
* here is execute BEFORE ROW triggers. We return false if one of them makes
* the delete a no-op; otherwise, return true.
*/
static bool
ExecDeletePrologue(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
ItemPointer tupleid, HeapTuple oldtuple,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
TupleTableSlot **epqreturnslot, TM_Result *result)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
if (result)
*result = TM_Ok;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/* BEFORE ROW DELETE triggers */
if (resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_delete_before_row)
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
{
/* Flush any pending inserts, so rows are visible to the triggers */
if (context->estate->es_insert_pending_result_relations != NIL)
ExecPendingInserts(context->estate);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return ExecBRDeleteTriggers(context->estate, context->epqstate,
resultRelInfo, tupleid, oldtuple,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
epqreturnslot, result, &context->tmfd);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
}
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return true;
}
/*
* ExecDeleteAct -- subroutine for ExecDelete
*
* Actually delete the tuple from a plain table.
*
* Caller is in charge of doing EvalPlanQual as necessary
*/
static TM_Result
ExecDeleteAct(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
ItemPointer tupleid, bool changingPart)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
EState *estate = context->estate;
return table_tuple_delete(resultRelInfo->ri_RelationDesc, tupleid,
estate->es_output_cid,
estate->es_snapshot,
estate->es_crosscheck_snapshot,
true /* wait for commit */ ,
&context->tmfd,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
changingPart);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
}
/*
* ExecDeleteEpilogue -- subroutine for ExecDelete
*
* Closing steps of tuple deletion; this invokes AFTER FOR EACH ROW triggers,
* including the UPDATE triggers if the deletion is being done as part of a
* cross-partition tuple move.
*/
static void
ExecDeleteEpilogue(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
ItemPointer tupleid, HeapTuple oldtuple, bool changingPart)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
ModifyTableState *mtstate = context->mtstate;
EState *estate = context->estate;
TransitionCaptureState *ar_delete_trig_tcs;
/*
* If this delete is the result of a partition key update that moved the
* tuple to a new partition, put this row into the transition OLD TABLE,
* if there is one. We need to do this separately for DELETE and INSERT
* because they happen on different tables.
*/
ar_delete_trig_tcs = mtstate->mt_transition_capture;
if (mtstate->operation == CMD_UPDATE && mtstate->mt_transition_capture &&
mtstate->mt_transition_capture->tcs_update_old_table)
{
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
ExecARUpdateTriggers(estate, resultRelInfo,
NULL, NULL,
tupleid, oldtuple,
NULL, NULL, mtstate->mt_transition_capture,
false);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
Fix copy-and-pasteo in comment.
2022-11-02 10:15:00 +01:00
* We've already captured the OLD TABLE row, so make sure any AR
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
* DELETE trigger fired below doesn't capture it again.
*/
ar_delete_trig_tcs = NULL;
}
/* AFTER ROW DELETE Triggers */
ExecARDeleteTriggers(estate, resultRelInfo, tupleid, oldtuple,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
ar_delete_trig_tcs, changingPart);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* ----------------------------------------------------------------
* ExecDelete
*
* DELETE is like UPDATE, except that we delete the tuple and no
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
* index modifications are needed.
*
* When deleting from a table, tupleid identifies the tuple to
* delete and oldtuple is NULL. When deleting from a view,
* oldtuple is passed to the INSTEAD OF triggers and identifies
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
* what to delete, and tupleid is invalid. When deleting from a
Offer triggers on foreign tables. This covers all the SQL-standard trigger types supported for regular tables; it does not cover constraint triggers. The approach for acquiring the old row mirrors that for view INSTEAD OF triggers. For AFTER ROW triggers, we spool the foreign tuples to a tuplestore. This changes the FDW API contract; when deciding which columns to populate in the slot returned from data modification callbacks, writable FDWs will need to check for AFTER ROW triggers in addition to checking for a RETURNING clause. In support of the feature addition, refactor the TriggerFlags bits and the assembly of old tuples in ModifyTable. Ronan Dunklau, reviewed by KaiGai Kohei; some additional hacking by me.
2014-03-23 07:16:34 +01:00
* foreign table, tupleid is invalid; the FDW has to figure out
* which row to delete using data from the planSlot. oldtuple is
* passed to foreign table triggers; it is NULL when the foreign
Allow using the updated tuple while moving it to a different partition. An update that causes the tuple to be moved to a different partition was missing out on re-constructing the to-be-updated tuple, based on the latest tuple in the update chain. Instead, it's simply deleting the latest tuple and inserting a new tuple in the new partition based on the old tuple. Commit 2f17844104 didn't consider this case, so some of the updates were getting lost. In passing, change the argument order for output parameter in ExecDelete and add some commentary about it. Reported-by: Pavan Deolasee Author: Amit Khandekar, with minor changes by me Reviewed-by: Dilip Kumar, Amit Kapila and Alvaro Herrera Backpatch-through: 11 Discussion: https://postgr.es/m/CAJ3gD9fRbEzDqdeDq1jxqZUb47kJn+tQ7=Bcgjc8quqKsDViKQ@mail.gmail.com
2018-07-12 09:21:39 +02:00
* table has no relevant triggers. We use tupleDeleted to indicate
* whether the tuple is actually deleted, callers can use it to
* decide whether to continue the operation. When this DELETE is a
* part of an UPDATE of partition-key, then the slot returned by
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
* EvalPlanQual() is passed back using output parameter epqreturnslot.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*
* Returns RETURNING result if any, otherwise NULL.
* ----------------------------------------------------------------
*/
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
static TupleTableSlot *
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecDelete(ModifyTableContext *context,
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
ResultRelInfo *resultRelInfo,
Fix transition tables for partition/inheritance. We disallow row-level triggers with transition tables on child tables. Transition tables for triggers on the parent table contain only those columns present in the parent. (We can't mix tuple formats in a single transition table.) Patch by Thomas Munro Discussion: https://postgr.es/m/CA%2BTgmoZzTBBAsEUh4MazAN7ga%3D8SsMC-Knp-6cetts9yNZUCcg%40mail.gmail.com
2017-06-28 19:55:03 +02:00
ItemPointer tupleid,
Offer triggers on foreign tables. This covers all the SQL-standard trigger types supported for regular tables; it does not cover constraint triggers. The approach for acquiring the old row mirrors that for view INSTEAD OF triggers. For AFTER ROW triggers, we spool the foreign tuples to a tuplestore. This changes the FDW API contract; when deciding which columns to populate in the slot returned from data modification callbacks, writable FDWs will need to check for AFTER ROW triggers in addition to checking for a RETURNING clause. In support of the feature addition, refactor the TriggerFlags bits and the assembly of old tuples in ModifyTable. Ronan Dunklau, reviewed by KaiGai Kohei; some additional hacking by me.
2014-03-23 07:16:34 +01:00
HeapTuple oldtuple,
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
bool processReturning,
Allow using the updated tuple while moving it to a different partition. An update that causes the tuple to be moved to a different partition was missing out on re-constructing the to-be-updated tuple, based on the latest tuple in the update chain. Instead, it's simply deleting the latest tuple and inserting a new tuple in the new partition based on the old tuple. Commit 2f17844104 didn't consider this case, so some of the updates were getting lost. In passing, change the argument order for output parameter in ExecDelete and add some commentary about it. Reported-by: Pavan Deolasee Author: Amit Khandekar, with minor changes by me Reviewed-by: Dilip Kumar, Amit Kapila and Alvaro Herrera Backpatch-through: 11 Discussion: https://postgr.es/m/CAJ3gD9fRbEzDqdeDq1jxqZUb47kJn+tQ7=Bcgjc8quqKsDViKQ@mail.gmail.com
2018-07-12 09:21:39 +02:00
bool changingPart,
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
bool canSetTag,
Allow using the updated tuple while moving it to a different partition. An update that causes the tuple to be moved to a different partition was missing out on re-constructing the to-be-updated tuple, based on the latest tuple in the update chain. Instead, it's simply deleting the latest tuple and inserting a new tuple in the new partition based on the old tuple. Commit 2f17844104 didn't consider this case, so some of the updates were getting lost. In passing, change the argument order for output parameter in ExecDelete and add some commentary about it. Reported-by: Pavan Deolasee Author: Amit Khandekar, with minor changes by me Reviewed-by: Dilip Kumar, Amit Kapila and Alvaro Herrera Backpatch-through: 11 Discussion: https://postgr.es/m/CAJ3gD9fRbEzDqdeDq1jxqZUb47kJn+tQ7=Bcgjc8quqKsDViKQ@mail.gmail.com
2018-07-12 09:21:39 +02:00
bool *tupleDeleted,
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
TupleTableSlot **epqreturnslot)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
EState *estate = context->estate;
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
Relation resultRelationDesc = resultRelInfo->ri_RelationDesc;
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
TupleTableSlot *slot = NULL;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
TM_Result result;
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
if (tupleDeleted)
*tupleDeleted = false;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* Prepare for the delete. This includes BEFORE ROW triggers, so we're
* done if it says we are.
*/
if (!ExecDeletePrologue(context, resultRelInfo, tupleid, oldtuple,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
epqreturnslot, NULL))
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
/* INSTEAD OF ROW DELETE Triggers */
if (resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_delete_instead_row)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
bool dodelete;
Assert(oldtuple != NULL);
Offer triggers on foreign tables. This covers all the SQL-standard trigger types supported for regular tables; it does not cover constraint triggers. The approach for acquiring the old row mirrors that for view INSTEAD OF triggers. For AFTER ROW triggers, we spool the foreign tuples to a tuplestore. This changes the FDW API contract; when deciding which columns to populate in the slot returned from data modification callbacks, writable FDWs will need to check for AFTER ROW triggers in addition to checking for a RETURNING clause. In support of the feature addition, refactor the TriggerFlags bits and the assembly of old tuples in ModifyTable. Ronan Dunklau, reviewed by KaiGai Kohei; some additional hacking by me.
2014-03-23 07:16:34 +01:00
dodelete = ExecIRDeleteTriggers(estate, resultRelInfo, oldtuple);
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
if (!dodelete) /* "do nothing" */
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
return NULL;
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
}
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
else if (resultRelInfo->ri_FdwRoutine)
{
/*
* delete from foreign table: let the FDW do it
*
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
* We offer the returning slot as a place to store RETURNING data,
* although the FDW can return some other slot if it wants.
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
*/
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
slot = ExecGetReturningSlot(estate, resultRelInfo);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
slot = resultRelInfo->ri_FdwRoutine->ExecForeignDelete(estate,
resultRelInfo,
slot,
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
context->planSlot);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
if (slot == NULL) /* "do nothing" */
return NULL;
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
/*
* RETURNING expressions might reference the tableoid column, so
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
* (re)initialize tts_tableOid before evaluating them.
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
*/
Move TupleTableSlots boolean member into one flag variable. There's several reasons for this change: 1) It reduces the total size of TupleTableSlot / reduces alignment padding, making the commonly accessed members fit into a single cacheline (but we currently do not force proper alignment, so that's not yet guaranteed to be helpful) 2) Combining the booleans into a flag allows to combine read/writes from memory. 3) With the upcoming slot abstraction changes, it allows to have core and extended flags, in a memory efficient way. Author: Ashutosh Bapat and Andres Freund Discussion: https://postgr.es/m/20180220224318.gw4oe5jadhpmcdnm@alap3.anarazel.de
2018-10-16 00:24:33 +02:00
if (TTS_EMPTY(slot))
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
ExecStoreAllNullTuple(slot);
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
slot->tts_tableOid = RelationGetRelid(resultRelationDesc);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
}
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
else
{
/*
* delete the tuple
*
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
* Note: if context->estate->es_crosscheck_snapshot isn't
* InvalidSnapshot, we check that the row to be deleted is visible to
* that snapshot, and throw a can't-serialize error if not. This is a
* special-case behavior needed for referential integrity updates in
* transaction-snapshot mode transactions.
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
*/
Remove unnecessary semicolons after goto labels According to the C standard, a label must followed by a statement. If there was ever a time we needed an empty statement here, it was a long time ago. Japin Li Reviewed by Julien Rouhaud Discussion: https://www.postgresql.org/message-id/MEYP282MB16690F40189A4F060B41D56DB65E9%40MEYP282MB1669.AUSP282.PROD.OUTLOOK.COM
2022-10-10 10:08:38 +02:00
ldelete:
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
result = ExecDeleteAct(context, resultRelInfo, tupleid, changingPart);
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
switch (result)
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_SelfModified:
pgindent run for release 9.3 This is the first run of the Perl-based pgindent script. Also update pgindent instructions.
2013-05-29 22:58:43 +02:00
Throw error if expiring tuple is again updated or deleted. This prevents surprising behavior when a FOR EACH ROW trigger BEFORE UPDATE or BEFORE DELETE directly or indirectly updates or deletes the the old row. Prior to this patch the requested action on the row could be silently ignored while all triggered actions based on the occurence of the requested action could be committed. One example of how this could happen is if the BEFORE DELETE trigger for a "parent" row deleted "children" which had trigger functions to update summary or status data on the parent. This also prevents similar surprising problems if the query has a volatile function which updates a target row while it is already being updated. There are related issues present in FOR UPDATE cursors and READ COMMITTED queries which are not handled by this patch. These issues need further evalution to determine what change, if any, is needed. Where the new error messages are generated, in most cases the best fix will be to move code from the BEFORE trigger to an AFTER trigger. Where this is not feasible, the trigger can avoid the error by re-issuing the triggering statement and returning NULL. Documentation changes will be submitted in a separate patch. Kevin Grittner and Tom Lane with input from Florian Pflug and Robert Haas, based on problems encountered during conversion of Wisconsin Circuit Court trigger logic to plpgsql triggers.
2012-10-26 21:55:36 +02:00
/*
* The target tuple was already updated or deleted by the
* current command, or by a later command in the current
* transaction. The former case is possible in a join DELETE
* where multiple tuples join to the same target tuple. This
* is somewhat questionable, but Postgres has always allowed
* it: we just ignore additional deletion attempts.
*
* The latter case arises if the tuple is modified by a
* command in a BEFORE trigger, or perhaps by a command in a
* volatile function used in the query. In such situations we
* should not ignore the deletion, but it is equally unsafe to
* proceed. We don't want to discard the original DELETE
* while keeping the triggered actions based on its deletion;
* and it would be no better to allow the original DELETE
* while discarding updates that it triggered. The row update
* carries some information that might be important according
* to business rules; so throwing an error is the only safe
* course.
*
* If a trigger actually intends this type of interaction, it
* can re-execute the DELETE and then return NULL to cancel
* the outer delete.
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
if (context->tmfd.cmax != estate->es_output_cid)
Throw error if expiring tuple is again updated or deleted. This prevents surprising behavior when a FOR EACH ROW trigger BEFORE UPDATE or BEFORE DELETE directly or indirectly updates or deletes the the old row. Prior to this patch the requested action on the row could be silently ignored while all triggered actions based on the occurence of the requested action could be committed. One example of how this could happen is if the BEFORE DELETE trigger for a "parent" row deleted "children" which had trigger functions to update summary or status data on the parent. This also prevents similar surprising problems if the query has a volatile function which updates a target row while it is already being updated. There are related issues present in FOR UPDATE cursors and READ COMMITTED queries which are not handled by this patch. These issues need further evalution to determine what change, if any, is needed. Where the new error messages are generated, in most cases the best fix will be to move code from the BEFORE trigger to an AFTER trigger. Where this is not feasible, the trigger can avoid the error by re-issuing the triggering statement and returning NULL. Documentation changes will be submitted in a separate patch. Kevin Grittner and Tom Lane with input from Florian Pflug and Robert Haas, based on problems encountered during conversion of Wisconsin Circuit Court trigger logic to plpgsql triggers.
2012-10-26 21:55:36 +02:00
ereport(ERROR,
(errcode(ERRCODE_TRIGGERED_DATA_CHANGE_VIOLATION),
Fix a number of issues around modifying a previously updated row. This commit fixes three, unfortunately related, issues: 1) Since 5db6df0c01, the introduction of DML via tableam, it was possible to trigger "ERROR: unexpected table_lock_tuple status: 1" when updating a row that was previously updated in the same transaction - but only when the previously updated row was before updated in a concurrent transaction (and READ COMMITTED was used). The reason for that was that that case simply wasn't expected. Fixing that lead to: 2) Even before the above commit, there were error checks (introduced in 6868ed7491b7) preventing a row being updated by different commands within the same statement (say in a function called by an UPDATE) - but that check wasn't performed when the row was first updated in a concurrent transaction - instead the second update was silently skipped in that case. After this change we throw the same error as we'd without the concurrent transaction. 3) The error messages (introduced in 6868ed7491b7) preventing such updates emitted the same error message for both DELETE and UPDATE ("tuple to be updated was already modified by an operation triggered by the current command"). While that could be changed separately, it made it hard to write tests that verify the correct correct behavior of the code. This commit changes heap's implementation of table_lock_tuple() to return TM_SelfModified instead of TM_Invisible (previously loosely modeled after EvalPlanQualFetch), and teaches nodeModifyTable.c to handle that in response to table_lock_tuple() and not just in response to table_(delete|update). Additionally it fixes the wrong error message (see 3 above). The comment for table_lock_tuple() is also adjusted to state that TM_Deleted won't return information in TM_FailureData - it'll not always be available. This also adds tests to ensure that DELETE/UPDATE correctly error out when affecting a row that concurrently was modified by another transaction. Author: Andres Freund Reported-By: Tom Lane, when investigating a bug bug fix to another bug by Amit Langote Discussion: https://postgr.es/m/19321.1554567786@sss.pgh.pa.us
2019-04-08 07:14:47 +02:00
errmsg("tuple to be deleted was already modified by an operation triggered by the current command"),
Throw error if expiring tuple is again updated or deleted. This prevents surprising behavior when a FOR EACH ROW trigger BEFORE UPDATE or BEFORE DELETE directly or indirectly updates or deletes the the old row. Prior to this patch the requested action on the row could be silently ignored while all triggered actions based on the occurence of the requested action could be committed. One example of how this could happen is if the BEFORE DELETE trigger for a "parent" row deleted "children" which had trigger functions to update summary or status data on the parent. This also prevents similar surprising problems if the query has a volatile function which updates a target row while it is already being updated. There are related issues present in FOR UPDATE cursors and READ COMMITTED queries which are not handled by this patch. These issues need further evalution to determine what change, if any, is needed. Where the new error messages are generated, in most cases the best fix will be to move code from the BEFORE trigger to an AFTER trigger. Where this is not feasible, the trigger can avoid the error by re-issuing the triggering statement and returning NULL. Documentation changes will be submitted in a separate patch. Kevin Grittner and Tom Lane with input from Florian Pflug and Robert Haas, based on problems encountered during conversion of Wisconsin Circuit Court trigger logic to plpgsql triggers.
2012-10-26 21:55:36 +02:00
errhint("Consider using an AFTER trigger instead of a BEFORE trigger to propagate changes to other rows.")));
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
/* Else, already deleted by self; nothing to do */
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
return NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Ok:
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
break;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Updated:
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
TupleTableSlot *inputslot;
TupleTableSlot *epqslot;
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent update")));
/*
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
* Already know that we're going to need to do EPQ, so
* fetch tuple directly into the right slot.
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
EvalPlanQualBegin(context->epqstate);
inputslot = EvalPlanQualSlot(context->epqstate, resultRelationDesc,
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
resultRelInfo->ri_RangeTableIndex);
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
result = table_tuple_lock(resultRelationDesc, tupleid,
estate->es_snapshot,
inputslot, estate->es_output_cid,
LockTupleExclusive, LockWaitBlock,
TUPLE_LOCK_FLAG_FIND_LAST_VERSION,
&context->tmfd);
switch (result)
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
{
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
case TM_Ok:
Assert(context->tmfd.traversed);
epqslot = EvalPlanQual(context->epqstate,
resultRelationDesc,
resultRelInfo->ri_RangeTableIndex,
inputslot);
if (TupIsNull(epqslot))
/* Tuple not passing quals anymore, exiting... */
return NULL;
/*
* If requested, skip delete and pass back the
* updated row.
*/
if (epqreturnslot)
{
*epqreturnslot = epqslot;
return NULL;
}
else
goto ldelete;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
case TM_SelfModified:
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
/*
* This can be reached when following an update
* chain from a tuple updated by another session,
* reaching a tuple that was already updated in
* this transaction. If previously updated by this
* command, ignore the delete, otherwise error
* out.
*
* See also TM_SelfModified response to
* table_tuple_delete() above.
*/
if (context->tmfd.cmax != estate->es_output_cid)
ereport(ERROR,
(errcode(ERRCODE_TRIGGERED_DATA_CHANGE_VIOLATION),
errmsg("tuple to be deleted was already modified by an operation triggered by the current command"),
errhint("Consider using an AFTER trigger instead of a BEFORE trigger to propagate changes to other rows.")));
return NULL;
case TM_Deleted:
/* tuple already deleted; nothing to do */
return NULL;
default:
/*
* TM_Invisible should be impossible because we're
* waiting for updated row versions, and would
* already have errored out if the first version
* is invisible.
*
* TM_Updated should be impossible, because we're
* locking the latest version via
* TUPLE_LOCK_FLAG_FIND_LAST_VERSION.
*/
elog(ERROR, "unexpected table_tuple_lock status: %u",
result);
return NULL;
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
}
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
Assert(false);
break;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Deleted:
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent delete")));
Fix outdated comment in ExecDelete This commend references a struct that disappeared before MERGE was merged ... and ExecDelete is not called by the committed MERGE anyway. Revert to the original wording. Backpatch to 15
2022-11-17 12:52:20 +01:00
/* tuple already deleted; nothing to do */
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
return NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
default:
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
elog(ERROR, "unrecognized table_tuple_delete status: %u",
result);
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
return NULL;
}
/*
* Note: Normally one would think that we have to delete index tuples
* associated with the heap tuple now...
*
* ... but in POSTGRES, we have no need to do this because VACUUM will
* take care of it later. We can't delete index tuples immediately
* anyway, since the tuple is still visible to other transactions.
*/
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
if (canSetTag)
(estate->es_processed)++;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
/* Tell caller that the delete actually happened. */
if (tupleDeleted)
*tupleDeleted = true;
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
ExecDeleteEpilogue(context, resultRelInfo, tupleid, oldtuple, changingPart);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
/* Process RETURNING if present and if requested */
if (processReturning && resultRelInfo->ri_projectReturning)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
/*
* We have to put the target tuple into a slot, which means first we
* gotta fetch it. We can use the trigger tuple slot.
*/
TupleTableSlot *rslot;
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
if (resultRelInfo->ri_FdwRoutine)
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
{
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
/* FDW must have provided a slot containing the deleted row */
Assert(!TupIsNull(slot));
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
}
else
{
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
slot = ExecGetReturningSlot(estate, resultRelInfo);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
if (oldtuple != NULL)
{
Fix two memory leaks around force-storing tuples in slots. As reported by Tom, when ExecStoreMinimalTuple() had to perform a conversion to store the minimal tuple in the slot, it forgot to respect the shouldFree flag, and leaked the tuple into the current memory context if true. Fix that by freeing the tuple in that case. Looking at the relevant code made me (Andres) realize that not having the shouldFree parameter to ExecForceStoreHeapTuple() was a bad idea. Some callers had to locally implement the necessary logic, and in one case it was missing, creating a potential per-group leak in non-hashed aggregation. The choice to not free the tuple in ExecComputeStoredGenerated() is not pretty, but not introduced by this commit - I'll start a separate discussion about it. Reported-By: Tom Lane Discussion: https://postgr.es/m/366.1555382816@sss.pgh.pa.us
2019-04-19 20:33:37 +02:00
ExecForceStoreHeapTuple(oldtuple, slot, false);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
}
Revert 764da7710b Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:55:09 +02:00
else
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
{
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
if (!table_tuple_fetch_row_version(resultRelationDesc, tupleid,
SnapshotAny, slot))
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
elog(ERROR, "failed to fetch deleted tuple for DELETE RETURNING");
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
}
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
rslot = ExecProcessReturning(resultRelInfo, slot, context->planSlot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
/*
* Before releasing the target tuple again, make sure rslot has a
* local copy of any pass-by-reference values.
*/
ExecMaterializeSlot(rslot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
ExecClearTuple(slot);
return rslot;
}
return NULL;
}
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
/*
* ExecCrossPartitionUpdate --- Move an updated tuple to another partition.
*
* This works by first deleting the old tuple from the current partition,
* followed by inserting the new tuple into the root parent table, that is,
* mtstate->rootResultRelInfo. It will be re-routed from there to the
* correct partition.
*
* Returns true if the tuple has been successfully moved, or if it's found
* that the tuple was concurrently deleted so there's nothing more to do
* for the caller.
*
* False is returned if the tuple we're trying to move is found to have been
Fix an outdated and grammatically wrong comment Authored by Amit Langote and myself independently Discussion: https://postgr.es/m/CA+HiwqGCjcH0gG-=tM7hhP7TEDmzrHMHJbPGSHtHgFmx9mnFkg@mail.gmail.com
2022-03-19 19:31:38 +01:00
* concurrently updated. In that case, the caller must check if the updated
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
* tuple that's returned in *retry_slot still needs to be re-routed, and call
* this function again or perform a regular update accordingly. For MERGE,
* the updated tuple is not returned in *retry_slot; it has its own retry
* logic.
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
*/
static bool
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecCrossPartitionUpdate(ModifyTableContext *context,
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
ResultRelInfo *resultRelInfo,
ItemPointer tupleid, HeapTuple oldtuple,
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
TupleTableSlot *slot,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
bool canSetTag,
UpdateContext *updateCxt,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
TupleTableSlot **retry_slot,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
TupleTableSlot **inserted_tuple,
ResultRelInfo **insert_destrel)
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
{
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ModifyTableState *mtstate = context->mtstate;
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
EState *estate = mtstate->ps.state;
TupleConversionMap *tupconv_map;
bool tuple_deleted;
TupleTableSlot *epqslot = NULL;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
context->cpUpdateReturningSlot = NULL;
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
*retry_slot = NULL;
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
/*
* Disallow an INSERT ON CONFLICT DO UPDATE that causes the original row
* to migrate to a different partition. Maybe this can be implemented
* some day, but it seems a fringe feature with little redeeming value.
*/
if (((ModifyTable *) mtstate->ps.plan)->onConflictAction == ONCONFLICT_UPDATE)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("invalid ON UPDATE specification"),
errdetail("The result tuple would appear in a different partition than the original tuple.")));
/*
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
* When an UPDATE is run directly on a leaf partition, simply fail with a
* partition constraint violation error.
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
*/
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
if (resultRelInfo == mtstate->rootResultRelInfo)
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
ExecPartitionCheckEmitError(resultRelInfo, slot, estate);
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/* Initialize tuple routing info if not already done. */
if (mtstate->mt_partition_tuple_routing == NULL)
{
Relation rootRel = mtstate->rootResultRelInfo->ri_RelationDesc;
MemoryContext oldcxt;
/* Things built here have to last for the query duration. */
oldcxt = MemoryContextSwitchTo(estate->es_query_cxt);
mtstate->mt_partition_tuple_routing =
ExecSetupPartitionTupleRouting(estate, rootRel);
/*
* Before a partition's tuple can be re-routed, it must first be
* converted to the root's format, so we'll need a slot for storing
* such tuples.
*/
Assert(mtstate->mt_root_tuple_slot == NULL);
mtstate->mt_root_tuple_slot = table_slot_create(rootRel, NULL);
MemoryContextSwitchTo(oldcxt);
}
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
/*
* Row movement, part 1. Delete the tuple, but skip RETURNING processing.
* We want to return rows from INSERT.
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecDelete(context, resultRelInfo,
tupleid, oldtuple,
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
false, /* processReturning */
true, /* changingPart */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
false, /* canSetTag */
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
&tuple_deleted, &epqslot);
/*
* For some reason if DELETE didn't happen (e.g. trigger prevented it, or
* it was already deleted by self, or it was concurrently deleted by
* another transaction), then we should skip the insert as well;
* otherwise, an UPDATE could cause an increase in the total number of
* rows across all partitions, which is clearly wrong.
*
* For a normal UPDATE, the case where the tuple has been the subject of a
* concurrent UPDATE or DELETE would be handled by the EvalPlanQual
* machinery, but for an UPDATE that we've translated into a DELETE from
* this partition and an INSERT into some other partition, that's not
* available, because CTID chains can't span relation boundaries. We
* mimic the semantics to a limited extent by skipping the INSERT if the
* DELETE fails to find a tuple. This ensures that two concurrent
* attempts to UPDATE the same tuple at the same time can't turn one tuple
* into two, and that an UPDATE of a just-deleted tuple can't resurrect
* it.
*/
if (!tuple_deleted)
{
/*
* epqslot will be typically NULL. But when ExecDelete() finds that
* another transaction has concurrently updated the same row, it
* re-fetches the row, skips the delete, and epqslot is set to the
* re-fetched tuple slot. In that case, we need to do all the checks
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
* again. For MERGE, we leave everything to the caller (it must do
* additional rechecking, and might end up executing a different
* action entirely).
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
*/
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
if (context->relaction != NULL)
return false;
else if (TupIsNull(epqslot))
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
return true;
else
{
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/* Fetch the most recent version of old tuple. */
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
TupleTableSlot *oldSlot;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/* ... but first, make sure ri_oldTupleSlot is initialized. */
if (unlikely(!resultRelInfo->ri_projectNewInfoValid))
ExecInitUpdateProjection(mtstate, resultRelInfo);
oldSlot = resultRelInfo->ri_oldTupleSlot;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
if (!table_tuple_fetch_row_version(resultRelInfo->ri_RelationDesc,
tupleid,
SnapshotAny,
oldSlot))
elog(ERROR, "failed to fetch tuple being updated");
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/* and project the new tuple to retry the UPDATE with */
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
*retry_slot = ExecGetUpdateNewTuple(resultRelInfo, epqslot,
oldSlot);
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
return false;
}
}
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* resultRelInfo is one of the per-relation resultRelInfos. So we should
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
* convert the tuple into root's tuple descriptor if needed, since
* ExecInsert() starts the search from root.
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
*/
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
tupconv_map = ExecGetChildToRootMap(resultRelInfo);
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
if (tupconv_map != NULL)
slot = execute_attr_map_slot(tupconv_map->attrMap,
slot,
mtstate->mt_root_tuple_slot);
/* Tuple routing starts from the root table. */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
context->cpUpdateReturningSlot =
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
ExecInsert(context, mtstate->rootResultRelInfo, slot, canSetTag,
inserted_tuple, insert_destrel);
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
/*
Remove PartitionRoutingInfo struct. The extra indirection neeeded to access its members via its enclosing ResultRelInfo seems pointless. Move all the fields from PartitionRoutingInfo to ResultRelInfo. Author: Amit Langote Reviewed-by: Alvaro Herrera Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:57 +02:00
* Reset the transition state that may possibly have been written by
* INSERT.
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
*/
Refactor code for cross-partition updates to a separate function. ExecUpdate() is very long, so extract the part of it that deals with cross-partition updates to a separate function to make it more readable. Per Andres Freund's suggestion. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqEUgb5RdUgxR7Sqco4S09jzJstHiaT2vnCRPGR4JCAPqA%40mail.gmail.com
2020-10-15 16:08:10 +02:00
if (mtstate->mt_transition_capture)
mtstate->mt_transition_capture->tcs_original_insert_tuple = NULL;
/* We're done moving. */
return true;
}
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* ExecUpdatePrologue -- subroutine for ExecUpdate
*
* Prepare executor state for UPDATE. This includes running BEFORE ROW
* triggers. We return false if one of them makes the update a no-op;
* otherwise, return true.
*/
static bool
ExecUpdatePrologue(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
ItemPointer tupleid, HeapTuple oldtuple, TupleTableSlot *slot,
TM_Result *result)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
Relation resultRelationDesc = resultRelInfo->ri_RelationDesc;
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
if (result)
*result = TM_Ok;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecMaterializeSlot(slot);
/*
* Open the table's indexes, if we have not done so already, so that we
* can add new index entries for the updated tuple.
*/
if (resultRelationDesc->rd_rel->relhasindex &&
resultRelInfo->ri_IndexRelationDescs == NULL)
ExecOpenIndices(resultRelInfo, false);
/* BEFORE ROW UPDATE triggers */
if (resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_update_before_row)
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
{
/* Flush any pending inserts, so rows are visible to the triggers */
if (context->estate->es_insert_pending_result_relations != NIL)
ExecPendingInserts(context->estate);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return ExecBRUpdateTriggers(context->estate, context->epqstate,
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
resultRelInfo, tupleid, oldtuple, slot,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
result, &context->tmfd);
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
}
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return true;
}
/*
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
* ExecUpdatePrepareSlot -- subroutine for ExecUpdateAct
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
*
* Apply the final modifications to the tuple slot before the update.
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
* (This is split out because we also need it in the foreign-table code path.)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
*/
static void
ExecUpdatePrepareSlot(ResultRelInfo *resultRelInfo,
TupleTableSlot *slot,
EState *estate)
{
Relation resultRelationDesc = resultRelInfo->ri_RelationDesc;
/*
* Constraints and GENERATED expressions might reference the tableoid
* column, so (re-)initialize tts_tableOid before evaluating them.
*/
slot->tts_tableOid = RelationGetRelid(resultRelationDesc);
/*
* Compute stored generated columns
*/
if (resultRelationDesc->rd_att->constr &&
resultRelationDesc->rd_att->constr->has_generated_stored)
ExecComputeStoredGenerated(resultRelInfo, estate, slot,
CMD_UPDATE);
}
/*
* ExecUpdateAct -- subroutine for ExecUpdate
*
* Actually update the tuple, when operating on a plain table. If the
* table is a partition, and the command was called referencing an ancestor
* partitioned table, this routine migrates the resulting tuple to another
* partition.
*
* The caller is in charge of keeping indexes current as necessary. The
* caller is also in charge of doing EvalPlanQual if the tuple is found to
* be concurrently updated. However, in case of a cross-partition update,
* this routine does it.
*/
static TM_Result
ExecUpdateAct(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
ItemPointer tupleid, HeapTuple oldtuple, TupleTableSlot *slot,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
bool canSetTag, UpdateContext *updateCxt)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
EState *estate = context->estate;
Relation resultRelationDesc = resultRelInfo->ri_RelationDesc;
bool partition_constraint_failed;
TM_Result result;
updateCxt->crossPartUpdate = false;
/*
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
* If we move the tuple to a new partition, we loop back here to recompute
* GENERATED values (which are allowed to be different across partitions)
* and recheck any RLS policies and constraints. We do not fire any
* BEFORE triggers of the new partition, however.
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
*/
Remove unnecessary semicolons after goto labels According to the C standard, a label must followed by a statement. If there was ever a time we needed an empty statement here, it was a long time ago. Japin Li Reviewed by Julien Rouhaud Discussion: https://www.postgresql.org/message-id/MEYP282MB16690F40189A4F060B41D56DB65E9%40MEYP282MB1669.AUSP282.PROD.OUTLOOK.COM
2022-10-10 10:08:38 +02:00
lreplace:
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
/* Fill in GENERATEd columns */
ExecUpdatePrepareSlot(resultRelInfo, slot, estate);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/* ensure slot is independent, consider e.g. EPQ */
ExecMaterializeSlot(slot);
/*
* If partition constraint fails, this row might get moved to another
* partition, in which case we should check the RLS CHECK policy just
* before inserting into the new partition, rather than doing it here.
* This is because a trigger on that partition might again change the row.
* So skip the WCO checks if the partition constraint fails.
*/
partition_constraint_failed =
resultRelationDesc->rd_rel->relispartition &&
!ExecPartitionCheck(resultRelInfo, slot, estate, false);
/* Check any RLS UPDATE WITH CHECK policies */
if (!partition_constraint_failed &&
resultRelInfo->ri_WithCheckOptions != NIL)
{
/*
* ExecWithCheckOptions() will skip any WCOs which are not of the kind
* we are looking for at this point.
*/
ExecWithCheckOptions(WCO_RLS_UPDATE_CHECK,
resultRelInfo, slot, estate);
}
/*
* If a partition check failed, try to move the row into the right
* partition.
*/
if (partition_constraint_failed)
{
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
TupleTableSlot *inserted_tuple,
*retry_slot;
Blind fix for uninitialized memory bug in ba9a7e392171 Valgrind animal skink shows a crash in this new code. I couldn't reproduce the problem locally, but going by blind code inspection, initializing insert_destrel should be sufficient to fix the problem.
2022-03-20 22:10:24 +01:00
ResultRelInfo *insert_destrel = NULL;
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* ExecCrossPartitionUpdate will first DELETE the row from the
* partition it's currently in and then insert it back into the root
* table, which will re-route it to the correct partition. However,
* if the tuple has been concurrently updated, a retry is needed.
*/
if (ExecCrossPartitionUpdate(context, resultRelInfo,
tupleid, oldtuple, slot,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
canSetTag, updateCxt,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
&retry_slot,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
&inserted_tuple,
&insert_destrel))
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
/* success! */
updateCxt->updated = true;
updateCxt->crossPartUpdate = true;
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
/*
* If the partitioned table being updated is referenced in foreign
* keys, queue up trigger events to check that none of them were
* violated. No special treatment is needed in
* non-cross-partition update situations, because the leaf
* partition's AR update triggers will take care of that. During
* cross-partition updates implemented as delete on the source
* partition followed by insert on the destination partition,
* AR-UPDATE triggers of the root table (that is, the table
* mentioned in the query) must be fired.
*
* NULL insert_destrel means that the move failed to occur, that
* is, the update failed, so no need to anything in that case.
*/
if (insert_destrel &&
resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_update_after_row)
ExecCrossPartitionUpdateForeignKey(context,
resultRelInfo,
insert_destrel,
tupleid, slot,
inserted_tuple);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return TM_Ok;
}
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/*
* No luck, a retry is needed. If running MERGE, we do not do so
* here; instead let it handle that on its own rules.
*/
if (context->relaction != NULL)
return TM_Updated;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* ExecCrossPartitionUpdate installed an updated version of the new
* tuple in the retry slot; start over.
*/
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
slot = retry_slot;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
goto lreplace;
}
/*
* Check the constraints of the tuple. We've already checked the
* partition constraint above; however, we must still ensure the tuple
* passes all other constraints, so we will call ExecConstraints() and
* have it validate all remaining checks.
*/
if (resultRelationDesc->rd_att->constr)
ExecConstraints(resultRelInfo, slot, estate);
/*
* replace the heap tuple
*
* Note: if es_crosscheck_snapshot isn't InvalidSnapshot, we check that
* the row to be updated is visible to that snapshot, and throw a
* can't-serialize error if not. This is a special-case behavior needed
* for referential integrity updates in transaction-snapshot mode
* transactions.
*/
result = table_tuple_update(resultRelationDesc, tupleid, slot,
estate->es_output_cid,
estate->es_snapshot,
estate->es_crosscheck_snapshot,
true /* wait for commit */ ,
Move ModifyTableContext->lockmode to UpdateContext Should have been done this way to start with, but I failed to notice This way we avoid some pointless initialization, and better contains the variable to exist in the scope where it is really used. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 11:18:04 +02:00
&context->tmfd, &updateCxt->lockmode,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
&updateCxt->updateIndexes);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
if (result == TM_Ok)
updateCxt->updated = true;
return result;
}
/*
* ExecUpdateEpilogue -- subroutine for ExecUpdate
*
* Closing steps of updating a tuple. Must be called if ExecUpdateAct
* returns indicating that the tuple was updated.
*/
static void
ExecUpdateEpilogue(ModifyTableContext *context, UpdateContext *updateCxt,
ResultRelInfo *resultRelInfo, ItemPointer tupleid,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
HeapTuple oldtuple, TupleTableSlot *slot)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
{
ModifyTableState *mtstate = context->mtstate;
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
List *recheckIndexes = NIL;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/* insert index entries for tuple if necessary */
Ignore BRIN indexes when checking for HOT updates When determining whether an index update may be skipped by using HOT, we can ignore attributes indexed by block summarizing indexes without references to individual tuples that need to be cleaned up. A new type TU_UpdateIndexes provides a signal to the executor to determine which indexes to update - no indexes, all indexes, or only the summarizing indexes. This also removes rd_indexattr list, and replaces it with rd_attrsvalid flag. The list was not used anywhere, and a simple flag is sufficient. This was originally committed as 5753d4ee32, but then got reverted by e3fcca0d0d because of correctness issues. Original patch by Josef Simanek, various fixes and improvements by Tomas Vondra and me. Authors: Matthias van de Meent, Josef Simanek, Tomas Vondra Reviewed-by: Tomas Vondra, Alvaro Herrera Discussion: https://postgr.es/m/05ebcb44-f383-86e3-4f31-0a97a55634cf@enterprisedb.com Discussion: https://postgr.es/m/CAFp7QwpMRGcDAQumN7onN9HjrJ3u4X3ZRXdGFT0K5G2JWvnbWg%40mail.gmail.com
2023-03-20 10:34:07 +01:00
if (resultRelInfo->ri_NumIndices > 0 && (updateCxt->updateIndexes != TU_None))
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
recheckIndexes = ExecInsertIndexTuples(resultRelInfo,
slot, context->estate,
true, false,
Ignore BRIN indexes when checking for HOT updates When determining whether an index update may be skipped by using HOT, we can ignore attributes indexed by block summarizing indexes without references to individual tuples that need to be cleaned up. A new type TU_UpdateIndexes provides a signal to the executor to determine which indexes to update - no indexes, all indexes, or only the summarizing indexes. This also removes rd_indexattr list, and replaces it with rd_attrsvalid flag. The list was not used anywhere, and a simple flag is sufficient. This was originally committed as 5753d4ee32, but then got reverted by e3fcca0d0d because of correctness issues. Original patch by Josef Simanek, various fixes and improvements by Tomas Vondra and me. Authors: Matthias van de Meent, Josef Simanek, Tomas Vondra Reviewed-by: Tomas Vondra, Alvaro Herrera Discussion: https://postgr.es/m/05ebcb44-f383-86e3-4f31-0a97a55634cf@enterprisedb.com Discussion: https://postgr.es/m/CAFp7QwpMRGcDAQumN7onN9HjrJ3u4X3ZRXdGFT0K5G2JWvnbWg%40mail.gmail.com
2023-03-20 10:34:07 +01:00
NULL, NIL,
(updateCxt->updateIndexes == TU_Summarizing));
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/* AFTER ROW UPDATE Triggers */
ExecARUpdateTriggers(context->estate, resultRelInfo,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
NULL, NULL,
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
tupleid, oldtuple, slot,
recheckIndexes,
mtstate->operation == CMD_INSERT ?
mtstate->mt_oc_transition_capture :
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
mtstate->mt_transition_capture,
false);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
list_free(recheckIndexes);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/*
* Check any WITH CHECK OPTION constraints from parent views. We are
* required to do this after testing all constraints and uniqueness
* violations per the SQL spec, so we do it after actually updating the
* record in the heap and all indexes.
*
* ExecWithCheckOptions() will skip any WCOs which are not of the kind we
* are looking for at this point.
*/
if (resultRelInfo->ri_WithCheckOptions != NIL)
ExecWithCheckOptions(WCO_VIEW_CHECK, resultRelInfo,
slot, context->estate);
}
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
/*
* Queues up an update event using the target root partitioned table's
* trigger to check that a cross-partition update hasn't broken any foreign
* keys pointing into it.
*/
static void
ExecCrossPartitionUpdateForeignKey(ModifyTableContext *context,
ResultRelInfo *sourcePartInfo,
ResultRelInfo *destPartInfo,
ItemPointer tupleid,
TupleTableSlot *oldslot,
TupleTableSlot *newslot)
{
ListCell *lc;
ResultRelInfo *rootRelInfo;
List *ancestorRels;
rootRelInfo = sourcePartInfo->ri_RootResultRelInfo;
ancestorRels = ExecGetAncestorResultRels(context->estate, sourcePartInfo);
/*
* For any foreign keys that point directly into a non-root ancestors of
* the source partition, we can in theory fire an update event to enforce
* those constraints using their triggers, if we could tell that both the
* source and the destination partitions are under the same ancestor. But
* for now, we simply report an error that those cannot be enforced.
*/
foreach(lc, ancestorRels)
{
ResultRelInfo *rInfo = lfirst(lc);
TriggerDesc *trigdesc = rInfo->ri_TrigDesc;
bool has_noncloned_fkey = false;
/* Root ancestor's triggers will be processed. */
if (rInfo == rootRelInfo)
continue;
if (trigdesc && trigdesc->trig_update_after_row)
{
for (int i = 0; i < trigdesc->numtriggers; i++)
{
Trigger *trig = &trigdesc->triggers[i];
if (!trig->tgisclone &&
RI_FKey_trigger_type(trig->tgfoid) == RI_TRIGGER_PK)
{
has_noncloned_fkey = true;
break;
}
}
}
if (has_noncloned_fkey)
ereport(ERROR,
(errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
errmsg("cannot move tuple across partitions when a non-root ancestor of the source partition is directly referenced in a foreign key"),
Message style improvements
2022-09-25 00:38:35 +02:00
errdetail("A foreign key points to ancestor \"%s\" but not the root ancestor \"%s\".",
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
RelationGetRelationName(rInfo->ri_RelationDesc),
RelationGetRelationName(rootRelInfo->ri_RelationDesc)),
Message style improvements
2022-09-25 00:38:35 +02:00
errhint("Consider defining the foreign key on table \"%s\".",
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
RelationGetRelationName(rootRelInfo->ri_RelationDesc))));
}
/* Perform the root table's triggers. */
ExecARUpdateTriggers(context->estate,
rootRelInfo, sourcePartInfo, destPartInfo,
tupleid, NULL, newslot, NIL, NULL, true);
}
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* ----------------------------------------------------------------
* ExecUpdate
*
* note: we can't run UPDATE queries with transactions
* off because UPDATEs are actually INSERTs and our
* scan will mistakenly loop forever, updating the tuple
* it just inserted.. This should be fixed but until it
* is, we don't want to get stuck in an infinite loop
* which corrupts your database..
*
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
* When updating a table, tupleid identifies the tuple to
* update and oldtuple is NULL. When updating a view, oldtuple
* is passed to the INSTEAD OF triggers and identifies what to
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
* update, and tupleid is invalid. When updating a foreign table,
Offer triggers on foreign tables. This covers all the SQL-standard trigger types supported for regular tables; it does not cover constraint triggers. The approach for acquiring the old row mirrors that for view INSTEAD OF triggers. For AFTER ROW triggers, we spool the foreign tuples to a tuplestore. This changes the FDW API contract; when deciding which columns to populate in the slot returned from data modification callbacks, writable FDWs will need to check for AFTER ROW triggers in addition to checking for a RETURNING clause. In support of the feature addition, refactor the TriggerFlags bits and the assembly of old tuples in ModifyTable. Ronan Dunklau, reviewed by KaiGai Kohei; some additional hacking by me.
2014-03-23 07:16:34 +01:00
* tupleid is invalid; the FDW has to figure out which row to
* update using data from the planSlot. oldtuple is passed to
* foreign table triggers; it is NULL when the foreign table has
* no relevant triggers.
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* slot contains the new tuple value to be stored.
* planSlot is the output of the ModifyTable's subplan; we use it
* to access values from other input tables (for RETURNING),
* row-ID junk columns, etc.
*
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* Returns RETURNING result if any, otherwise NULL.
* ----------------------------------------------------------------
*/
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
static TupleTableSlot *
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecUpdate(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
ItemPointer tupleid, HeapTuple oldtuple, TupleTableSlot *slot,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
bool canSetTag)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
EState *estate = context->estate;
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
Relation resultRelationDesc = resultRelInfo->ri_RelationDesc;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
UpdateContext updateCxt = {0};
TM_Result result;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* abort the operation if not running transactions
*/
if (IsBootstrapProcessingMode())
elog(ERROR, "cannot UPDATE during bootstrap");
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/*
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
* Prepare for the update. This includes BEFORE ROW triggers, so we're
* done if it says we are.
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
*/
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
if (!ExecUpdatePrologue(context, resultRelInfo, tupleid, oldtuple, slot, NULL))
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
/* INSTEAD OF ROW UPDATE Triggers */
if (resultRelInfo->ri_TrigDesc &&
resultRelInfo->ri_TrigDesc->trig_update_instead_row)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Use slots in trigger infrastructure, except for the actual invocation. In preparation for abstracting table storage, convert trigger.c to track tuples in slots. Which also happens to make code calling triggers simpler. As the calling interface for triggers themselves is not changed in this patch, HeapTuples still are extracted from the slot at that time. But that's handled solely inside trigger.c, not visible to callers. It's quite likely that we'll want to revise the external trigger interface, but that's a separate large project. As part of this work the slots used for old/new/return tuples are moved from EState into ResultRelInfo, as different updated tables might need different slots. The slots are now also now created on-demand, which is good both from an efficiency POV, but also makes the modifying code simpler. Author: Andres Freund, Amit Khandekar and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 05:30:28 +01:00
if (!ExecIRUpdateTriggers(estate, resultRelInfo,
oldtuple, slot))
return NULL; /* "do nothing" */
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
}
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
else if (resultRelInfo->ri_FdwRoutine)
{
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
/* Fill in GENERATEd columns */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecUpdatePrepareSlot(resultRelInfo, slot, estate);
Generated columns This is an SQL-standard feature that allows creating columns that are computed from expressions rather than assigned, similar to a view or materialized view but on a column basis. This implements one kind of generated column: stored (computed on write). Another kind, virtual (computed on read), is planned for the future, and some room is left for it. Reviewed-by: Michael Paquier <michael@paquier.xyz> Reviewed-by: Pavel Stehule <pavel.stehule@gmail.com> Discussion: https://www.postgresql.org/message-id/flat/b151f851-4019-bdb1-699e-ebab07d2f40a@2ndquadrant.com
2019-03-30 08:13:09 +01:00
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
/*
* update in foreign table: let the FDW do it
*/
slot = resultRelInfo->ri_FdwRoutine->ExecForeignUpdate(estate,
resultRelInfo,
slot,
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
context->planSlot);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
if (slot == NULL) /* "do nothing" */
return NULL;
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
/*
* AFTER ROW Triggers or RETURNING expressions might reference the
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
* tableoid column, so (re-)initialize tts_tableOid before evaluating
Fix usage of "tableoid" in GENERATED expressions. We consider this supported (though I've got my doubts that it's a good idea, because tableoid is not immutable). However, several code paths failed to fill the field in soon enough, causing such a GENERATED expression to see zero or the wrong value. This occurred when ALTER TABLE adds a new GENERATED column to a table with existing rows, and during regular INSERT or UPDATE on a foreign table with GENERATED columns. Noted during investigation of a report from Vitaly Ustinov. Back-patch to v12 where GENERATED came in. Discussion: https://postgr.es/m/CAM_DEiWR2DPT6U4xb-Ehigozzd3n3G37ZB1+867zbsEVtYoJww@mail.gmail.com
2021-05-21 21:02:06 +02:00
* them. (This covers the case where the FDW replaced the slot.)
When modifying a foreign table, initialize tableoid field properly. Failure to do this can cause AFTER ROW triggers or RETURNING expressions that reference this field to misbehave. Etsuro Fujita, reviewed by Thom Brown
2016-02-05 03:15:57 +01:00
*/
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
slot->tts_tableOid = RelationGetRelid(resultRelationDesc);
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
}
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
else
{
Fix some more cases of missed GENERATED-column updates. If UPDATE is forced to retry after an EvalPlanQual check, it neglected to repeat GENERATED-column computations, even though those might well have changed since we're dealing with a different tuple than before. Fixing this is mostly a matter of looping back a bit further when we retry. In v15 and HEAD that's most easily done by altering the API of ExecUpdateAct so that it includes computing GENERATED expressions. Also, if an UPDATE in a partitioned table turns into a cross-partition INSERT operation, we failed to recompute GENERATED columns. That's a bug since 8bf6ec3ba allowed partitions to have different generation expressions; although it seems to have no ill effects before that. Fixing this is messier because we can now have situations where the same query needs both the UPDATE-aligned set of GENERATED columns and the INSERT-aligned set, and it's unclear which set will be generated first (else we could hack things by forcing the INSERT-aligned set to be generated, which is indeed how fe9e658f4 made it work for MERGE). The best fix seems to be to build and store separate sets of expressions for the INSERT and UPDATE cases. That would create ABI issues in the back branches, but so far it seems we can leave this alone in the back branches. Per bug #17823 from Hisahiro Kauchi. The first part of this affects all branches back to v12 where GENERATED columns were added. Discussion: https://postgr.es/m/17823-b64909cf7d63de84@postgresql.org
2023-03-07 00:31:16 +01:00
/*
* If we generate a new candidate tuple after EvalPlanQual testing, we
* must loop back here to try again. (We don't need to redo triggers,
* however. If there are any BEFORE triggers then trigger.c will have
* done table_tuple_lock to lock the correct tuple, so there's no need
* to do them again.)
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
redo_act:
result = ExecUpdateAct(context, resultRelInfo, tupleid, oldtuple, slot,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
canSetTag, &updateCxt);
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
/*
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
* If ExecUpdateAct reports that a cross-partition update was done,
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* then the RETURNING tuple (if any) has been projected and there's
* nothing else for us to do.
Perform RLS WITH CHECK before constraints, etc The RLS capability is built on top of the WITH CHECK OPTION system which was added for auto-updatable views, however, unlike WCOs on views (which are mandated by the SQL spec to not fire until after all other constraints and checks are done), it makes much more sense for RLS checks to happen earlier than constraint and uniqueness checks. This patch reworks the structure which holds the WCOs a bit to be explicitly either VIEW or RLS checks and the RLS-related checks are done prior to the constraint and uniqueness checks. This also allows better error reporting as we are now reporting when a violation is due to a WITH CHECK OPTION and when it's due to an RLS policy violation, which was independently noted by Craig Ringer as being confusing. The documentation is also updated to include a paragraph about when RLS WITH CHECK handling is performed, as there have been a number of questions regarding that and the documentation was previously silent on the matter. Author: Dean Rasheed, with some kabitzing and comment changes by me.
2015-04-25 02:34:26 +02:00
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
if (updateCxt.crossPartUpdate)
return context->cpUpdateReturningSlot;
Store table oid and tuple's tid in tuple slots directly. After the introduction of tuple table slots all table AMs need to support returning the table oid of the tuple stored in a slot created by said AM. It does not make sense to re-implement that in every AM, therefore move handling of table OIDs into the TupleTableSlot structure itself. It's possible that we, at a later date, might want to get rid of HeapTupleData.t_tableOid entirely, but doing so before the abstractions for table AMs are integrated turns out to be too hard, so delay that for now. Similarly, every AM needs to support the concept of a tuple identifier (tid / item pointer) for its tuples. It's quite possible that we'll generalize the exact form of a tid at a future point (to allow for things like index organized tables), but for now many parts of the code know about tids, so there's not much point in abstracting tids away. Therefore also move into slot (rather than providing API to set/get the tid associated with the tuple in a slot). Once table AM includes insert/updating/deleting tuples, the responsibility to set the correct tid after such an action will move into that. After that change, code doing such modifications, should not have to deal with HeapTuples directly anymore. Author: Andres Freund, Haribabu Kommi and Ashutosh Bapat Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-02-27 03:21:44 +01:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
switch (result)
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_SelfModified:
pgindent run for release 9.3 This is the first run of the Perl-based pgindent script. Also update pgindent instructions.
2013-05-29 22:58:43 +02:00
Throw error if expiring tuple is again updated or deleted. This prevents surprising behavior when a FOR EACH ROW trigger BEFORE UPDATE or BEFORE DELETE directly or indirectly updates or deletes the the old row. Prior to this patch the requested action on the row could be silently ignored while all triggered actions based on the occurence of the requested action could be committed. One example of how this could happen is if the BEFORE DELETE trigger for a "parent" row deleted "children" which had trigger functions to update summary or status data on the parent. This also prevents similar surprising problems if the query has a volatile function which updates a target row while it is already being updated. There are related issues present in FOR UPDATE cursors and READ COMMITTED queries which are not handled by this patch. These issues need further evalution to determine what change, if any, is needed. Where the new error messages are generated, in most cases the best fix will be to move code from the BEFORE trigger to an AFTER trigger. Where this is not feasible, the trigger can avoid the error by re-issuing the triggering statement and returning NULL. Documentation changes will be submitted in a separate patch. Kevin Grittner and Tom Lane with input from Florian Pflug and Robert Haas, based on problems encountered during conversion of Wisconsin Circuit Court trigger logic to plpgsql triggers.
2012-10-26 21:55:36 +02:00
/*
* The target tuple was already updated or deleted by the
* current command, or by a later command in the current
* transaction. The former case is possible in a join UPDATE
* where multiple tuples join to the same target tuple. This
* is pretty questionable, but Postgres has always allowed it:
* we just execute the first update action and ignore
* additional update attempts.
*
* The latter case arises if the tuple is modified by a
* command in a BEFORE trigger, or perhaps by a command in a
* volatile function used in the query. In such situations we
* should not ignore the update, but it is equally unsafe to
* proceed. We don't want to discard the original UPDATE
* while keeping the triggered actions based on it; and we
* have no principled way to merge this update with the
* previous ones. So throwing an error is the only safe
* course.
*
* If a trigger actually intends this type of interaction, it
* can re-execute the UPDATE (assuming it can figure out how)
* and then return NULL to cancel the outer update.
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
if (context->tmfd.cmax != estate->es_output_cid)
Throw error if expiring tuple is again updated or deleted. This prevents surprising behavior when a FOR EACH ROW trigger BEFORE UPDATE or BEFORE DELETE directly or indirectly updates or deletes the the old row. Prior to this patch the requested action on the row could be silently ignored while all triggered actions based on the occurence of the requested action could be committed. One example of how this could happen is if the BEFORE DELETE trigger for a "parent" row deleted "children" which had trigger functions to update summary or status data on the parent. This also prevents similar surprising problems if the query has a volatile function which updates a target row while it is already being updated. There are related issues present in FOR UPDATE cursors and READ COMMITTED queries which are not handled by this patch. These issues need further evalution to determine what change, if any, is needed. Where the new error messages are generated, in most cases the best fix will be to move code from the BEFORE trigger to an AFTER trigger. Where this is not feasible, the trigger can avoid the error by re-issuing the triggering statement and returning NULL. Documentation changes will be submitted in a separate patch. Kevin Grittner and Tom Lane with input from Florian Pflug and Robert Haas, based on problems encountered during conversion of Wisconsin Circuit Court trigger logic to plpgsql triggers.
2012-10-26 21:55:36 +02:00
ereport(ERROR,
(errcode(ERRCODE_TRIGGERED_DATA_CHANGE_VIOLATION),
errmsg("tuple to be updated was already modified by an operation triggered by the current command"),
errhint("Consider using an AFTER trigger instead of a BEFORE trigger to propagate changes to other rows.")));
/* Else, already updated by self; nothing to do */
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
return NULL;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Ok:
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
break;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Updated:
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
TupleTableSlot *inputslot;
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
TupleTableSlot *epqslot;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
TupleTableSlot *oldSlot;
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent update")));
/*
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
* Already know that we're going to need to do EPQ, so
* fetch tuple directly into the right slot.
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
inputslot = EvalPlanQualSlot(context->epqstate, resultRelationDesc,
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
resultRelInfo->ri_RangeTableIndex);
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
result = table_tuple_lock(resultRelationDesc, tupleid,
estate->es_snapshot,
inputslot, estate->es_output_cid,
updateCxt.lockmode, LockWaitBlock,
TUPLE_LOCK_FLAG_FIND_LAST_VERSION,
&context->tmfd);
switch (result)
{
case TM_Ok:
Assert(context->tmfd.traversed);
epqslot = EvalPlanQual(context->epqstate,
resultRelationDesc,
resultRelInfo->ri_RangeTableIndex,
inputslot);
if (TupIsNull(epqslot))
/* Tuple not passing quals anymore, exiting... */
return NULL;
Revert 764da7710b Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:55:09 +02:00
/* Make sure ri_oldTupleSlot is initialized. */
if (unlikely(!resultRelInfo->ri_projectNewInfoValid))
ExecInitUpdateProjection(context->mtstate,
resultRelInfo);
/* Fetch the most recent version of old tuple. */
oldSlot = resultRelInfo->ri_oldTupleSlot;
if (!table_tuple_fetch_row_version(resultRelationDesc,
tupleid,
SnapshotAny,
oldSlot))
elog(ERROR, "failed to fetch tuple being updated");
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
slot = ExecGetUpdateNewTuple(resultRelInfo,
epqslot, oldSlot);
goto redo_act;
case TM_Deleted:
/* tuple already deleted; nothing to do */
return NULL;
case TM_SelfModified:
/*
* This can be reached when following an update
* chain from a tuple updated by another session,
* reaching a tuple that was already updated in
* this transaction. If previously modified by
* this command, ignore the redundant update,
* otherwise error out.
*
* See also TM_SelfModified response to
* table_tuple_update() above.
*/
if (context->tmfd.cmax != estate->es_output_cid)
ereport(ERROR,
(errcode(ERRCODE_TRIGGERED_DATA_CHANGE_VIOLATION),
errmsg("tuple to be updated was already modified by an operation triggered by the current command"),
errhint("Consider using an AFTER trigger instead of a BEFORE trigger to propagate changes to other rows.")));
return NULL;
default:
/* see table_tuple_lock call in ExecDelete() */
elog(ERROR, "unexpected table_tuple_lock status: %u",
result);
return NULL;
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
break;
case TM_Deleted:
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent delete")));
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
/* tuple already deleted; nothing to do */
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
return NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
default:
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
elog(ERROR, "unrecognized table_tuple_update status: %u",
result);
Support triggers on views. This patch adds the SQL-standard concept of an INSTEAD OF trigger, which is fired instead of performing a physical insert/update/delete. The trigger function is passed the entire old and/or new rows of the view, and must figure out what to do to the underlying tables to implement the update. So this feature can be used to implement updatable views using trigger programming style rather than rule hacking. In passing, this patch corrects the names of some columns in the information_schema.triggers view. It seems the SQL committee renamed them somewhere between SQL:99 and SQL:2003. Dean Rasheed, reviewed by Bernd Helmle; some additional hacking by me.
2010-10-10 19:43:33 +02:00
return NULL;
}
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
if (canSetTag)
(estate->es_processed)++;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecUpdateEpilogue(context, &updateCxt, resultRelInfo, tupleid, oldtuple,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
slot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* Process RETURNING if present */
if (resultRelInfo->ri_projectReturning)
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
return ExecProcessReturning(resultRelInfo, slot, context->planSlot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
return NULL;
}
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* ExecOnConflictUpdate --- execute UPDATE of INSERT ON CONFLICT DO UPDATE
*
* Try to lock tuple for update as part of speculative insertion. If
* a qual originating from ON CONFLICT DO UPDATE is satisfied, update
* (but still lock row, even though it may not satisfy estate's
* snapshot).
*
Fix misc typos, mostly in comments. A collection of typos I happened to spot while reading code, as well as grepping for common mistakes. Backpatch to all supported versions, as applicable, to avoid conflicts when backporting other commits in the future.
2018-07-09 15:10:44 +02:00
* Returns true if we're done (with or without an update), or false if
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
* the caller must retry the INSERT from scratch.
*/
static bool
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecOnConflictUpdate(ModifyTableContext *context,
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ResultRelInfo *resultRelInfo,
ItemPointer conflictTid,
TupleTableSlot *excludedSlot,
bool canSetTag,
TupleTableSlot **returning)
{
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ModifyTableState *mtstate = context->mtstate;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ExprContext *econtext = mtstate->ps.ps_ExprContext;
Relation relation = resultRelInfo->ri_RelationDesc;
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
ExprState *onConflictSetWhere = resultRelInfo->ri_onConflict->oc_WhereClause;
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
TupleTableSlot *existing = resultRelInfo->ri_onConflict->oc_Existing;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
TM_FailureData tmfd;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
LockTupleMode lockmode;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
TM_Result test;
Datum xminDatum;
TransactionId xmin;
bool isnull;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* Determine lock mode to use */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
lockmode = ExecUpdateLockMode(context->estate, resultRelInfo);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* Lock tuple for update. Don't follow updates when tuple cannot be
* locked without doing so. A row locking conflict here means our
* previous conclusion that the tuple is conclusively committed is not
* true anymore.
*/
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
test = table_tuple_lock(relation, conflictTid,
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
context->estate->es_snapshot,
existing, context->estate->es_output_cid,
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
lockmode, LockWaitBlock, 0,
&tmfd);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
switch (test)
{
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Ok:
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* success! */
break;
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Invisible:
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* This can occur when a just inserted tuple is updated again in
* the same command. E.g. because multiple rows with the same
* conflicting key values are inserted.
*
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
* This is somewhat similar to the ExecUpdate() TM_SelfModified
* case. We do not want to proceed because it would lead to the
* same row being updated a second time in some unspecified order,
* and in contrast to plain UPDATEs there's no historical behavior
* to break.
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
*
* It is the user's responsibility to prevent this situation from
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* occurring. These problems are why the SQL standard similarly
* specifies that for SQL MERGE, an exception must be raised in
* the event of an attempt to update the same row twice.
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
*/
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
xminDatum = slot_getsysattr(existing,
MinTransactionIdAttributeNumber,
&isnull);
Assert(!isnull);
xmin = DatumGetTransactionId(xminDatum);
if (TransactionIdIsCurrentTransactionId(xmin))
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ereport(ERROR,
(errcode(ERRCODE_CARDINALITY_VIOLATION),
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/* translator: %s is a SQL command name */
errmsg("%s command cannot affect row a second time",
"ON CONFLICT DO UPDATE"),
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
errhint("Ensure that no rows proposed for insertion within the same command have duplicate constrained values.")));
/* This shouldn't happen */
elog(ERROR, "attempted to lock invisible tuple");
Clean up warnings from -Wimplicit-fallthrough. Recent gcc can warn about switch-case fall throughs that are not explicitly labeled as intentional. This seems like a good thing, so clean up the warnings exposed thereby by labeling all such cases with comments that gcc will recognize. In files that already had one or more suitable comments, I generally matched the existing style of those. Otherwise I went with /* FALLTHROUGH */, which is one of the spellings approved at the more-restrictive-than-default level -Wimplicit-fallthrough=4. (At the default level you can also spell it /* FALL ?THRU */, and it's not picky about case. What you can't do is include additional text in the same comment, so some existing comments containing versions of this aren't good enough.) Testing with gcc 8.0.1 (Fedora 28's current version), I found that I also had to put explicit "break"s after elog(ERROR) or ereport(ERROR); apparently, for this purpose gcc doesn't recognize that those don't return. That seems like possibly a gcc bug, but it's fine because in most places we did that anyway; so this amounts to a visit from the style police. Discussion: https://postgr.es/m/15083.1525207729@sss.pgh.pa.us
2018-05-02 01:35:08 +02:00
break;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_SelfModified:
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* This state should never be reached. As a dirty snapshot is used
* to find conflicting tuples, speculative insertion wouldn't have
* seen this row to conflict with.
*/
elog(ERROR, "unexpected self-updated tuple");
Clean up warnings from -Wimplicit-fallthrough. Recent gcc can warn about switch-case fall throughs that are not explicitly labeled as intentional. This seems like a good thing, so clean up the warnings exposed thereby by labeling all such cases with comments that gcc will recognize. In files that already had one or more suitable comments, I generally matched the existing style of those. Otherwise I went with /* FALLTHROUGH */, which is one of the spellings approved at the more-restrictive-than-default level -Wimplicit-fallthrough=4. (At the default level you can also spell it /* FALL ?THRU */, and it's not picky about case. What you can't do is include additional text in the same comment, so some existing comments containing versions of this aren't good enough.) Testing with gcc 8.0.1 (Fedora 28's current version), I found that I also had to put explicit "break"s after elog(ERROR) or ereport(ERROR); apparently, for this purpose gcc doesn't recognize that those don't return. That seems like possibly a gcc bug, but it's fine because in most places we did that anyway; so this amounts to a visit from the style police. Discussion: https://postgr.es/m/15083.1525207729@sss.pgh.pa.us
2018-05-02 01:35:08 +02:00
break;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
case TM_Updated:
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent update")));
Raise error when affecting tuple moved into different partition. When an update moves a row between partitions (supported since 2f178441044b), our normal logic for following update chains in READ COMMITTED mode doesn't work anymore. Cross partition updates are modeled as an delete from the old and insert into the new partition. No ctid chain exists across partitions, and there's no convenient space to introduce that link. Not throwing an error in a partitioned context when one would have been thrown without partitioning is obviously problematic. This commit introduces infrastructure to detect when a tuple has been moved, not just plainly deleted. That allows to throw an error when encountering a deletion that's actually a move, while attempting to following a ctid chain. The row deleted as part of a cross partition update is marked by pointing it's t_ctid to an invalid block, instead of self as a normal update would. That was deemed to be the least invasive and most future proof way to represent the knowledge, given how few infomask bits are there to be recycled (there's also some locking issues with using infomask bits). External code following ctid chains should be updated to check for moved tuples. The most likely consequence of not doing so is a missed error. Author: Amul Sul, editorialized by me Reviewed-By: Amit Kapila, Pavan Deolasee, Andres Freund, Robert Haas Discussion: http://postgr.es/m/CAAJ_b95PkwojoYfz0bzXU8OokcTVGzN6vYGCNVUukeUDrnF3dw@mail.gmail.com
2018-04-07 22:24:10 +02:00
/*
* As long as we don't support an UPDATE of INSERT ON CONFLICT for
* a partitioned table we shouldn't reach to a case where tuple to
* be lock is moved to another partition due to concurrent update
* of the partition key.
*/
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
Assert(!ItemPointerIndicatesMovedPartitions(&tmfd.ctid));
Raise error when affecting tuple moved into different partition. When an update moves a row between partitions (supported since 2f178441044b), our normal logic for following update chains in READ COMMITTED mode doesn't work anymore. Cross partition updates are modeled as an delete from the old and insert into the new partition. No ctid chain exists across partitions, and there's no convenient space to introduce that link. Not throwing an error in a partitioned context when one would have been thrown without partitioning is obviously problematic. This commit introduces infrastructure to detect when a tuple has been moved, not just plainly deleted. That allows to throw an error when encountering a deletion that's actually a move, while attempting to following a ctid chain. The row deleted as part of a cross partition update is marked by pointing it's t_ctid to an invalid block, instead of self as a normal update would. That was deemed to be the least invasive and most future proof way to represent the knowledge, given how few infomask bits are there to be recycled (there's also some locking issues with using infomask bits). External code following ctid chains should be updated to check for moved tuples. The most likely consequence of not doing so is a missed error. Author: Amul Sul, editorialized by me Reviewed-By: Amit Kapila, Pavan Deolasee, Andres Freund, Robert Haas Discussion: http://postgr.es/m/CAAJ_b95PkwojoYfz0bzXU8OokcTVGzN6vYGCNVUukeUDrnF3dw@mail.gmail.com
2018-04-07 22:24:10 +02:00
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* Tell caller to try again from the very start.
*
* It does not make sense to use the usual EvalPlanQual() style
* loop here, as the new version of the row might not conflict
* anymore, or the conflicting tuple has actually been deleted.
*/
tableam: Add tuple_{insert, delete, update, lock} and use. This adds new, required, table AM callbacks for insert/delete/update and lock_tuple. To be able to reasonably use those, the EvalPlanQual mechanism had to be adapted, moving more logic into the AM. Previously both delete/update/lock call-sites and the EPQ mechanism had to have awareness of the specific tuple format to be able to fetch the latest version of a tuple. Obviously that needs to be abstracted away. To do so, move the logic that find the latest row version into the AM. lock_tuple has a new flag argument, TUPLE_LOCK_FLAG_FIND_LAST_VERSION, that forces it to lock the last version, rather than the current one. It'd have been possible to do so via a separate callback as well, but finding the last version usually also necessitates locking the newest version, making it sensible to combine the two. This replaces the previous use of EvalPlanQualFetch(). Additionally HeapTupleUpdated, which previously signaled either a concurrent update or delete, is now split into two, to avoid callers needing AM specific knowledge to differentiate. The move of finding the latest row version into tuple_lock means that encountering a row concurrently moved into another partition will now raise an error about "tuple to be locked" rather than "tuple to be updated/deleted" - which is accurate, as that always happens when locking rows. While possible slightly less helpful for users, it seems like an acceptable trade-off. As part of this commit HTSU_Result has been renamed to TM_Result, and its members been expanded to differentiated between updating and deleting. HeapUpdateFailureData has been renamed to TM_FailureData. The interface to speculative insertion is changed so nodeModifyTable.c does not have to set the speculative token itself anymore. Instead there's a version of tuple_insert, tuple_insert_speculative, that performs the speculative insertion (without requiring a flag to signal that fact), and the speculative insertion is either made permanent with table_complete_speculative(succeeded = true) or aborted with succeeded = false). Note that multi_insert is not yet routed through tableam, nor is COPY. Changing multi_insert requires changes to copy.c that are large enough to better be done separately. Similarly, although simpler, CREATE TABLE AS and CREATE MATERIALIZED VIEW are also only going to be adjusted in a later commit. Author: Andres Freund and Haribabu Kommi Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20190313003903.nwvrxi7rw3ywhdel@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-24 03:55:57 +01:00
ExecClearTuple(existing);
return false;
case TM_Deleted:
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent delete")));
/* see TM_Updated case */
Assert(!ItemPointerIndicatesMovedPartitions(&tmfd.ctid));
ExecClearTuple(existing);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
return false;
default:
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
elog(ERROR, "unrecognized table_tuple_lock status: %u", test);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
}
Rationalize expression context reset in ExecModifyTable(). The current pattern of reseting expressions both in ExecProcessReturning() and ExecOnConflictUpdate() makes it harder than necessary to reason about memory lifetimes. It also requires materializing slots unnecessarily, although this patch doesn't take advantage of the fact that that's not necessary anymore. Instead reset the expression context once for each input tuple. Author: Ashutosh Bapat Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-13 21:18:25 +01:00
/* Success, the tuple is locked. */
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* Verify that the tuple is visible to our MVCC snapshot if the current
* isolation level mandates that.
*
* It's not sufficient to rely on the check within ExecUpdate() as e.g.
* CONFLICT ... WHERE clause may prevent us from reaching that.
*
* This means we only ever continue when a new command in the current
* transaction could see the row, even though in READ COMMITTED mode the
* tuple will not be visible according to the current statement's
* snapshot. This is in line with the way UPDATE deals with newer tuple
* versions.
*/
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ExecCheckTupleVisible(context->estate, relation, existing);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* Make tuple and any needed join variables available to ExecQual and
* ExecProject. The EXCLUDED tuple is installed in ecxt_innertuple, while
* the target's existing tuple is installed in the scantuple. EXCLUDED
* has been made to reference INNER_VAR in setrefs.c, but there is no
* other redirection.
*/
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
econtext->ecxt_scantuple = existing;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
econtext->ecxt_innertuple = excludedSlot;
econtext->ecxt_outertuple = NULL;
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
if (!ExecQual(onConflictSetWhere, econtext))
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
{
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
ExecClearTuple(existing); /* see return below */
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
InstrCountFiltered1(&mtstate->ps, 1);
return true; /* done with the tuple */
}
if (resultRelInfo->ri_WithCheckOptions != NIL)
{
/*
* Check target's existing tuple against UPDATE-applicable USING
* security barrier quals (if any), enforced here as RLS checks/WCOs.
*
* The rewriter creates UPDATE RLS checks/WCOs for UPDATE security
* quals, and stores them as WCOs of "kind" WCO_RLS_CONFLICT_CHECK,
* but that's almost the extent of its special handling for ON
* CONFLICT DO UPDATE.
*
* The rewriter will also have associated UPDATE applicable straight
* RLS checks/WCOs for the benefit of the ExecUpdate() call that
* follows. INSERTs and UPDATEs naturally have mutually exclusive WCO
* kinds, so there is no danger of spurious over-enforcement in the
* INSERT or UPDATE path.
*/
ExecWithCheckOptions(WCO_RLS_CONFLICT_CHECK, resultRelInfo,
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
existing,
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
mtstate->ps.state);
}
/* Project the new tuple version */
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
ExecProject(resultRelInfo->ri_onConflict->oc_ProjInfo);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
Fix ON CONFLICT UPDATE bug breaking AFTER UPDATE triggers. ExecOnConflictUpdate() passed t_ctid of the to-be-updated tuple to ExecUpdate(). That's problematic primarily because of two reason: First and foremost t_ctid could point to a different tuple. Secondly, and that's what triggered the complaint by Stanislav, t_ctid is changed by heap_update() to point to the new tuple version. The behavior of AFTER UPDATE triggers was therefore broken, with NEW.* and OLD.* tuples spuriously identical within AFTER UPDATE triggers. To fix both issues, pass a pointer to t_self of a on-stack HeapTuple instead. Fixing this bug lead to one change in regression tests, which previously failed due to the first issue mentioned above. There's a reasonable expectation that test fails, as it updates one row repeatedly within one INSERT ... ON CONFLICT statement. That is only possible if the second update is triggered via ON CONFLICT ... SET, ON CONFLICT ... WHERE, or by a WITH CHECK expression, as those are executed after ExecOnConflictUpdate() does a visibility check. That could easily be prohibited, but given it's allowed for plain UPDATEs and a rare corner case, it doesn't seem worthwhile. Reported-By: Stanislav Grozev Author: Andres Freund and Peter Geoghegan Discussion: CAA78GVqy1+LisN-8DygekD_Ldfy=BJLarSpjGhytOsgkpMavfQ@mail.gmail.com Backpatch: 9.5, where ON CONFLICT was introduced
2015-12-10 16:26:45 +01:00
/*
* Note that it is possible that the target tuple has been modified in
tableam: Rename wrapper functions to match callback names. Some of the wrapper functions didn't match the callback names. Many of them due to staying "consistent" with historic naming of the wrapped functionality. We decided that for most cases it's more important to be for tableam to be consistent going forward, than with the past. The one exception is beginscan/endscan/... because it'd have looked odd to have systable_beginscan/endscan/... with a different naming scheme, and changing the systable_* APIs would have caused way too much churn (including breaking a lot of external users). Author: Ashwin Agrawal, with some small additions by Andres Freund Reviewed-By: Andres Freund Discussion: https://postgr.es/m/CALfoeiugyrXZfX7n0ORCa4L-m834dzmaE8eFdbNR6PMpetU4Ww@mail.gmail.com
2019-05-24 01:25:48 +02:00
* this session, after the above table_tuple_lock. We choose to not error
Fix ON CONFLICT UPDATE bug breaking AFTER UPDATE triggers. ExecOnConflictUpdate() passed t_ctid of the to-be-updated tuple to ExecUpdate(). That's problematic primarily because of two reason: First and foremost t_ctid could point to a different tuple. Secondly, and that's what triggered the complaint by Stanislav, t_ctid is changed by heap_update() to point to the new tuple version. The behavior of AFTER UPDATE triggers was therefore broken, with NEW.* and OLD.* tuples spuriously identical within AFTER UPDATE triggers. To fix both issues, pass a pointer to t_self of a on-stack HeapTuple instead. Fixing this bug lead to one change in regression tests, which previously failed due to the first issue mentioned above. There's a reasonable expectation that test fails, as it updates one row repeatedly within one INSERT ... ON CONFLICT statement. That is only possible if the second update is triggered via ON CONFLICT ... SET, ON CONFLICT ... WHERE, or by a WITH CHECK expression, as those are executed after ExecOnConflictUpdate() does a visibility check. That could easily be prohibited, but given it's allowed for plain UPDATEs and a rare corner case, it doesn't seem worthwhile. Reported-By: Stanislav Grozev Author: Andres Freund and Peter Geoghegan Discussion: CAA78GVqy1+LisN-8DygekD_Ldfy=BJLarSpjGhytOsgkpMavfQ@mail.gmail.com Backpatch: 9.5, where ON CONFLICT was introduced
2015-12-10 16:26:45 +01:00
* out in that case, in line with ExecUpdate's treatment of similar cases.
* This can happen if an UPDATE is triggered from within ExecQual(),
* ExecWithCheckOptions() or ExecProject() above, e.g. by selecting from a
* wCTE in the ON CONFLICT's SET.
*/
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* Execute UPDATE with projection */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
*returning = ExecUpdate(context, resultRelInfo,
conflictTid, NULL,
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
resultRelInfo->ri_onConflict->oc_ProjSlot,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
canSetTag);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
/*
* Clear out existing tuple, as there might not be another conflict among
* the next input rows. Don't want to hold resources till the end of the
* query.
*/
ExecClearTuple(existing);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
return true;
}
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/*
* Perform MERGE.
*/
static TupleTableSlot *
ExecMerge(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
ItemPointer tupleid, bool canSetTag)
{
bool matched;
/*-----
* If we are dealing with a WHEN MATCHED case (tupleid is valid), we
* execute the first action for which the additional WHEN MATCHED AND
* quals pass. If an action without quals is found, that action is
* executed.
*
* Similarly, if we are dealing with WHEN NOT MATCHED case, we look at
* the given WHEN NOT MATCHED actions in sequence until one passes.
*
* Things get interesting in case of concurrent update/delete of the
* target tuple. Such concurrent update/delete is detected while we are
* executing a WHEN MATCHED action.
*
* A concurrent update can:
*
* 1. modify the target tuple so that it no longer satisfies the
* additional quals attached to the current WHEN MATCHED action
*
* In this case, we are still dealing with a WHEN MATCHED case.
* We recheck the list of WHEN MATCHED actions from the start and
* choose the first one that satisfies the new target tuple.
*
* 2. modify the target tuple so that the join quals no longer pass and
* hence the source tuple no longer has a match.
*
* In this case, the source tuple no longer matches the target tuple,
* so we now instead find a qualifying WHEN NOT MATCHED action to
* execute.
*
* XXX Hmmm, what if the updated tuple would now match one that was
* considered NOT MATCHED so far?
*
* A concurrent delete changes a WHEN MATCHED case to WHEN NOT MATCHED.
*
* ExecMergeMatched takes care of following the update chain and
* re-finding the qualifying WHEN MATCHED action, as long as the updated
* target tuple still satisfies the join quals, i.e., it remains a WHEN
* MATCHED case. If the tuple gets deleted or the join quals fail, it
* returns and we try ExecMergeNotMatched. Given that ExecMergeMatched
* always make progress by following the update chain and we never switch
* from ExecMergeNotMatched to ExecMergeMatched, there is no risk of a
* livelock.
*/
matched = tupleid != NULL;
if (matched)
matched = ExecMergeMatched(context, resultRelInfo, tupleid, canSetTag);
/*
* Either we were dealing with a NOT MATCHED tuple or ExecMergeMatched()
* returned "false", indicating the previously MATCHED tuple no longer
* matches.
*/
if (!matched)
ExecMergeNotMatched(context, resultRelInfo, canSetTag);
/* No RETURNING support yet */
return NULL;
}
/*
* Check and execute the first qualifying MATCHED action. The current target
* tuple is identified by tupleid.
*
* We start from the first WHEN MATCHED action and check if the WHEN quals
* pass, if any. If the WHEN quals for the first action do not pass, we
* check the second, then the third and so on. If we reach to the end, no
* action is taken and we return true, indicating that no further action is
* required for this tuple.
*
* If we do find a qualifying action, then we attempt to execute the action.
*
* If the tuple is concurrently updated, EvalPlanQual is run with the updated
* tuple to recheck the join quals. Note that the additional quals associated
* with individual actions are evaluated by this routine via ExecQual, while
* EvalPlanQual checks for the join quals. If EvalPlanQual tells us that the
* updated tuple still passes the join quals, then we restart from the first
* action to look for a qualifying action. Otherwise, we return false --
* meaning that a NOT MATCHED action must now be executed for the current
* source tuple.
*/
static bool
ExecMergeMatched(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
ItemPointer tupleid, bool canSetTag)
{
ModifyTableState *mtstate = context->mtstate;
TupleTableSlot *newslot;
EState *estate = context->estate;
ExprContext *econtext = mtstate->ps.ps_ExprContext;
bool isNull;
EPQState *epqstate = &mtstate->mt_epqstate;
ListCell *l;
/*
* If there are no WHEN MATCHED actions, we are done.
*/
if (resultRelInfo->ri_matchedMergeAction == NIL)
return true;
/*
* Make tuple and any needed join variables available to ExecQual and
* ExecProject. The target's existing tuple is installed in the scantuple.
* Again, this target relation's slot is required only in the case of a
* MATCHED tuple and UPDATE/DELETE actions.
*/
econtext->ecxt_scantuple = resultRelInfo->ri_oldTupleSlot;
econtext->ecxt_innertuple = context->planSlot;
econtext->ecxt_outertuple = NULL;
Remove unnecessary semicolons after goto labels According to the C standard, a label must followed by a statement. If there was ever a time we needed an empty statement here, it was a long time ago. Japin Li Reviewed by Julien Rouhaud Discussion: https://www.postgresql.org/message-id/MEYP282MB16690F40189A4F060B41D56DB65E9%40MEYP282MB1669.AUSP282.PROD.OUTLOOK.COM
2022-10-10 10:08:38 +02:00
lmerge_matched:
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/*
* This routine is only invoked for matched rows, and we must have found
* the tupleid of the target row in that case; fetch that tuple.
*
* We use SnapshotAny for this because we might get called again after
* EvalPlanQual returns us a new tuple, which may not be visible to our
* MVCC snapshot.
*/
if (!table_tuple_fetch_row_version(resultRelInfo->ri_RelationDesc,
tupleid,
SnapshotAny,
resultRelInfo->ri_oldTupleSlot))
elog(ERROR, "failed to fetch the target tuple");
foreach(l, resultRelInfo->ri_matchedMergeAction)
{
MergeActionState *relaction = (MergeActionState *) lfirst(l);
CmdType commandType = relaction->mas_action->commandType;
TM_Result result;
UpdateContext updateCxt = {0};
/*
* Test condition, if any.
*
* In the absence of any condition, we perform the action
* unconditionally (no need to check separately since ExecQual() will
* return true if there are no conditions to evaluate).
*/
if (!ExecQual(relaction->mas_whenqual, econtext))
continue;
/*
* Check if the existing target tuple meets the USING checks of
* UPDATE/DELETE RLS policies. If those checks fail, we throw an
* error.
*
* The WITH CHECK quals are applied in ExecUpdate() and hence we need
* not do anything special to handle them.
*
* NOTE: We must do this after WHEN quals are evaluated, so that we
* check policies only when they matter.
*/
if (resultRelInfo->ri_WithCheckOptions)
{
ExecWithCheckOptions(commandType == CMD_UPDATE ?
WCO_RLS_MERGE_UPDATE_CHECK : WCO_RLS_MERGE_DELETE_CHECK,
resultRelInfo,
resultRelInfo->ri_oldTupleSlot,
context->mtstate->ps.state);
}
/* Perform stated action */
switch (commandType)
{
case CMD_UPDATE:
/*
* Project the output tuple, and use that to update the table.
* We don't need to filter out junk attributes, because the
* UPDATE action's targetlist doesn't have any.
*/
newslot = ExecProject(relaction->mas_proj);
context->relaction = relaction;
if (!ExecUpdatePrologue(context, resultRelInfo,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
tupleid, NULL, newslot, &result))
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
{
Fix MERGE command tag for actions blocked by BEFORE ROW triggers. This ensures that the row count in the command tag for a MERGE is correctly computed in the case where UPDATEs or DELETEs are skipped due to a BEFORE ROW trigger returning NULL (the INSERT case was already handled correctly by ExecMergeNotMatched() calling ExecInsert()). Back-patch to v15, where MERGE was introduced. Discussion: https://postgr.es/m/CAEZATCU8XEmR0JWKDtyb7iZ%3DqCffxS9uyJt0iOZ4TV4RT%2Bow1w%40mail.gmail.com
2023-03-13 12:09:39 +01:00
if (result == TM_Ok)
return true; /* "do nothing" */
break; /* concurrent update/delete */
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
}
result = ExecUpdateAct(context, resultRelInfo, tupleid, NULL,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
newslot, false, &updateCxt);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
if (result == TM_Ok && updateCxt.updated)
{
ExecUpdateEpilogue(context, &updateCxt, resultRelInfo,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
tupleid, NULL, newslot);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
mtstate->mt_merge_updated += 1;
}
break;
case CMD_DELETE:
context->relaction = relaction;
if (!ExecDeletePrologue(context, resultRelInfo, tupleid,
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
NULL, NULL, &result))
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
{
Fix MERGE command tag for actions blocked by BEFORE ROW triggers. This ensures that the row count in the command tag for a MERGE is correctly computed in the case where UPDATEs or DELETEs are skipped due to a BEFORE ROW trigger returning NULL (the INSERT case was already handled correctly by ExecMergeNotMatched() calling ExecInsert()). Back-patch to v15, where MERGE was introduced. Discussion: https://postgr.es/m/CAEZATCU8XEmR0JWKDtyb7iZ%3DqCffxS9uyJt0iOZ4TV4RT%2Bow1w%40mail.gmail.com
2023-03-13 12:09:39 +01:00
if (result == TM_Ok)
return true; /* "do nothing" */
break; /* concurrent update/delete */
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
}
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
result = ExecDeleteAct(context, resultRelInfo, tupleid, false);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
if (result == TM_Ok)
{
ExecDeleteEpilogue(context, resultRelInfo, tupleid, NULL,
false);
mtstate->mt_merge_deleted += 1;
}
break;
case CMD_NOTHING:
/* Doing nothing is always OK */
result = TM_Ok;
break;
default:
elog(ERROR, "unknown action in MERGE WHEN MATCHED clause");
}
switch (result)
{
case TM_Ok:
/* all good; perform final actions */
Fix MERGE tuple count with DO NOTHING Reporting tuples for which nothing is done is useless and goes against the documented behavior, so don't do it. Backpatch to 15. Reported by: Luca Ferrari Discussion: https://postgr.es/m/CAKoxK+42MmACUh6s8XzASQKizbzrtOGA6G1UjzCP75NcXHsiNw@mail.gmail.com
2022-11-17 18:56:11 +01:00
if (canSetTag && commandType != CMD_NOTHING)
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
(estate->es_processed)++;
break;
case TM_SelfModified:
/*
* The SQL standard disallows this for MERGE.
*/
if (TransactionIdIsCurrentTransactionId(context->tmfd.xmax))
ereport(ERROR,
(errcode(ERRCODE_CARDINALITY_VIOLATION),
/* translator: %s is a SQL command name */
errmsg("%s command cannot affect row a second time",
"MERGE"),
errhint("Ensure that not more than one source row matches any one target row.")));
/* This shouldn't happen */
elog(ERROR, "attempted to update or delete invisible tuple");
break;
case TM_Deleted:
if (IsolationUsesXactSnapshot())
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("could not serialize access due to concurrent delete")));
/*
* If the tuple was already deleted, return to let caller
* handle it under NOT MATCHED clauses.
*/
return false;
case TM_Updated:
{
Relation resultRelationDesc;
TupleTableSlot *epqslot,
*inputslot;
LockTupleMode lockmode;
/*
* The target tuple was concurrently updated by some other
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
* transaction. Run EvalPlanQual() with the new version of
* the tuple. If it does not return a tuple, then we
* switch to the NOT MATCHED list of actions. If it does
* return a tuple and the join qual is still satisfied,
* then we just need to recheck the MATCHED actions,
* starting from the top, and execute the first qualifying
* action.
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
*/
resultRelationDesc = resultRelInfo->ri_RelationDesc;
lockmode = ExecUpdateLockMode(estate, resultRelInfo);
inputslot = EvalPlanQualSlot(epqstate, resultRelationDesc,
resultRelInfo->ri_RangeTableIndex);
result = table_tuple_lock(resultRelationDesc, tupleid,
estate->es_snapshot,
inputslot, estate->es_output_cid,
lockmode, LockWaitBlock,
TUPLE_LOCK_FLAG_FIND_LAST_VERSION,
&context->tmfd);
switch (result)
{
case TM_Ok:
epqslot = EvalPlanQual(epqstate,
resultRelationDesc,
resultRelInfo->ri_RangeTableIndex,
inputslot);
/*
* If we got no tuple, or the tuple we get has a
* NULL ctid, go back to caller: this one is not a
* MATCHED tuple anymore, so they can retry with
* NOT MATCHED actions.
*/
if (TupIsNull(epqslot))
return false;
(void) ExecGetJunkAttribute(epqslot,
resultRelInfo->ri_RowIdAttNo,
&isNull);
if (isNull)
return false;
/*
* When a tuple was updated and migrated to
* another partition concurrently, the current
* MERGE implementation can't follow. There's
* probably a better way to handle this case, but
* it'd require recognizing the relation to which
* the tuple moved, and setting our current
* resultRelInfo to that.
*/
if (ItemPointerIndicatesMovedPartitions(&context->tmfd.ctid))
ereport(ERROR,
(errcode(ERRCODE_T_R_SERIALIZATION_FAILURE),
errmsg("tuple to be deleted was already moved to another partition due to concurrent update")));
/*
* A non-NULL ctid means that we are still dealing
* with MATCHED case. Restart the loop so that we
* apply all the MATCHED rules again, to ensure
* that the first qualifying WHEN MATCHED action
* is executed.
*
* Update tupleid to that of the new tuple, for
* the refetch we do at the top.
*/
ItemPointerCopy(&context->tmfd.ctid, tupleid);
goto lmerge_matched;
case TM_Deleted:
/*
* tuple already deleted; tell caller to run NOT
* MATCHED actions
*/
return false;
case TM_SelfModified:
/*
* This can be reached when following an update
* chain from a tuple updated by another session,
* reaching a tuple that was already updated in
* this transaction. If previously modified by
* this command, ignore the redundant update,
* otherwise error out.
*
* See also response to TM_SelfModified in
* ExecUpdate().
*/
if (context->tmfd.cmax != estate->es_output_cid)
ereport(ERROR,
(errcode(ERRCODE_TRIGGERED_DATA_CHANGE_VIOLATION),
errmsg("tuple to be updated or deleted was already modified by an operation triggered by the current command"),
errhint("Consider using an AFTER trigger instead of a BEFORE trigger to propagate changes to other rows.")));
return false;
default:
/* see table_tuple_lock call in ExecDelete() */
elog(ERROR, "unexpected table_tuple_lock status: %u",
result);
return false;
}
}
case TM_Invisible:
case TM_WouldBlock:
case TM_BeingModified:
/* these should not occur */
elog(ERROR, "unexpected tuple operation result: %d", result);
break;
}
/*
* We've activated one of the WHEN clauses, so we don't search
* further. This is required behaviour, not an optimization.
*/
break;
}
/*
* Successfully executed an action or no qualifying action was found.
*/
return true;
}
/*
* Execute the first qualifying NOT MATCHED action.
*/
static void
ExecMergeNotMatched(ModifyTableContext *context, ResultRelInfo *resultRelInfo,
bool canSetTag)
{
ModifyTableState *mtstate = context->mtstate;
ExprContext *econtext = mtstate->ps.ps_ExprContext;
List *actionStates = NIL;
ListCell *l;
/*
* For INSERT actions, the root relation's merge action is OK since the
* INSERT's targetlist and the WHEN conditions can only refer to the
* source relation and hence it does not matter which result relation we
* work with.
*
* XXX does this mean that we can avoid creating copies of actionStates on
* partitioned tables, for not-matched actions?
*/
actionStates = resultRelInfo->ri_notMatchedMergeAction;
/*
* Make source tuple available to ExecQual and ExecProject. We don't need
* the target tuple, since the WHEN quals and targetlist can't refer to
* the target columns.
*/
econtext->ecxt_scantuple = NULL;
econtext->ecxt_innertuple = context->planSlot;
econtext->ecxt_outertuple = NULL;
foreach(l, actionStates)
{
MergeActionState *action = (MergeActionState *) lfirst(l);
CmdType commandType = action->mas_action->commandType;
TupleTableSlot *newslot;
/*
* Test condition, if any.
*
* In the absence of any condition, we perform the action
* unconditionally (no need to check separately since ExecQual() will
* return true if there are no conditions to evaluate).
*/
if (!ExecQual(action->mas_whenqual, econtext))
continue;
/* Perform stated action */
switch (commandType)
{
case CMD_INSERT:
/*
* Project the tuple. In case of a partitioned table, the
* projection was already built to use the root's descriptor,
* so we don't need to map the tuple here.
*/
newslot = ExecProject(action->mas_proj);
context->relaction = action;
(void) ExecInsert(context, mtstate->rootResultRelInfo, newslot,
canSetTag, NULL, NULL);
mtstate->mt_merge_inserted += 1;
break;
case CMD_NOTHING:
/* Do nothing */
break;
default:
elog(ERROR, "unknown action in MERGE WHEN NOT MATCHED clause");
}
/*
* We've activated one of the WHEN clauses, so we don't search
* further. This is required behaviour, not an optimization.
*/
break;
}
}
/*
* Initialize state for execution of MERGE.
*/
void
ExecInitMerge(ModifyTableState *mtstate, EState *estate)
{
ModifyTable *node = (ModifyTable *) mtstate->ps.plan;
ResultRelInfo *rootRelInfo = mtstate->rootResultRelInfo;
ResultRelInfo *resultRelInfo;
ExprContext *econtext;
ListCell *lc;
int i;
if (node->mergeActionLists == NIL)
return;
mtstate->mt_merge_subcommands = 0;
if (mtstate->ps.ps_ExprContext == NULL)
ExecAssignExprContext(estate, &mtstate->ps);
econtext = mtstate->ps.ps_ExprContext;
/*
* Create a MergeActionState for each action on the mergeActionList and
* add it to either a list of matched actions or not-matched actions.
*
* Similar logic appears in ExecInitPartitionInfo(), so if changing
* anything here, do so there too.
*/
i = 0;
foreach(lc, node->mergeActionLists)
{
List *mergeActionList = lfirst(lc);
TupleDesc relationDesc;
ListCell *l;
resultRelInfo = mtstate->resultRelInfo + i;
i++;
relationDesc = RelationGetDescr(resultRelInfo->ri_RelationDesc);
/* initialize slots for MERGE fetches from this rel */
if (unlikely(!resultRelInfo->ri_projectNewInfoValid))
ExecInitMergeTupleSlots(mtstate, resultRelInfo);
foreach(l, mergeActionList)
{
MergeAction *action = (MergeAction *) lfirst(l);
MergeActionState *action_state;
TupleTableSlot *tgtslot;
TupleDesc tgtdesc;
List **list;
/*
* Build action merge state for this rel. (For partitions,
* equivalent code exists in ExecInitPartitionInfo.)
*/
action_state = makeNode(MergeActionState);
action_state->mas_action = action;
action_state->mas_whenqual = ExecInitQual((List *) action->qual,
&mtstate->ps);
/*
* We create two lists - one for WHEN MATCHED actions and one for
* WHEN NOT MATCHED actions - and stick the MergeActionState into
* the appropriate list.
*/
if (action_state->mas_action->matched)
list = &resultRelInfo->ri_matchedMergeAction;
else
list = &resultRelInfo->ri_notMatchedMergeAction;
*list = lappend(*list, action_state);
switch (action->commandType)
{
case CMD_INSERT:
ExecCheckPlanOutput(rootRelInfo->ri_RelationDesc,
action->targetList);
/*
* If the MERGE targets a partitioned table, any INSERT
* actions must be routed through it, not the child
* relations. Initialize the routing struct and the root
* table's "new" tuple slot for that, if not already done.
* The projection we prepare, for all relations, uses the
* root relation descriptor, and targets the plan's root
* slot. (This is consistent with the fact that we
* checked the plan output to match the root relation,
* above.)
*/
if (rootRelInfo->ri_RelationDesc->rd_rel->relkind ==
RELKIND_PARTITIONED_TABLE)
{
if (mtstate->mt_partition_tuple_routing == NULL)
{
/*
* Initialize planstate for routing if not already
* done.
*
* Note that the slot is managed as a standalone
* slot belonging to ModifyTableState, so we pass
* NULL for the 2nd argument.
*/
mtstate->mt_root_tuple_slot =
table_slot_create(rootRelInfo->ri_RelationDesc,
NULL);
mtstate->mt_partition_tuple_routing =
ExecSetupPartitionTupleRouting(estate,
rootRelInfo->ri_RelationDesc);
}
tgtslot = mtstate->mt_root_tuple_slot;
tgtdesc = RelationGetDescr(rootRelInfo->ri_RelationDesc);
}
else
{
/* not partitioned? use the stock relation and slot */
tgtslot = resultRelInfo->ri_newTupleSlot;
tgtdesc = RelationGetDescr(resultRelInfo->ri_RelationDesc);
}
action_state->mas_proj =
ExecBuildProjectionInfo(action->targetList, econtext,
tgtslot,
&mtstate->ps,
tgtdesc);
mtstate->mt_merge_subcommands |= MERGE_INSERT;
break;
case CMD_UPDATE:
action_state->mas_proj =
ExecBuildUpdateProjection(action->targetList,
true,
action->updateColnos,
relationDesc,
econtext,
resultRelInfo->ri_newTupleSlot,
&mtstate->ps);
mtstate->mt_merge_subcommands |= MERGE_UPDATE;
break;
case CMD_DELETE:
mtstate->mt_merge_subcommands |= MERGE_DELETE;
break;
case CMD_NOTHING:
break;
default:
elog(ERROR, "unknown operation");
break;
}
}
}
}
/*
* Initializes the tuple slots in a ResultRelInfo for any MERGE action.
*
* We mark 'projectNewInfoValid' even though the projections themselves
* are not initialized here.
*/
void
ExecInitMergeTupleSlots(ModifyTableState *mtstate,
ResultRelInfo *resultRelInfo)
{
EState *estate = mtstate->ps.state;
Assert(!resultRelInfo->ri_projectNewInfoValid);
resultRelInfo->ri_oldTupleSlot =
table_slot_create(resultRelInfo->ri_RelationDesc,
&estate->es_tupleTable);
resultRelInfo->ri_newTupleSlot =
table_slot_create(resultRelInfo->ri_RelationDesc,
&estate->es_tupleTable);
resultRelInfo->ri_projectNewInfoValid = true;
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* Process BEFORE EACH STATEMENT triggers
*/
static void
fireBSTriggers(ModifyTableState *node)
{
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
ModifyTable *plan = (ModifyTable *) node->ps.plan;
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
ResultRelInfo *resultRelInfo = node->rootResultRelInfo;
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
switch (node->operation)
{
case CMD_INSERT:
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
ExecBSInsertTriggers(node->ps.state, resultRelInfo);
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
if (plan->onConflictAction == ONCONFLICT_UPDATE)
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ExecBSUpdateTriggers(node->ps.state,
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
resultRelInfo);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
case CMD_UPDATE:
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
ExecBSUpdateTriggers(node->ps.state, resultRelInfo);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
case CMD_DELETE:
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
ExecBSDeleteTriggers(node->ps.state, resultRelInfo);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
case CMD_MERGE:
if (node->mt_merge_subcommands & MERGE_INSERT)
ExecBSInsertTriggers(node->ps.state, resultRelInfo);
if (node->mt_merge_subcommands & MERGE_UPDATE)
ExecBSUpdateTriggers(node->ps.state, resultRelInfo);
if (node->mt_merge_subcommands & MERGE_DELETE)
ExecBSDeleteTriggers(node->ps.state, resultRelInfo);
break;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
default:
elog(ERROR, "unknown operation");
break;
}
}
Fix transition tables for partition/inheritance. We disallow row-level triggers with transition tables on child tables. Transition tables for triggers on the parent table contain only those columns present in the parent. (We can't mix tuple formats in a single transition table.) Patch by Thomas Munro Discussion: https://postgr.es/m/CA%2BTgmoZzTBBAsEUh4MazAN7ga%3D8SsMC-Knp-6cetts9yNZUCcg%40mail.gmail.com
2017-06-28 19:55:03 +02:00
/*
* Process AFTER EACH STATEMENT triggers
*/
static void
fireASTriggers(ModifyTableState *node)
{
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
ModifyTable *plan = (ModifyTable *) node->ps.plan;
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
ResultRelInfo *resultRelInfo = node->rootResultRelInfo;
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
switch (node->operation)
{
case CMD_INSERT:
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
if (plan->onConflictAction == ONCONFLICT_UPDATE)
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ExecASUpdateTriggers(node->ps.state,
Fix transition tables for wCTEs. The original coding didn't handle this case properly; each separate DML substatement needs its own set of transitions. Patch by Thomas Munro Discussion: https://postgr.es/m/CAL9smLCDQ%3D2o024rBgtD4WihzX8B3C6u_oSQ2K3%2BR5grJrV0bg%40mail.gmail.com
2017-06-28 19:59:01 +02:00
resultRelInfo,
Fix SQL-spec incompatibilities in new transition table feature. The standard says that all changes of the same kind (insert, update, or delete) caused in one table by a single SQL statement should be reported in a single transition table; and by that, they mean to include foreign key enforcement actions cascading from the statement's direct effects. It's also reasonable to conclude that if the standard had wCTEs, they would say that effects of wCTEs applying to the same table as each other or the outer statement should be merged into one transition table. We weren't doing it like that. Hence, arrange to merge tuples from multiple update actions into a single transition table as much as we can. There is a problem, which is that if the firing of FK enforcement triggers and after-row triggers with transition tables is interspersed, we might need to report more tuples after some triggers have already seen the transition table. It seems like a bad idea for the transition table to be mutable between trigger calls. There's no good way around this without a major redesign of the FK logic, so for now, resolve it by opening a new transition table each time this happens. Also, ensure that AFTER STATEMENT triggers fire just once per statement, or once per transition table when we're forced to make more than one. Previous versions of Postgres have allowed each FK enforcement query to cause an additional firing of the AFTER STATEMENT triggers for the referencing table, but that's certainly not per spec. (We're still doing multiple firings of BEFORE STATEMENT triggers, though; is that something worth changing?) Also, forbid using transition tables with column-specific UPDATE triggers. The spec requires such transition tables to show only the tuples for which the UPDATE trigger would have fired, which means maintaining multiple transition tables or else somehow filtering the contents at readout. Maybe someday we'll bother to support that option, but it looks like a lot of trouble for a marginal feature. The transition tables are now managed by the AfterTriggers data structures, rather than being directly the responsibility of ModifyTable nodes. This removes a subtransaction-lifespan memory leak introduced by my previous band-aid patch 3c4359521. In passing, refactor the AfterTriggers data structures to reduce the management overhead for them, by using arrays of structs rather than several parallel arrays for per-query-level and per-subtransaction state. I failed to resist the temptation to do some copy-editing on the SGML docs about triggers, above and beyond merely documenting the effects of this patch. Back-patch to v10, because we don't want the semantics of transition tables to change post-release. Patch by me, with help and review from Thomas Munro. Discussion: https://postgr.es/m/20170909064853.25630.12825@wrigleys.postgresql.org
2017-09-16 19:20:32 +02:00
node->mt_oc_transition_capture);
Fix transition tables for wCTEs. The original coding didn't handle this case properly; each separate DML substatement needs its own set of transitions. Patch by Thomas Munro Discussion: https://postgr.es/m/CAL9smLCDQ%3D2o024rBgtD4WihzX8B3C6u_oSQ2K3%2BR5grJrV0bg%40mail.gmail.com
2017-06-28 19:59:01 +02:00
ExecASInsertTriggers(node->ps.state, resultRelInfo,
node->mt_transition_capture);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
case CMD_UPDATE:
Fix transition tables for wCTEs. The original coding didn't handle this case properly; each separate DML substatement needs its own set of transitions. Patch by Thomas Munro Discussion: https://postgr.es/m/CAL9smLCDQ%3D2o024rBgtD4WihzX8B3C6u_oSQ2K3%2BR5grJrV0bg%40mail.gmail.com
2017-06-28 19:59:01 +02:00
ExecASUpdateTriggers(node->ps.state, resultRelInfo,
node->mt_transition_capture);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
case CMD_DELETE:
Fix transition tables for wCTEs. The original coding didn't handle this case properly; each separate DML substatement needs its own set of transitions. Patch by Thomas Munro Discussion: https://postgr.es/m/CAL9smLCDQ%3D2o024rBgtD4WihzX8B3C6u_oSQ2K3%2BR5grJrV0bg%40mail.gmail.com
2017-06-28 19:59:01 +02:00
ExecASDeleteTriggers(node->ps.state, resultRelInfo,
node->mt_transition_capture);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
case CMD_MERGE:
if (node->mt_merge_subcommands & MERGE_DELETE)
ExecASDeleteTriggers(node->ps.state, resultRelInfo,
node->mt_transition_capture);
if (node->mt_merge_subcommands & MERGE_UPDATE)
ExecASUpdateTriggers(node->ps.state, resultRelInfo,
node->mt_transition_capture);
if (node->mt_merge_subcommands & MERGE_INSERT)
ExecASInsertTriggers(node->ps.state, resultRelInfo,
node->mt_transition_capture);
break;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
default:
elog(ERROR, "unknown operation");
break;
}
}
Fix transition tables for partition/inheritance. We disallow row-level triggers with transition tables on child tables. Transition tables for triggers on the parent table contain only those columns present in the parent. (We can't mix tuple formats in a single transition table.) Patch by Thomas Munro Discussion: https://postgr.es/m/CA%2BTgmoZzTBBAsEUh4MazAN7ga%3D8SsMC-Knp-6cetts9yNZUCcg%40mail.gmail.com
2017-06-28 19:55:03 +02:00
/*
* Set up the state needed for collecting transition tuples for AFTER
* triggers.
*/
static void
ExecSetupTransitionCaptureState(ModifyTableState *mtstate, EState *estate)
{
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
ModifyTable *plan = (ModifyTable *) mtstate->ps.plan;
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
ResultRelInfo *targetRelInfo = mtstate->rootResultRelInfo;
Fix transition tables for partition/inheritance. We disallow row-level triggers with transition tables on child tables. Transition tables for triggers on the parent table contain only those columns present in the parent. (We can't mix tuple formats in a single transition table.) Patch by Thomas Munro Discussion: https://postgr.es/m/CA%2BTgmoZzTBBAsEUh4MazAN7ga%3D8SsMC-Knp-6cetts9yNZUCcg%40mail.gmail.com
2017-06-28 19:55:03 +02:00
/* Check for transition tables on the directly targeted relation. */
mtstate->mt_transition_capture =
Fix SQL-spec incompatibilities in new transition table feature. The standard says that all changes of the same kind (insert, update, or delete) caused in one table by a single SQL statement should be reported in a single transition table; and by that, they mean to include foreign key enforcement actions cascading from the statement's direct effects. It's also reasonable to conclude that if the standard had wCTEs, they would say that effects of wCTEs applying to the same table as each other or the outer statement should be merged into one transition table. We weren't doing it like that. Hence, arrange to merge tuples from multiple update actions into a single transition table as much as we can. There is a problem, which is that if the firing of FK enforcement triggers and after-row triggers with transition tables is interspersed, we might need to report more tuples after some triggers have already seen the transition table. It seems like a bad idea for the transition table to be mutable between trigger calls. There's no good way around this without a major redesign of the FK logic, so for now, resolve it by opening a new transition table each time this happens. Also, ensure that AFTER STATEMENT triggers fire just once per statement, or once per transition table when we're forced to make more than one. Previous versions of Postgres have allowed each FK enforcement query to cause an additional firing of the AFTER STATEMENT triggers for the referencing table, but that's certainly not per spec. (We're still doing multiple firings of BEFORE STATEMENT triggers, though; is that something worth changing?) Also, forbid using transition tables with column-specific UPDATE triggers. The spec requires such transition tables to show only the tuples for which the UPDATE trigger would have fired, which means maintaining multiple transition tables or else somehow filtering the contents at readout. Maybe someday we'll bother to support that option, but it looks like a lot of trouble for a marginal feature. The transition tables are now managed by the AfterTriggers data structures, rather than being directly the responsibility of ModifyTable nodes. This removes a subtransaction-lifespan memory leak introduced by my previous band-aid patch 3c4359521. In passing, refactor the AfterTriggers data structures to reduce the management overhead for them, by using arrays of structs rather than several parallel arrays for per-query-level and per-subtransaction state. I failed to resist the temptation to do some copy-editing on the SGML docs about triggers, above and beyond merely documenting the effects of this patch. Back-patch to v10, because we don't want the semantics of transition tables to change post-release. Patch by me, with help and review from Thomas Munro. Discussion: https://postgr.es/m/20170909064853.25630.12825@wrigleys.postgresql.org
2017-09-16 19:20:32 +02:00
MakeTransitionCaptureState(targetRelInfo->ri_TrigDesc,
RelationGetRelid(targetRelInfo->ri_RelationDesc),
mtstate->operation);
Remove unnecessary members from ModifyTableState and ExecInsert These values can be obtained from the ModifyTable node which is already a part of both the ModifyTableState and ExecInsert. Author: Álvaro Herrera, Amit Langote Reviewed-by: Peter Geoghegan Discussion: https://postgr.es/m/20180316151303.rml2p5wffn3o6qy6@alvherre.pgsql
2018-03-19 22:09:43 +01:00
if (plan->operation == CMD_INSERT &&
plan->onConflictAction == ONCONFLICT_UPDATE)
Fix SQL-spec incompatibilities in new transition table feature. The standard says that all changes of the same kind (insert, update, or delete) caused in one table by a single SQL statement should be reported in a single transition table; and by that, they mean to include foreign key enforcement actions cascading from the statement's direct effects. It's also reasonable to conclude that if the standard had wCTEs, they would say that effects of wCTEs applying to the same table as each other or the outer statement should be merged into one transition table. We weren't doing it like that. Hence, arrange to merge tuples from multiple update actions into a single transition table as much as we can. There is a problem, which is that if the firing of FK enforcement triggers and after-row triggers with transition tables is interspersed, we might need to report more tuples after some triggers have already seen the transition table. It seems like a bad idea for the transition table to be mutable between trigger calls. There's no good way around this without a major redesign of the FK logic, so for now, resolve it by opening a new transition table each time this happens. Also, ensure that AFTER STATEMENT triggers fire just once per statement, or once per transition table when we're forced to make more than one. Previous versions of Postgres have allowed each FK enforcement query to cause an additional firing of the AFTER STATEMENT triggers for the referencing table, but that's certainly not per spec. (We're still doing multiple firings of BEFORE STATEMENT triggers, though; is that something worth changing?) Also, forbid using transition tables with column-specific UPDATE triggers. The spec requires such transition tables to show only the tuples for which the UPDATE trigger would have fired, which means maintaining multiple transition tables or else somehow filtering the contents at readout. Maybe someday we'll bother to support that option, but it looks like a lot of trouble for a marginal feature. The transition tables are now managed by the AfterTriggers data structures, rather than being directly the responsibility of ModifyTable nodes. This removes a subtransaction-lifespan memory leak introduced by my previous band-aid patch 3c4359521. In passing, refactor the AfterTriggers data structures to reduce the management overhead for them, by using arrays of structs rather than several parallel arrays for per-query-level and per-subtransaction state. I failed to resist the temptation to do some copy-editing on the SGML docs about triggers, above and beyond merely documenting the effects of this patch. Back-patch to v10, because we don't want the semantics of transition tables to change post-release. Patch by me, with help and review from Thomas Munro. Discussion: https://postgr.es/m/20170909064853.25630.12825@wrigleys.postgresql.org
2017-09-16 19:20:32 +02:00
mtstate->mt_oc_transition_capture =
MakeTransitionCaptureState(targetRelInfo->ri_TrigDesc,
RelationGetRelid(targetRelInfo->ri_RelationDesc),
CMD_UPDATE);
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
}
Use ResultRelInfo ** rather than ResultRelInfo * for tuple routing. The previous convention doesn't lend itself to creating ResultRelInfos lazily, as we already do in ExecGetTriggerResultRel. This patch doesn't make anything lazier than before, but the pending patch for UPDATE tuple routing proposes to do so (and there might be other opportunities as well). Amit Khandekar with some adjustments by me. Discussion: http://postgr.es/m/CA+TgmoYPVP9Lyf6vUFA5DwxS4c--x6LOj2y36BsJaYtp62eXPQ@mail.gmail.com
2017-10-12 22:50:53 +02:00
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
/*
* ExecPrepareTupleRouting --- prepare for routing one tuple
*
* Determine the partition in which the tuple in slot is to be inserted,
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
* and return its ResultRelInfo in *partRelInfo. The return value is
* a slot holding the tuple of the partition rowtype.
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
*
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
* This also sets the transition table information in mtstate based on the
* selected partition.
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
*/
Revert MERGE patch This reverts commits d204ef63776b8a00ca220adec23979091564e465, 83454e3c2b28141c0db01c7d2027e01040df5249 and a few more commits thereafter (complete list at the end) related to MERGE feature. While the feature was fully functional, with sufficient test coverage and necessary documentation, it was felt that some parts of the executor and parse-analyzer can use a different design and it wasn't possible to do that in the available time. So it was decided to revert the patch for PG11 and retry again in the future. Thanks again to all reviewers and bug reporters. List of commits reverted, in reverse chronological order: f1464c5380 Improve parse representation for MERGE ddb4158579 MERGE syntax diagram correction 530e69e59b Allow cpluspluscheck to pass by renaming variable 01b88b4df5 MERGE minor errata 3af7b2b0d4 MERGE fix variable warning in non-assert builds a5d86181ec MERGE INSERT allows only one VALUES clause 4b2d44031f MERGE post-commit review 4923550c20 Tab completion for MERGE aa3faa3c7a WITH support in MERGE 83454e3c2b New files for MERGE d204ef6377 MERGE SQL Command following SQL:2016 Author: Pavan Deolasee Reviewed-by: Michael Paquier
2018-04-12 12:22:56 +02:00
static TupleTableSlot *
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
ExecPrepareTupleRouting(ModifyTableState *mtstate,
EState *estate,
PartitionTupleRouting *proute,
ResultRelInfo *targetRelInfo,
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
TupleTableSlot *slot,
ResultRelInfo **partRelInfo)
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
{
ResultRelInfo *partrel;
Use slots more widely in tuple mapping code and make naming more consistent. It's inefficient to use a single slot for mapping between tuple descriptors for multiple tuples, as previously done when using ConvertPartitionTupleSlot(), as that means the slot's tuple descriptors change for every tuple. Previously we also, via ConvertPartitionTupleSlot(), built new tuples after the mapping even in cases where we, immediately afterwards, access individual columns again. Refactor the code so one slot, on demand, is used for each partition. That avoids having to change the descriptor (and allows to use the more efficient "fixed" tuple slots). Then use slot->slot mapping, to avoid unnecessarily forming a tuple. As the naming between the tuple and slot mapping functions wasn't consistent, rename them to execute_attr_map_{tuple,slot}. It's likely that we'll also rename convert_tuples_by_* to denote that these functions "only" build a map, but that's left for later. Author: Amit Khandekar and Amit Langote, editorialized by me Reviewed-By: Amit Langote, Amit Khandekar, Andres Freund Discussion: https://postgr.es/m/CAJ3gD9fR0wRNeAE8VqffNTyONS_UfFPRpqxhnD9Q42vZB+Jvpg@mail.gmail.com https://postgr.es/m/e4f9d743-cd4b-efb0-7574-da21d86a7f36%40lab.ntt.co.jp Backpatch: -
2018-10-02 20:14:26 +02:00
TupleConversionMap *map;
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
/*
Redesign initialization of partition routing structures This speeds up write operations (INSERT, UPDATE, DELETE, COPY, as well as the future MERGE) on partitioned tables. This changes the setup for tuple routing so that it does far less work during the initial setup and pushes more work out to when partitions receive tuples. PartitionDispatchData structs for sub-partitioned tables are only created when a tuple gets routed through it. The possibly large arrays in the PartitionTupleRouting struct have largely been removed. The partitions[] array remains but now never contains any NULL gaps. Previously the NULLs had to be skipped during ExecCleanupTupleRouting(), which could add a large overhead to the cleanup when the number of partitions was large. The partitions[] array is allocated small to start with and only enlarged when we route tuples to enough partitions that it runs out of space. This allows us to keep simple single-row partition INSERTs running quickly. Redesign The arrays in PartitionTupleRouting which stored the tuple translation maps have now been removed. These have been moved out into a PartitionRoutingInfo struct which is an additional field in ResultRelInfo. The find_all_inheritors() call still remains by far the slowest part of ExecSetupPartitionTupleRouting(). This commit just removes the other slow parts. In passing also rename the tuple translation maps from being ParentToChild and ChildToParent to being RootToPartition and PartitionToRoot. The old names mislead you into thinking that a partition of some sub-partitioned table would translate to the rowtype of the sub-partitioned table rather than the root partitioned table. Authors: David Rowley and Amit Langote, heavily revised by Álvaro Herrera Testing help from Jesper Pedersen and Kato Sho. Discussion: https://postgr.es/m/CAKJS1f_1RJyFquuCKRFHTdcXqoPX-PYqAd7nz=GVBwvGh4a6xA@mail.gmail.com
2018-11-16 18:54:15 +01:00
* Lookup the target partition's ResultRelInfo. If ExecFindPartition does
* not find a valid partition for the tuple in 'slot' then an error is
* raised. An error may also be raised if the found partition is not a
* valid target for INSERTs. This is required since a partitioned table
* UPDATE to another partition becomes a DELETE+INSERT.
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
*/
Redesign initialization of partition routing structures This speeds up write operations (INSERT, UPDATE, DELETE, COPY, as well as the future MERGE) on partitioned tables. This changes the setup for tuple routing so that it does far less work during the initial setup and pushes more work out to when partitions receive tuples. PartitionDispatchData structs for sub-partitioned tables are only created when a tuple gets routed through it. The possibly large arrays in the PartitionTupleRouting struct have largely been removed. The partitions[] array remains but now never contains any NULL gaps. Previously the NULLs had to be skipped during ExecCleanupTupleRouting(), which could add a large overhead to the cleanup when the number of partitions was large. The partitions[] array is allocated small to start with and only enlarged when we route tuples to enough partitions that it runs out of space. This allows us to keep simple single-row partition INSERTs running quickly. Redesign The arrays in PartitionTupleRouting which stored the tuple translation maps have now been removed. These have been moved out into a PartitionRoutingInfo struct which is an additional field in ResultRelInfo. The find_all_inheritors() call still remains by far the slowest part of ExecSetupPartitionTupleRouting(). This commit just removes the other slow parts. In passing also rename the tuple translation maps from being ParentToChild and ChildToParent to being RootToPartition and PartitionToRoot. The old names mislead you into thinking that a partition of some sub-partitioned table would translate to the rowtype of the sub-partitioned table rather than the root partitioned table. Authors: David Rowley and Amit Langote, heavily revised by Álvaro Herrera Testing help from Jesper Pedersen and Kato Sho. Discussion: https://postgr.es/m/CAKJS1f_1RJyFquuCKRFHTdcXqoPX-PYqAd7nz=GVBwvGh4a6xA@mail.gmail.com
2018-11-16 18:54:15 +01:00
partrel = ExecFindPartition(mtstate, targetRelInfo, proute, slot, estate);
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
/*
* If we're capturing transition tuples, we might need to convert from the
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
* partition rowtype to root partitioned table's rowtype. But if there
* are no BEFORE triggers on the partition that could change the tuple, we
* can just remember the original unconverted tuple to avoid a needless
* round trip conversion.
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
*/
if (mtstate->mt_transition_capture != NULL)
{
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
bool has_before_insert_row_trig;
has_before_insert_row_trig = (partrel->ri_TrigDesc &&
partrel->ri_TrigDesc->trig_insert_before_row);
mtstate->mt_transition_capture->tcs_original_insert_tuple =
!has_before_insert_row_trig ? slot : NULL;
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
}
/*
* Convert the tuple, if necessary.
*/
Generalize ri_RootToPartitionMap to use for non-partition children ri_RootToPartitionMap is currently only initialized for tuple routing target partitions, though a future commit will need the ability to use it even for the non-partition child tables, so make adjustments to the decouple it from the partitioning code. Also, make it lazily initialized via ExecGetRootToChildMap(), making that function its preferred access path. Existing third-party code accessing it directly should no longer do so; consequently, it's been renamed to ri_RootToChildMap, which also makes it consistent with ri_ChildToRootMap. ExecGetRootToChildMap() houses the logic of setting the map appropriately depending on whether a given child relation is partition or not. To support this, also add a separate entry point for TupleConversionMap creation that receives an AttrMap. No new code here, just split an existing function in two. Author: Amit Langote <amitlangote09@gmail.com> Discussion: https://postgr.es/m/CA+HiwqEYUhDXSK5BTvG_xk=eaAEJCD4GS3C6uH7ybBvv+Z_Tmg@mail.gmail.com
2022-12-02 10:35:55 +01:00
map = ExecGetRootToChildMap(partrel, estate);
Use slots more widely in tuple mapping code and make naming more consistent. It's inefficient to use a single slot for mapping between tuple descriptors for multiple tuples, as previously done when using ConvertPartitionTupleSlot(), as that means the slot's tuple descriptors change for every tuple. Previously we also, via ConvertPartitionTupleSlot(), built new tuples after the mapping even in cases where we, immediately afterwards, access individual columns again. Refactor the code so one slot, on demand, is used for each partition. That avoids having to change the descriptor (and allows to use the more efficient "fixed" tuple slots). Then use slot->slot mapping, to avoid unnecessarily forming a tuple. As the naming between the tuple and slot mapping functions wasn't consistent, rename them to execute_attr_map_{tuple,slot}. It's likely that we'll also rename convert_tuples_by_* to denote that these functions "only" build a map, but that's left for later. Author: Amit Khandekar and Amit Langote, editorialized by me Reviewed-By: Amit Langote, Amit Khandekar, Andres Freund Discussion: https://postgr.es/m/CAJ3gD9fR0wRNeAE8VqffNTyONS_UfFPRpqxhnD9Q42vZB+Jvpg@mail.gmail.com https://postgr.es/m/e4f9d743-cd4b-efb0-7574-da21d86a7f36%40lab.ntt.co.jp Backpatch: -
2018-10-02 20:14:26 +02:00
if (map != NULL)
{
Remove PartitionRoutingInfo struct. The extra indirection neeeded to access its members via its enclosing ResultRelInfo seems pointless. Move all the fields from PartitionRoutingInfo to ResultRelInfo. Author: Amit Langote Reviewed-by: Alvaro Herrera Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:57 +02:00
TupleTableSlot *new_slot = partrel->ri_PartitionTupleSlot;
Use slots more widely in tuple mapping code and make naming more consistent. It's inefficient to use a single slot for mapping between tuple descriptors for multiple tuples, as previously done when using ConvertPartitionTupleSlot(), as that means the slot's tuple descriptors change for every tuple. Previously we also, via ConvertPartitionTupleSlot(), built new tuples after the mapping even in cases where we, immediately afterwards, access individual columns again. Refactor the code so one slot, on demand, is used for each partition. That avoids having to change the descriptor (and allows to use the more efficient "fixed" tuple slots). Then use slot->slot mapping, to avoid unnecessarily forming a tuple. As the naming between the tuple and slot mapping functions wasn't consistent, rename them to execute_attr_map_{tuple,slot}. It's likely that we'll also rename convert_tuples_by_* to denote that these functions "only" build a map, but that's left for later. Author: Amit Khandekar and Amit Langote, editorialized by me Reviewed-By: Amit Langote, Amit Khandekar, Andres Freund Discussion: https://postgr.es/m/CAJ3gD9fR0wRNeAE8VqffNTyONS_UfFPRpqxhnD9Q42vZB+Jvpg@mail.gmail.com https://postgr.es/m/e4f9d743-cd4b-efb0-7574-da21d86a7f36%40lab.ntt.co.jp Backpatch: -
2018-10-02 20:14:26 +02:00
slot = execute_attr_map_slot(map->attrMap, slot, new_slot);
}
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
Remove es_result_relation_info from EState. Maintaining 'es_result_relation_info' correctly at all times has become cumbersome, especially with partitioning where each partition gets its own result relation info. Having to set and reset it across arbitrary operations has caused bugs in the past. This changes all the places that used 'es_result_relation_info', to receive the currently active ResultRelInfo via function parameters instead. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-14 10:41:40 +02:00
*partRelInfo = partrel;
Fix state reversal after partition tuple routing We make some changes to ModifyTableState and the EState it uses whenever we route tuples to partitions; but we weren't restoring properly in all cases, possibly causing crashes when partitions with different tuple descriptors are targeted by tuples inserted in the same command. Refactor some code, creating ExecPrepareTupleRouting, to encapsulate the needed state changing logic, and have it invoked one level above its current place (ie. put it in ExecModifyTable instead of ExecInsert); this makes it all more readable. Add a test case to exercise this. We don't support having views as partitions; and since only views can have INSTEAD OF triggers, there is no point in testing for INSTEAD OF when processing insertions into a partitioned table. Remove code that appears to support this (but which is actually never relevant.) In passing, fix location of some very confusing comments in ModifyTableState. Reported-by: Amit Langote Author: Etsuro Fujita, Amit Langote Discussion: https://postgr/es/m/0473bf5c-57b1-f1f7-3d58-455c2230bc5f@lab.ntt.co.jp
2018-03-19 21:43:57 +01:00
return slot;
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* ----------------------------------------------------------------
* ExecModifyTable
*
* Perform table modifications as required, and return RETURNING results
* if needed.
* ----------------------------------------------------------------
*/
Move ExecProcNode from dispatch to function pointer based model. This allows us to add stack-depth checks the first time an executor node is called, and skip that overhead on following calls. Additionally it yields a nice speedup. While it'd probably have been a good idea to have that check all along, it has become more important after the new expression evaluation framework in b8d7f053c5c2bf2a7e - there's no stack depth check in common paths anymore now. We previously relied on ExecEvalExpr() being executed somewhere. We should move towards that model for further routines, but as this is required for v10, it seems better to only do the necessary (which already is quite large). Author: Andres Freund, Tom Lane Reported-By: Julien Rouhaud Discussion: https://postgr.es/m/22833.1490390175@sss.pgh.pa.us https://postgr.es/m/b0af9eaa-130c-60d0-9e4e-7a135b1e0c76@dalibo.com
2017-07-17 09:33:49 +02:00
static TupleTableSlot *
ExecModifyTable(PlanState *pstate)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Move ExecProcNode from dispatch to function pointer based model. This allows us to add stack-depth checks the first time an executor node is called, and skip that overhead on following calls. Additionally it yields a nice speedup. While it'd probably have been a good idea to have that check all along, it has become more important after the new expression evaluation framework in b8d7f053c5c2bf2a7e - there's no stack depth check in common paths anymore now. We previously relied on ExecEvalExpr() being executed somewhere. We should move towards that model for further routines, but as this is required for v10, it seems better to only do the necessary (which already is quite large). Author: Andres Freund, Tom Lane Reported-By: Julien Rouhaud Discussion: https://postgr.es/m/22833.1490390175@sss.pgh.pa.us https://postgr.es/m/b0af9eaa-130c-60d0-9e4e-7a135b1e0c76@dalibo.com
2017-07-17 09:33:49 +02:00
ModifyTableState *node = castNode(ModifyTableState, pstate);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ModifyTableContext context;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
EState *estate = node->ps.state;
CmdType operation = node->operation;
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
ResultRelInfo *resultRelInfo;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
PlanState *subplanstate;
TupleTableSlot *slot;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
TupleTableSlot *oldSlot;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
ItemPointerData tuple_ctid;
Offer triggers on foreign tables. This covers all the SQL-standard trigger types supported for regular tables; it does not cover constraint triggers. The approach for acquiring the old row mirrors that for view INSTEAD OF triggers. For AFTER ROW triggers, we spool the foreign tuples to a tuplestore. This changes the FDW API contract; when deciding which columns to populate in the slot returned from data modification callbacks, writable FDWs will need to check for AFTER ROW triggers in addition to checking for a RETURNING clause. In support of the feature addition, refactor the TriggerFlags bits and the assembly of old tuples in ModifyTable. Ronan Dunklau, reviewed by KaiGai Kohei; some additional hacking by me.
2014-03-23 07:16:34 +01:00
HeapTupleData oldtupdata;
HeapTuple oldtuple;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
ItemPointer tupleid;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Move interrupt checking from ExecProcNode() to executor nodes. In a followup commit ExecProcNode(), and especially the large switch it contains, will largely be replaced by a function pointer directly to the correct node. The node functions will then get invoked by a thin inline function wrapper. To avoid having to include miscadmin.h in headers - CHECK_FOR_INTERRUPTS() - move the interrupt checks into the individual executor routines. While looking through all executor nodes, I noticed a number of arguably missing interrupt checks, add these too. Author: Andres Freund, Tom Lane Reviewed-By: Tom Lane Discussion: https://postgr.es/m/22833.1490390175@sss.pgh.pa.us
2017-07-26 02:37:17 +02:00
CHECK_FOR_INTERRUPTS();
Fix handling of data-modifying CTE subplans in EvalPlanQual. We can't just skip initializing such subplans, because the referencing CTE node will expect to find the subplan available when it initializes. That in turn means that ExecInitModifyTable must allow the case (which actually it needed to do anyway, since there's no guarantee that ModifyTable is exactly at the top of the CTE plan tree). So move the complaint about not being allowed in EvalPlanQual mode to execution instead of initialization. Testing turned up yet another problem, which is that we'd try to re-initialize the result relation's index list, leading to leaks and dangling pointers. Per report from Phil Sorber. Back-patch to 9.1 where data-modifying CTEs were introduced.
2012-01-28 23:43:57 +01:00
/*
* This should NOT get called during EvalPlanQual; we should have passed a
* subplan tree to EvalPlanQual, instead. Use a runtime test not just
* Assert because this condition is easy to miss in testing. (Note:
* although ModifyTable should not get executed within an EvalPlanQual
* operation, we do have to allow it to be initialized and shut down in
* case it is within a CTE subplan. Hence this test must be here, not in
* ExecInitModifyTable.)
*/
Reorder EPQ work, to fix rowmark related bugs and improve efficiency. In ad0bda5d24ea I changed the EvalPlanQual machinery to store substitution tuples in slot, instead of using plain HeapTuples. The main motivation for that was that using HeapTuples will be inefficient for future tableams. But it turns out that that conversion was buggy for non-locking rowmarks - the wrong tuple descriptor was used to create the slot. As a secondary issue 5db6df0c0 changed ExecLockRows() to begin EPQ earlier, to allow to fetch the locked rows directly into the EPQ slots, instead of having to copy tuples around. Unfortunately, as Tom complained, that forces some expensive initialization to happen earlier. As a third issue, the test coverage for EPQ was clearly insufficient. Fixing the first issue is unfortunately not trivial: Non-locked row marks were fetched at the start of EPQ, and we don't have the type information for the rowmarks available at that point. While we could change that, it's not easy. It might be worthwhile to change that at some point, but to fix this bug, it seems better to delay fetching non-locking rowmarks when they're actually needed, rather than eagerly. They're referenced at most once, and in cases where EPQ fails, might never be referenced. Fetching them when needed also increases locality a bit. To be able to fetch rowmarks during execution, rather than initialization, we need to be able to access the active EPQState, as that contains necessary data. To do so move EPQ related data from EState to EPQState, and, only for EStates creates as part of EPQ, reference the associated EPQState from EState. To fix the second issue, change EPQ initialization to allow use of EvalPlanQualSlot() to be used before EvalPlanQualBegin() (but obviously still requiring EvalPlanQualInit() to have been done). As these changes made struct EState harder to understand, e.g. by adding multiple EStates, significantly reorder the members, and add a lot more comments. Also add a few more EPQ tests, including one that fails for the first issue above. More is needed. Reported-By: yi huang Author: Andres Freund Reviewed-By: Tom Lane Discussion: https://postgr.es/m/CAHU7rYZo_C4ULsAx_LAj8az9zqgrD8WDd4hTegDTMM1LMqrBsg@mail.gmail.com https://postgr.es/m/24530.1562686693@sss.pgh.pa.us Backpatch: 12-, where the EPQ changes were introduced
2019-09-05 22:00:20 +02:00
if (estate->es_epq_active != NULL)
Fix handling of data-modifying CTE subplans in EvalPlanQual. We can't just skip initializing such subplans, because the referencing CTE node will expect to find the subplan available when it initializes. That in turn means that ExecInitModifyTable must allow the case (which actually it needed to do anyway, since there's no guarantee that ModifyTable is exactly at the top of the CTE plan tree). So move the complaint about not being allowed in EvalPlanQual mode to execution instead of initialization. Testing turned up yet another problem, which is that we'd try to re-initialize the result relation's index list, leading to leaks and dangling pointers. Per report from Phil Sorber. Back-patch to 9.1 where data-modifying CTEs were introduced.
2012-01-28 23:43:57 +01:00
elog(ERROR, "ModifyTable should not be called during EvalPlanQual");
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
/*
* If we've already completed processing, don't try to do more. We need
* this test because ExecPostprocessPlan might call us an extra time, and
* our subplan's nodes aren't necessarily robust against being called
* extra times.
*/
if (node->mt_done)
return NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* On first call, fire BEFORE STATEMENT triggers before proceeding.
*/
if (node->fireBSTriggers)
{
fireBSTriggers(node);
node->fireBSTriggers = false;
}
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
/* Preload local variables */
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
resultRelInfo = node->resultRelInfo + node->mt_lastResultIndex;
subplanstate = outerPlanState(node);
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
/* Set global context */
context.mtstate = node;
context.epqstate = &node->mt_epqstate;
context.estate = estate;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* Fetch rows from subplan, and execute the required table modification
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
* for each row.
*/
for (;;)
{
Reset the per-output-tuple exprcontext each time through the main loop in ExecModifyTable(). This avoids memory leakage when trigger functions leave junk behind in that context (as they more or less must). Problem and solution identified by Dean Rasheed. I'm a bit concerned about the longevity of this solution --- once a plan can have multiple ModifyTable nodes, we are very possibly going to have to do something different. But it should hold up for 9.0.
2010-08-18 23:52:24 +02:00
/*
* Reset the per-output-tuple exprcontext. This is needed because
* triggers expect to use that context as workspace. It's a bit ugly
* to do this below the top level of the plan, however. We might need
* to rethink this later.
*/
ResetPerTupleExprContext(estate);
Rationalize expression context reset in ExecModifyTable(). The current pattern of reseting expressions both in ExecProcessReturning() and ExecOnConflictUpdate() makes it harder than necessary to reason about memory lifetimes. It also requires materializing slots unnecessarily, although this patch doesn't take advantage of the fact that that's not necessary anymore. Instead reset the expression context once for each input tuple. Author: Ashutosh Bapat Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-13 21:18:25 +01:00
/*
* Reset per-tuple memory context used for processing on conflict and
* returning clauses, to free any expression evaluation storage
* allocated in the previous cycle.
*/
if (pstate->ps_ExprContext)
ResetExprContext(pstate->ps_ExprContext);
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
context.planSlot = ExecProcNode(subplanstate);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/* No more tuples to process? */
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
if (TupIsNull(context.planSlot))
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
break;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Introduce notion of different types of slots (without implementing them). Upcoming work intends to allow pluggable ways to introduce new ways of storing table data. Accessing those table access methods from the executor requires TupleTableSlots to be carry tuples in the native format of such storage methods; otherwise there'll be a significant conversion overhead. Different access methods will require different data to store tuples efficiently (just like virtual, minimal, heap already require fields in TupleTableSlot). To allow that without requiring additional pointer indirections, we want to have different structs (embedding TupleTableSlot) for different types of slots. Thus different types of slots are needed, which requires adapting creators of slots. The slot that most efficiently can represent a type of tuple in an executor node will often depend on the type of slot a child node uses. Therefore we need to track the type of slot is returned by nodes, so parent slots can create slots based on that. Relatedly, JIT compilation of tuple deforming needs to know which type of slot a certain expression refers to, so it can create an appropriate deforming function for the type of tuple in the slot. But not all nodes will only return one type of slot, e.g. an append node will potentially return different types of slots for each of its subplans. Therefore add function that allows to query the type of a node's result slot, and whether it'll always be the same type (whether it's fixed). This can be queried using ExecGetResultSlotOps(). The scan, result, inner, outer type of slots are automatically inferred from ExecInitScanTupleSlot(), ExecInitResultSlot(), left/right subtrees respectively. If that's not correct for a node, that can be overwritten using new fields in PlanState. This commit does not introduce the actually abstracted implementation of different kind of TupleTableSlots, that will be left for a followup commit. The different types of slots introduced will, for now, still use the same backing implementation. While this already partially invalidates the big comment in tuptable.h, it seems to make more sense to update it later, when the different TupleTableSlot implementations actually exist. Author: Ashutosh Bapat and Andres Freund, with changes by Amit Khandekar Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-16 07:00:30 +01:00
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* When there are multiple result relations, each tuple contains a
* junk column that gives the OID of the rel from which it came.
* Extract it and select the correct result relation.
Introduce notion of different types of slots (without implementing them). Upcoming work intends to allow pluggable ways to introduce new ways of storing table data. Accessing those table access methods from the executor requires TupleTableSlots to be carry tuples in the native format of such storage methods; otherwise there'll be a significant conversion overhead. Different access methods will require different data to store tuples efficiently (just like virtual, minimal, heap already require fields in TupleTableSlot). To allow that without requiring additional pointer indirections, we want to have different structs (embedding TupleTableSlot) for different types of slots. Thus different types of slots are needed, which requires adapting creators of slots. The slot that most efficiently can represent a type of tuple in an executor node will often depend on the type of slot a child node uses. Therefore we need to track the type of slot is returned by nodes, so parent slots can create slots based on that. Relatedly, JIT compilation of tuple deforming needs to know which type of slot a certain expression refers to, so it can create an appropriate deforming function for the type of tuple in the slot. But not all nodes will only return one type of slot, e.g. an append node will potentially return different types of slots for each of its subplans. Therefore add function that allows to query the type of a node's result slot, and whether it'll always be the same type (whether it's fixed). This can be queried using ExecGetResultSlotOps(). The scan, result, inner, outer type of slots are automatically inferred from ExecInitScanTupleSlot(), ExecInitResultSlot(), left/right subtrees respectively. If that's not correct for a node, that can be overwritten using new fields in PlanState. This commit does not introduce the actually abstracted implementation of different kind of TupleTableSlots, that will be left for a followup commit. The different types of slots introduced will, for now, still use the same backing implementation. While this already partially invalidates the big comment in tuptable.h, it seems to make more sense to update it later, when the different TupleTableSlot implementations actually exist. Author: Ashutosh Bapat and Andres Freund, with changes by Amit Khandekar Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-16 07:00:30 +01:00
*/
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
if (AttributeNumberIsValid(node->mt_resultOidAttno))
Introduce notion of different types of slots (without implementing them). Upcoming work intends to allow pluggable ways to introduce new ways of storing table data. Accessing those table access methods from the executor requires TupleTableSlots to be carry tuples in the native format of such storage methods; otherwise there'll be a significant conversion overhead. Different access methods will require different data to store tuples efficiently (just like virtual, minimal, heap already require fields in TupleTableSlot). To allow that without requiring additional pointer indirections, we want to have different structs (embedding TupleTableSlot) for different types of slots. Thus different types of slots are needed, which requires adapting creators of slots. The slot that most efficiently can represent a type of tuple in an executor node will often depend on the type of slot a child node uses. Therefore we need to track the type of slot is returned by nodes, so parent slots can create slots based on that. Relatedly, JIT compilation of tuple deforming needs to know which type of slot a certain expression refers to, so it can create an appropriate deforming function for the type of tuple in the slot. But not all nodes will only return one type of slot, e.g. an append node will potentially return different types of slots for each of its subplans. Therefore add function that allows to query the type of a node's result slot, and whether it'll always be the same type (whether it's fixed). This can be queried using ExecGetResultSlotOps(). The scan, result, inner, outer type of slots are automatically inferred from ExecInitScanTupleSlot(), ExecInitResultSlot(), left/right subtrees respectively. If that's not correct for a node, that can be overwritten using new fields in PlanState. This commit does not introduce the actually abstracted implementation of different kind of TupleTableSlots, that will be left for a followup commit. The different types of slots introduced will, for now, still use the same backing implementation. While this already partially invalidates the big comment in tuptable.h, it seems to make more sense to update it later, when the different TupleTableSlot implementations actually exist. Author: Ashutosh Bapat and Andres Freund, with changes by Amit Khandekar Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-16 07:00:30 +01:00
{
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
Datum datum;
bool isNull;
Oid resultoid;
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
datum = ExecGetJunkAttribute(context.planSlot, node->mt_resultOidAttno,
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
&isNull);
if (isNull)
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
{
/*
* For commands other than MERGE, any tuples having InvalidOid
* for tableoid are errors. For MERGE, we may need to handle
* them as WHEN NOT MATCHED clauses if any, so do that.
*
* Note that we use the node's toplevel resultRelInfo, not any
* specific partition's.
*/
if (operation == CMD_MERGE)
{
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
EvalPlanQualSetSlot(&node->mt_epqstate, context.planSlot);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
ExecMerge(&context, node->resultRelInfo, NULL, node->canSetTag);
continue; /* no RETURNING support yet */
}
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
elog(ERROR, "tableoid is NULL");
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
}
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
resultoid = DatumGetObjectId(datum);
/* If it's not the same as last time, we need to locate the rel */
if (resultoid != node->mt_lastResultOid)
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
resultRelInfo = ExecLookupResultRelByOid(node, resultoid,
false, true);
Introduce notion of different types of slots (without implementing them). Upcoming work intends to allow pluggable ways to introduce new ways of storing table data. Accessing those table access methods from the executor requires TupleTableSlots to be carry tuples in the native format of such storage methods; otherwise there'll be a significant conversion overhead. Different access methods will require different data to store tuples efficiently (just like virtual, minimal, heap already require fields in TupleTableSlot). To allow that without requiring additional pointer indirections, we want to have different structs (embedding TupleTableSlot) for different types of slots. Thus different types of slots are needed, which requires adapting creators of slots. The slot that most efficiently can represent a type of tuple in an executor node will often depend on the type of slot a child node uses. Therefore we need to track the type of slot is returned by nodes, so parent slots can create slots based on that. Relatedly, JIT compilation of tuple deforming needs to know which type of slot a certain expression refers to, so it can create an appropriate deforming function for the type of tuple in the slot. But not all nodes will only return one type of slot, e.g. an append node will potentially return different types of slots for each of its subplans. Therefore add function that allows to query the type of a node's result slot, and whether it'll always be the same type (whether it's fixed). This can be queried using ExecGetResultSlotOps(). The scan, result, inner, outer type of slots are automatically inferred from ExecInitScanTupleSlot(), ExecInitResultSlot(), left/right subtrees respectively. If that's not correct for a node, that can be overwritten using new fields in PlanState. This commit does not introduce the actually abstracted implementation of different kind of TupleTableSlots, that will be left for a followup commit. The different types of slots introduced will, for now, still use the same backing implementation. While this already partially invalidates the big comment in tuptable.h, it seems to make more sense to update it later, when the different TupleTableSlot implementations actually exist. Author: Ashutosh Bapat and Andres Freund, with changes by Amit Khandekar Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-16 07:00:30 +01:00
}
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
/*
* If resultRelInfo->ri_usesFdwDirectModify is true, all we need to do
* here is compute the RETURNING expressions.
*/
if (resultRelInfo->ri_usesFdwDirectModify)
{
Assert(resultRelInfo->ri_projectReturning);
/*
* A scan slot containing the data that was actually inserted,
* updated or deleted has already been made available to
* ExecProcessReturning by IterateDirectModify, so no need to
* provide it here.
*/
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
slot = ExecProcessReturning(resultRelInfo, NULL, context.planSlot);
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
return slot;
}
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
EvalPlanQualSetSlot(&node->mt_epqstate, context.planSlot);
slot = context.planSlot;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Fix creation of resjunk tlist entries for inherited mixed UPDATE/DELETE. rewriteTargetListUD's processing is dependent on the relkind of the query's target table. That was fine at the time it was made to act that way, even for queries on inheritance trees, because all tables in an inheritance tree would necessarily be plain tables. However, the 9.5 feature addition allowing some members of an inheritance tree to be foreign tables broke the assumption that rewriteTargetListUD's output tlist could be applied to all child tables with nothing more than column-number mapping. This led to visible failures if foreign child tables had row-level triggers, and would also break in cases where child tables belonged to FDWs that used methods other than CTID for row identification. To fix, delay running rewriteTargetListUD until after the planner has expanded inheritance, so that it is applied separately to the (already mapped) tlist for each child table. We can conveniently call it from preprocess_targetlist. Refactor associated code slightly to avoid the need to heap_open the target relation multiple times during preprocess_targetlist. (The APIs remain a bit ugly, particularly around the point of which steps scribble on parse->targetList and which don't. But avoiding such scribbling would require a change in FDW callback APIs, which is more pain than it's worth.) Also fix ExecModifyTable to ensure that "tupleid" is reset to NULL when we transition from rows providing a CTID to rows that don't. (That's really an independent bug, but it manifests in much the same cases.) Add a regression test checking one manifestation of this problem, which was that row-level triggers on a foreign child table did not work right. Back-patch to 9.5 where the problem was introduced. Etsuro Fujita, reviewed by Ildus Kurbangaliev and Ashutosh Bapat Discussion: https://postgr.es/m/20170514150525.0346ba72@postgrespro.ru
2017-11-27 23:53:56 +01:00
tupleid = NULL;
Offer triggers on foreign tables. This covers all the SQL-standard trigger types supported for regular tables; it does not cover constraint triggers. The approach for acquiring the old row mirrors that for view INSTEAD OF triggers. For AFTER ROW triggers, we spool the foreign tuples to a tuplestore. This changes the FDW API contract; when deciding which columns to populate in the slot returned from data modification callbacks, writable FDWs will need to check for AFTER ROW triggers in addition to checking for a RETURNING clause. In support of the feature addition, refactor the TriggerFlags bits and the assembly of old tuples in ModifyTable. Ronan Dunklau, reviewed by KaiGai Kohei; some additional hacking by me.
2014-03-23 07:16:34 +01:00
oldtuple = NULL;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/*
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* For UPDATE/DELETE/MERGE, fetch the row identity info for the tuple
* to be updated/deleted/merged. For a heap relation, that's a TID;
* otherwise we may have a wholerow junk attr that carries the old
* tuple in toto. Keep this in step with the part of
* ExecInitModifyTable that sets up ri_RowIdAttNo.
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
*/
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
if (operation == CMD_UPDATE || operation == CMD_DELETE ||
operation == CMD_MERGE)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
char relkind;
Datum datum;
bool isNull;
relkind = resultRelInfo->ri_RelationDesc->rd_rel->relkind;
if (relkind == RELKIND_RELATION ||
relkind == RELKIND_MATVIEW ||
relkind == RELKIND_PARTITIONED_TABLE)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/* ri_RowIdAttNo refers to a ctid attribute */
Assert(AttributeNumberIsValid(resultRelInfo->ri_RowIdAttNo));
datum = ExecGetJunkAttribute(slot,
resultRelInfo->ri_RowIdAttNo,
&isNull);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/*
* For commands other than MERGE, any tuples having a null row
* identifier are errors. For MERGE, we may need to handle
* them as WHEN NOT MATCHED clauses if any, so do that.
*
* Note that we use the node's toplevel resultRelInfo, not any
* specific partition's.
*/
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
if (isNull)
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
{
if (operation == CMD_MERGE)
{
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
EvalPlanQualSetSlot(&node->mt_epqstate, context.planSlot);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
ExecMerge(&context, node->resultRelInfo, NULL, node->canSetTag);
continue; /* no RETURNING support yet */
}
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
elog(ERROR, "ctid is NULL");
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
}
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
tupleid = (ItemPointer) DatumGetPointer(datum);
tuple_ctid = *tupleid; /* be sure we don't free ctid!! */
tupleid = &tuple_ctid;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* Use the wholerow attribute, when available, to reconstruct the
* old relation tuple. The old tuple serves one or both of two
* purposes: 1) it serves as the OLD tuple for row triggers, 2) it
* provides values for any unchanged columns for the NEW tuple of
* an UPDATE, because the subplan does not produce all the columns
* of the target table.
*
* Note that the wholerow attribute does not carry system columns,
* so foreign table triggers miss seeing those, except that we
* know enough here to set t_tableOid. Quite separately from
* this, the FDW may fetch its own junk attrs to identify the row.
*
* Other relevant relkinds, currently limited to views, always
* have a wholerow attribute.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*/
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
else if (AttributeNumberIsValid(resultRelInfo->ri_RowIdAttNo))
{
datum = ExecGetJunkAttribute(slot,
resultRelInfo->ri_RowIdAttNo,
&isNull);
/* shouldn't ever get a null result... */
if (isNull)
elog(ERROR, "wholerow is NULL");
oldtupdata.t_data = DatumGetHeapTupleHeader(datum);
oldtupdata.t_len =
HeapTupleHeaderGetDatumLength(oldtupdata.t_data);
ItemPointerSetInvalid(&(oldtupdata.t_self));
/* Historically, view triggers see invalid t_tableOid. */
oldtupdata.t_tableOid =
(relkind == RELKIND_VIEW) ? InvalidOid :
RelationGetRelid(resultRelInfo->ri_RelationDesc);
oldtuple = &oldtupdata;
}
else
{
/* Only foreign tables are allowed to omit a row-ID attr */
Assert(relkind == RELKIND_FOREIGN_TABLE);
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
switch (operation)
{
case CMD_INSERT:
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/* Initialize projection info if first time for this table */
if (unlikely(!resultRelInfo->ri_projectNewInfoValid))
ExecInitInsertProjection(node, resultRelInfo);
ExecModifyTable: use context.planSlot instead of planSlot There's no reason to keep a separate local variable when we have a place for it elsewhere. This allows to simplify some code. Reviewed-by: Michaël Paquier <michael@paquier.xyz> Discussion: https://postgr.es/m/202204191345.qerjy3kxi3eb@alvherre.pgsql
2022-04-20 10:34:58 +02:00
slot = ExecGetInsertNewTuple(resultRelInfo, context.planSlot);
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
slot = ExecInsert(&context, resultRelInfo, slot,
Enforce foreign key correctly during cross-partition updates When an update on a partitioned table referenced in foreign key constraints causes a row to move from one partition to another, the fact that the move is implemented as a delete followed by an insert on the target partition causes the foreign key triggers to have surprising behavior. For example, a given foreign key's delete trigger which implements the ON DELETE CASCADE clause of that key will delete any referencing rows when triggered for that internal DELETE, although it should not, because the referenced row is simply being moved from one partition of the referenced root partitioned table into another, not being deleted from it. This commit teaches trigger.c to skip queuing such delete trigger events on the leaf partitions in favor of an UPDATE event fired on the root target relation. Doing so is sensible because both the old and the new tuple "logically" belong to the root relation. The after trigger event queuing interface now allows passing the source and the target partitions of a particular cross-partition update when registering the update event for the root partitioned table. Along with the two ctids of the old and the new tuple, the after trigger event now also stores the OIDs of those partitions. The tuples fetched from the source and the target partitions are converted into the root table format, if necessary, before they are passed to the trigger function. The implementation currently has a limitation that only the foreign keys pointing into the query's target relation are considered, not those of its sub-partitioned partitions. That seems like a reasonable limitation, because it sounds rare to have distinct foreign keys pointing to sub-partitioned partitions instead of to the root table. This misbehavior stems from commit f56f8f8da6af (which added support for foreign keys to reference partitioned tables) not paying sufficient attention to commit 2f178441044b (which had introduced cross-partition updates a year earlier). Even though the former commit goes back to Postgres 12, we're not backpatching this fix at this time for fear of destabilizing things too much, and because there are a few ABI breaks in it that we'd have to work around in older branches. It also depends on commit f4566345cf40, which had its own share of backpatchability issues as well. Author: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Masahiko Sawada <sawada.mshk@gmail.com> Reviewed-by: Álvaro Herrera <alvherre@alvh.no-ip.org> Reported-by: Eduard Català <eduard.catala@gmail.com> Discussion: https://postgr.es/m/CA+HiwqFvkBCmfwkQX_yBqv2Wz8ugUGiBDxum8=WvVbfU1TXaNg@mail.gmail.com Discussion: https://postgr.es/m/CAL54xNZsLwEM1XCk5yW9EqaRzsZYHuWsHQkA2L5MOSKXAwviCQ@mail.gmail.com
2022-03-20 18:43:40 +01:00
node->canSetTag, NULL, NULL);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
case CMD_UPDATE:
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/* Initialize projection info if first time for this table */
if (unlikely(!resultRelInfo->ri_projectNewInfoValid))
ExecInitUpdateProjection(node, resultRelInfo);
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/*
* Make the new tuple by combining plan's output tuple with
* the old tuple being updated.
*/
oldSlot = resultRelInfo->ri_oldTupleSlot;
if (oldtuple != NULL)
{
/* Use the wholerow junk attr as the old tuple. */
ExecForceStoreHeapTuple(oldtuple, oldSlot, false);
}
else
{
/* Fetch the most recent version of old tuple. */
Relation relation = resultRelInfo->ri_RelationDesc;
if (!table_tuple_fetch_row_version(relation, tupleid,
SnapshotAny,
oldSlot))
elog(ERROR, "failed to fetch tuple being updated");
}
Fix concurrent update issues with MERGE. If MERGE attempts an UPDATE or DELETE on a table with BEFORE ROW triggers, or a cross-partition UPDATE (with or without triggers), and a concurrent UPDATE or DELETE happens, the merge code would fail. In some cases this would lead to a crash, while in others it would cause the wrong merge action to be executed, or no action at all. The immediate cause of the crash was the trigger code calling ExecGetUpdateNewTuple() as part of the EPQ mechanism, which fails because during a merge ri_projectNew is NULL, since merge has its own per-action projection information, which ExecGetUpdateNewTuple() knows nothing about. Fix by arranging for the trigger code to exit early, returning the TM_Result and TM_FailureData information, if a concurrent modification is detected, allowing the merge code to do the necessary EPQ handling in its own way. Similarly, prevent the cross-partition update code from doing any EPQ processing for a merge, allowing the merge code to work out what it needs to do. This leads to a number of simplifications in nodeModifyTable.c. Most notably, the ModifyTableContext->GetUpdateNewTuple() callback is no longer needed, and mergeGetUpdateNewTuple() can be deleted, since there is no longer any requirement for get-update-new-tuple during a merge. Similarly, ModifyTableContext->cpUpdateRetrySlot is no longer needed. Thus ExecGetUpdateNewTuple() and the retry_slot handling of ExecCrossPartitionUpdate() can be restored to how they were in v14, before the merge code was added, and ExecMergeMatched() no longer needs any special-case handling for cross-partition updates. While at it, tidy up ExecUpdateEpilogue() a bit, making it handle recheckIndexes locally, rather than passing it in as a parameter, ensuring that it is freed properly. This dates back to when it was split off from ExecUpdate() to support merge. Per bug #17809 from Alexander Lakhin, and follow-up investigation of bug #17792, also from Alexander Lakhin. Back-patch to v15, where MERGE was introduced, taking care to preserve backwards-compatibility of the trigger API in v15 for any extensions that might use it. Discussion: https://postgr.es/m/17809-9e6650bef133f0fe%40postgresql.org https://postgr.es/m/17792-0f89452029662c36%40postgresql.org
2023-03-13 11:22:22 +01:00
slot = ExecGetUpdateNewTuple(resultRelInfo, context.planSlot,
oldSlot);
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
context.relaction = NULL;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/* Now apply the update. */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
slot = ExecUpdate(&context, resultRelInfo, tupleid, oldtuple,
Revert 11470f544e Discussion: https://postgr.es/m/20230323003003.plgaxjqahjgkuxrk%40awork3.anarazel.de
2023-04-03 15:54:31 +02:00
slot, node->canSetTag);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
case CMD_DELETE:
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
slot = ExecDelete(&context, resultRelInfo, tupleid, oldtuple,
true, false, node->canSetTag, NULL, NULL);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
break;
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
case CMD_MERGE:
slot = ExecMerge(&context, resultRelInfo, tupleid, node->canSetTag);
break;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
default:
elog(ERROR, "unknown operation");
break;
}
/*
* If we got a RETURNING result, return it to caller. We'll continue
* the work on next call.
*/
if (slot)
return slot;
}
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
/*
* Insert remaining tuples for batch insert.
*/
Fix handling of pending inserts in nodeModifyTable.c. Commit b663a4136, which allowed FDWs to INSERT rows in bulk, added to nodeModifyTable.c code to flush pending inserts to the foreign-table result relation(s) before completing processing of the ModifyTable node, but the code failed to take into account the case where the INSERT query has modifying CTEs, leading to incorrect results. Also, that commit failed to flush pending inserts before firing BEFORE ROW triggers so that rows are visible to such triggers. In that commit we scanned through EState's es_tuple_routing_result_relations or es_opened_result_relations list to find the foreign-table result relations to which pending inserts are flushed, but that would be inefficient in some cases. So to fix, 1) add a List member to EState to record the insert-pending result relations, and 2) modify nodeModifyTable.c so that it adds the foreign-table result relation to the list in ExecInsert() if appropriate, and flushes pending inserts properly using the list where needed. While here, fix a copy-and-pasteo in a comment in ExecBatchInsert(), which was added by that commit. Back-patch to v14 where that commit appeared. Discussion: https://postgr.es/m/CAPmGK16qutyCmyJJzgQOhfBq%3DNoGDqTB6O0QBZTihrbqre%2BoxA%40mail.gmail.com
2022-11-25 09:45:00 +01:00
if (estate->es_insert_pending_result_relations != NIL)
ExecPendingInserts(estate);
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* We're done, but fire AFTER STATEMENT triggers before exiting.
*/
fireASTriggers(node);
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
node->mt_done = true;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
return NULL;
}
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/*
* ExecLookupResultRelByOid
* If the table with given OID is among the result relations to be
* updated by the given ModifyTable node, return its ResultRelInfo.
*
* If not found, return NULL if missing_ok, else raise error.
*
* If update_cache is true, then upon successful lookup, update the node's
* one-element cache. ONLY ExecModifyTable may pass true for this.
*/
ResultRelInfo *
ExecLookupResultRelByOid(ModifyTableState *node, Oid resultoid,
bool missing_ok, bool update_cache)
{
if (node->mt_resultOidHash)
{
/* Use the pre-built hash table to locate the rel */
MTTargetRelLookup *mtlookup;
mtlookup = (MTTargetRelLookup *)
hash_search(node->mt_resultOidHash, &resultoid, HASH_FIND, NULL);
if (mtlookup)
{
if (update_cache)
{
node->mt_lastResultOid = resultoid;
node->mt_lastResultIndex = mtlookup->relationIndex;
}
return node->resultRelInfo + mtlookup->relationIndex;
}
}
else
{
/* With few target rels, just search the ResultRelInfo array */
for (int ndx = 0; ndx < node->mt_nrels; ndx++)
{
ResultRelInfo *rInfo = node->resultRelInfo + ndx;
if (RelationGetRelid(rInfo->ri_RelationDesc) == resultoid)
{
if (update_cache)
{
node->mt_lastResultOid = resultoid;
node->mt_lastResultIndex = ndx;
}
return rInfo;
}
}
}
if (!missing_ok)
elog(ERROR, "incorrect result relation OID %u", resultoid);
return NULL;
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* ----------------------------------------------------------------
* ExecInitModifyTable
* ----------------------------------------------------------------
*/
ModifyTableState *
ExecInitModifyTable(ModifyTable *node, EState *estate, int eflags)
{
ModifyTableState *mtstate;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
Plan *subplan = outerPlan(node);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
CmdType operation = node->operation;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
int nrels = list_length(node->resultRelations);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
ResultRelInfo *resultRelInfo;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
List *arowmarks;
ListCell *l;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
int i;
Implement table partitioning. Table partitioning is like table inheritance and reuses much of the existing infrastructure, but there are some important differences. The parent is called a partitioned table and is always empty; it may not have indexes or non-inherited constraints, since those make no sense for a relation with no data of its own. The children are called partitions and contain all of the actual data. Each partition has an implicit partitioning constraint. Multiple inheritance is not allowed, and partitioning and inheritance can't be mixed. Partitions can't have extra columns and may not allow nulls unless the parent does. Tuples inserted into the parent are automatically routed to the correct partition, so tuple-routing ON INSERT triggers are not needed. Tuple routing isn't yet supported for partitions which are foreign tables, and it doesn't handle updates that cross partition boundaries. Currently, tables can be range-partitioned or list-partitioned. List partitioning is limited to a single column, but range partitioning can involve multiple columns. A partitioning "column" can be an expression. Because table partitioning is less general than table inheritance, it is hoped that it will be easier to reason about properties of partitions, and therefore that this will serve as a better foundation for a variety of possible optimizations, including query planner optimizations. The tuple routing based which this patch does based on the implicit partitioning constraints is an example of this, but it seems likely that many other useful optimizations are also possible. Amit Langote, reviewed and tested by Robert Haas, Ashutosh Bapat, Amit Kapila, Rajkumar Raghuwanshi, Corey Huinker, Jaime Casanova, Rushabh Lathia, Erik Rijkers, among others. Minor revisions by me.
2016-12-07 19:17:43 +01:00
Relation rel;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* check for unsupported flags */
Assert(!(eflags & (EXEC_FLAG_BACKWARD | EXEC_FLAG_MARK)));
/*
* create state structure
*/
mtstate = makeNode(ModifyTableState);
mtstate->ps.plan = (Plan *) node;
mtstate->ps.state = estate;
Move ExecProcNode from dispatch to function pointer based model. This allows us to add stack-depth checks the first time an executor node is called, and skip that overhead on following calls. Additionally it yields a nice speedup. While it'd probably have been a good idea to have that check all along, it has become more important after the new expression evaluation framework in b8d7f053c5c2bf2a7e - there's no stack depth check in common paths anymore now. We previously relied on ExecEvalExpr() being executed somewhere. We should move towards that model for further routines, but as this is required for v10, it seems better to only do the necessary (which already is quite large). Author: Andres Freund, Tom Lane Reported-By: Julien Rouhaud Discussion: https://postgr.es/m/22833.1490390175@sss.pgh.pa.us https://postgr.es/m/b0af9eaa-130c-60d0-9e4e-7a135b1e0c76@dalibo.com
2017-07-17 09:33:49 +02:00
mtstate->ps.ExecProcNode = ExecModifyTable;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
mtstate->operation = operation;
mtstate->canSetTag = node->canSetTag;
mtstate->mt_done = false;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
mtstate->mt_nrels = nrels;
Create ResultRelInfos later in InitPlan, index them by RT index. Instead of allocating all the ResultRelInfos upfront in one big array, allocate them in ExecInitModifyTable(). es_result_relations is now an array of ResultRelInfo pointers, rather than an array of structs, and it is indexed by the RT index. This simplifies things: we get rid of the separate concept of a "result rel index", and don't need to set it in setrefs.c anymore. This also allows follow-up optimizations (not included in this commit yet) to skip initializing ResultRelInfos for target relations that were not needed at runtime, and removal of the es_result_relation_info pointer. The EState arrays of regular result rels and root result rels are merged into one array. Similarly, the resultRelations and rootResultRelations lists in PlannedStmt are merged into one. It's not actually clear to me why they were kept separate in the first place, but now that the es_result_relations array is indexed by RT index, it certainly seems pointless. The PlannedStmt->resultRelations list is now only needed for ExecRelationIsTargetRelation(). One visible effect of this change is that ExecRelationIsTargetRelation() will now return 'true' also for the partition root, if a partitioned table is updated. That seems like a good thing, although the function isn't used in core code, and I don't see any reason for an FDW to call it on a partition root. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-13 11:57:02 +02:00
mtstate->resultRelInfo = (ResultRelInfo *)
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
palloc(nrels * sizeof(ResultRelInfo));
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
mtstate->mt_merge_inserted = 0;
mtstate->mt_merge_updated = 0;
mtstate->mt_merge_deleted = 0;
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
/*----------
* Resolve the target relation. This is the same as:
*
* - the relation for which we will fire FOR STATEMENT triggers,
* - the relation into whose tuple format all captured transition tuples
* must be converted, and
* - the root partitioned table used for tuple routing.
*
* If it's a partitioned table, the root partition doesn't appear
* elsewhere in the plan and its RT index is given explicitly in
* node->rootRelation. Otherwise (i.e. table inheritance) the target
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
* relation is the first relation in the node->resultRelations list.
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
*----------
*/
Create ResultRelInfos later in InitPlan, index them by RT index. Instead of allocating all the ResultRelInfos upfront in one big array, allocate them in ExecInitModifyTable(). es_result_relations is now an array of ResultRelInfo pointers, rather than an array of structs, and it is indexed by the RT index. This simplifies things: we get rid of the separate concept of a "result rel index", and don't need to set it in setrefs.c anymore. This also allows follow-up optimizations (not included in this commit yet) to skip initializing ResultRelInfos for target relations that were not needed at runtime, and removal of the es_result_relation_info pointer. The EState arrays of regular result rels and root result rels are merged into one array. Similarly, the resultRelations and rootResultRelations lists in PlannedStmt are merged into one. It's not actually clear to me why they were kept separate in the first place, but now that the es_result_relations array is indexed by RT index, it certainly seems pointless. The PlannedStmt->resultRelations list is now only needed for ExecRelationIsTargetRelation(). One visible effect of this change is that ExecRelationIsTargetRelation() will now return 'true' also for the partition root, if a partitioned table is updated. That seems like a good thing, although the function isn't used in core code, and I don't see any reason for an FDW to call it on a partition root. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-13 11:57:02 +02:00
if (node->rootRelation > 0)
{
mtstate->rootResultRelInfo = makeNode(ResultRelInfo);
ExecInitResultRelation(estate, mtstate->rootResultRelInfo,
node->rootRelation);
}
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
else
{
mtstate->rootResultRelInfo = mtstate->resultRelInfo;
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
ExecInitResultRelation(estate, mtstate->resultRelInfo,
linitial_int(node->resultRelations));
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
}
Fire per-statement triggers on partitioned tables. Even though no actual tuples are ever inserted into a partitioned table (the actual tuples are in the partitions, not the partitioned table itself), we still need to have a ResultRelInfo for the partitioned table, or per-statement triggers won't get fired. Amit Langote, per a report from Rajkumar Raghuwanshi. Reviewed by me. Discussion: http://postgr.es/m/CAKcux6%3DwYospCRY2J4XEFuVy0L41S%3Dfic7rmkbsU-GXhhSbmBg%40mail.gmail.com
2017-05-01 14:23:01 +02:00
Fix PlanRowMark/ExecRowMark structures to handle inheritance correctly. In an inherited UPDATE/DELETE, each target table has its own subplan, because it might have a column set different from other targets. This means that the resjunk columns we add to support EvalPlanQual might be at different physical column numbers in each subplan. The EvalPlanQual rewrite I did for 9.0 failed to account for this, resulting in possible misbehavior or even crashes during concurrent updates to the same row, as seen in a recent report from Gordon Shannon. Revise the data structure so that we track resjunk column numbers separately for each subplan. I also chose to move responsibility for identifying the physical column numbers back to executor startup, instead of assuming that numbers derived during preprocess_targetlist would stay valid throughout subsequent massaging of the plan. That's a bit slower, so we might want to consider undoing it someday; but it would complicate the patch considerably and didn't seem justifiable in a bug fix that has to be back-patched to 9.0.
2011-01-13 02:47:02 +01:00
/* set up epqstate with dummy subplan data for the moment */
Fix misbehavior of EvalPlanQual checks with multiple result relations. The idea of EvalPlanQual is that we replace the query's scan of the result relation with a single injected tuple, and see if we get a tuple out, thereby implying that the injected tuple still passes the query quals. (In join cases, other relations in the query are still scanned normally.) This logic was not updated when commit 86dc90056 made it possible for a single DML query plan to have multiple result relations, when the query target relation has inheritance or partition children. We replaced the output for the current result relation successfully, but other result relations were still scanned normally; thus, if any other result relation contained a tuple satisfying the quals, we'd think the EPQ check passed, even if it did not pass for the injected tuple itself. This would lead to update or delete actions getting performed when they should have been skipped due to a conflicting concurrent update in READ COMMITTED isolation mode. Fix by blocking all sibling result relations from emitting tuples during an EvalPlanQual recheck. In the back branches, the fix is complicated a bit by the need to not change the size of struct EPQState (else we'd have ABI-breaking changes in offsets in struct ModifyTableState). Like the back-patches of 3f7836ff6 and 4b3e37993, add a separately palloc'd struct to avoid that. The logic is the same as in HEAD otherwise. This is only a live bug back to v14 where 86dc90056 came in. However, I chose to back-patch the test cases further, on the grounds that this whole area is none too well tested. I skipped doing so in v11 though because none of the test applied cleanly, and it didn't quite seem worth extra work for a branch with only six months to live. Per report from Ante Krešić (via Aleksander Alekseev) Discussion: https://postgr.es/m/CAJ7c6TMBTN3rcz4=AjYhLPD_w3FFT0Wq_C15jxCDn8U4tZnH1g@mail.gmail.com
2023-05-19 20:26:34 +02:00
EvalPlanQualInit(&mtstate->mt_epqstate, estate, NULL, NIL,
node->epqParam, node->resultRelations);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
mtstate->fireBSTriggers = true;
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
/*
* Build state for collecting transition tuples. This requires having a
* valid trigger query context, so skip it in explain-only mode.
*/
if (!(eflags & EXEC_FLAG_EXPLAIN_ONLY))
ExecSetupTransitionCaptureState(mtstate, estate);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* Open all the result relations and initialize the ResultRelInfo structs.
* (But root relation was initialized above, if it's part of the array.)
* We must do this before initializing the subplan, because direct-modify
* FDWs expect their ResultRelInfos to be available.
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*/
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
resultRelInfo = mtstate->resultRelInfo;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
i = 0;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
foreach(l, node->resultRelations)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Create ResultRelInfos later in InitPlan, index them by RT index. Instead of allocating all the ResultRelInfos upfront in one big array, allocate them in ExecInitModifyTable(). es_result_relations is now an array of ResultRelInfo pointers, rather than an array of structs, and it is indexed by the RT index. This simplifies things: we get rid of the separate concept of a "result rel index", and don't need to set it in setrefs.c anymore. This also allows follow-up optimizations (not included in this commit yet) to skip initializing ResultRelInfos for target relations that were not needed at runtime, and removal of the es_result_relation_info pointer. The EState arrays of regular result rels and root result rels are merged into one array. Similarly, the resultRelations and rootResultRelations lists in PlannedStmt are merged into one. It's not actually clear to me why they were kept separate in the first place, but now that the es_result_relations array is indexed by RT index, it certainly seems pointless. The PlannedStmt->resultRelations list is now only needed for ExecRelationIsTargetRelation(). One visible effect of this change is that ExecRelationIsTargetRelation() will now return 'true' also for the partition root, if a partitioned table is updated. That seems like a good thing, although the function isn't used in core code, and I don't see any reason for an FDW to call it on a partition root. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqGEmiib8FLiHMhKB%2BCH5dRgHSLc5N5wnvc4kym%2BZYpQEQ%40mail.gmail.com
2020-10-13 11:57:02 +02:00
Index resultRelation = lfirst_int(l);
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
if (resultRelInfo != mtstate->rootResultRelInfo)
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
{
Revise child-to-root tuple conversion map management. Store the tuple conversion map to convert a tuple from a child table's format to the root format in a new ri_ChildToRootMap field in ResultRelInfo. It is initialized if transition tuple capture for FOR STATEMENT triggers or INSERT tuple routing on a partitioned table is needed. Previously, ModifyTable kept the maps in the per-subplan ModifyTableState->mt_per_subplan_tupconv_maps array, or when tuple routing was used, in ResultRelInfo->ri_Partitioninfo->pi_PartitionToRootMap. The new field replaces both of those. Now that the child-to-root tuple conversion map is always available in ResultRelInfo (when needed), remove the TransitionCaptureState.tcs_map field. The callers of Exec*Trigger() functions no longer need to set or save it, which is much less confusing and bug-prone. Also, as a future optimization, this will allow us to delay creating the map for a given result relation until the relation is actually processed during execution. Author: Amit Langote Discussion: https://www.postgresql.org/message-id/CA%2BHiwqHtCWLdK-LO%3DNEsvOdHx%2B7yv4mE_zYK0i3BH7dXb-wxog%40mail.gmail.com
2020-10-19 13:11:54 +02:00
ExecInitResultRelation(estate, resultRelInfo, resultRelation);
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/*
* For child result relations, store the root result relation
* pointer. We do so for the convenience of places that want to
* look at the query's original target relation but don't have the
* mtstate handy.
*/
resultRelInfo->ri_RootResultRelInfo = mtstate->rootResultRelInfo;
}
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
/* Initialize the usesFdwDirectModify flag */
Split ExecUpdate and ExecDelete into reusable pieces Create subroutines ExecUpdatePrologue / ExecUpdateAct / ExecUpdateEpilogue, and similar for ExecDelete. Introduce a new struct to be used internally in nodeModifyTable.c, dubbed ModifyTableContext, which contains all context information needed to perform these operations, as well as ExecInsert and others. This allows using a different schedule and a different way of evaluating the results of these operations, which can be exploited by a later commit introducing support for MERGE. It also makes ExecUpdate and ExecDelete proper shorter and (hopefully) simpler. Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Reviewed-by: Amit Langote <amitlangote09@gmail.com> Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/202202271724.4z7xv3cf46kv@alvherre.pgsql
2022-03-17 11:47:04 +01:00
resultRelInfo->ri_usesFdwDirectModify =
bms_is_member(i, node->fdwDirectModifyPlans);
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
/*
* Verify result relation is a valid target for the current operation
*/
Even if some partitions are foreign, allow tuple routing. This doesn't allow routing tuple to the foreign partitions themselves, but it permits tuples to be routed to regular partitions despite the presence of foreign partitions in the same inheritance hierarchy. Etsuro Fujita, reviewed by Amit Langote and by me. Discussion: http://postgr.es/m/bc3db4c1-1693-3b8a-559f-33ad2b50b7ad@lab.ntt.co.jp
2017-09-07 16:55:45 +02:00
CheckValidResultRel(resultRelInfo, operation);
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
resultRelInfo++;
i++;
}
/*
* Now we may initialize the subplan.
*/
outerPlanState(mtstate) = ExecInitNode(subplan, estate, eflags);
/*
* Do additional per-result-relation initialization.
*/
for (i = 0; i < nrels; i++)
{
resultRelInfo = &mtstate->resultRelInfo[i];
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/* Let FDWs init themselves for foreign-table result rels */
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
if (!resultRelInfo->ri_usesFdwDirectModify &&
resultRelInfo->ri_FdwRoutine != NULL &&
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
resultRelInfo->ri_FdwRoutine->BeginForeignModify != NULL)
{
List *fdw_private = (List *) list_nth(node->fdwPrivLists, i);
resultRelInfo->ri_FdwRoutine->BeginForeignModify(mtstate,
resultRelInfo,
fdw_private,
i,
eflags);
}
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/*
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* For UPDATE/DELETE/MERGE, find the appropriate junk attr now, either
* a 'ctid' or 'wholerow' attribute depending on relkind. For foreign
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
* tables, the FDW might have created additional junk attr(s), but
* those are no concern of ours.
*/
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
if (operation == CMD_UPDATE || operation == CMD_DELETE ||
operation == CMD_MERGE)
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
{
char relkind;
relkind = resultRelInfo->ri_RelationDesc->rd_rel->relkind;
if (relkind == RELKIND_RELATION ||
relkind == RELKIND_MATVIEW ||
relkind == RELKIND_PARTITIONED_TABLE)
{
resultRelInfo->ri_RowIdAttNo =
ExecFindJunkAttributeInTlist(subplan->targetlist, "ctid");
if (!AttributeNumberIsValid(resultRelInfo->ri_RowIdAttNo))
elog(ERROR, "could not find junk ctid column");
}
else if (relkind == RELKIND_FOREIGN_TABLE)
{
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/*
* We don't support MERGE with foreign tables for now. (It's
* problematic because the implementation uses CTID.)
*/
Assert(operation != CMD_MERGE);
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/*
* When there is a row-level trigger, there should be a
* wholerow attribute. We also require it to be present in
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* UPDATE and MERGE, so we can get the values of unchanged
* columns.
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
*/
resultRelInfo->ri_RowIdAttNo =
ExecFindJunkAttributeInTlist(subplan->targetlist,
"wholerow");
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
if ((mtstate->operation == CMD_UPDATE || mtstate->operation == CMD_MERGE) &&
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
!AttributeNumberIsValid(resultRelInfo->ri_RowIdAttNo))
elog(ERROR, "could not find junk wholerow column");
}
else
{
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/* No support for MERGE */
Assert(operation != CMD_MERGE);
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/* Other valid target relkinds must provide wholerow */
resultRelInfo->ri_RowIdAttNo =
ExecFindJunkAttributeInTlist(subplan->targetlist,
"wholerow");
if (!AttributeNumberIsValid(resultRelInfo->ri_RowIdAttNo))
elog(ERROR, "could not find junk wholerow column");
}
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
/*
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* If this is an inherited update/delete/merge, there will be a junk
* attribute named "tableoid" present in the subplan's targetlist. It
* will be used to identify the result relation for a given tuple to be
* updated/deleted/merged.
Postpone some more stuff out of ExecInitModifyTable. Delay creation of the projections for INSERT and UPDATE tuples until they're needed. This saves a pretty fair amount of work when only some of the partitions are actually touched. The logic associated with identifying junk columns in UPDATE/DELETE is moved to another loop, allowing removal of one loop over the target relations; but it didn't actually change at all. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-07 00:13:05 +02:00
*/
mtstate->mt_resultOidAttno =
ExecFindJunkAttributeInTlist(subplan->targetlist, "tableoid");
Assert(AttributeNumberIsValid(mtstate->mt_resultOidAttno) || nrels == 1);
mtstate->mt_lastResultOid = InvalidOid; /* force lookup at first tuple */
mtstate->mt_lastResultIndex = 0; /* must be zero if no such attr */
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/* Get the root target relation */
Clean up code to resolve the "root target relation" in nodeModifyTable.c When executing DDL on a partitioned table or on a table with inheritance children, statement-level triggers must be fired against the table given in the original statement. The code to look that up was a bit messy and duplicative. Commit 501ed02cf6 added a helper function, getASTriggerResultRelInfo() (later renamed to getTargetResultRelInfo()) for it, but for some reason it was only used when firing AFTER STATEMENT triggers and the code to fire BEFORE STATEMENT triggers duplicated the logic. Determine the target relation in ExecInitModifyTable(), and set it always in ModifyTableState. Code that used to call getTargetResultRelInfo() can now use ModifyTableState->rootResultRelInfo directly. Discussion: https://www.postgresql.org/message-id/CA%2BHiwqFViT47Zbr_ASBejiK7iDG8%3DQ1swQ-tjM6caRPQ67pT%3Dw%40mail.gmail.com
2020-10-19 13:11:44 +02:00
rel = mtstate->rootResultRelInfo->ri_RelationDesc;
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
/*
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
* Build state for tuple routing if it's a partitioned INSERT. An UPDATE
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
* or MERGE might need this too, but only if it actually moves tuples
* between partitions; in that case setup is done by
* ExecCrossPartitionUpdate.
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
*/
if (rel->rd_rel->relkind == RELKIND_PARTITIONED_TABLE &&
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
operation == CMD_INSERT)
Be lazier about partition tuple routing. It's not necessary to fully initialize the executor data structures for partitions to which no tuples are ever routed. Consider, for example, an INSERT statement that inserts only one row: it only cares about the partition to which that one row is routed. The new function ExecInitPartitionInfo performs the initialization in question only when a particular partition is about to receive a tuple. This includes creating, validating, and saving a pointer to the ResultRelInfo, setting up for speculative insertions, translating WCOs and initializing the resulting expressions, translating returning lists and building the appropriate projection information, and setting up a tuple conversion map. One thing that's not deferred is locking the child partitions; that seems desirable but would need more thought. Still, testing shows that this makes single-row inserts significantly faster on a table with many partitions without harming the bulk-insert case. Amit Langote, reviewed by Etsuro Fujita, with a few changes by me Discussion: http://postgr.es/m/8975331d-d961-cbdd-f862-fdd3d97dc2d0@lab.ntt.co.jp
2018-02-22 16:55:54 +01:00
mtstate->mt_partition_tuple_routing =
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
ExecSetupPartitionTupleRouting(estate, rel);
Allow UPDATE to move rows between partitions. When an UPDATE causes a row to no longer match the partition constraint, try to move it to a different partition where it does match the partition constraint. In essence, the UPDATE is split into a DELETE from the old partition and an INSERT into the new one. This can lead to surprising behavior in concurrency scenarios because EvalPlanQual rechecks won't work as they normally did; the known problems are documented. (There is a pending patch to improve the situation further, but it needs more review.) Amit Khandekar, reviewed and tested by Amit Langote, David Rowley, Rajkumar Raghuwanshi, Dilip Kumar, Amul Sul, Thomas Munro, Álvaro Herrera, Amit Kapila, and me. A few final revisions by me. Discussion: http://postgr.es/m/CAJ3gD9do9o2ccQ7j7+tSgiE1REY65XRiMb=yJO3u3QhyP8EEPQ@mail.gmail.com
2018-01-19 21:33:06 +01:00
WITH CHECK OPTION support for auto-updatable VIEWs For simple views which are automatically updatable, this patch allows the user to specify what level of checking should be done on records being inserted or updated. For 'LOCAL CHECK', new tuples are validated against the conditionals of the view they are being inserted into, while for 'CASCADED CHECK' the new tuples are validated against the conditionals for all views involved (from the top down). This option is part of the SQL specification. Dean Rasheed, reviewed by Pavel Stehule
2013-07-18 23:10:16 +02:00
/*
* Initialize any WITH CHECK OPTION constraints if needed.
*/
resultRelInfo = mtstate->resultRelInfo;
foreach(l, node->withCheckOptionLists)
{
List *wcoList = (List *) lfirst(l);
List *wcoExprs = NIL;
ListCell *ll;
foreach(ll, wcoList)
{
WithCheckOption *wco = (WithCheckOption *) lfirst(ll);
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
ExprState *wcoExpr = ExecInitQual((List *) wco->qual,
Fix parent of WCO qual. The parent of some WCO expressions was, apparently by accident, set to the the source of DML queries, rather than the target table. This causes problems for the upcoming pluggable storage work, because the target and source table might be of different storage types. It's possible that this is already problematic, but neither experimenting nor inquiries on -hackers have found them. So don't backpatch for now. Author: Andres Freund Discussion: https://postgr.es/m/20181205225213.hiwa3kgoxeybqcqv@alap3.anarazel.de
2019-01-15 20:59:32 +01:00
&mtstate->ps);
pgindent run for 9.4 This includes removing tabs after periods in C comments, which was applied to back branches, so this change should not effect backpatching.
2014-05-06 18:12:18 +02:00
WITH CHECK OPTION support for auto-updatable VIEWs For simple views which are automatically updatable, this patch allows the user to specify what level of checking should be done on records being inserted or updated. For 'LOCAL CHECK', new tuples are validated against the conditionals of the view they are being inserted into, while for 'CASCADED CHECK' the new tuples are validated against the conditionals for all views involved (from the top down). This option is part of the SQL specification. Dean Rasheed, reviewed by Pavel Stehule
2013-07-18 23:10:16 +02:00
wcoExprs = lappend(wcoExprs, wcoExpr);
}
resultRelInfo->ri_WithCheckOptions = wcoList;
resultRelInfo->ri_WithCheckOptionExprs = wcoExprs;
resultRelInfo++;
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* Initialize RETURNING projections if needed.
*/
if (node->returningLists)
{
TupleTableSlot *slot;
ExprContext *econtext;
/*
* Initialize result tuple slot and assign its rowtype using the first
* RETURNING list. We assume the rest will look the same.
*/
Rely on executor utils to build targetlist for DML RETURNING. This is useful because it gets rid of the sole direct user of ExecAssignResultType(). A future commit will likely make use of that and combine creating the targetlist with the initialization of the result slot. But it seems like good code hygiene anyway. Author: Andres Freund Discussion: https://postgr.es/m/20170901064131.tazjxwus3k2w3ybh@alap3.anarazel.de
2017-12-29 21:26:29 +01:00
mtstate->ps.plan->targetlist = (List *) linitial(node->returningLists);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/* Set up a slot for the output of the RETURNING projection(s) */
Introduce notion of different types of slots (without implementing them). Upcoming work intends to allow pluggable ways to introduce new ways of storing table data. Accessing those table access methods from the executor requires TupleTableSlots to be carry tuples in the native format of such storage methods; otherwise there'll be a significant conversion overhead. Different access methods will require different data to store tuples efficiently (just like virtual, minimal, heap already require fields in TupleTableSlot). To allow that without requiring additional pointer indirections, we want to have different structs (embedding TupleTableSlot) for different types of slots. Thus different types of slots are needed, which requires adapting creators of slots. The slot that most efficiently can represent a type of tuple in an executor node will often depend on the type of slot a child node uses. Therefore we need to track the type of slot is returned by nodes, so parent slots can create slots based on that. Relatedly, JIT compilation of tuple deforming needs to know which type of slot a certain expression refers to, so it can create an appropriate deforming function for the type of tuple in the slot. But not all nodes will only return one type of slot, e.g. an append node will potentially return different types of slots for each of its subplans. Therefore add function that allows to query the type of a node's result slot, and whether it'll always be the same type (whether it's fixed). This can be queried using ExecGetResultSlotOps(). The scan, result, inner, outer type of slots are automatically inferred from ExecInitScanTupleSlot(), ExecInitResultSlot(), left/right subtrees respectively. If that's not correct for a node, that can be overwritten using new fields in PlanState. This commit does not introduce the actually abstracted implementation of different kind of TupleTableSlots, that will be left for a followup commit. The different types of slots introduced will, for now, still use the same backing implementation. While this already partially invalidates the big comment in tuptable.h, it seems to make more sense to update it later, when the different TupleTableSlot implementations actually exist. Author: Ashutosh Bapat and Andres Freund, with changes by Amit Khandekar Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-16 07:00:30 +01:00
ExecInitResultTupleSlotTL(&mtstate->ps, &TTSOpsVirtual);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
slot = mtstate->ps.ps_ResultTupleSlot;
/* Need an econtext too */
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
if (mtstate->ps.ps_ExprContext == NULL)
ExecAssignExprContext(estate, &mtstate->ps);
econtext = mtstate->ps.ps_ExprContext;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* Build a projection for each result rel.
*/
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
resultRelInfo = mtstate->resultRelInfo;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
foreach(l, node->returningLists)
{
List *rlist = (List *) lfirst(l);
Allow insert and update tuple routing and COPY for foreign tables. Also enable this for postgres_fdw. Etsuro Fujita, based on an earlier patch by Amit Langote. The larger patch series of which this is a part has been reviewed by Amit Langote, David Fetter, Maksim Milyutin, Álvaro Herrera, Stephen Frost, and me. Minor documentation changes to the final version by me. Discussion: http://postgr.es/m/29906a26-da12-8c86-4fb9-d8f88442f2b9@lab.ntt.co.jp
2018-04-07 01:16:11 +02:00
resultRelInfo->ri_returningList = rlist;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
resultRelInfo->ri_projectReturning =
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
ExecBuildProjectionInfo(rlist, econtext, slot, &mtstate->ps,
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
resultRelInfo->ri_RelationDesc->rd_att);
resultRelInfo++;
}
}
else
{
/*
* We still must construct a dummy result tuple type, because InitPlan
* expects one (maybe should change that?).
*/
Rely on executor utils to build targetlist for DML RETURNING. This is useful because it gets rid of the sole direct user of ExecAssignResultType(). A future commit will likely make use of that and combine creating the targetlist with the initialization of the result slot. But it seems like good code hygiene anyway. Author: Andres Freund Discussion: https://postgr.es/m/20170901064131.tazjxwus3k2w3ybh@alap3.anarazel.de
2017-12-29 21:26:29 +01:00
mtstate->ps.plan->targetlist = NIL;
Don't require return slots for nodes without projection. In a lot of nodes the return slot is not required. That can either be because the node doesn't do any projection (say an Append node), or because the node does perform projections but the projection is optimized away because the projection would yield an identical row. Slots aren't that small, especially for wide rows, so it's worthwhile to avoid creating them. It's not possible to just skip creating the slot - it's currently used to determine the tuple descriptor returned by ExecGetResultType(). So separate the determination of the result type from the slot creation. The work previously done internally ExecInitResultTupleSlotTL() can now also be done separately with ExecInitResultTypeTL() and ExecInitResultSlot(). That way nodes that aren't guaranteed to need a result slot, can use ExecInitResultTypeTL() to determine the result type of the node, and ExecAssignScanProjectionInfo() (via ExecConditionalAssignProjectionInfo()) determines that a result slot is needed, it is created with ExecInitResultSlot(). Besides the advantage of avoiding to create slots that then are unused, this is necessary preparation for later patches around tuple table slot abstraction. In particular separating the return descriptor and slot is a prerequisite to allow JITing of tuple deforming with knowledge of the underlying tuple format, and to avoid unnecessarily creating JITed tuple deforming for virtual slots. This commit removes a redundant argument from ExecInitResultTupleSlotTL(). While this commit touches a lot of the relevant lines anyway, it'd normally still not worthwhile to cause breakage, except that aforementioned later commits will touch *all* ExecInitResultTupleSlotTL() callers anyway (but fits worse thematically). Author: Andres Freund Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-10 02:19:39 +01:00
ExecInitResultTypeTL(&mtstate->ps);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
mtstate->ps.ps_ExprContext = NULL;
}
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
/* Set the list of arbiter indexes if needed for ON CONFLICT */
resultRelInfo = mtstate->resultRelInfo;
if (node->onConflictAction != ONCONFLICT_NONE)
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
{
/* insert may only have one relation, inheritance is not expanded */
Assert(nrels == 1);
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
resultRelInfo->ri_onConflictArbiterIndexes = node->arbiterIndexes;
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
}
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/*
* If needed, Initialize target list, projection and qual for ON CONFLICT
* DO UPDATE.
*/
if (node->onConflictAction == ONCONFLICT_UPDATE)
{
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
OnConflictSetState *onconfl = makeNode(OnConflictSetState);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
ExprContext *econtext;
Allow tupleslots to have a fixed tupledesc, use in executor nodes. The reason for doing so is that it will allow expression evaluation to optimize based on the underlying tupledesc. In particular it will allow to JIT tuple deforming together with the expression itself. For that expression initialization needs to be moved after the relevant slots are initialized - mostly unproblematic, except in the case of nodeWorktablescan.c. After doing so there's no need for ExecAssignResultType() and ExecAssignResultTypeFromTL() anymore, as all former callers have been converted to create a slot with a fixed descriptor. When creating a slot with a fixed descriptor, tts_values/isnull can be allocated together with the main slot, reducing allocation overhead and increasing cache density a bit. Author: Andres Freund Discussion: https://postgr.es/m/20171206093717.vqdxe5icqttpxs3p@alap3.anarazel.de
2018-02-17 06:17:38 +01:00
TupleDesc relationDesc;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
/* already exists if created by RETURNING processing above */
if (mtstate->ps.ps_ExprContext == NULL)
ExecAssignExprContext(estate, &mtstate->ps);
econtext = mtstate->ps.ps_ExprContext;
Allow tupleslots to have a fixed tupledesc, use in executor nodes. The reason for doing so is that it will allow expression evaluation to optimize based on the underlying tupledesc. In particular it will allow to JIT tuple deforming together with the expression itself. For that expression initialization needs to be moved after the relevant slots are initialized - mostly unproblematic, except in the case of nodeWorktablescan.c. After doing so there's no need for ExecAssignResultType() and ExecAssignResultTypeFromTL() anymore, as all former callers have been converted to create a slot with a fixed descriptor. When creating a slot with a fixed descriptor, tts_values/isnull can be allocated together with the main slot, reducing allocation overhead and increasing cache density a bit. Author: Andres Freund Discussion: https://postgr.es/m/20171206093717.vqdxe5icqttpxs3p@alap3.anarazel.de
2018-02-17 06:17:38 +01:00
relationDesc = resultRelInfo->ri_RelationDesc->rd_att;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
/* create state for DO UPDATE SET operation */
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
resultRelInfo->ri_onConflict = onconfl;
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
/* initialize slot for the existing tuple */
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
onconfl->oc_Existing =
tableam: Add and use scan APIs. Too allow table accesses to be not directly dependent on heap, several new abstractions are needed. Specifically: 1) Heap scans need to be generalized into table scans. Do this by introducing TableScanDesc, which will be the "base class" for individual AMs. This contains the AM independent fields from HeapScanDesc. The previous heap_{beginscan,rescan,endscan} et al. have been replaced with a table_ version. There's no direct replacement for heap_getnext(), as that returned a HeapTuple, which is undesirable for a other AMs. Instead there's table_scan_getnextslot(). But note that heap_getnext() lives on, it's still used widely to access catalog tables. This is achieved by new scan_begin, scan_end, scan_rescan, scan_getnextslot callbacks. 2) The portion of parallel scans that's shared between backends need to be able to do so without the user doing per-AM work. To achieve that new parallelscan_{estimate, initialize, reinitialize} callbacks are introduced, which operate on a new ParallelTableScanDesc, which again can be subclassed by AMs. As it is likely that several AMs are going to be block oriented, block oriented callbacks that can be shared between such AMs are provided and used by heap. table_block_parallelscan_{estimate, intiialize, reinitialize} as callbacks, and table_block_parallelscan_{nextpage, init} for use in AMs. These operate on a ParallelBlockTableScanDesc. 3) Index scans need to be able to access tables to return a tuple, and there needs to be state across individual accesses to the heap to store state like buffers. That's now handled by introducing a sort-of-scan IndexFetchTable, which again is intended to be subclassed by individual AMs (for heap IndexFetchHeap). The relevant callbacks for an AM are index_fetch_{end, begin, reset} to create the necessary state, and index_fetch_tuple to retrieve an indexed tuple. Note that index_fetch_tuple implementations need to be smarter than just blindly fetching the tuples for AMs that have optimizations similar to heap's HOT - the currently alive tuple in the update chain needs to be fetched if appropriate. Similar to table_scan_getnextslot(), it's undesirable to continue to return HeapTuples. Thus index_fetch_heap (might want to rename that later) now accepts a slot as an argument. Core code doesn't have a lot of call sites performing index scans without going through the systable_* API (in contrast to loads of heap_getnext calls and working directly with HeapTuples). Index scans now store the result of a search in IndexScanDesc->xs_heaptid, rather than xs_ctup->t_self. As the target is not generally a HeapTuple anymore that seems cleaner. To be able to sensible adapt code to use the above, two further callbacks have been introduced: a) slot_callbacks returns a TupleTableSlotOps* suitable for creating slots capable of holding a tuple of the AMs type. table_slot_callbacks() and table_slot_create() are based upon that, but have additional logic to deal with views, foreign tables, etc. While this change could have been done separately, nearly all the call sites that needed to be adapted for the rest of this commit also would have been needed to be adapted for table_slot_callbacks(), making separation not worthwhile. b) tuple_satisfies_snapshot checks whether the tuple in a slot is currently visible according to a snapshot. That's required as a few places now don't have a buffer + HeapTuple around, but a slot (which in heap's case internally has that information). Additionally a few infrastructure changes were needed: I) SysScanDesc, as used by systable_{beginscan, getnext} et al. now internally uses a slot to keep track of tuples. While systable_getnext() still returns HeapTuples, and will so for the foreseeable future, the index API (see 1) above) now only deals with slots. The remainder, and largest part, of this commit is then adjusting all scans in postgres to use the new APIs. Author: Andres Freund, Haribabu Kommi, Alvaro Herrera Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de https://postgr.es/m/20160812231527.GA690404@alvherre.pgsql
2019-03-11 20:46:41 +01:00
table_slot_create(resultRelInfo->ri_RelationDesc,
&mtstate->ps.state->es_tupleTable);
Don't reuse slots between root and partition in ON CONFLICT ... UPDATE. Until now the the slot to store the conflicting tuple, and the result of the ON CONFLICT SET, where reused between partitions. That necessitated changing slots descriptor when switching partitions. Besides the overhead of switching descriptors on a slot (which requires memory allocations and prevents JITing), that's importantly also problematic for tableam. There individual partitions might belong to different tableams, needing different kinds of slots. In passing also fix ExecOnConflictUpdate to clear the existing slot at exit. Otherwise that slot could continue to hold a pin till the query ends, which could be far too long if the input data set is large, and there's no further conflicts. While previously also problematic, it's now more important as there will be more such slots when partitioned. Author: Andres Freund Reviewed-By: Robert Haas, David Rowley Discussion: https://postgr.es/m/20180703070645.wchpu5muyto5n647@alap3.anarazel.de
2019-03-07 00:43:33 +01:00
Fix system column accesses in ON CONFLICT ... RETURNING. After 277cb789836 ON CONFLICT ... SET ... RETURNING failed with ERROR: virtual tuple table slot does not have system attributes when taking the update path, as the slot used to insert into the table (and then process RETURNING) was defined to be a virtual slot in that commit. Virtual slots don't support system columns except for tableoid and ctid, as the other system columns are AM dependent. Fix that by using a slot of the table's type. Add tests for system column accesses in ON CONFLICT ... RETURNING. Reported-By: Roby, bisected to the relevant commit by Jeff Janes Author: Andres Freund Discussion: https://postgr.es/m/73436355-6432-49B1-92ED-1FE4F7E7E100@finefun.com.au Backpatch: 12-, where the bug was introduced in 277cb789836
2019-07-25 03:45:58 +02:00
/*
* Create the tuple slot for the UPDATE SET projection. We want a slot
* of the table's type here, because the slot will be used to insert
* into the table, and for RETURNING processing - which may access
* system attributes.
*/
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
onconfl->oc_ProjSlot =
table_slot_create(resultRelInfo->ri_RelationDesc,
&mtstate->ps.state->es_tupleTable);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
/* build UPDATE SET projection state */
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
onconfl->oc_ProjInfo =
ExecBuildUpdateProjection(node->onConflictSet,
true,
node->onConflictCols,
relationDesc,
econtext,
onconfl->oc_ProjSlot,
&mtstate->ps);
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
Handle INSERT .. ON CONFLICT with partitioned tables Commit eb7ed3f30634 enabled unique constraints on partitioned tables, but one thing that was not working properly is INSERT/ON CONFLICT. This commit introduces a new node keeps state related to the ON CONFLICT clause per partition, and fills it when that partition is about to be used for tuple routing. Author: Amit Langote, Álvaro Herrera Reviewed-by: Etsuro Fujita, Pavan Deolasee Discussion: https://postgr.es/m/20180228004602.cwdyralmg5ejdqkq@alvherre.pgsql
2018-03-26 15:43:54 +02:00
/* initialize state to evaluate the WHERE clause, if any */
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
if (node->onConflictWhere)
{
ExprState *qualexpr;
Faster expression evaluation and targetlist projection. This replaces the old, recursive tree-walk based evaluation, with non-recursive, opcode dispatch based, expression evaluation. Projection is now implemented as part of expression evaluation. This both leads to significant performance improvements, and makes future just-in-time compilation of expressions easier. The speed gains primarily come from: - non-recursive implementation reduces stack usage / overhead - simple sub-expressions are implemented with a single jump, without function calls - sharing some state between different sub-expressions - reduced amount of indirect/hard to predict memory accesses by laying out operation metadata sequentially; including the avoidance of nearly all of the previously used linked lists - more code has been moved to expression initialization, avoiding constant re-checks at evaluation time Future just-in-time compilation (JIT) has become easier, as demonstrated by released patches intended to be merged in a later release, for primarily two reasons: Firstly, due to a stricter split between expression initialization and evaluation, less code has to be handled by the JIT. Secondly, due to the non-recursive nature of the generated "instructions", less performance-critical code-paths can easily be shared between interpreted and compiled evaluation. The new framework allows for significant future optimizations. E.g.: - basic infrastructure for to later reduce the per executor-startup overhead of expression evaluation, by caching state in prepared statements. That'd be helpful in OLTPish scenarios where initialization overhead is measurable. - optimizing the generated "code". A number of proposals for potential work has already been made. - optimizing the interpreter. Similarly a number of proposals have been made here too. The move of logic into the expression initialization step leads to some backward-incompatible changes: - Function permission checks are now done during expression initialization, whereas previously they were done during execution. In edge cases this can lead to errors being raised that previously wouldn't have been, e.g. a NULL array being coerced to a different array type previously didn't perform checks. - The set of domain constraints to be checked, is now evaluated once during expression initialization, previously it was re-built every time a domain check was evaluated. For normal queries this doesn't change much, but e.g. for plpgsql functions, which caches ExprStates, the old set could stick around longer. The behavior around might still change. Author: Andres Freund, with significant changes by Tom Lane, changes by Heikki Linnakangas Reviewed-By: Tom Lane, Heikki Linnakangas Discussion: https://postgr.es/m/20161206034955.bh33paeralxbtluv@alap3.anarazel.de
2017-03-14 23:45:36 +01:00
qualexpr = ExecInitQual((List *) node->onConflictWhere,
Attach ON CONFLICT SET ... WHERE to the correct planstate. The previous coding was a leftover from attempting to hang all the on conflict logic onto modify table's child nodes. It appears to not have actually caused problems except for explain. Add test exercising the broken and some other code paths. Author: Peter Geoghegan and Andres Freund
2015-05-19 01:55:10 +02:00
&mtstate->ps);
Fix mishandling of resjunk columns in ON CONFLICT ... UPDATE tlists. It's unusual to have any resjunk columns in an ON CONFLICT ... UPDATE list, but it can happen when MULTIEXPR_SUBLINK SubPlans are present. If it happens, the ON CONFLICT UPDATE code path would end up storing tuples that include the values of the extra resjunk columns. That's fairly harmless in the short run, but if new columns are added to the table then the values would become accessible, possibly leading to malfunctions if they don't match the datatypes of the new columns. This had escaped notice through a confluence of missing sanity checks, including * There's no cross-check that a tuple presented to heap_insert or heap_update matches the table rowtype. While it's difficult to check that fully at reasonable cost, we can easily add assertions that there aren't too many columns. * The output-column-assignment cases in execExprInterp.c lacked any sanity checks on the output column numbers, which seems like an oversight considering there are plenty of assertion checks on input column numbers. Add assertions there too. * We failed to apply nodeModifyTable's ExecCheckPlanOutput() to the ON CONFLICT UPDATE tlist. That wouldn't have caught this specific error, since that function is chartered to ignore resjunk columns; but it sure seems like a bad omission now that we've seen this bug. In HEAD, the right way to fix this is to make the processing of ON CONFLICT UPDATE tlists work the same as regular UPDATE tlists now do, that is don't add "SET x = x" entries, and use ExecBuildUpdateProjection to evaluate the tlist and combine it with old values of the not-set columns. This adds a little complication to ExecBuildUpdateProjection, but allows removal of a comparable amount of now-dead code from the planner. In the back branches, the most expedient solution seems to be to (a) use an output slot for the ON CONFLICT UPDATE projection that actually matches the target table, and then (b) invent a variant of ExecBuildProjectionInfo that can be told to not store values resulting from resjunk columns, so it doesn't try to store into nonexistent columns of the output slot. (We can't simply ignore the resjunk columns altogether; they have to be evaluated for MULTIEXPR_SUBLINK to work.) This works back to v10. In 9.6, projections work much differently and we can't cheaply give them such an option. The 9.6 version of this patch works by inserting a JunkFilter when it's necessary to get rid of resjunk columns. In addition, v11 and up have the reverse problem when trying to perform ON CONFLICT UPDATE on a partitioned table. Through a further oversight, adjust_partition_tlist() discarded resjunk columns when re-ordering the ON CONFLICT UPDATE tlist to match a partition. This accidentally prevented the storing-bogus-tuples problem, but at the cost that MULTIEXPR_SUBLINK cases didn't work, typically crashing if more than one row has to be updated. Fix by preserving resjunk columns in that routine. (I failed to resist the temptation to add more assertions there too, and to do some minor code beautification.) Per report from Andres Freund. Back-patch to all supported branches. Security: CVE-2021-32028
2021-05-10 17:02:29 +02:00
onconfl->oc_WhereClause = qualexpr;
Add support for INSERT ... ON CONFLICT DO NOTHING/UPDATE. The newly added ON CONFLICT clause allows to specify an alternative to raising a unique or exclusion constraint violation error when inserting. ON CONFLICT refers to constraints that can either be specified using a inference clause (by specifying the columns of a unique constraint) or by naming a unique or exclusion constraint. DO NOTHING avoids the constraint violation, without touching the pre-existing row. DO UPDATE SET ... [WHERE ...] updates the pre-existing tuple, and has access to both the tuple proposed for insertion and the existing tuple; the optional WHERE clause can be used to prevent an update from being executed. The UPDATE SET and WHERE clauses have access to the tuple proposed for insertion using the "magic" EXCLUDED alias, and to the pre-existing tuple using the table name or its alias. This feature is often referred to as upsert. This is implemented using a new infrastructure called "speculative insertion". It is an optimistic variant of regular insertion that first does a pre-check for existing tuples and then attempts an insert. If a violating tuple was inserted concurrently, the speculatively inserted tuple is deleted and a new attempt is made. If the pre-check finds a matching tuple the alternative DO NOTHING or DO UPDATE action is taken. If the insertion succeeds without detecting a conflict, the tuple is deemed inserted. To handle the possible ambiguity between the excluded alias and a table named excluded, and for convenience with long relation names, INSERT INTO now can alias its target table. Bumps catversion as stored rules change. Author: Peter Geoghegan, with significant contributions from Heikki Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes. Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs, Dean Rasheed, Stephen Frost and many others.
2015-05-08 05:31:36 +02:00
}
}
Re-implement EvalPlanQual processing to improve its performance and eliminate a lot of strange behaviors that occurred in join cases. We now identify the "current" row for every joined relation in UPDATE, DELETE, and SELECT FOR UPDATE/SHARE queries. If an EvalPlanQual recheck is necessary, we jam the appropriate row into each scan node in the rechecking plan, forcing it to emit only that one row. The former behavior could rescan the whole of each joined relation for each recheck, which was terrible for performance, and what's much worse could result in duplicated output tuples. Also, the original implementation of EvalPlanQual could not re-use the recheck execution tree --- it had to go through a full executor init and shutdown for every row to be tested. To avoid this overhead, I've associated a special runtime Param with each LockRows or ModifyTable plan node, and arranged to make every scan node below such a node depend on that Param. Thus, by signaling a change in that Param, the EPQ machinery can just rescan the already-built test plan. This patch also adds a prohibition on set-returning functions in the targetlist of SELECT FOR UPDATE/SHARE. This is needed to avoid the duplicate-output-tuple problem. It seems fairly reasonable since the other restrictions on SELECT FOR UPDATE are meant to ensure that there is a unique correspondence between source tuples and result tuples, which an output SRF destroys as much as anything else does.
2009-10-26 03:26:45 +01:00
/*
* If we have any secondary relations in an UPDATE or DELETE, they need to
* be treated like non-locked relations in SELECT FOR UPDATE, ie, the
* EvalPlanQual mechanism needs to be told about them. Locate the
* relevant ExecRowMarks.
*/
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
arowmarks = NIL;
Re-implement EvalPlanQual processing to improve its performance and eliminate a lot of strange behaviors that occurred in join cases. We now identify the "current" row for every joined relation in UPDATE, DELETE, and SELECT FOR UPDATE/SHARE queries. If an EvalPlanQual recheck is necessary, we jam the appropriate row into each scan node in the rechecking plan, forcing it to emit only that one row. The former behavior could rescan the whole of each joined relation for each recheck, which was terrible for performance, and what's much worse could result in duplicated output tuples. Also, the original implementation of EvalPlanQual could not re-use the recheck execution tree --- it had to go through a full executor init and shutdown for every row to be tested. To avoid this overhead, I've associated a special runtime Param with each LockRows or ModifyTable plan node, and arranged to make every scan node below such a node depend on that Param. Thus, by signaling a change in that Param, the EPQ machinery can just rescan the already-built test plan. This patch also adds a prohibition on set-returning functions in the targetlist of SELECT FOR UPDATE/SHARE. This is needed to avoid the duplicate-output-tuple problem. It seems fairly reasonable since the other restrictions on SELECT FOR UPDATE are meant to ensure that there is a unique correspondence between source tuples and result tuples, which an output SRF destroys as much as anything else does.
2009-10-26 03:26:45 +01:00
foreach(l, node->rowMarks)
{
Improve castNode notation by introducing list-extraction-specific variants. This extends the castNode() notation introduced by commit 5bcab1114 to provide, in one step, extraction of a list cell's pointer and coercion to a concrete node type. For example, "lfirst_node(Foo, lc)" is the same as "castNode(Foo, lfirst(lc))". Almost half of the uses of castNode that have appeared so far include a list extraction call, so this is pretty widely useful, and it saves a few more keystrokes compared to the old way. As with the previous patch, back-patch the addition of these macros to pg_list.h, so that the notation will be available when back-patching. Patch by me, after an idea of Andrew Gierth's. Discussion: https://postgr.es/m/14197.1491841216@sss.pgh.pa.us
2017-04-10 19:51:29 +02:00
PlanRowMark *rc = lfirst_node(PlanRowMark, l);
Fix PlanRowMark/ExecRowMark structures to handle inheritance correctly. In an inherited UPDATE/DELETE, each target table has its own subplan, because it might have a column set different from other targets. This means that the resjunk columns we add to support EvalPlanQual might be at different physical column numbers in each subplan. The EvalPlanQual rewrite I did for 9.0 failed to account for this, resulting in possible misbehavior or even crashes during concurrent updates to the same row, as seen in a recent report from Gordon Shannon. Revise the data structure so that we track resjunk column numbers separately for each subplan. I also chose to move responsibility for identifying the physical column numbers back to executor startup, instead of assuming that numbers derived during preprocess_targetlist would stay valid throughout subsequent massaging of the plan. That's a bit slower, so we might want to consider undoing it someday; but it would complicate the patch considerably and didn't seem justifiable in a bug fix that has to be back-patched to 9.0.
2011-01-13 02:47:02 +01:00
ExecRowMark *erm;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
ExecAuxRowMark *aerm;
Re-implement EvalPlanQual processing to improve its performance and eliminate a lot of strange behaviors that occurred in join cases. We now identify the "current" row for every joined relation in UPDATE, DELETE, and SELECT FOR UPDATE/SHARE queries. If an EvalPlanQual recheck is necessary, we jam the appropriate row into each scan node in the rechecking plan, forcing it to emit only that one row. The former behavior could rescan the whole of each joined relation for each recheck, which was terrible for performance, and what's much worse could result in duplicated output tuples. Also, the original implementation of EvalPlanQual could not re-use the recheck execution tree --- it had to go through a full executor init and shutdown for every row to be tested. To avoid this overhead, I've associated a special runtime Param with each LockRows or ModifyTable plan node, and arranged to make every scan node below such a node depend on that Param. Thus, by signaling a change in that Param, the EPQ machinery can just rescan the already-built test plan. This patch also adds a prohibition on set-returning functions in the targetlist of SELECT FOR UPDATE/SHARE. This is needed to avoid the duplicate-output-tuple problem. It seems fairly reasonable since the other restrictions on SELECT FOR UPDATE are meant to ensure that there is a unique correspondence between source tuples and result tuples, which an output SRF destroys as much as anything else does.
2009-10-26 03:26:45 +01:00
/* ignore "parent" rowmarks; they are irrelevant at runtime */
if (rc->isParent)
continue;
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/* Find ExecRowMark and build ExecAuxRowMark */
Add support for doing late row locking in FDWs. Previously, FDWs could only do "early row locking", that is lock a row as soon as it's fetched, even though local restriction/join conditions might discard the row later. This patch adds callbacks that allow FDWs to do late locking in the same way that it's done for regular tables. To make use of this feature, an FDW must support the "ctid" column as a unique row identifier. Currently, since ctid has to be of type TID, the feature is of limited use, though in principle it could be used by postgres_fdw. We may eventually allow FDWs to specify another data type for ctid, which would make it possible for more FDWs to use this feature. This commit does not modify postgres_fdw to use late locking. We've tested some prototype code for that, but it's not in committable shape, and besides it's quite unclear whether it actually makes sense to do late locking against a remote server. The extra round trips required are likely to outweigh any benefit from improved concurrency. Etsuro Fujita, reviewed by Ashutosh Bapat, and hacked up a lot by me
2015-05-12 20:10:10 +02:00
erm = ExecFindRowMark(estate, rc->rti, false);
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
aerm = ExecBuildAuxRowMark(erm, subplan->targetlist);
arowmarks = lappend(arowmarks, aerm);
Re-implement EvalPlanQual processing to improve its performance and eliminate a lot of strange behaviors that occurred in join cases. We now identify the "current" row for every joined relation in UPDATE, DELETE, and SELECT FOR UPDATE/SHARE queries. If an EvalPlanQual recheck is necessary, we jam the appropriate row into each scan node in the rechecking plan, forcing it to emit only that one row. The former behavior could rescan the whole of each joined relation for each recheck, which was terrible for performance, and what's much worse could result in duplicated output tuples. Also, the original implementation of EvalPlanQual could not re-use the recheck execution tree --- it had to go through a full executor init and shutdown for every row to be tested. To avoid this overhead, I've associated a special runtime Param with each LockRows or ModifyTable plan node, and arranged to make every scan node below such a node depend on that Param. Thus, by signaling a change in that Param, the EPQ machinery can just rescan the already-built test plan. This patch also adds a prohibition on set-returning functions in the targetlist of SELECT FOR UPDATE/SHARE. This is needed to avoid the duplicate-output-tuple problem. It seems fairly reasonable since the other restrictions on SELECT FOR UPDATE are meant to ensure that there is a unique correspondence between source tuples and result tuples, which an output SRF destroys as much as anything else does.
2009-10-26 03:26:45 +01:00
}
Add support for MERGE SQL command MERGE performs actions that modify rows in the target table using a source table or query. MERGE provides a single SQL statement that can conditionally INSERT/UPDATE/DELETE rows -- a task that would otherwise require multiple PL statements. For example, MERGE INTO target AS t USING source AS s ON t.tid = s.sid WHEN MATCHED AND t.balance > s.delta THEN UPDATE SET balance = t.balance - s.delta WHEN MATCHED THEN DELETE WHEN NOT MATCHED AND s.delta > 0 THEN INSERT VALUES (s.sid, s.delta) WHEN NOT MATCHED THEN DO NOTHING; MERGE works with regular tables, partitioned tables and inheritance hierarchies, including column and row security enforcement, as well as support for row and statement triggers and transition tables therein. MERGE is optimized for OLTP and is parameterizable, though also useful for large scale ETL/ELT. MERGE is not intended to be used in preference to existing single SQL commands for INSERT, UPDATE or DELETE since there is some overhead. MERGE can be used from PL/pgSQL. MERGE does not support targetting updatable views or foreign tables, and RETURNING clauses are not allowed either. These limitations are likely fixable with sufficient effort. Rewrite rules are also not supported, but it's not clear that we'd want to support them. Author: Pavan Deolasee <pavan.deolasee@gmail.com> Author: Álvaro Herrera <alvherre@alvh.no-ip.org> Author: Amit Langote <amitlangote09@gmail.com> Author: Simon Riggs <simon.riggs@enterprisedb.com> Reviewed-by: Peter Eisentraut <peter.eisentraut@enterprisedb.com> Reviewed-by: Andres Freund <andres@anarazel.de> (earlier versions) Reviewed-by: Peter Geoghegan <pg@bowt.ie> (earlier versions) Reviewed-by: Robert Haas <robertmhaas@gmail.com> (earlier versions) Reviewed-by: Japin Li <japinli@hotmail.com> Reviewed-by: Justin Pryzby <pryzby@telsasoft.com> Reviewed-by: Tomas Vondra <tomas.vondra@enterprisedb.com> Reviewed-by: Zhihong Yu <zyu@yugabyte.com> Discussion: https://postgr.es/m/CANP8+jKitBSrB7oTgT9CY2i1ObfOt36z0XMraQc+Xrz8QB0nXA@mail.gmail.com Discussion: https://postgr.es/m/CAH2-WzkJdBuxj9PO=2QaO9-3h3xGbQPZ34kJH=HukRekwM-GZg@mail.gmail.com Discussion: https://postgr.es/m/20201231134736.GA25392@alvherre.pgsql
2022-03-28 16:45:58 +02:00
/* For a MERGE command, initialize its state */
if (mtstate->operation == CMD_MERGE)
ExecInitMerge(mtstate, estate);
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
EvalPlanQualSetPlan(&mtstate->mt_epqstate, subplan, arowmarks);
Fix PlanRowMark/ExecRowMark structures to handle inheritance correctly. In an inherited UPDATE/DELETE, each target table has its own subplan, because it might have a column set different from other targets. This means that the resjunk columns we add to support EvalPlanQual might be at different physical column numbers in each subplan. The EvalPlanQual rewrite I did for 9.0 failed to account for this, resulting in possible misbehavior or even crashes during concurrent updates to the same row, as seen in a recent report from Gordon Shannon. Revise the data structure so that we track resjunk column numbers separately for each subplan. I also chose to move responsibility for identifying the physical column numbers back to executor startup, instead of assuming that numbers derived during preprocess_targetlist would stay valid throughout subsequent massaging of the plan. That's a bit slower, so we might want to consider undoing it someday; but it would complicate the patch considerably and didn't seem justifiable in a bug fix that has to be back-patched to 9.0.
2011-01-13 02:47:02 +01:00
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
/*
* If there are a lot of result relations, use a hash table to speed the
* lookups. If there are not a lot, a simple linear search is faster.
*
* It's not clear where the threshold is, but try 64 for starters. In a
* debugging build, use a small threshold so that we get some test
* coverage of both code paths.
*/
#ifdef USE_ASSERT_CHECKING
#define MT_NRELS_HASH 4
#else
#define MT_NRELS_HASH 64
#endif
if (nrels >= MT_NRELS_HASH)
{
HASHCTL hash_ctl;
hash_ctl.keysize = sizeof(Oid);
hash_ctl.entrysize = sizeof(MTTargetRelLookup);
hash_ctl.hcxt = CurrentMemoryContext;
mtstate->mt_resultOidHash =
hash_create("ModifyTable target hash",
nrels, &hash_ctl,
HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);
for (i = 0; i < nrels; i++)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
Oid hashkey;
MTTargetRelLookup *mtlookup;
bool found;
resultRelInfo = &mtstate->resultRelInfo[i];
hashkey = RelationGetRelid(resultRelInfo->ri_RelationDesc);
mtlookup = (MTTargetRelLookup *)
hash_search(mtstate->mt_resultOidHash, &hashkey,
HASH_ENTER, &found);
Assert(!found);
mtlookup->relationIndex = i;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
}
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
else
mtstate->mt_resultOidHash = NULL;
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
/*
* Determine if the FDW supports batch insert and determine the batch size
* (a FDW may support batching, but it may be disabled for the
* server/table).
Fix initialization of FDW batching in ExecInitModifyTable ExecInitModifyTable has to initialize batching for all result relations, not just the first one. Furthermore, when junk filters were necessary, the pointer pointed past the mtstate->resultRelInfo array. Per reports from multiple non-x86 animals (florican, locust, ...). Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-21 03:23:24 +01:00
*
* We only do this for INSERT, so that for UPDATE/DELETE the batch size
* remains set to 0.
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
*/
Fix initialization of FDW batching in ExecInitModifyTable ExecInitModifyTable has to initialize batching for all result relations, not just the first one. Furthermore, when junk filters were necessary, the pointer pointed past the mtstate->resultRelInfo array. Per reports from multiple non-x86 animals (florican, locust, ...). Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-21 03:23:24 +01:00
if (operation == CMD_INSERT)
{
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
/* insert may only have one relation, inheritance is not expanded */
Assert(nrels == 1);
Fix initialization of FDW batching in ExecInitModifyTable ExecInitModifyTable has to initialize batching for all result relations, not just the first one. Furthermore, when junk filters were necessary, the pointer pointed past the mtstate->resultRelInfo array. Per reports from multiple non-x86 animals (florican, locust, ...). Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-21 03:23:24 +01:00
resultRelInfo = mtstate->resultRelInfo;
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
if (!resultRelInfo->ri_usesFdwDirectModify &&
resultRelInfo->ri_FdwRoutine != NULL &&
resultRelInfo->ri_FdwRoutine->GetForeignModifyBatchSize &&
resultRelInfo->ri_FdwRoutine->ExecForeignBatchInsert)
Fix initialization of FDW batching in ExecInitModifyTable ExecInitModifyTable has to initialize batching for all result relations, not just the first one. Furthermore, when junk filters were necessary, the pointer pointed past the mtstate->resultRelInfo array. Per reports from multiple non-x86 animals (florican, locust, ...). Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-21 03:23:24 +01:00
{
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
resultRelInfo->ri_BatchSize =
resultRelInfo->ri_FdwRoutine->GetForeignModifyBatchSize(resultRelInfo);
Fix initialization of FDW batching in ExecInitModifyTable ExecInitModifyTable has to initialize batching for all result relations, not just the first one. Furthermore, when junk filters were necessary, the pointer pointed past the mtstate->resultRelInfo array. Per reports from multiple non-x86 animals (florican, locust, ...). Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-21 03:23:24 +01:00
Assert(resultRelInfo->ri_BatchSize >= 1);
}
Postpone some stuff out of ExecInitModifyTable. Arrange to do some things on-demand, rather than immediately during executor startup, because there's a fair chance of never having to do them at all: * Don't open result relations' indexes until needed. * Don't initialize partition tuple routing, nor the child-to-root tuple conversion map, until needed. This wins in UPDATEs on partitioned tables when only some of the partitions will actually receive updates; with larger partition counts the savings is quite noticeable. Also, we can remove some sketchy heuristics in ExecInitModifyTable about whether to set up tuple routing. Also, remove execPartition.c's private hash table tracking which partitions were already opened by the ModifyTable node. Instead use the hash added to ModifyTable itself by commit 86dc90056. To allow lazy computation of the conversion maps, we now set ri_RootResultRelInfo in all child ResultRelInfos. We formerly set it only in some, not terribly well-defined, cases. This has user-visible side effects in that now more error messages refer to the root relation instead of some partition (and provide error data in the root's column order, too). It looks to me like this is a strict improvement in consistency, so I don't have a problem with the output changes visible in this commit. Extracted from a larger patch, which seemed to me to be too messy to push in one commit. Amit Langote, reviewed at different times by Heikki Linnakangas and myself Discussion: https://postgr.es/m/CA+HiwqG7ZruBmmih3wPsBZ4s0H2EhywrnXEduckY5Hr3fWzPWA@mail.gmail.com
2021-04-06 21:56:55 +02:00
else
resultRelInfo->ri_BatchSize = 1;
Fix initialization of FDW batching in ExecInitModifyTable ExecInitModifyTable has to initialize batching for all result relations, not just the first one. Furthermore, when junk filters were necessary, the pointer pointed past the mtstate->resultRelInfo array. Per reports from multiple non-x86 animals (florican, locust, ...). Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-21 03:23:24 +01:00
}
Implement support for bulk inserts in postgres_fdw Extends the FDW API to allow batching inserts into foreign tables. That is usually much more efficient than inserting individual rows, due to high latency for each round-trip to the foreign server. It was possible to implement something similar in the regular FDW API, but it was inconvenient and there were issues with reporting the number of actually inserted rows etc. This extends the FDW API with two new functions: * GetForeignModifyBatchSize - allows the FDW picking optimal batch size * ExecForeignBatchInsert - inserts a batch of rows at once Currently, only INSERT queries support batching. Support for DELETE and UPDATE may be added in the future. This also implements batching for postgres_fdw. The batch size may be specified using "batch_size" option both at the server and table level. The initial patch version was written by me, but it was rewritten and improved in many ways by Takayuki Tsunakawa. Author: Takayuki Tsunakawa Reviewed-by: Tomas Vondra, Amit Langote Discussion: https://postgr.es/m/20200628151002.7x5laxwpgvkyiu3q@development
2021-01-20 23:05:46 +01:00
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
/*
* Lastly, if this is not the primary (canSetTag) ModifyTable node, add it
* to estate->es_auxmodifytables so that it will be run to completion by
* ExecPostprocessPlan. (It'd actually work fine to add the primary
Fix order of shutdown processing when CTEs contain inter-references. We need ExecutorEnd to run the ModifyTable nodes to completion in reverse order of initialization, not forward order. Easily done by constructing the list back-to-front.
2011-02-26 05:53:34 +01:00
* ModifyTable node too, but there's no need.) Note the use of lcons not
* lappend: we need later-initialized ModifyTable nodes to be shut down
* before earlier ones. This ensures that we don't throw away RETURNING
* rows that need to be seen by a later CTE subplan.
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
*/
if (!mtstate->canSetTag)
Fix order of shutdown processing when CTEs contain inter-references. We need ExecutorEnd to run the ModifyTable nodes to completion in reverse order of initialization, not forward order. Easily done by constructing the list back-to-front.
2011-02-26 05:53:34 +01:00
estate->es_auxmodifytables = lcons(mtstate,
estate->es_auxmodifytables);
Support data-modifying commands (INSERT/UPDATE/DELETE) in WITH. This patch implements data-modifying WITH queries according to the semantics that the updates all happen with the same command counter value, and in an unspecified order. Therefore one WITH clause can't see the effects of another, nor can the outer query see the effects other than through the RETURNING values. And attempts to do conflicting updates will have unpredictable results. We'll need to document all that. This commit just fixes the code; documentation updates are waiting on author. Marko Tiikkaja and Hitoshi Harada
2011-02-26 00:56:23 +01:00
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
return mtstate;
}
/* ----------------------------------------------------------------
* ExecEndModifyTable
*
* Shuts down the plan.
*
* Returns nothing of interest.
* ----------------------------------------------------------------
*/
void
ExecEndModifyTable(ModifyTableState *node)
{
int i;
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
/*
* Allow any FDWs to shut down
*/
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
for (i = 0; i < node->mt_nrels; i++)
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
{
Optimize creation of slots for FDW bulk inserts Commit b663a41363 introduced bulk inserts for FDW, but the handling of tuple slots turned out to be problematic for two reasons. Firstly, the slots were re-created for each individual batch. Secondly, all slots referenced the same tuple descriptor - with reasonably small batches this is not an issue, but with large batches this triggers O(N^2) behavior in the resource owner code. These two issues work against each other - to reduce the number of times a slot has to be created/dropped, larger batches are needed. However, the larger the batch, the more expensive the resource owner gets. For practical batch sizes (100 - 1000) this would not be a big problem, as the benefits (latency savings) greatly exceed the resource owner costs. But for extremely large batches it might be much worse, possibly even losing with non-batching mode. Fixed by initializing tuple slots only once (and reusing them across batches) and by using a new tuple descriptor copy for each slot. Discussion: https://postgr.es/m/ebbbcc7d-4286-8c28-0272-61b4753af761%40enterprisedb.com
2021-06-11 20:19:48 +02:00
int j;
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
ResultRelInfo *resultRelInfo = node->resultRelInfo + i;
Directly modify foreign tables. postgres_fdw can now sent an UPDATE or DELETE statement directly to the foreign server in simple cases, rather than sending a SELECT FOR UPDATE statement and then updating or deleting rows one-by-one. Etsuro Fujita, reviewed by Rushabh Lathia, Shigeru Hanada, Kyotaro Horiguchi, Albe Laurenz, Thom Brown, and me.
2016-03-18 18:48:58 +01:00
if (!resultRelInfo->ri_usesFdwDirectModify &&
resultRelInfo->ri_FdwRoutine != NULL &&
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
resultRelInfo->ri_FdwRoutine->EndForeignModify != NULL)
resultRelInfo->ri_FdwRoutine->EndForeignModify(node->ps.state,
resultRelInfo);
Optimize creation of slots for FDW bulk inserts Commit b663a41363 introduced bulk inserts for FDW, but the handling of tuple slots turned out to be problematic for two reasons. Firstly, the slots were re-created for each individual batch. Secondly, all slots referenced the same tuple descriptor - with reasonably small batches this is not an issue, but with large batches this triggers O(N^2) behavior in the resource owner code. These two issues work against each other - to reduce the number of times a slot has to be created/dropped, larger batches are needed. However, the larger the batch, the more expensive the resource owner gets. For practical batch sizes (100 - 1000) this would not be a big problem, as the benefits (latency savings) greatly exceed the resource owner costs. But for extremely large batches it might be much worse, possibly even losing with non-batching mode. Fixed by initializing tuple slots only once (and reusing them across batches) and by using a new tuple descriptor copy for each slot. Discussion: https://postgr.es/m/ebbbcc7d-4286-8c28-0272-61b4753af761%40enterprisedb.com
2021-06-11 20:19:48 +02:00
/*
* Cleanup the initialized batch slots. This only matters for FDWs
* with batching, but the other cases will have ri_NumSlotsInitialized
* == 0.
*/
for (j = 0; j < resultRelInfo->ri_NumSlotsInitialized; j++)
{
ExecDropSingleTupleTableSlot(resultRelInfo->ri_Slots[j]);
ExecDropSingleTupleTableSlot(resultRelInfo->ri_PlanSlots[j]);
}
Support writable foreign tables. This patch adds the core-system infrastructure needed to support updates on foreign tables, and extends contrib/postgres_fdw to allow updates against remote Postgres servers. There's still a great deal of room for improvement in optimization of remote updates, but at least there's basic functionality there now. KaiGai Kohei, reviewed by Alexander Korotkov and Laurenz Albe, and rather heavily revised by Tom Lane.
2013-03-10 19:14:53 +01:00
}
Redesign initialization of partition routing structures This speeds up write operations (INSERT, UPDATE, DELETE, COPY, as well as the future MERGE) on partitioned tables. This changes the setup for tuple routing so that it does far less work during the initial setup and pushes more work out to when partitions receive tuples. PartitionDispatchData structs for sub-partitioned tables are only created when a tuple gets routed through it. The possibly large arrays in the PartitionTupleRouting struct have largely been removed. The partitions[] array remains but now never contains any NULL gaps. Previously the NULLs had to be skipped during ExecCleanupTupleRouting(), which could add a large overhead to the cleanup when the number of partitions was large. The partitions[] array is allocated small to start with and only enlarged when we route tuples to enough partitions that it runs out of space. This allows us to keep simple single-row partition INSERTs running quickly. Redesign The arrays in PartitionTupleRouting which stored the tuple translation maps have now been removed. These have been moved out into a PartitionRoutingInfo struct which is an additional field in ResultRelInfo. The find_all_inheritors() call still remains by far the slowest part of ExecSetupPartitionTupleRouting(). This commit just removes the other slow parts. In passing also rename the tuple translation maps from being ParentToChild and ChildToParent to being RootToPartition and PartitionToRoot. The old names mislead you into thinking that a partition of some sub-partitioned table would translate to the rowtype of the sub-partitioned table rather than the root partitioned table. Authors: David Rowley and Amit Langote, heavily revised by Álvaro Herrera Testing help from Jesper Pedersen and Kato Sho. Discussion: https://postgr.es/m/CAKJS1f_1RJyFquuCKRFHTdcXqoPX-PYqAd7nz=GVBwvGh4a6xA@mail.gmail.com
2018-11-16 18:54:15 +01:00
/*
* Close all the partitioned tables, leaf partitions, and their indices
* and release the slot used for tuple routing, if set.
*/
Simplify and encapsulate tuple routing support code. Instead of having ExecSetupPartitionTupleRouting return multiple out parameters, have it return a pointer to a structure containing all of those different things. Also, provide and use a cleanup function, ExecCleanupTupleRouting, instead of cleaning up all of the resources allocated by ExecSetupPartitionTupleRouting individually. Amit Khandekar, reviewed by Amit Langote, David Rowley, and me Discussion: http://postgr.es/m/CAJ3gD9fWfxgKC+PfJZF3hkgAcNOy-LpfPxVYitDEXKHjeieWQQ@mail.gmail.com
2018-01-04 21:48:15 +01:00
if (node->mt_partition_tuple_routing)
Redesign initialization of partition routing structures This speeds up write operations (INSERT, UPDATE, DELETE, COPY, as well as the future MERGE) on partitioned tables. This changes the setup for tuple routing so that it does far less work during the initial setup and pushes more work out to when partitions receive tuples. PartitionDispatchData structs for sub-partitioned tables are only created when a tuple gets routed through it. The possibly large arrays in the PartitionTupleRouting struct have largely been removed. The partitions[] array remains but now never contains any NULL gaps. Previously the NULLs had to be skipped during ExecCleanupTupleRouting(), which could add a large overhead to the cleanup when the number of partitions was large. The partitions[] array is allocated small to start with and only enlarged when we route tuples to enough partitions that it runs out of space. This allows us to keep simple single-row partition INSERTs running quickly. Redesign The arrays in PartitionTupleRouting which stored the tuple translation maps have now been removed. These have been moved out into a PartitionRoutingInfo struct which is an additional field in ResultRelInfo. The find_all_inheritors() call still remains by far the slowest part of ExecSetupPartitionTupleRouting(). This commit just removes the other slow parts. In passing also rename the tuple translation maps from being ParentToChild and ChildToParent to being RootToPartition and PartitionToRoot. The old names mislead you into thinking that a partition of some sub-partitioned table would translate to the rowtype of the sub-partitioned table rather than the root partitioned table. Authors: David Rowley and Amit Langote, heavily revised by Álvaro Herrera Testing help from Jesper Pedersen and Kato Sho. Discussion: https://postgr.es/m/CAKJS1f_1RJyFquuCKRFHTdcXqoPX-PYqAd7nz=GVBwvGh4a6xA@mail.gmail.com
2018-11-16 18:54:15 +01:00
{
Allow insert and update tuple routing and COPY for foreign tables. Also enable this for postgres_fdw. Etsuro Fujita, based on an earlier patch by Amit Langote. The larger patch series of which this is a part has been reviewed by Amit Langote, David Fetter, Maksim Milyutin, Álvaro Herrera, Stephen Frost, and me. Minor documentation changes to the final version by me. Discussion: http://postgr.es/m/29906a26-da12-8c86-4fb9-d8f88442f2b9@lab.ntt.co.jp
2018-04-07 01:16:11 +02:00
ExecCleanupTupleRouting(node, node->mt_partition_tuple_routing);
Move partition_tuple_slot out of EState. Commit 2ac3ef7a01df859c62d0a02333b646d65eaec5ff added a TupleTapleSlot for partition tuple slot to EState (es_partition_tuple_slot) but it's more logical to have it as part of ModifyTableState (mt_partition_tuple_slot) and CopyState (partition_tuple_slot). Discussion: http://postgr.es/m/1bd459d9-4c0c-197a-346e-e5e59e217d97@lab.ntt.co.jp Amit Langote, per a gripe from me
2017-01-04 19:05:29 +01:00
Redesign initialization of partition routing structures This speeds up write operations (INSERT, UPDATE, DELETE, COPY, as well as the future MERGE) on partitioned tables. This changes the setup for tuple routing so that it does far less work during the initial setup and pushes more work out to when partitions receive tuples. PartitionDispatchData structs for sub-partitioned tables are only created when a tuple gets routed through it. The possibly large arrays in the PartitionTupleRouting struct have largely been removed. The partitions[] array remains but now never contains any NULL gaps. Previously the NULLs had to be skipped during ExecCleanupTupleRouting(), which could add a large overhead to the cleanup when the number of partitions was large. The partitions[] array is allocated small to start with and only enlarged when we route tuples to enough partitions that it runs out of space. This allows us to keep simple single-row partition INSERTs running quickly. Redesign The arrays in PartitionTupleRouting which stored the tuple translation maps have now been removed. These have been moved out into a PartitionRoutingInfo struct which is an additional field in ResultRelInfo. The find_all_inheritors() call still remains by far the slowest part of ExecSetupPartitionTupleRouting(). This commit just removes the other slow parts. In passing also rename the tuple translation maps from being ParentToChild and ChildToParent to being RootToPartition and PartitionToRoot. The old names mislead you into thinking that a partition of some sub-partitioned table would translate to the rowtype of the sub-partitioned table rather than the root partitioned table. Authors: David Rowley and Amit Langote, heavily revised by Álvaro Herrera Testing help from Jesper Pedersen and Kato Sho. Discussion: https://postgr.es/m/CAKJS1f_1RJyFquuCKRFHTdcXqoPX-PYqAd7nz=GVBwvGh4a6xA@mail.gmail.com
2018-11-16 18:54:15 +01:00
if (node->mt_root_tuple_slot)
ExecDropSingleTupleTableSlot(node->mt_root_tuple_slot);
}
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
* Free the exprcontext
*/
ExecFreeExprContext(&node->ps);
/*
* clean out the tuple table
*/
Don't require return slots for nodes without projection. In a lot of nodes the return slot is not required. That can either be because the node doesn't do any projection (say an Append node), or because the node does perform projections but the projection is optimized away because the projection would yield an identical row. Slots aren't that small, especially for wide rows, so it's worthwhile to avoid creating them. It's not possible to just skip creating the slot - it's currently used to determine the tuple descriptor returned by ExecGetResultType(). So separate the determination of the result type from the slot creation. The work previously done internally ExecInitResultTupleSlotTL() can now also be done separately with ExecInitResultTypeTL() and ExecInitResultSlot(). That way nodes that aren't guaranteed to need a result slot, can use ExecInitResultTypeTL() to determine the result type of the node, and ExecAssignScanProjectionInfo() (via ExecConditionalAssignProjectionInfo()) determines that a result slot is needed, it is created with ExecInitResultSlot(). Besides the advantage of avoiding to create slots that then are unused, this is necessary preparation for later patches around tuple table slot abstraction. In particular separating the return descriptor and slot is a prerequisite to allow JITing of tuple deforming with knowledge of the underlying tuple format, and to avoid unnecessarily creating JITed tuple deforming for virtual slots. This commit removes a redundant argument from ExecInitResultTupleSlotTL(). While this commit touches a lot of the relevant lines anyway, it'd normally still not worthwhile to cause breakage, except that aforementioned later commits will touch *all* ExecInitResultTupleSlotTL() callers anyway (but fits worse thematically). Author: Andres Freund Discussion: https://postgr.es/m/20181105210039.hh4vvi4vwoq5ba2q@alap3.anarazel.de
2018-11-10 02:19:39 +01:00
if (node->ps.ps_ResultTupleSlot)
ExecClearTuple(node->ps.ps_ResultTupleSlot);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
Re-implement EvalPlanQual processing to improve its performance and eliminate a lot of strange behaviors that occurred in join cases. We now identify the "current" row for every joined relation in UPDATE, DELETE, and SELECT FOR UPDATE/SHARE queries. If an EvalPlanQual recheck is necessary, we jam the appropriate row into each scan node in the rechecking plan, forcing it to emit only that one row. The former behavior could rescan the whole of each joined relation for each recheck, which was terrible for performance, and what's much worse could result in duplicated output tuples. Also, the original implementation of EvalPlanQual could not re-use the recheck execution tree --- it had to go through a full executor init and shutdown for every row to be tested. To avoid this overhead, I've associated a special runtime Param with each LockRows or ModifyTable plan node, and arranged to make every scan node below such a node depend on that Param. Thus, by signaling a change in that Param, the EPQ machinery can just rescan the already-built test plan. This patch also adds a prohibition on set-returning functions in the targetlist of SELECT FOR UPDATE/SHARE. This is needed to avoid the duplicate-output-tuple problem. It seems fairly reasonable since the other restrictions on SELECT FOR UPDATE are meant to ensure that there is a unique correspondence between source tuples and result tuples, which an output SRF destroys as much as anything else does.
2009-10-26 03:26:45 +01:00
/*
* Terminate EPQ execution if active
*/
EvalPlanQualEnd(&node->mt_epqstate);
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
/*
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
* shut down subplan
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
*/
Rework planning and execution of UPDATE and DELETE. This patch makes two closely related sets of changes: 1. For UPDATE, the subplan of the ModifyTable node now only delivers the new values of the changed columns (i.e., the expressions computed in the query's SET clause) plus row identity information such as CTID. ModifyTable must re-fetch the original tuple to merge in the old values of any unchanged columns. The core advantage of this is that the changed columns are uniform across all tables of an inherited or partitioned target relation, whereas the other columns might not be. A secondary advantage, when the UPDATE involves joins, is that less data needs to pass through the plan tree. The disadvantage of course is an extra fetch of each tuple to be updated. However, that seems to be very nearly free in context; even worst-case tests don't show it to add more than a couple percent to the total query cost. At some point it might be interesting to combine the re-fetch with the tuple access that ModifyTable must do anyway to mark the old tuple dead; but that would require a good deal of refactoring and it seems it wouldn't buy all that much, so this patch doesn't attempt it. 2. For inherited UPDATE/DELETE, instead of generating a separate subplan for each target relation, we now generate a single subplan that is just exactly like a SELECT's plan, then stick ModifyTable on top of that. To let ModifyTable know which target relation a given incoming row refers to, a tableoid junk column is added to the row identity information. This gets rid of the horrid hack that was inheritance_planner(), eliminating O(N^2) planning cost and memory consumption in cases where there were many unprunable target relations. Point 2 of course requires point 1, so that there is a uniform definition of the non-junk columns to be returned by the subplan. We can't insist on uniform definition of the row identity junk columns however, if we want to keep the ability to have both plain and foreign tables in a partitioning hierarchy. Since it wouldn't scale very far to have every child table have its own row identity column, this patch includes provisions to merge similar row identity columns into one column of the subplan result. In particular, we can merge the whole-row Vars typically used as row identity by FDWs into one column by pretending they are type RECORD. (It's still okay for the actual composite Datums to be labeled with the table's rowtype OID, though.) There is more that can be done to file down residual inefficiencies in this patch, but it seems to be committable now. FDW authors should note several API changes: * The argument list for AddForeignUpdateTargets() has changed, and so has the method it must use for adding junk columns to the query. Call add_row_identity_var() instead of manipulating the parse tree directly. You might want to reconsider exactly what you're adding, too. * PlanDirectModify() must now work a little harder to find the ForeignScan plan node; if the foreign table is part of a partitioning hierarchy then the ForeignScan might not be the direct child of ModifyTable. See postgres_fdw for sample code. * To check whether a relation is a target relation, it's no longer sufficient to compare its relid to root->parse->resultRelation. Instead, check it against all_result_relids or leaf_result_relids, as appropriate. Amit Langote and Tom Lane Discussion: https://postgr.es/m/CA+HiwqHpHdqdDn48yCEhynnniahH78rwcrv1rEX65-fsZGBOLQ@mail.gmail.com
2021-03-31 17:52:34 +02:00
ExecEndNode(outerPlanState(node));
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
}
void
Make NestLoop plan nodes pass outer-relation variables into their inner relation using the general PARAM_EXEC executor parameter mechanism, rather than the ad-hoc kluge of passing the outer tuple down through ExecReScan. The previous method was hard to understand and could never be extended to handle parameters coming from multiple join levels. This patch doesn't change the set of possible plans nor have any significant performance effect, but it's necessary infrastructure for future generalization of the concept of an inner indexscan plan. ExecReScan's second parameter is now unused, so it's removed.
2010-07-12 19:01:06 +02:00
ExecReScanModifyTable(ModifyTableState *node)
Split the processing of INSERT/UPDATE/DELETE operations out of execMain.c. They are now handled by a new plan node type called ModifyTable, which is placed at the top of the plan tree. In itself this change doesn't do much, except perhaps make the handling of RETURNING lists and inherited UPDATEs a tad less klugy. But it is necessary preparation for the intended extension of allowing RETURNING queries inside WITH. Marko Tiikkaja
2009-10-10 03:43:50 +02:00
{
/*
* Currently, we don't need to support rescan on ModifyTable nodes. The
* semantics of that would be a bit debatable anyway.
*/
elog(ERROR, "ExecReScanModifyTable is not implemented");
}