rdelaage/postgresql
1
0
Fork 0
mirror of https://git.postgresql.org/git/postgresql.git synced 2024-09-07 23:09:22 +02:00
Code Issues Packages Projects Releases Wiki Activity
1e4c050b24
postgresql/src/backend/utils/cache/relcache.c

4611 lines
140 KiB
C
Raw Normal View History

Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
/*-------------------------------------------------------------------------
*
Change my-function-name-- to my_function_name, and optimizer renames.
1999-02-14 00:22:53 +01:00
* relcache.c
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* POSTGRES relation descriptor cache code
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Update copyright for the year 2010.
2010-01-02 17:58:17 +01:00
* Portions Copyright (c) 1996-2010, PostgreSQL Global Development Group
Add: * Portions Copyright (c) 1996-2000, PostgreSQL, Inc to all files copyright Regents of Berkeley. Man, that's a lot of files.
2000-01-26 06:58:53 +01:00
* Portions Copyright (c) 1994, Regents of the University of California
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
*
* IDENTIFICATION
pgindent run for 9.0, second run
2010-07-06 21:19:02 +02:00
* $PostgreSQL: pgsql/src/backend/utils/cache/relcache.c,v 1.311 2010/07/06 19:18:58 momjian Exp $
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
*-------------------------------------------------------------------------
*/
/*
* INTERFACE ROUTINES
Further minor simplification of relcache startup: don't need a static needNewCacheFile flag anymore, it can just be local in RelationCacheInitializePhase2.
2006-05-06 17:51:07 +02:00
* RelationCacheInitialize - initialize relcache (to empty)
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* RelationCacheInitializePhase2 - initialize shared-catalog entries
* RelationCacheInitializePhase3 - finish initializing relcache
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* RelationIdGetRelation - get a reldesc by relation id
* RelationClose - close an open relation
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
* NOTES
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* The following code contains many undocumented hacks. Please be
* careful....
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Clean up syscache so that recursive invocation is safe, and remove error message about recursive use of a syscache. Also remove most of the specialized indexscan routines in indexing.c --- it turns out that catcache.c is perfectly able to perform the indexscan for itself, in fact has already looked up all the information needed to do so! This should be faster as well as needing far less boilerplate code.
2000-11-10 01:33:12 +01:00
#include "postgres.h"
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
#include <sys/file.h>
More stuff to make it compile on various ports.
1996-11-27 09:16:44 +01:00
#include <fcntl.h>
Fix low-probability bug in relcache startup: write_irels wrote the pg_internal.init file in-place, which meant that if another backend started at about the same time, it might read the incomplete file. init_irels tries to guard against that, but I have now seen a crash due to reading bad data from a partly-written file. (This may indicate a kernel bug on my platform? Not sure.) Anyway, clearly the safest course is to write the new pg_internal.init file under a unique temporary filename, and rename it into place only after it's all written.
2000-03-31 21:39:22 +02:00
#include <unistd.h>
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
#include "access/genam.h"
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
#include "access/reloptions.h"
Restructure some header files a bit, in particular heapam.h, by removing some unnecessary #include lines in it. Also, move some tuple routine prototypes and macros to htup.h, which allows removal of heapam.h inclusion from some .c files. For this to work, a new header file access/sysattr.h needed to be created, initially containing attribute numbers of system columns, for pg_dump usage. While at it, make contrib ltree, intarray and hstore header files more consistent with our header style.
2008-05-12 02:00:54 +02:00
#include "access/sysattr.h"
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
#include "access/transam.h"
Allow include files to compile own their own. Strip unused include files out unused include files, and add needed includes to C files. The next step is to remove unused include files in C files.
2006-07-13 18:49:20 +02:00
#include "access/xact.h"
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
#include "catalog/catalog.h"
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
#include "catalog/index.h"
This patch... 1. Removes the unnecessary "#define AbcRegProcedure 123"'s from pg_proc.h. 2. Changes those #defines to use the names already defined in fmgr.h. 3. Forces the make of fmgr.h in backend/Makefile instead of having it made as a dependency in access/common/Makefile *hack*hack*hack* 4. Rearranged the #includes to a less helter-skelter arrangement, also changing <file.h> to "file.h" to signify a non-system header. 5. Removed "pg_proc.h" from files where its only purpose was for the #defines removed in item #1. 6. Added "fmgr.h" to each file changed for completeness sake. Turns out that #6 was not necessary for some files because fmgr.h was being included in a roundabout way SIX levels deep by the first include. "access/genam.h" ->"access/relscan.h" ->"utils/rel.h" ->"access/strat.h" ->"access/skey.h" ->"fmgr.h" So adding fmgr.h really didn't add anything to the compile, hopefully just made it clearer to the programmer. S Darren.
1998-04-27 06:08:07 +02:00
#include "catalog/indexing.h"
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
#include "catalog/namespace.h"
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
#include "catalog/pg_amop.h"
#include "catalog/pg_amproc.h"
This patch... 1. Removes the unnecessary "#define AbcRegProcedure 123"'s from pg_proc.h. 2. Changes those #defines to use the names already defined in fmgr.h. 3. Forces the make of fmgr.h in backend/Makefile instead of having it made as a dependency in access/common/Makefile *hack*hack*hack* 4. Rearranged the #includes to a less helter-skelter arrangement, also changing <file.h> to "file.h" to signify a non-system header. 5. Removed "pg_proc.h" from files where its only purpose was for the #defines removed in item #1. 6. Added "fmgr.h" to each file changed for completeness sake. Turns out that #6 was not necessary for some files because fmgr.h was being included in a roundabout way SIX levels deep by the first include. "access/genam.h" ->"access/relscan.h" ->"utils/rel.h" ->"access/strat.h" ->"access/skey.h" ->"fmgr.h" So adding fmgr.h really didn't add anything to the compile, hopefully just made it clearer to the programmer. S Darren.
1998-04-27 06:08:07 +02:00
#include "catalog/pg_attrdef.h"
Arrange for indexes and toast tables to inherit their ownership from the parent table, even if the command that creates them is executed by someone else (such as a superuser or a member of the owning role). Per gripe from Michael Fuhr.
2005-08-26 05:08:15 +02:00
#include "catalog/pg_authid.h"
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
#include "catalog/pg_auth_members.h"
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
#include "catalog/pg_constraint.h"
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
#include "catalog/pg_database.h"
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
#include "catalog/pg_namespace.h"
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
#include "catalog/pg_opclass.h"
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
#include "catalog/pg_operator.h"
Final cleanup
1999-07-16 07:23:30 +02:00
#include "catalog/pg_proc.h"
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
#include "catalog/pg_rewrite.h"
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
#include "catalog/pg_tablespace.h"
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
#include "catalog/pg_trigger.h"
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
#include "catalog/pg_type.h"
Get rid of the need for manual maintenance of the initial contents of pg_attribute, by having genbki.pl derive the information from the various catalog header files. This greatly simplifies modification of the "bootstrapped" catalogs. This patch finally kills genbki.sh and Gen_fmgrtab.sh; we now rely entirely on Perl scripts for those build steps. To avoid creating a Perl build dependency where there was not one before, the output files generated by these scripts are now treated as distprep targets, ie, they will be built and shipped in tarballs. But you will need a reasonably modern Perl (probably at least 5.6) if you want to build from a CVS pull. The changes to the MSVC build process are untested, and may well break --- we'll soon find out from the buildfarm. John Naylor, based on ideas from Robert Haas and others
2010-01-05 02:06:57 +01:00
#include "catalog/schemapg.h"
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
#include "catalog/storage.h"
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
#include "commands/trigger.h"
This patch... 1. Removes the unnecessary "#define AbcRegProcedure 123"'s from pg_proc.h. 2. Changes those #defines to use the names already defined in fmgr.h. 3. Forces the make of fmgr.h in backend/Makefile instead of having it made as a dependency in access/common/Makefile *hack*hack*hack* 4. Rearranged the #includes to a less helter-skelter arrangement, also changing <file.h> to "file.h" to signify a non-system header. 5. Removed "pg_proc.h" from files where its only purpose was for the #defines removed in item #1. 6. Added "fmgr.h" to each file changed for completeness sake. Turns out that #6 was not necessary for some files because fmgr.h was being included in a roundabout way SIX levels deep by the first include. "access/genam.h" ->"access/relscan.h" ->"utils/rel.h" ->"access/strat.h" ->"access/skey.h" ->"fmgr.h" So adding fmgr.h really didn't add anything to the compile, hopefully just made it clearer to the programmer. S Darren.
1998-04-27 06:08:07 +02:00
#include "miscadmin.h"
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
#include "optimizer/clauses.h"
#include "optimizer/planmain.h"
Clean up the usage of canonicalize_qual(): in particular, be consistent about whether it is applied before or after eval_const_expressions(). I believe there were some corner cases where the system would fail to recognize that a partial index is applicable because of the previous inconsistency. Store normal rather than 'implicit AND' representations of constraints and index predicates in the catalogs. initdb forced due to representation change of constraints/predicates.
2003-12-28 22:57:37 +01:00
#include "optimizer/prep.h"
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
#include "optimizer/var.h"
Get rid of the separate RULE privilege for tables: now only a table's owner can create or modify rules for the table. Do setRuleCheckAsUser() while loading rules into the relcache, rather than when defining a rule. This ensures that permission checks for tables referenced in a rule are done with respect to the current owner of the rule's table, whereas formerly ALTER TABLE OWNER would fail to update the permission checking for associated rules. Removal of separate RULE privilege is needed to prevent various scenarios in which a grantee of RULE privilege could effectively have any privilege of the table owner. For backwards compatibility, GRANT/REVOKE RULE is still accepted, but it doesn't do anything. Per discussion here: http://archives.postgresql.org/pgsql-hackers/2006-04/msg01138.php
2006-09-05 23:08:36 +02:00
#include "rewrite/rewriteDefine.h"
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
#include "storage/fd.h"
Restructure some header files a bit, in particular heapam.h, by removing some unnecessary #include lines in it. Also, move some tuple routine prototypes and macros to htup.h, which allows removal of heapam.h inclusion from some .c files. For this to work, a new header file access/sysattr.h needed to be created, initially containing attribute numbers of system columns, for pg_dump usage. While at it, make contrib ltree, intarray and hstore header files more consistent with our header style.
2008-05-12 02:00:54 +02:00
#include "storage/lmgr.h"
This patch... 1. Removes the unnecessary "#define AbcRegProcedure 123"'s from pg_proc.h. 2. Changes those #defines to use the names already defined in fmgr.h. 3. Forces the make of fmgr.h in backend/Makefile instead of having it made as a dependency in access/common/Makefile *hack*hack*hack* 4. Rearranged the #includes to a less helter-skelter arrangement, also changing <file.h> to "file.h" to signify a non-system header. 5. Removed "pg_proc.h" from files where its only purpose was for the #defines removed in item #1. 6. Added "fmgr.h" to each file changed for completeness sake. Turns out that #6 was not necessary for some files because fmgr.h was being included in a roundabout way SIX levels deep by the first include. "access/genam.h" ->"access/relscan.h" ->"utils/rel.h" ->"access/strat.h" ->"access/skey.h" ->"fmgr.h" So adding fmgr.h really didn't add anything to the compile, hopefully just made it clearer to the programmer. S Darren.
1998-04-27 06:08:07 +02:00
#include "storage/smgr.h"
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
#include "utils/array.h"
Clean up syscache so that recursive invocation is safe, and remove error message about recursive use of a syscache. Also remove most of the specialized indexscan routines in indexing.c --- it turns out that catcache.c is perfectly able to perform the indexscan for itself, in fact has already looked up all the information needed to do so! This should be faster as well as needing far less boilerplate code.
2000-11-10 01:33:12 +01:00
#include "utils/builtins.h"
First round of changes for new fmgr interface. fmgr itself and the key call sites are changed, but most called functions are still oldstyle. An exception is that the PL managers are updated (so, for example, NULL handling now behaves as expected in plperl and plpgsql functions). NOTE initdb is forced due to added column in pg_proc.
2000-05-28 19:56:29 +02:00
#include "utils/fmgroids.h"
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
#include "utils/inval.h"
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
#include "utils/lsyscache.h"
For some reason access/tupmacs.h has been #including utils/memutils.h, which is neither needed by nor related to that header. Remove the bogus inclusion and instead include the header in those C files that actually need it. Also fix unnecessary inclusions and bad inclusion order in tsearch2 files.
2005-05-06 19:24:55 +02:00
#include "utils/memutils.h"
This patch... 1. Removes the unnecessary "#define AbcRegProcedure 123"'s from pg_proc.h. 2. Changes those #defines to use the names already defined in fmgr.h. 3. Forces the make of fmgr.h in backend/Makefile instead of having it made as a dependency in access/common/Makefile *hack*hack*hack* 4. Rearranged the #includes to a less helter-skelter arrangement, also changing <file.h> to "file.h" to signify a non-system header. 5. Removed "pg_proc.h" from files where its only purpose was for the #defines removed in item #1. 6. Added "fmgr.h" to each file changed for completeness sake. Turns out that #6 was not necessary for some files because fmgr.h was being included in a roundabout way SIX levels deep by the first include. "access/genam.h" ->"access/relscan.h" ->"utils/rel.h" ->"access/strat.h" ->"access/skey.h" ->"fmgr.h" So adding fmgr.h really didn't add anything to the compile, hopefully just made it clearer to the programmer. S Darren.
1998-04-27 06:08:07 +02:00
#include "utils/relcache.h"
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
#include "utils/relmapper.h"
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
#include "utils/resowner.h"
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
#include "utils/syscache.h"
Move the HTSU_Result enum definition into snapshot.h, to avoid including tqual.h into heapam.h. This makes all inclusion of tqual.h explicit. I also sorted alphabetically the includes on some source files.
2008-03-26 22:10:39 +01:00
#include "utils/tqual.h"
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
#include "utils/typcache.h"
This patch... 1. Removes the unnecessary "#define AbcRegProcedure 123"'s from pg_proc.h. 2. Changes those #defines to use the names already defined in fmgr.h. 3. Forces the make of fmgr.h in backend/Makefile instead of having it made as a dependency in access/common/Makefile *hack*hack*hack* 4. Rearranged the #includes to a less helter-skelter arrangement, also changing <file.h> to "file.h" to signify a non-system header. 5. Removed "pg_proc.h" from files where its only purpose was for the #defines removed in item #1. 6. Added "fmgr.h" to each file changed for completeness sake. Turns out that #6 was not necessary for some files because fmgr.h was being included in a roundabout way SIX levels deep by the first include. "access/genam.h" ->"access/relscan.h" ->"utils/rel.h" ->"access/strat.h" ->"access/skey.h" ->"fmgr.h" So adding fmgr.h really didn't add anything to the compile, hopefully just made it clearer to the programmer. S Darren.
1998-04-27 06:08:07 +02:00
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* name of relcache init file(s), used to speed up backend startup
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
#define RELCACHE_INIT_FILENAME "pg_internal.init"
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
#define RELCACHE_INIT_FILEMAGIC 0x573265 /* version ID value */
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* hardcoded tuple descriptors, generated by genbki.pl
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
static const FormData_pg_attribute Desc_pg_class[Natts_pg_class] = {Schema_pg_class};
static const FormData_pg_attribute Desc_pg_attribute[Natts_pg_attribute] = {Schema_pg_attribute};
static const FormData_pg_attribute Desc_pg_proc[Natts_pg_proc] = {Schema_pg_proc};
static const FormData_pg_attribute Desc_pg_type[Natts_pg_type] = {Schema_pg_type};
static const FormData_pg_attribute Desc_pg_database[Natts_pg_database] = {Schema_pg_database};
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
static const FormData_pg_attribute Desc_pg_authid[Natts_pg_authid] = {Schema_pg_authid};
static const FormData_pg_attribute Desc_pg_auth_members[Natts_pg_auth_members] = {Schema_pg_auth_members};
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
static const FormData_pg_attribute Desc_pg_index[Natts_pg_index] = {Schema_pg_index};
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
* Hash tables that index the relation cache
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
* We used to index the cache by both name and OID, but now there
* is only an index by OID.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
typedef struct relidcacheent
{
Oid reloid;
Relation reldesc;
} RelIdCacheEnt;
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
static HTAB *RelationIdCache;
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
/*
* This flag is false until we have prepared the critical relcache entries
* that are needed to do indexscans on the tables read by relcache building.
*/
pgindent run.
2002-09-04 22:31:48 +02:00
bool criticalRelcachesBuilt = false;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
* This flag is false until we have prepared the critical relcache entries
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* for shared catalogs (which are the tables needed for login).
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*/
bool criticalSharedRelcachesBuilt = false;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* This counter counts relcache inval events received since backend startup
pgindent run.
2002-09-04 22:31:48 +02:00
* (but only for rels that are actually in cache). Presently, we use it only
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* to detect whether data about to be written by write_relcache_init_file()
* might already be obsolete.
*/
static long relcacheInvalsReceived = 0L;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* This list remembers the OIDs of the non-shared relations cached in the
* database's local relcache init file. Note that there is no corresponding
* list for the shared relcache init file, for reasons explained in the
* comments for RelationCacheInitFileRemove.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
static List *initFileRelationIds = NIL;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
/*
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
* This flag lets us optimize away work in AtEO(Sub)Xact_RelationCache().
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
*/
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
static bool need_eoxact_work = false;
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Another round of cleanups for dynahash.c (maybe it's finally clean of portability issues). Caller-visible data structures are now allocated on MAXALIGN boundaries, allowing safe use of datatypes wider than 'long'. Rejigger hash_create API so that caller specifies size of key and total size of entry, not size of key and size of rest of entry. This simplifies life considerably since each number is just a sizeof(), and padding issues etc. are taken care of automatically.
2001-10-01 07:36:17 +02:00
* macros to manipulate the lookup hashtables
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
#define RelationCacheInsert(RELATION) \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
do { \
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
RelIdCacheEnt *idhentry; bool found; \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
idhentry = (RelIdCacheEnt*)hash_search(RelationIdCache, \
Another round of cleanups for dynahash.c (maybe it's finally clean of portability issues). Caller-visible data structures are now allocated on MAXALIGN boundaries, allowing safe use of datatypes wider than 'long'. Rejigger hash_create API so that caller specifies size of key and total size of entry, not size of key and size of rest of entry. This simplifies life considerably since each number is just a sizeof(), and padding issues etc. are taken care of automatically.
2001-10-01 07:36:17 +02:00
(void *) &(RELATION->rd_id), \
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
HASH_ENTER, &found); \
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
/* used to give notice if found -- now just keep quiet */ \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
idhentry->reldesc = RELATION; \
} while(0)
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
#define RelationIdCacheLookup(ID, RELATION) \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
do { \
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
RelIdCacheEnt *hentry; \
hentry = (RelIdCacheEnt*)hash_search(RelationIdCache, \
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
(void *) &(ID), \
HASH_FIND, NULL); \
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
if (hentry) \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
RELATION = hentry->reldesc; \
else \
RELATION = NULL; \
} while(0)
#define RelationCacheDelete(RELATION) \
do { \
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
RelIdCacheEnt *idhentry; \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
idhentry = (RelIdCacheEnt*)hash_search(RelationIdCache, \
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
(void *) &(RELATION->rd_id), \
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
HASH_REMOVE, NULL); \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
if (idhentry == NULL) \
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(WARNING, "trying to delete a rd_id reldesc that does not exist"); \
Fix macros that were not properly surrounded by parens or braces.
1998-06-15 20:40:05 +02:00
} while(0)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* Special cache for opclass-related information
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
*
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
* Note: only default operators and support procs get cached, ie, those with
* lefttype = righttype = opcintype.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
typedef struct opclasscacheent
{
Oid opclassoid; /* lookup key: OID of opclass */
bool valid; /* set TRUE after successful fill-in */
StrategyNumber numStrats; /* max # of strategies (from pg_am) */
StrategyNumber numSupport; /* max # of support procs (from pg_am) */
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Oid opcfamily; /* OID of opclass's family */
Oid opcintype; /* OID of opclass's declared input type */
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Oid *operatorOids; /* strategy operators' OIDs */
pgindent run.
2002-09-04 22:31:48 +02:00
RegProcedure *supportProcs; /* support procs */
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
} OpClassCacheEnt;
static HTAB *OpClassCache = NULL;
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
/* non-export function prototypes */
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
static void RelationDestroyRelation(Relation relation);
Improve relcache.c error reporting for the next guy who has to debug this thing.
2002-07-15 03:57:51 +02:00
static void RelationClearRelation(Relation relation, bool rebuild);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
static void RelationReloadIndexInfo(Relation relation);
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
static void RelationFlushRelation(Relation relation);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
static bool load_relcache_init_file(bool shared);
static void write_relcache_init_file(bool shared);
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
static void write_item(const void *data, Size len, FILE *fp);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
static void formrdesc(const char *relationName, Oid relationReltype,
bool isshared, bool hasoids,
int natts, const FormData_pg_attribute *attrs);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
static HeapTuple ScanPgRelation(Oid targetRelId, bool indexOK);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
static Relation AllocateRelationDesc(Form_pg_class relp);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
static void RelationParseRelOptions(Relation relation, HeapTuple tuple);
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
static void RelationBuildTupleDesc(Relation relation);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
static Relation RelationBuildDesc(Oid targetRelId, bool insertIt);
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
static void RelationInitPhysicalAddr(Relation relation);
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
static void load_critical_index(Oid indexoid, Oid heapoid);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
static TupleDesc GetPgClassDescriptor(void);
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
static TupleDesc GetPgIndexDescriptor(void);
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
static void AttrDefaultFetch(Relation relation);
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
static void CheckConstraintFetch(Relation relation);
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
static List *insert_ordered_oid(List *list, Oid datum);
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
static void IndexSupportInitialize(oidvector *indclass,
pgindent run.
2002-09-04 22:31:48 +02:00
Oid *indexOperator,
RegProcedure *indexSupport,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Oid *opFamily,
Oid *opcInType,
pgindent run.
2002-09-04 22:31:48 +02:00
StrategyNumber maxStrategyNumber,
StrategyNumber maxSupportNumber,
AttrNumber maxAttributeNumber);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
static OpClassCacheEnt *LookupOpclassInfo(Oid operatorClassOid,
pgindent run.
2002-09-04 22:31:48 +02:00
StrategyNumber numStrats,
StrategyNumber numSupport);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
static void RelationCacheInitFileRemoveInDir(const char *tblspcpath);
static void unlink_initfile(const char *initfilename);
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* ScanPgRelation
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Fix LOAD_CRIT_INDEX() macro to take out AccessShareLock on the system index it is trying to build a relcache entry for. This is an oversight in my 8.2 patch that tried to ensure we always took a lock on a relation before trying to build its relcache entry. The implication is that if someone committed a reindex of a critical system index at about the same time that some other backend were starting up without a valid pg_internal.init file, the second one might PANIC due to not seeing any valid version of the index's pg_class row. Improbable case, but definitely not impossible.
2008-04-16 20:23:04 +02:00
* This is used by RelationBuildDesc to find a pg_class
* tuple matching targetRelId. The caller must hold at least
* AccessShareLock on the target relid to prevent concurrent-update
* scenarios --- else our SnapshotNow scan might fail to find any
* version that it thinks is live.
Repair memory leaks that caused CacheCxt to grow without bound. We really ought to fix relcache entry construction so that it does not do so much with CurrentMemoryContext = CacheCxt. As is, relatively harmless leaks in either sequential or index scanning translate to permanent leaks if they occur when called from relcache build. For the moment, however, the path of least resistance is to repair all such leaks...
2000-05-21 04:28:55 +02:00
*
* NB: the returned tuple has been copied into palloc'd storage
* and must eventually be freed with heap_freetuple.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
static HeapTuple
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
ScanPgRelation(Oid targetRelId, bool indexOK)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
HeapTuple pg_class_tuple;
Relation pg_class_desc;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
SysScanDesc pg_class_scan;
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
ScanKeyData key[1];
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
* If something goes wrong during backend startup, we might find ourselves
* trying to read pg_class before we've selected a database. That ain't
* gonna work, so bail out with a useful error message. If this happens,
* it probably means a relcache entry that needs to be nailed isn't.
*/
if (!OidIsValid(MyDatabaseId))
elog(FATAL, "cannot read pg_class without having selected a database");
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* form a scan key
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
ScanKeyInit(&key[0],
ObjectIdAttributeNumber,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(targetRelId));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* Open pg_class and fetch a tuple. Force heap scan if we haven't yet
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* built the critical relcache entries (this includes initdb and startup
* without a pg_internal.init file). The caller can also force a heap
* scan by setting indexOK == false.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
pg_class_desc = heap_open(RelationRelationId, AccessShareLock);
pg_class_scan = systable_beginscan(pg_class_desc, ClassOidIndexId,
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
indexOK && criticalRelcachesBuilt,
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
SnapshotNow,
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
1, key);
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
pg_class_tuple = systable_getnext(pg_class_scan);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* Must copy tuple before releasing buffer.
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (HeapTupleIsValid(pg_class_tuple))
pg_class_tuple = heap_copytuple(pg_class_tuple);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* all done */
systable_endscan(pg_class_scan);
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
heap_close(pg_class_desc, AccessShareLock);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
return pg_class_tuple;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* AllocateRelationDesc
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* This is used to allocate memory for a new relation descriptor
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* and initialize the rd_rel field from the given pg_class tuple.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
static Relation
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
AllocateRelationDesc(Form_pg_class relp)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
Relation relation;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
MemoryContext oldcxt;
Renaming cleanup, no pgindent yet.
1998-09-01 05:29:17 +02:00
Form_pg_class relationForm;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
/* Relcache entries must live in CacheMemoryContext */
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* allocate and zero space for new relation descriptor
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
relation = (Relation) palloc0(sizeof(RelationData));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
/* make sure relation is marked as having no open file yet */
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
relation->rd_smgr = NULL;
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* Copy the relation tuple form
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* We only allocate space for the fixed fields, ie, CLASS_TUPLE_SIZE. The
* variable-length fields (relacl, reloptions) are NOT stored in the
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
* relcache --- there'd be little point in it, since we don't copy the
* tuple's nulls bitmap and hence wouldn't know if the values are valid.
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Bottom line is that relacl *cannot* be retrieved from the relcache. Get
* it from the syscache if you need it. The same goes for the original
* form of reloptions (however, we do store the parsed form of reloptions
* in rd_options).
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*/
relationForm = (Form_pg_class) palloc(CLASS_TUPLE_SIZE);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Code cleanup: in C89, there is no point casting the first argument to memset() or MemSet() to a char *. For one, memset()'s first argument is a void *, and further void * can be implicitly coerced to/from any other pointer type.
2005-05-11 03:26:02 +02:00
memcpy(relationForm, relp, CLASS_TUPLE_SIZE);
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
/* initialize relation tuple form */
Renaming cleanup, no pgindent yet.
1998-09-01 05:29:17 +02:00
relation->rd_rel = relationForm;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
/* and allocate attribute tuple form storage */
Code review for HeapTupleHeader changes. Add version number to page headers (overlaying low byte of page size) and add HEAP_HASOID bit to t_infomask, per earlier discussion. Simplify scheme for overlaying fields in tuple header (no need for cmax to live in more than one place). Don't try to clear infomask status bits in tqual.c --- not safe to do it there. Don't try to force output table of a SELECT INTO to have OIDs, either. Get rid of unnecessarily complex three-state scheme for TupleDesc.tdhasoids, which has already caused one recent failure. Improve documentation.
2002-09-02 03:05:06 +02:00
relation->rd_att = CreateTemplateTupleDesc(relationForm->relnatts,
relationForm->relhasoids);
Fix problems with cached tuple descriptors disappearing while still in use by creating a reference-count mechanism, similar to what we did a long time ago for catcache entries. The back branches have an ugly solution involving lots of extra copies, but this way is more efficient. Reference counting is only applied to tupdescs that are actually in caches --- there seems no need to use it for tupdescs that are generated in the executor, since they'll go away during plan shutdown by virtue of being in the per-query memory context. Neil Conway and Tom Lane
2006-06-16 20:42:24 +02:00
/* which we mark as a reference-counted tupdesc */
relation->rd_att->tdrefcount = 1;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
MemoryContextSwitchTo(oldcxt);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
return relation;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
/*
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
* RelationParseRelOptions
* Convert pg_class.reloptions into pre-parsed rd_options
*
* tuple is the real pg_class tuple (not rd_rel!) for relation
*
* Note: rd_rel and (if an index) rd_am must be valid already
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
*/
static void
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
RelationParseRelOptions(Relation relation, HeapTuple tuple)
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
{
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
bytea *options;
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
relation->rd_options = NULL;
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
/* Fall out if relkind should not have options */
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
switch (relation->rd_rel->relkind)
{
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
case RELKIND_RELATION:
case RELKIND_TOASTVALUE:
case RELKIND_INDEX:
break;
default:
return;
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
}
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Fetch reloptions from tuple; have to use a hardwired descriptor because
* we might not have any other for pg_class yet (consider executing this
* code for pg_class itself)
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
*/
Allow extracting and parsing of reloptions from a bare pg_class tuple, and refactor the relcache code that used to do that. This allows other callers (particularly autovacuum) to do the same without necessarily having to open and lock a table.
2009-01-26 20:41:06 +01:00
options = extractRelOptions(tuple,
GetPgClassDescriptor(),
relation->rd_rel->relkind == RELKIND_INDEX ?
relation->rd_am->amoptions : InvalidOid);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
Document why we copy reloptions into CacheMemoryContext after-the-fact.
2010-01-07 21:39:45 +01:00
/*
* Copy parsed data into CacheMemoryContext. To guard against the
* possibility of leaks in the reloptions code, we want to do the actual
* parsing in the caller's memory context and copy the results into
* CacheMemoryContext after the fact.
*/
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
if (options)
{
relation->rd_options = MemoryContextAlloc(CacheMemoryContext,
VARSIZE(options));
memcpy(relation->rd_options, options, VARSIZE(options));
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
pfree(options);
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
}
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* RelationBuildTupleDesc
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* Form the relation's tuple descriptor from information in
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
* the pg_attribute, pg_attrdef & pg_constraint system catalogs.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
static void
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
RelationBuildTupleDesc(Relation relation)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
HeapTuple pg_attribute_tuple;
Relation pg_attribute_desc;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
SysScanDesc pg_attribute_scan;
ScanKeyData skey[2];
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
int need;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
TupleConstr *constr;
Implement reindex command
2000-02-18 10:30:20 +01:00
AttrDefault *attrdef = NULL;
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
int ndef = 0;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
/* copy some fields from pg_class row to rd_att */
relation->rd_att->tdtypeid = relation->rd_rel->reltype;
relation->rd_att->tdtypmod = -1; /* unnecessary, but... */
relation->rd_att->tdhasoid = relation->rd_rel->relhasoids;
Code review for HeapTupleHeader changes. Add version number to page headers (overlaying low byte of page size) and add HEAP_HASOID bit to t_infomask, per earlier discussion. Simplify scheme for overlaying fields in tuple header (no need for cmax to live in more than one place). Don't try to clear infomask status bits in tqual.c --- not safe to do it there. Don't try to force output table of a SELECT INTO to have OIDs, either. Get rid of unnecessarily complex three-state scheme for TupleDesc.tdhasoids, which has already caused one recent failure. Improve documentation.
2002-09-02 03:05:06 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
constr = (TupleConstr *) MemoryContextAlloc(CacheMemoryContext,
sizeof(TupleConstr));
Implement reindex command
2000-02-18 10:30:20 +01:00
constr->has_not_null = false;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* Form a scan key that selects only user attributes (attnum > 0).
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* (Eliminating system attribute rows at the index level is lots faster
* than fetching them.)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Anum_pg_attribute_attrelid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(RelationGetRelid(relation)));
ScanKeyInit(&skey[1],
Anum_pg_attribute_attnum,
BTGreaterStrategyNumber, F_INT2GT,
Int16GetDatum(0));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Open pg_attribute and begin a scan. Force heap scan if we haven't yet
* built the critical relcache entries (this includes initdb and startup
* without a pg_internal.init file).
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
pg_attribute_desc = heap_open(AttributeRelationId, AccessShareLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
pg_attribute_scan = systable_beginscan(pg_attribute_desc,
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
AttributeRelidNumIndexId,
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
criticalRelcachesBuilt,
SnapshotNow,
2, skey);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* add attribute data to relation->rd_att
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
need = relation->rd_rel->relnatts;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
while (HeapTupleIsValid(pg_attribute_tuple = systable_getnext(pg_attribute_scan)))
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Form_pg_attribute attp;
Renaming cleanup, no pgindent yet.
1998-09-01 05:29:17 +02:00
attp = (Form_pg_attribute) GETSTRUCT(pg_attribute_tuple);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (attp->attnum <= 0 ||
attp->attnum > relation->rd_rel->relnatts)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "invalid attribute number %d for %s",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
attp->attnum, RelationGetRelationName(relation));
Adjust creation/destruction of TupleDesc data structure to reduce the number of palloc calls. This has a salutory impact on plpgsql operations with record variables (which create and destroy tupdescs constantly) and probably helps a bit in some other cases too.
2005-03-07 05:42:17 +01:00
memcpy(relation->rd_att->attrs[attp->attnum - 1],
attp,
Support column-level privileges, as required by SQL standard. Stephen Frost, with help from KaiGai Kohei and others
2009-01-22 21:16:10 +01:00
ATTRIBUTE_FIXED_PART_SIZE);
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
/* Update constraint/default info */
if (attp->attnotnull)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
constr->has_not_null = true;
Implement reindex command
2000-02-18 10:30:20 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (attp->atthasdef)
{
if (attrdef == NULL)
attrdef = (AttrDefault *)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
MemoryContextAllocZero(CacheMemoryContext,
relation->rd_rel->relnatts *
sizeof(AttrDefault));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
attrdef[ndef].adnum = attp->attnum;
attrdef[ndef].adbin = NULL;
ndef++;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
need--;
if (need == 0)
break;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* end the scan and close the attribute relation
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
systable_endscan(pg_attribute_scan);
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
heap_close(pg_attribute_desc, AccessShareLock);
Implement reindex command
2000-02-18 10:30:20 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (need != 0)
elog(ERROR, "catalog is missing %d attribute(s) for relid %u",
need, RelationGetRelid(relation));
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* The attcacheoff values we read from pg_attribute should all be -1
* ("unknown"). Verify this if assert checking is on. They will be
* computed when and if needed during tuple access.
Remove VARLENA_FIXED_SIZE hack, which is irreversibly broken now that both MULTIBYTE and TOAST prevent char(n) from being truly fixed-size. Simplify and speed up fastgetattr() and index_getattr() macros by eliminating special cases for attnum=1. It's just as fast to handle the first attribute by presetting its attcacheoff to zero; so do that instead when loading the tupledesc in relcache.c.
2000-11-30 19:38:47 +01:00
*/
#ifdef USE_ASSERT_CHECKING
{
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
int i;
Remove VARLENA_FIXED_SIZE hack, which is irreversibly broken now that both MULTIBYTE and TOAST prevent char(n) from being truly fixed-size. Simplify and speed up fastgetattr() and index_getattr() macros by eliminating special cases for attnum=1. It's just as fast to handle the first attribute by presetting its attcacheoff to zero; so do that instead when loading the tupledesc in relcache.c.
2000-11-30 19:38:47 +01:00
for (i = 0; i < relation->rd_rel->relnatts; i++)
Assert(relation->rd_att->attrs[i]->attcacheoff == -1);
}
#endif
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* However, we can easily set the attcacheoff value for the first
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* attribute: it must be zero. This eliminates the need for special cases
* for attnum=1 that used to exist in fastgetattr() and index_getattr().
Remove VARLENA_FIXED_SIZE hack, which is irreversibly broken now that both MULTIBYTE and TOAST prevent char(n) from being truly fixed-size. Simplify and speed up fastgetattr() and index_getattr() macros by eliminating special cases for attnum=1. It's just as fast to handle the first attribute by presetting its attcacheoff to zero; so do that instead when loading the tupledesc in relcache.c.
2000-11-30 19:38:47 +01:00
*/
Fix ALTER TABLE ADD COLUMN to disallow the same column types that are disallowed by CREATE TABLE (eg, pseudo-types); also disallow these types from being introduced by the range-function syntax. While at it, allow CREATE TABLE to create zero-column tables, per recent pghackers discussion. I am back-patching this into 7.3 since failure to disallow pseudo-types is arguably a security hole.
2002-12-16 19:39:22 +01:00
if (relation->rd_rel->relnatts > 0)
relation->rd_att->attrs[0]->attcacheoff = 0;
Remove VARLENA_FIXED_SIZE hack, which is irreversibly broken now that both MULTIBYTE and TOAST prevent char(n) from being truly fixed-size. Simplify and speed up fastgetattr() and index_getattr() macros by eliminating special cases for attnum=1. It's just as fast to handle the first attribute by presetting its attcacheoff to zero; so do that instead when loading the tupledesc in relcache.c.
2000-11-30 19:38:47 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* Set up constraint/default info
*/
if (constr->has_not_null || ndef > 0 || relation->rd_rel->relchecks)
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
relation->rd_att->constr = constr;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (ndef > 0) /* DEFAULTs */
Trial implementation of ALTER DROP COLUMN. They are #ifdef'd. Add -D_DROP_COLUMN_HACK__ compile option to evaluate it.
2000-03-09 06:00:26 +01:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (ndef < relation->rd_rel->relnatts)
constr->defval = (AttrDefault *)
repalloc(attrdef, ndef * sizeof(AttrDefault));
else
constr->defval = attrdef;
constr->num_defval = ndef;
AttrDefaultFetch(relation);
Repair memory leaks that caused CacheCxt to grow without bound. We really ought to fix relcache entry construction so that it does not do so much with CurrentMemoryContext = CacheCxt. As is, relatively harmless leaks in either sequential or index scanning translate to permanent leaks if they occur when called from relcache build. For the moment, however, the path of least resistance is to repair all such leaks...
2000-05-21 04:28:55 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
else
constr->num_defval = 0;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (relation->rd_rel->relchecks > 0) /* CHECKs */
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
constr->num_check = relation->rd_rel->relchecks;
constr->check = (ConstrCheck *)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
MemoryContextAllocZero(CacheMemoryContext,
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
constr->num_check * sizeof(ConstrCheck));
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
CheckConstraintFetch(relation);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
else
constr->num_check = 0;
}
else
{
pfree(constr);
relation->rd_att->constr = NULL;
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
}
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* RelationBuildRuleLock
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* Form the relation's rewrite rules from information in
* the pg_rewrite system catalog.
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*
* Note: The rule parsetrees are potentially very complex node structures.
* To allow these trees to be freed when the relcache entry is flushed,
* we make a private memory context to hold the RuleLock information for
* each relcache entry that has associated rules. The context is used
* just for rule info, not for any other subsidiary data of the relcache
* entry, because that keeps the update logic in RelationClearRelation()
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* manageable. The other subsidiary data structures are simple enough
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
* to be easy to free explicitly, anyway.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
static void
RelationBuildRuleLock(Relation relation)
{
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
MemoryContext rulescxt;
MemoryContext oldcxt;
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
HeapTuple rewrite_tuple;
Relation rewrite_desc;
TupleDesc rewrite_tupdesc;
SysScanDesc rewrite_scan;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
ScanKeyData key;
RuleLock *rulelock;
int numlocks;
RewriteRule **rules;
int maxlocks;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Make the private context. Parameters are set on the assumption that
* it'll probably not contain much data.
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*/
rulescxt = AllocSetContextCreate(CacheMemoryContext,
RelationGetRelationName(relation),
Tweak default memory context allocation policy so that a context is not given any malloc block until something is first allocated in it; but thereafter, MemoryContextReset won't release that first malloc block. This preserves the quick-reset property of the original policy, without forcing 8K to be allocated to every context whether any of it is ever used or not. Also, remove some more no-longer-needed explicit freeing during ExecEndPlan.
2002-12-15 22:01:34 +01:00
ALLOCSET_SMALL_MINSIZE,
ALLOCSET_SMALL_INITSIZE,
ALLOCSET_SMALL_MAXSIZE);
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
relation->rd_rulescxt = rulescxt;
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* allocate an array to hold the rewrite rules (the array is extended if
* necessary)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
maxlocks = 4;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
rules = (RewriteRule **)
MemoryContextAlloc(rulescxt, sizeof(RewriteRule *) * maxlocks);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
numlocks = 0;
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* form a scan key
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&key,
Anum_pg_rewrite_ev_class,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(RelationGetRelid(relation)));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* open pg_rewrite and begin a scan
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* Note: since we scan the rules using RewriteRelRulenameIndexId, we will
* be reading the rules in name order, except possibly during
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* emergency-recovery operations (ie, IgnoreSystemIndexes). This in turn
* ensures that rules will be fired in name order.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
rewrite_desc = heap_open(RewriteRelationId, AccessShareLock);
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
rewrite_tupdesc = RelationGetDescr(rewrite_desc);
pgindent run.
2002-09-04 22:31:48 +02:00
rewrite_scan = systable_beginscan(rewrite_desc,
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
RewriteRelRulenameIndexId,
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
true, SnapshotNow,
1, &key);
while (HeapTupleIsValid(rewrite_tuple = systable_getnext(rewrite_scan)))
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
Form_pg_rewrite rewrite_form = (Form_pg_rewrite) GETSTRUCT(rewrite_tuple);
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
bool isnull;
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
Datum rule_datum;
char *rule_str;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
RewriteRule *rule;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
rule = (RewriteRule *) MemoryContextAlloc(rulescxt,
sizeof(RewriteRule));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
oid is needed, it is added at the end of the struct (after the null bitmap, if present). Per Tom Lane's suggestion the information whether a tuple has an oid or not is carried in the tuple descriptor. For debugging reasons tdhasoid is of type char, not bool. There are predefined values for WITHOID, WITHOUTOID and UNDEFOID. This patch has been generated against a cvs snapshot from last week and I don't expect it to apply cleanly to current sources. While I post it here for public review, I'm working on a new version against a current snapshot. (There's been heavy activity recently; hope to catch up some day ...) This is a long patch; if it is too hard to swallow, I can provide it in smaller pieces: Part 1: Accessor macros Part 2: tdhasoid in TupDesc Part 3: Regression test Part 4: Parameter withoid to heap_addheader Part 5: Eliminate t_oid from HeapTupleHeader Part 2 is the most hairy part because of changes in the executor and even in the parser; the other parts are straightforward. Up to part 4 the patched postmaster stays binary compatible to databases created with an unpatched version. Part 5 is small (100 lines) and finally breaks compatibility. Manfred Koizar
2002-07-20 07:16:59 +02:00
rule->ruleId = HeapTupleGetOid(rewrite_tuple);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Rule names are now unique per-relation, rather than unique globally. DROP RULE and COMMENT ON RULE syntax adds an 'ON tablename' clause, similar to TRIGGER syntaxes. To allow loading of existing pg_dump files containing COMMENT ON RULE, the COMMENT code will still accept the old syntax --- but only if the target rulename is unique across the whole database.
2002-04-18 22:01:11 +02:00
rule->event = rewrite_form->ev_type - '0';
rule->attrno = rewrite_form->ev_attr;
Changes pg_trigger and extend pg_rewrite in order to allow triggers and rules to be defined with different, per session controllable, behaviors for replication purposes. This will allow replication systems like Slony-I and, as has been stated on pgsql-hackers, other products to control the firing mechanism of triggers and rewrite rules without modifying the system catalog directly. The firing mechanisms are controlled by a new superuser-only GUC variable, session_replication_role, together with a change to pg_trigger.tgenabled and a new column pg_rewrite.ev_enabled. Both columns are a single char data type now (tgenabled was a bool before). The possible values in these attributes are: 'O' - Trigger/Rule fires when session_replication_role is "origin" (default) or "local". This is the default behavior. 'D' - Trigger/Rule is disabled and fires never 'A' - Trigger/Rule fires always regardless of the setting of session_replication_role 'R' - Trigger/Rule fires when session_replication_role is "replica" The GUC variable can only be changed as long as the system does not have any cached query plans. This will prevent changing the session role and accidentally executing stored procedures or functions that have plans cached that expand to the wrong query set due to differences in the rule firing semantics. The SQL syntax for changing a triggers/rules firing semantics is ALTER TABLE <tabname> <when> TRIGGER|RULE <name>; <when> ::= ENABLE | ENABLE ALWAYS | ENABLE REPLICA | DISABLE psql's \d command as well as pg_dump are extended in a backward compatible fashion. Jan
2007-03-20 00:38:32 +01:00
rule->enabled = rewrite_form->ev_enabled;
Rule names are now unique per-relation, rather than unique globally. DROP RULE and COMMENT ON RULE syntax adds an 'ON tablename' clause, similar to TRIGGER syntaxes. To allow loading of existing pg_dump files containing COMMENT ON RULE, the COMMENT code will still accept the old syntax --- but only if the target rulename is unique across the whole database.
2002-04-18 22:01:11 +02:00
rule->isInstead = rewrite_form->is_instead;
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
/*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* Must use heap_getattr to fetch ev_action and ev_qual. Also, the
* rule strings are often large enough to be toasted. To avoid
* leaking memory in the caller's context, do the detoasting here so
* we can free the detoasted version.
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
*/
rule_datum = heap_getattr(rewrite_tuple,
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
Anum_pg_rewrite_ev_action,
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
rewrite_tupdesc,
pgindent run over code.
1999-05-25 18:15:34 +02:00
&isnull);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Assert(!isnull);
Simplify and standardize conversions between TEXT datums and ordinary C strings. This patch introduces four support functions cstring_to_text, cstring_to_text_with_len, text_to_cstring, and text_to_cstring_buffer, and two macros CStringGetTextDatum and TextDatumGetCString. A number of existing macros that provided variants on these themes were removed. Most of the places that need to make such conversions now require just one function or macro call, in place of the multiple notational layers that used to be needed. There are no longer any direct calls of textout or textin, and we got most of the places that were using handmade conversions via memcpy (there may be a few still lurking, though). This commit doesn't make any serious effort to eliminate transient memory leaks caused by detoasting toasted text objects before they reach text_to_cstring. We changed PG_GETARG_TEXT_P to PG_GETARG_TEXT_PP in a few places where it was easy, but much more could be done. Brendan Jurd and Tom Lane
2008-03-25 23:42:46 +01:00
rule_str = TextDatumGetCString(rule_datum);
Fix memory leak in relcache handling of rules: allocate rule parsetrees in per-entry sub-memory-context, where they were supposed to go, rather than in CacheMemoryContext where the code was putting them. Must've suffered a severe brain fade when I wrote this :-(
2001-01-06 22:53:18 +01:00
oldcxt = MemoryContextSwitchTo(rulescxt);
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
rule->actions = (List *) stringToNode(rule_str);
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
MemoryContextSwitchTo(oldcxt);
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
pfree(rule_str);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
rule_datum = heap_getattr(rewrite_tuple,
Anum_pg_rewrite_ev_qual,
rewrite_tupdesc,
&isnull);
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Assert(!isnull);
Simplify and standardize conversions between TEXT datums and ordinary C strings. This patch introduces four support functions cstring_to_text, cstring_to_text_with_len, text_to_cstring, and text_to_cstring_buffer, and two macros CStringGetTextDatum and TextDatumGetCString. A number of existing macros that provided variants on these themes were removed. Most of the places that need to make such conversions now require just one function or macro call, in place of the multiple notational layers that used to be needed. There are no longer any direct calls of textout or textin, and we got most of the places that were using handmade conversions via memcpy (there may be a few still lurking, though). This commit doesn't make any serious effort to eliminate transient memory leaks caused by detoasting toasted text objects before they reach text_to_cstring. We changed PG_GETARG_TEXT_P to PG_GETARG_TEXT_PP in a few places where it was easy, but much more could be done. Brendan Jurd and Tom Lane
2008-03-25 23:42:46 +01:00
rule_str = TextDatumGetCString(rule_datum);
Fix memory leak in relcache handling of rules: allocate rule parsetrees in per-entry sub-memory-context, where they were supposed to go, rather than in CacheMemoryContext where the code was putting them. Must've suffered a severe brain fade when I wrote this :-(
2001-01-06 22:53:18 +01:00
oldcxt = MemoryContextSwitchTo(rulescxt);
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
rule->qual = (Node *) stringToNode(rule_str);
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
MemoryContextSwitchTo(oldcxt);
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
pfree(rule_str);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
Get rid of the separate RULE privilege for tables: now only a table's owner can create or modify rules for the table. Do setRuleCheckAsUser() while loading rules into the relcache, rather than when defining a rule. This ensures that permission checks for tables referenced in a rule are done with respect to the current owner of the rule's table, whereas formerly ALTER TABLE OWNER would fail to update the permission checking for associated rules. Removal of separate RULE privilege is needed to prevent various scenarios in which a grantee of RULE privilege could effectively have any privilege of the table owner. For backwards compatibility, GRANT/REVOKE RULE is still accepted, but it doesn't do anything. Per discussion here: http://archives.postgresql.org/pgsql-hackers/2006-04/msg01138.php
2006-09-05 23:08:36 +02:00
/*
* We want the rule's table references to be checked as though by the
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* table owner, not the user referencing the rule. Therefore, scan
Get rid of the separate RULE privilege for tables: now only a table's owner can create or modify rules for the table. Do setRuleCheckAsUser() while loading rules into the relcache, rather than when defining a rule. This ensures that permission checks for tables referenced in a rule are done with respect to the current owner of the rule's table, whereas formerly ALTER TABLE OWNER would fail to update the permission checking for associated rules. Removal of separate RULE privilege is needed to prevent various scenarios in which a grantee of RULE privilege could effectively have any privilege of the table owner. For backwards compatibility, GRANT/REVOKE RULE is still accepted, but it doesn't do anything. Per discussion here: http://archives.postgresql.org/pgsql-hackers/2006-04/msg01138.php
2006-09-05 23:08:36 +02:00
* through the rule's actions and set the checkAsUser field on all
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* rtable entries. We have to look at the qual as well, in case it
Get rid of the separate RULE privilege for tables: now only a table's owner can create or modify rules for the table. Do setRuleCheckAsUser() while loading rules into the relcache, rather than when defining a rule. This ensures that permission checks for tables referenced in a rule are done with respect to the current owner of the rule's table, whereas formerly ALTER TABLE OWNER would fail to update the permission checking for associated rules. Removal of separate RULE privilege is needed to prevent various scenarios in which a grantee of RULE privilege could effectively have any privilege of the table owner. For backwards compatibility, GRANT/REVOKE RULE is still accepted, but it doesn't do anything. Per discussion here: http://archives.postgresql.org/pgsql-hackers/2006-04/msg01138.php
2006-09-05 23:08:36 +02:00
* contains sublinks.
*
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* The reason for doing this when the rule is loaded, rather than when
* it is stored, is that otherwise ALTER TABLE OWNER would have to
* grovel through stored rules to update checkAsUser fields. Scanning
* the rule tree during load is relatively cheap (compared to
* constructing it in the first place), so we do it here.
Get rid of the separate RULE privilege for tables: now only a table's owner can create or modify rules for the table. Do setRuleCheckAsUser() while loading rules into the relcache, rather than when defining a rule. This ensures that permission checks for tables referenced in a rule are done with respect to the current owner of the rule's table, whereas formerly ALTER TABLE OWNER would fail to update the permission checking for associated rules. Removal of separate RULE privilege is needed to prevent various scenarios in which a grantee of RULE privilege could effectively have any privilege of the table owner. For backwards compatibility, GRANT/REVOKE RULE is still accepted, but it doesn't do anything. Per discussion here: http://archives.postgresql.org/pgsql-hackers/2006-04/msg01138.php
2006-09-05 23:08:36 +02:00
*/
setRuleCheckAsUser((Node *) rule->actions, relation->rd_rel->relowner);
setRuleCheckAsUser(rule->qual, relation->rd_rel->relowner);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
if (numlocks >= maxlocks)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
maxlocks *= 2;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
rules = (RewriteRule **)
repalloc(rules, sizeof(RewriteRule *) * maxlocks);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
rules[numlocks++] = rule;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* end the scan and close the attribute relation
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
systable_endscan(rewrite_scan);
heap_close(rewrite_desc, AccessShareLock);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Make relhasrules and relhastriggers work like relhasindex, namely we let VACUUM reset them to false rather than trying to clean 'em up during DROP.
2008-11-10 01:49:37 +01:00
/*
* there might not be any rules (if relhasrules is out-of-date)
*/
if (numlocks == 0)
{
relation->rd_rules = NULL;
relation->rd_rulescxt = NULL;
MemoryContextDelete(rulescxt);
return;
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* form a RuleLock and insert into relation
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
rulelock = (RuleLock *) MemoryContextAlloc(rulescxt, sizeof(RuleLock));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
rulelock->numLocks = numlocks;
rulelock->rules = rules;
relation->rd_rules = rulelock;
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
* equalRuleLocks
*
* Determine whether two RuleLocks are equivalent
*
* Probably this should be in the rules code someplace...
*/
static bool
equalRuleLocks(RuleLock *rlock1, RuleLock *rlock2)
{
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
int i;
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* As of 7.3 we assume the rule ordering is repeatable, because
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* RelationBuildRuleLock should read 'em in a consistent order. So just
* compare corresponding slots.
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
*/
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
if (rlock1 != NULL)
{
if (rlock2 == NULL)
return false;
if (rlock1->numLocks != rlock2->numLocks)
return false;
for (i = 0; i < rlock1->numLocks; i++)
{
RewriteRule *rule1 = rlock1->rules[i];
pg_trigger's index on tgrelid is replaced by a unique index on (tgrelid, tgname). This provides an additional check on trigger name uniqueness per-table (which was already enforced by the code anyway). With this change, RelationBuildTriggers will read the triggers in order by tgname, since it's scanning using this index. Since a predictable trigger ordering has been requested for some time, document this behavior as a feature. Also document that rules fire in name order, since yesterday's changes to pg_rewrite indexing cause that too.
2002-04-19 18:36:08 +02:00
RewriteRule *rule2 = rlock2->rules[i];
if (rule1->ruleId != rule2->ruleId)
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
return false;
if (rule1->event != rule2->event)
return false;
if (rule1->attrno != rule2->attrno)
return false;
Fix oversight in ALTER TABLE ENABLE/DISABLE RULE patch: the new enabled field needs to be included in equalRuleLocks() comparisons, else updates will fail to propagate into relcache entries when they have positive reference count (ie someone is using the relcache entry). Per report from Alex Hunsaker.
2008-12-30 04:59:19 +01:00
if (rule1->enabled != rule2->enabled)
return false;
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
if (rule1->isInstead != rule2->isInstead)
return false;
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
if (!equal(rule1->qual, rule2->qual))
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
return false;
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
if (!equal(rule1->actions, rule2->actions))
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
return false;
}
}
else if (rlock2 != NULL)
return false;
return true;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Fix LOAD_CRIT_INDEX() macro to take out AccessShareLock on the system index it is trying to build a relcache entry for. This is an oversight in my 8.2 patch that tried to ensure we always took a lock on a relation before trying to build its relcache entry. The implication is that if someone committed a reindex of a critical system index at about the same time that some other backend were starting up without a valid pg_internal.init file, the second one might PANIC due to not seeing any valid version of the index's pg_class row. Improbable case, but definitely not impossible.
2008-04-16 20:23:04 +02:00
/*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* RelationBuildDesc
*
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* Build a relation descriptor. The caller must hold at least
Fix LOAD_CRIT_INDEX() macro to take out AccessShareLock on the system index it is trying to build a relcache entry for. This is an oversight in my 8.2 patch that tried to ensure we always took a lock on a relation before trying to build its relcache entry. The implication is that if someone committed a reindex of a critical system index at about the same time that some other backend were starting up without a valid pg_internal.init file, the second one might PANIC due to not seeing any valid version of the index's pg_class row. Improbable case, but definitely not impossible.
2008-04-16 20:23:04 +02:00
* AccessShareLock on the target relid.
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
*
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* The new descriptor is inserted into the hash table if insertIt is true.
*
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
* Returns NULL if no pg_class row could be found for the given relid
* (suggesting we are trying to access a just-deleted relation).
* Any other error is reported via elog.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
static Relation
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
RelationBuildDesc(Oid targetRelId, bool insertIt)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Relation relation;
Oid relid;
Repair bugs discussed in pghackers thread of 15 May 1999: creation of a relcache entry no longer leaks a small amount of memory. index_endscan now releases all the memory acquired by index_beginscan, so callers of it should NOT pfree the scan descriptor anymore.
1999-12-30 06:05:13 +01:00
HeapTuple pg_class_tuple;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Form_pg_class relp;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* find the tuple in pg_class corresponding to the given relation id
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
pg_class_tuple = ScanPgRelation(targetRelId, true);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* if no such tuple exists, return NULL
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
if (!HeapTupleIsValid(pg_class_tuple))
return NULL;
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* get information from the pg_class_tuple
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
oid is needed, it is added at the end of the struct (after the null bitmap, if present). Per Tom Lane's suggestion the information whether a tuple has an oid or not is carried in the tuple descriptor. For debugging reasons tdhasoid is of type char, not bool. There are predefined values for WITHOID, WITHOUTOID and UNDEFOID. This patch has been generated against a cvs snapshot from last week and I don't expect it to apply cleanly to current sources. While I post it here for public review, I'm working on a new version against a current snapshot. (There's been heavy activity recently; hope to catch up some day ...) This is a long patch; if it is too hard to swallow, I can provide it in smaller pieces: Part 1: Accessor macros Part 2: tdhasoid in TupDesc Part 3: Regression test Part 4: Parameter withoid to heap_addheader Part 5: Eliminate t_oid from HeapTupleHeader Part 2 is the most hairy part because of changes in the executor and even in the parser; the other parts are straightforward. Up to part 4 the patched postmaster stays binary compatible to databases created with an unpatched version. Part 5 is small (100 lines) and finally breaks compatibility. Manfred Koizar
2002-07-20 07:16:59 +02:00
relid = HeapTupleGetOid(pg_class_tuple);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
relp = (Form_pg_class) GETSTRUCT(pg_class_tuple);
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
Assert(relid == targetRelId);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* allocate storage for the relation descriptor, and copy pg_class_tuple
* to relation->rd_rel.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
relation = AllocateRelationDesc(relp);
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize the relation's relation id (relation->rd_id)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
heap_fetch requires buffer pointer, must be released; heap_getnext no longer returns buffer pointer, can be gotten from scan; descriptor; bootstrap can create multi-key indexes; pg_procname index now is multi-key index; oidint2, oidint4, oidname are gone (must be removed from regression tests); use System Cache rather than sequential scan in many places; heap_modifytuple no longer takes buffer parameter; remove unused buffer parameter in a few other functions; oid8 is not index-able; remove some use of single-character variable names; cleanup Buffer variables usage and scan descriptor looping; cleaned up allocation and freeing of tuples; 18k lines of diff;
1998-08-19 04:04:17 +02:00
RelationGetRelid(relation) = relid;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* normal relations are not nailed into the cache; nor can a pre-existing
* relation be new. It could be temp though. (Actually, it could be new
* too, but it's okay to forget that fact if forced to flush the entry.)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
relation->rd_refcnt = 0;
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
relation->rd_isnailed = false;
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
relation->rd_createSubid = InvalidSubTransactionId;
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
relation->rd_newRelfilenodeSubid = InvalidSubTransactionId;
Modify the relcache to record the temp status of both local and nonlocal temp relations; this is no more expensive than before, now that we have pg_class.relistemp. Insert tests into bufmgr.c to prevent attempting to fetch pages from nonlocal temp relations. This provides a low-level defense against bugs-of-omission allowing temp pages to be loaded into shared buffers, as in the contrib/pgstattuple problem reported by Stuart Bishop. While at it, tweak a bunch of places to use new relcache tests (instead of expensive probes into pg_namespace) to detect local or nonlocal temp tables.
2009-04-01 00:12:48 +02:00
relation->rd_istemp = relation->rd_rel->relistemp;
if (relation->rd_istemp)
relation->rd_islocaltemp = isTempOrToastNamespace(relation->rd_rel->relnamespace);
else
relation->rd_islocaltemp = false;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize the tuple descriptor (relation->rd_att).
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
RelationBuildTupleDesc(relation);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* Fetch rules and triggers that affect this relation
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
if (relation->rd_rel->relhasrules)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationBuildRuleLock(relation);
else
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
{
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
relation->rd_rules = NULL;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
relation->rd_rulescxt = NULL;
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Replace pg_class.reltriggers with relhastriggers, which is just a boolean hint ("there might be triggers") rather than an exact count. This is necessary catalog infrastructure for the upcoming patch to reduce the strength of locking needed for trigger addition/removal. Split out and committed separately for ease of reviewing/testing. In passing, also get rid of the unused pg_class columns relukeys, relfkeys, and relrefs, which haven't been maintained in many years and now have no chance of ever being maintained (because of wishing to avoid locking). Simon Riggs
2008-11-09 22:24:33 +01:00
if (relation->rd_rel->relhastriggers)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationBuildTriggers(relation);
else
relation->trigdesc = NULL;
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* if it's an index, initialize index-related information
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (OidIsValid(relation->rd_rel->relam))
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
RelationInitIndexAccessInfo(relation);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
/* extract reloptions if any */
RelationParseRelOptions(relation, pg_class_tuple);
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize the relation lock manager information
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
RelationInitLockInfo(relation); /* see lmgr.c */
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
/*
* initialize physical addressing information for the relation
*/
RelationInitPhysicalAddr(relation);
New file naming. Database OID is used as "tablespace" id and relation OID is used as file node on creation but may be changed later if required. Regression Tests Approved (c) -:)))
2000-10-16 16:52:28 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* make sure relation is marked as having no open file yet */
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
relation->rd_smgr = NULL;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
/*
* now we can free the memory allocated for pg_class_tuple
*/
heap_freetuple(pg_class_tuple);
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* Insert newly created relation into relcache hash table, if requested.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
if (insertIt)
RelationCacheInsert(relation);
Repair bugs discussed in pghackers thread of 15 May 1999: creation of a relcache entry no longer leaks a small amount of memory. index_endscan now releases all the memory acquired by index_beginscan, so callers of it should NOT pfree the scan descriptor anymore.
1999-12-30 06:05:13 +01:00
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
/* It's fully valid */
relation->rd_isvalid = true;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
return relation;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
/*
* Initialize the physical addressing info (RelFileNode) for a relcache entry
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
*
* Note: at the physical level, relations in the pg_global tablespace must
* be treated as shared, even if relisshared isn't set. Hence we do not
* look at relisshared here.
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
*/
static void
RelationInitPhysicalAddr(Relation relation)
{
if (relation->rd_rel->reltablespace)
relation->rd_node.spcNode = relation->rd_rel->reltablespace;
else
relation->rd_node.spcNode = MyDatabaseTableSpace;
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
if (relation->rd_node.spcNode == GLOBALTABLESPACE_OID)
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
relation->rd_node.dbNode = InvalidOid;
else
relation->rd_node.dbNode = MyDatabaseId;
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
if (relation->rd_rel->relfilenode)
relation->rd_node.relNode = relation->rd_rel->relfilenode;
else
{
/* Consult the relation mapper */
relation->rd_node.relNode =
RelationMapOidToFilenode(relation->rd_id,
relation->rd_rel->relisshared);
if (!OidIsValid(relation->rd_node.relNode))
elog(ERROR, "could not find relation mapping for relation \"%s\", OID %u",
RelationGetRelationName(relation), relation->rd_id);
}
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
}
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
/*
* Initialize index-access-method support data for an index relation
*/
void
RelationInitIndexAccessInfo(Relation relation)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
HeapTuple tuple;
Form_pg_am aform;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
Datum indclassDatum;
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
Datum indoptionDatum;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
bool isnull;
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
oidvector *indclass;
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int2vector *indoption;
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
MemoryContext indexcxt;
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
MemoryContext oldcontext;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
int natts;
Clean up some minor problems exposed by further thought about Panon's bug report on old-style functions invoked by RI triggers. We had a number of other places that were being sloppy about which memory context FmgrInfo subsidiary data will be allocated in. Turns out none of them actually cause a problem in 7.1, but this is for arcane reasons such as the fact that old-style triggers aren't supported anyway. To avoid getting burnt later, I've restructured the trigger support so that we don't keep trigger FmgrInfo structs in relcache memory. Some other related cleanups too: it's not really necessary to call fmgr_info at all while setting up the index support info in relcache entries, because those ScanKeyEntry structs are never used to invoke the functions. This should speed up relcache initialization a tiny bit.
2001-06-01 04:41:36 +02:00
uint16 amstrategies;
uint16 amsupport;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
* Make a copy of the pg_index entry for the index. Since pg_index
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* contains variable-length and possibly-null fields, we have to do this
* honestly rather than just treating it as a Form_pg_index struct.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
Wrap calls to SearchSysCache and related functions using macros. The purpose of this change is to eliminate the need for every caller of SearchSysCache, SearchSysCacheCopy, SearchSysCacheExists, GetSysCacheOid, and SearchSysCacheList to know the maximum number of allowable keys for a syscache entry (currently 4). This will make it far easier to increase the maximum number of keys in a future release should we choose to do so, and it makes the code shorter, too. Design and review by Tom Lane.
2010-02-14 19:42:19 +01:00
tuple = SearchSysCache1(INDEXRELID,
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
ObjectIdGetDatum(RelationGetRelid(relation)));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (!HeapTupleIsValid(tuple))
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "cache lookup failed for index %u",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
RelationGetRelid(relation));
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
oldcontext = MemoryContextSwitchTo(CacheMemoryContext);
relation->rd_indextuple = heap_copytuple(tuple);
relation->rd_index = (Form_pg_index) GETSTRUCT(relation->rd_indextuple);
MemoryContextSwitchTo(oldcontext);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
ReleaseSysCache(tuple);
/*
* Make a copy of the pg_am entry for the index's access method
*/
Wrap calls to SearchSysCache and related functions using macros. The purpose of this change is to eliminate the need for every caller of SearchSysCache, SearchSysCacheCopy, SearchSysCacheExists, GetSysCacheOid, and SearchSysCacheList to know the maximum number of allowable keys for a syscache entry (currently 4). This will make it far easier to increase the maximum number of keys in a future release should we choose to do so, and it makes the code shorter, too. Design and review by Tom Lane.
2010-02-14 19:42:19 +01:00
tuple = SearchSysCache1(AMOID, ObjectIdGetDatum(relation->rd_rel->relam));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (!HeapTupleIsValid(tuple))
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "cache lookup failed for access method %u",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
relation->rd_rel->relam);
aform = (Form_pg_am) MemoryContextAlloc(CacheMemoryContext, sizeof *aform);
memcpy(aform, GETSTRUCT(tuple), sizeof *aform);
ReleaseSysCache(tuple);
relation->rd_am = aform;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
natts = relation->rd_rel->relnatts;
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
if (natts != relation->rd_index->indnatts)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "relnatts disagrees with indnatts for index %u",
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
RelationGetRelid(relation));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
amstrategies = aform->amstrategies;
amsupport = aform->amsupport;
Clean up some minor problems exposed by further thought about Panon's bug report on old-style functions invoked by RI triggers. We had a number of other places that were being sloppy about which memory context FmgrInfo subsidiary data will be allocated in. Turns out none of them actually cause a problem in 7.1, but this is for arcane reasons such as the fact that old-style triggers aren't supported anyway. To avoid getting burnt later, I've restructured the trigger support so that we don't keep trigger FmgrInfo structs in relcache memory. Some other related cleanups too: it's not really necessary to call fmgr_info at all while setting up the index support info in relcache entries, because those ScanKeyEntry structs are never used to invoke the functions. This should speed up relcache initialization a tiny bit.
2001-06-01 04:41:36 +02:00
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Make the private context to hold index access info. The reason we need
* a context, and not just a couple of pallocs, is so that we won't leak
* any subsidiary info attached to fmgr lookup records.
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
*
* Context parameters are set on the assumption that it'll probably not
* contain much data.
*/
indexcxt = AllocSetContextCreate(CacheMemoryContext,
RelationGetRelationName(relation),
Tweak default memory context allocation policy so that a context is not given any malloc block until something is first allocated in it; but thereafter, MemoryContextReset won't release that first malloc block. This preserves the quick-reset property of the original policy, without forcing 8K to be allocated to every context whether any of it is ever used or not. Also, remove some more no-longer-needed explicit freeing during ExecEndPlan.
2002-12-15 22:01:34 +01:00
ALLOCSET_SMALL_MINSIZE,
ALLOCSET_SMALL_INITSIZE,
ALLOCSET_SMALL_MAXSIZE);
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
relation->rd_indexcxt = indexcxt;
/*
* Allocate arrays to hold data
*/
Arrange to cache fmgr lookup information for an index's access method routines in the index's relcache entry, instead of doing a fresh fmgr_info on every index access. We were already doing this for the index's opclass support functions; not sure why we didn't think to do it for the AM functions too. This supersedes the former method of caching (only) amgettuple in indexscan scan descriptors; it's an improvement because the function lookup can be amortized across multiple statements instead of being repeated for each statement. Even though lookup for builtin functions is pretty cheap, this seems to drop a percent or two off some simple benchmarks.
2005-05-28 01:31:21 +02:00
relation->rd_aminfo = (RelationAmInfo *)
MemoryContextAllocZero(indexcxt, sizeof(RelationAmInfo));
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
relation->rd_opfamily = (Oid *)
MemoryContextAllocZero(indexcxt, natts * sizeof(Oid));
relation->rd_opcintype = (Oid *)
MemoryContextAllocZero(indexcxt, natts * sizeof(Oid));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (amstrategies > 0)
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
relation->rd_operator = (Oid *)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
MemoryContextAllocZero(indexcxt,
natts * amstrategies * sizeof(Oid));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
else
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
relation->rd_operator = NULL;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Clean up some minor problems exposed by further thought about Panon's bug report on old-style functions invoked by RI triggers. We had a number of other places that were being sloppy about which memory context FmgrInfo subsidiary data will be allocated in. Turns out none of them actually cause a problem in 7.1, but this is for arcane reasons such as the fact that old-style triggers aren't supported anyway. To avoid getting burnt later, I've restructured the trigger support so that we don't keep trigger FmgrInfo structs in relcache memory. Some other related cleanups too: it's not really necessary to call fmgr_info at all while setting up the index support info in relcache entries, because those ScanKeyEntry structs are never used to invoke the functions. This should speed up relcache initialization a tiny bit.
2001-06-01 04:41:36 +02:00
if (amsupport > 0)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
pgindent run on all C files. Java run to follow. initdb/regression tests pass.
2001-10-25 07:50:21 +02:00
int nsupport = natts * amsupport;
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
relation->rd_support = (RegProcedure *)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
MemoryContextAllocZero(indexcxt, nsupport * sizeof(RegProcedure));
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
relation->rd_supportinfo = (FmgrInfo *)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
MemoryContextAllocZero(indexcxt, nsupport * sizeof(FmgrInfo));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
else
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
{
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
relation->rd_support = NULL;
relation->rd_supportinfo = NULL;
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
relation->rd_indoption = (int16 *)
MemoryContextAllocZero(indexcxt, natts * sizeof(int16));
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
/*
* indclass cannot be referenced directly through the C struct, because it
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* comes after the variable-width indkey field. Must extract the datum
* the hard way...
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
*/
indclassDatum = fastgetattr(relation->rd_indextuple,
Anum_pg_index_indclass,
GetPgIndexDescriptor(),
&isnull);
Assert(!isnull);
indclass = (oidvector *) DatumGetPointer(indclassDatum);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
/*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* Fill the operator and support procedure OID arrays, as well as the info
* about opfamilies and opclass input types. (aminfo and supportinfo are
* left as zeroes, and are filled on-the-fly when used)
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
*/
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
IndexSupportInitialize(indclass,
relation->rd_operator, relation->rd_support,
relation->rd_opfamily, relation->rd_opcintype,
Clean up some minor problems exposed by further thought about Panon's bug report on old-style functions invoked by RI triggers. We had a number of other places that were being sloppy about which memory context FmgrInfo subsidiary data will be allocated in. Turns out none of them actually cause a problem in 7.1, but this is for arcane reasons such as the fact that old-style triggers aren't supported anyway. To avoid getting burnt later, I've restructured the trigger support so that we don't keep trigger FmgrInfo structs in relcache memory. Some other related cleanups too: it's not really necessary to call fmgr_info at all while setting up the index support info in relcache entries, because those ScanKeyEntry structs are never used to invoke the functions. This should speed up relcache initialization a tiny bit.
2001-06-01 04:41:36 +02:00
amstrategies, amsupport, natts);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
/*
* Similarly extract indoption and copy it to the cache entry
*/
indoptionDatum = fastgetattr(relation->rd_indextuple,
Anum_pg_index_indoption,
GetPgIndexDescriptor(),
&isnull);
Assert(!isnull);
indoption = (int2vector *) DatumGetPointer(indoptionDatum);
memcpy(relation->rd_indoption, indoption->values, natts * sizeof(int16));
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/*
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
* expressions, predicate, exclusion caches will be filled later
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
*/
relation->rd_indexprs = NIL;
relation->rd_indpred = NIL;
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
relation->rd_exclops = NULL;
relation->rd_exclprocs = NULL;
relation->rd_exclstrats = NULL;
Arrange to cache btree metapage data in the relcache entry for the index, thereby saving a visit to the metapage in most index searches/updates. This wouldn't actually save any I/O (since in the old regime the metapage generally stayed in cache anyway), but it does provide a useful decrease in bufmgr traffic in high-contention scenarios. Per my recent proposal.
2006-04-26 00:46:05 +02:00
relation->rd_amcache = NULL;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* IndexSupportInitialize
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
* Initializes an index's cached opclass information,
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
* given the index's pg_index.indclass entry.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
* Data is returned into *indexOperator, *indexSupport, *opFamily, and
* *opcInType, which are arrays allocated by the caller.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
* The caller also passes maxStrategyNumber, maxSupportNumber, and
* maxAttributeNumber, since these indicate the size of the arrays
* it has allocated --- but in practice these numbers must always match
* those obtainable from the system catalog entries for the index and
* access method.
*/
static void
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
IndexSupportInitialize(oidvector *indclass,
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Oid *indexOperator,
RegProcedure *indexSupport,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Oid *opFamily,
Oid *opcInType,
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
StrategyNumber maxStrategyNumber,
StrategyNumber maxSupportNumber,
AttrNumber maxAttributeNumber)
{
int attIndex;
for (attIndex = 0; attIndex < maxAttributeNumber; attIndex++)
{
OpClassCacheEnt *opcentry;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
if (!OidIsValid(indclass->values[attIndex]))
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "bogus pg_index tuple");
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* look up the info for this opclass, using a cache */
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
opcentry = LookupOpclassInfo(indclass->values[attIndex],
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
maxStrategyNumber,
maxSupportNumber);
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
/* copy cached data into relcache entry */
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
opFamily[attIndex] = opcentry->opcfamily;
opcInType[attIndex] = opcentry->opcintype;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (maxStrategyNumber > 0)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
memcpy(&indexOperator[attIndex * maxStrategyNumber],
opcentry->operatorOids,
maxStrategyNumber * sizeof(Oid));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (maxSupportNumber > 0)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
memcpy(&indexSupport[attIndex * maxSupportNumber],
opcentry->supportProcs,
maxSupportNumber * sizeof(RegProcedure));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
}
/*
* LookupOpclassInfo
*
* This routine maintains a per-opclass cache of the information needed
* by IndexSupportInitialize(). This is more efficient than relying on
* the catalog cache, because we can load all the info about a particular
* opclass in a single indexscan of pg_amproc or pg_amop.
*
* The information from pg_am about expected range of strategy and support
* numbers is passed in, rather than being looked up, mainly because the
* caller will have it already.
*
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
* Note there is no provision for flushing the cache. This is OK at the
* moment because there is no way to ALTER any interesting properties of an
* existing opclass --- all you can do is drop it, which will result in
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* a useless but harmless dead entry in the cache. To support altering
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
* opclass membership (not the same as opfamily membership!), we'd need to
* be able to flush this cache as well as the contents of relcache entries
* for indexes.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
static OpClassCacheEnt *
LookupOpclassInfo(Oid operatorClassOid,
StrategyNumber numStrats,
StrategyNumber numSupport)
{
OpClassCacheEnt *opcentry;
bool found;
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
Relation rel;
SysScanDesc scan;
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
ScanKeyData skey[3];
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
HeapTuple htup;
bool indexOK;
if (OpClassCache == NULL)
{
/* First time through: initialize the opclass cache */
HASHCTL ctl;
MemSet(&ctl, 0, sizeof(ctl));
ctl.keysize = sizeof(Oid);
ctl.entrysize = sizeof(OpClassCacheEnt);
Marginal hack to use a specialized hash function for dynahash hashtables whose keys are OIDs. The only one that looks particularly performance critical is the relcache hashtable, but as long as we've got the function we may as well use it wherever it's applicable.
2005-04-14 22:32:43 +02:00
ctl.hash = oid_hash;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
OpClassCache = hash_create("Operator class cache", 64,
&ctl, HASH_ELEM | HASH_FUNCTION);
Remove a couple of unnecessary calls of CreateCacheMemoryContext. These probably got there via blind copy-and-paste from one of the legitimate callers, so rearrange and comment that code a bit to make it clearer that this isn't a necessary prerequisite to hash_create. Per observation from Robert Haas.
2009-12-27 19:55:52 +01:00
/* Also make sure CacheMemoryContext exists */
if (!CacheMemoryContext)
CreateCacheMemoryContext();
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
opcentry = (OpClassCacheEnt *) hash_search(OpClassCache,
(void *) &operatorClassOid,
HASH_ENTER, &found);
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
if (!found)
{
/* Need to allocate memory for new entry */
opcentry->valid = false; /* until known OK */
opcentry->numStrats = numStrats;
opcentry->numSupport = numSupport;
if (numStrats > 0)
opcentry->operatorOids = (Oid *)
MemoryContextAllocZero(CacheMemoryContext,
numStrats * sizeof(Oid));
else
opcentry->operatorOids = NULL;
if (numSupport > 0)
opcentry->supportProcs = (RegProcedure *)
MemoryContextAllocZero(CacheMemoryContext,
numSupport * sizeof(RegProcedure));
else
opcentry->supportProcs = NULL;
}
else
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
Assert(numStrats == opcentry->numStrats);
Assert(numSupport == opcentry->numSupport);
}
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
/*
* When testing for cache-flush hazards, we intentionally disable the
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* operator class cache and force reloading of the info on each call. This
* is helpful because we want to test the case where a cache flush occurs
* while we are loading the info, and it's very hard to provoke that if
* this happens only once per opclass per backend.
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
*/
#if defined(CLOBBER_CACHE_ALWAYS)
opcentry->valid = false;
#endif
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
if (opcentry->valid)
return opcentry;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Improve test coverage of CLOBBER_CACHE_ALWAYS by having it also force reloading of operator class information on each use of LookupOpclassInfo. Had this been in place a year ago, it would have helped me find a bug in the then-new 'operator family' code. Now that we have a build farm member testing CLOBBER_CACHE_ALWAYS on a regular basis, it seems worth expending a little bit of effort here.
2007-11-28 21:44:26 +01:00
* Need to fill in new entry.
*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* To avoid infinite recursion during startup, force heap scans if we're
* looking up info for the opclasses used by the indexes we would like to
* reference here.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
indexOK = criticalRelcachesBuilt ||
(operatorClassOid != OID_BTREE_OPS_OID &&
operatorClassOid != INT2_BTREE_OPS_OID);
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
/*
* We have to fetch the pg_opclass row to determine its opfamily and
* opcintype, which are needed to look up the operators and functions.
* It'd be convenient to use the syscache here, but that probably doesn't
* work while bootstrapping.
*/
ScanKeyInit(&skey[0],
ObjectIdAttributeNumber,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(operatorClassOid));
rel = heap_open(OperatorClassRelationId, AccessShareLock);
scan = systable_beginscan(rel, OpclassOidIndexId, indexOK,
SnapshotNow, 1, skey);
if (HeapTupleIsValid(htup = systable_getnext(scan)))
{
Form_pg_opclass opclassform = (Form_pg_opclass) GETSTRUCT(htup);
opcentry->opcfamily = opclassform->opcfamily;
opcentry->opcintype = opclassform->opcintype;
}
else
elog(ERROR, "could not find tuple for opclass %u", operatorClassOid);
systable_endscan(scan);
heap_close(rel, AccessShareLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Scan pg_amop to obtain operators for the opclass. We only fetch the
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
* default ones (those with lefttype = righttype = opcintype).
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
if (numStrats > 0)
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Anum_pg_amop_amopfamily,
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
BTEqualStrategyNumber, F_OIDEQ,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
ObjectIdGetDatum(opcentry->opcfamily));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[1],
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Anum_pg_amop_amoplefttype,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(opcentry->opcintype));
ScanKeyInit(&skey[2],
Anum_pg_amop_amoprighttype,
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
BTEqualStrategyNumber, F_OIDEQ,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
ObjectIdGetDatum(opcentry->opcintype));
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
rel = heap_open(AccessMethodOperatorRelationId, AccessShareLock);
scan = systable_beginscan(rel, AccessMethodStrategyIndexId, indexOK,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
SnapshotNow, 3, skey);
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
while (HeapTupleIsValid(htup = systable_getnext(scan)))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
Form_pg_amop amopform = (Form_pg_amop) GETSTRUCT(htup);
if (amopform->amopstrategy <= 0 ||
(StrategyNumber) amopform->amopstrategy > numStrats)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "invalid amopstrategy number %d for opclass %u",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
amopform->amopstrategy, operatorClassOid);
opcentry->operatorOids[amopform->amopstrategy - 1] =
amopform->amopopr;
}
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
systable_endscan(scan);
heap_close(rel, AccessShareLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Scan pg_amproc to obtain support procs for the opclass. We only fetch
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
* the default ones (those with lefttype = righttype = opcintype).
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
if (numSupport > 0)
{
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Anum_pg_amproc_amprocfamily,
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
BTEqualStrategyNumber, F_OIDEQ,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
ObjectIdGetDatum(opcentry->opcfamily));
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[1],
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Anum_pg_amproc_amproclefttype,
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
BTEqualStrategyNumber, F_OIDEQ,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
ObjectIdGetDatum(opcentry->opcintype));
ScanKeyInit(&skey[2],
Anum_pg_amproc_amprocrighttype,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(opcentry->opcintype));
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
rel = heap_open(AccessMethodProcedureRelationId, AccessShareLock);
scan = systable_beginscan(rel, AccessMethodProcedureIndexId, indexOK,
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
SnapshotNow, 3, skey);
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
while (HeapTupleIsValid(htup = systable_getnext(scan)))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
Form_pg_amproc amprocform = (Form_pg_amproc) GETSTRUCT(htup);
if (amprocform->amprocnum <= 0 ||
(StrategyNumber) amprocform->amprocnum > numSupport)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "invalid amproc number %d for opclass %u",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
amprocform->amprocnum, operatorClassOid);
opcentry->supportProcs[amprocform->amprocnum - 1] =
amprocform->amproc;
}
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
systable_endscan(scan);
heap_close(rel, AccessShareLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
opcentry->valid = true;
return opcentry;
}
/*
* formrdesc
*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* This is a special cut-down version of RelationBuildDesc(),
* used while initializing the relcache.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* The relation descriptor is built just from the supplied parameters,
Clean up syscache so that recursive invocation is safe, and remove error message about recursive use of a syscache. Also remove most of the specialized indexscan routines in indexing.c --- it turns out that catcache.c is perfectly able to perform the indexscan for itself, in fact has already looked up all the information needed to do so! This should be faster as well as needing far less boilerplate code.
2000-11-10 01:33:12 +01:00
* without actually looking at any system table entries. We cheat
* quite a lot since we only need to work for a few basic system
Support toasting of shared system relations, and provide toast tables for pg_database, pg_shadow, pg_group, all of which now have potentially-long fields. Along the way, get rid of SharedSystemRelationNames list: shared rels are now identified in their include/pg_catalog/*.h files by a BKI_SHARED_RELATION macro, while indexes and toast rels inherit sharedness automatically from their parent table. Fix some bugs with failure to detoast pg_group.grolist during ALTER GROUP.
2002-04-27 23:24:34 +02:00
* catalogs.
*
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* formrdesc is currently used for: pg_database, pg_authid, pg_auth_members,
* pg_class, pg_attribute, pg_proc, and pg_type
* (see RelationCacheInitializePhase2/3).
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
* Note that these catalogs can't have constraints (except attnotnull),
* default values, rules, or triggers, since we don't cope with any of that.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* (Well, actually, this only matters for properties that need to be valid
* during bootstrap or before RelationCacheInitializePhase3 runs, and none of
* these properties matter then...)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
* NOTE: we assume we are already switched into CacheMemoryContext.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
static void
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
formrdesc(const char *relationName, Oid relationReltype,
bool isshared, bool hasoids,
int natts, const FormData_pg_attribute *attrs)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Relation relation;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
int i;
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
bool has_not_null;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
* allocate new relation desc, clear all fields of reldesc
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Add new palloc0 call as merge of palloc and MemSet(0).
2002-11-13 01:39:48 +01:00
relation = (Relation) palloc0(sizeof(RelationData));
Install infrastructure for shared-memory free space map. Doesn't actually do anything yet, but it has the necessary connections to initialization and so forth. Make some gestures towards allowing number of blocks in a relation to be BlockNumber, ie, unsigned int, rather than signed int. (I doubt I got all the places that are sloppy about it, yet.) On the way, replace the hardwired NLOCKS_PER_XACT fudge factor with a GUC variable.
2001-06-28 01:31:40 +02:00
/* make sure relation is marked as having no open file yet */
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
relation->rd_smgr = NULL;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
* initialize reference count: 1 because it is nailed in cache
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
relation->rd_refcnt = 1;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* all entries built with this routine are nailed-in-cache; none are for
* new or temp relations.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
relation->rd_isnailed = true;
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
relation->rd_createSubid = InvalidSubTransactionId;
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
relation->rd_newRelfilenodeSubid = InvalidSubTransactionId;
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
relation->rd_istemp = false;
Modify the relcache to record the temp status of both local and nonlocal temp relations; this is no more expensive than before, now that we have pg_class.relistemp. Insert tests into bufmgr.c to prevent attempting to fetch pages from nonlocal temp relations. This provides a low-level defense against bugs-of-omission allowing temp pages to be loaded into shared buffers, as in the contrib/pgstattuple problem reported by Stuart Bishop. While at it, tweak a bunch of places to use new relcache tests (instead of expensive probes into pg_namespace) to detect local or nonlocal temp tables.
2009-04-01 00:12:48 +02:00
relation->rd_islocaltemp = false;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize relation tuple form
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* The data we insert here is pretty incomplete/bogus, but it'll serve to
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* get us launched. RelationCacheInitializePhase3() will read the real
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* data from pg_class and replace what we've done here. Note in
* particular that relowner is left as zero; this cues
* RelationCacheInitializePhase3 that the real data isn't there yet.
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*/
Add new palloc0 call as merge of palloc and MemSet(0).
2002-11-13 01:39:48 +01:00
relation->rd_rel = (Form_pg_class) palloc0(CLASS_TUPLE_SIZE);
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
namestrcpy(&relation->rd_rel->relname, relationName);
relation->rd_rel->relnamespace = PG_CATALOG_NAMESPACE;
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
relation->rd_rel->reltype = relationReltype;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* It's important to distinguish between shared and non-shared relations,
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* even at bootstrap time, to make sure we know where they are stored.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
relation->rd_rel->relisshared = isshared;
if (isshared)
relation->rd_rel->reltablespace = GLOBALTABLESPACE_OID;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Add a "relistemp" boolean column to pg_class, which is true for temporary relations (including a temp table's indexes and toast table/index), and false for normal relations. For ease of checking, this commit just adds the column and fills it correctly --- revising the relation access machinery to use it will come separately.
2009-03-31 19:59:56 +02:00
/*
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
* Likewise, we must know if a relation is temp ... but formrdesc is not
* used for any temp relations.
Add a "relistemp" boolean column to pg_class, which is true for temporary relations (including a temp table's indexes and toast table/index), and false for normal relations. For ease of checking, this commit just adds the column and fills it correctly --- revising the relation access machinery to use it will come separately.
2009-03-31 19:59:56 +02:00
*/
relation->rd_rel->relistemp = false;
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
relation->rd_rel->relpages = 1;
relation->rd_rel->reltuples = 1;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
relation->rd_rel->relkind = RELKIND_RELATION;
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
relation->rd_rel->relhasoids = hasoids;
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
relation->rd_rel->relnatts = (int16) natts;
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize attribute tuple form
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
pgindent run.
2002-09-04 22:31:48 +02:00
* Unlike the case with the relation tuple, this data had better be right
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* because it will never be replaced. The data comes from
* src/include/catalog/ headers via genbki.pl.
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
*/
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
relation->rd_att = CreateTemplateTupleDesc(natts, hasoids);
Fix problems with cached tuple descriptors disappearing while still in use by creating a reference-count mechanism, similar to what we did a long time ago for catcache entries. The back branches have an ugly solution involving lots of extra copies, but this way is more efficient. Reference counting is only applied to tupdescs that are actually in caches --- there seems no need to use it for tupdescs that are generated in the executor, since they'll go away during plan shutdown by virtue of being in the per-query memory context. Neil Conway and Tom Lane
2006-06-16 20:42:24 +02:00
relation->rd_att->tdrefcount = 1; /* mark as refcounted */
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
relation->rd_att->tdtypeid = relationReltype;
relation->rd_att->tdtypmod = -1; /* unnecessary, but... */
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize tuple desc info
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
has_not_null = false;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
for (i = 0; i < natts; i++)
{
Adjust creation/destruction of TupleDesc data structure to reduce the number of palloc calls. This has a salutory impact on plpgsql operations with record variables (which create and destroy tupdescs constantly) and probably helps a bit in some other cases too.
2005-03-07 05:42:17 +01:00
memcpy(relation->rd_att->attrs[i],
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
&attrs[i],
Support column-level privileges, as required by SQL standard. Stephen Frost, with help from KaiGai Kohei and others
2009-01-22 21:16:10 +01:00
ATTRIBUTE_FIXED_PART_SIZE);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
has_not_null |= attrs[i].attnotnull;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* make sure attcacheoff is valid */
relation->rd_att->attrs[i]->attcacheoff = -1;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* initialize first attribute's attcacheoff, cf RelationBuildTupleDesc */
relation->rd_att->attrs[0]->attcacheoff = 0;
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
/* mark not-null status */
if (has_not_null)
{
TupleConstr *constr = (TupleConstr *) palloc0(sizeof(TupleConstr));
constr->has_not_null = true;
relation->rd_att->constr = constr;
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Clean up some minor problems exposed by further thought about Panon's bug report on old-style functions invoked by RI triggers. We had a number of other places that were being sloppy about which memory context FmgrInfo subsidiary data will be allocated in. Turns out none of them actually cause a problem in 7.1, but this is for arcane reasons such as the fact that old-style triggers aren't supported anyway. To avoid getting burnt later, I've restructured the trigger support so that we don't keep trigger FmgrInfo structs in relcache memory. Some other related cleanups too: it's not really necessary to call fmgr_info at all while setting up the index support info in relcache entries, because those ScanKeyEntry structs are never used to invoke the functions. This should speed up relcache initialization a tiny bit.
2001-06-01 04:41:36 +02:00
* initialize relation id from info in att array (my, this is ugly)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
heap_fetch requires buffer pointer, must be released; heap_getnext no longer returns buffer pointer, can be gotten from scan; descriptor; bootstrap can create multi-key indexes; pg_procname index now is multi-key index; oidint2, oidint4, oidname are gone (must be removed from regression tests); use System Cache rather than sequential scan in many places; heap_modifytuple no longer takes buffer parameter; remove unused buffer parameter in a few other functions; oid8 is not index-able; remove some use of single-character variable names; cleanup Buffer variables usage and scan descriptor looping; cleaned up allocation and freeing of tuples; 18k lines of diff;
1998-08-19 04:04:17 +02:00
RelationGetRelid(relation) = relation->rd_att->attrs[0]->attrelid;
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/*
* All relations made with formrdesc are mapped. This is necessarily so
* because there is no other way to know what filenode they currently
* have. In bootstrap mode, add them to the initial relation mapper data,
* specifying that the initial filenode is the same as the OID.
*/
relation->rd_rel->relfilenode = InvalidOid;
if (IsBootstrapProcessingMode())
RelationMapUpdateMap(RelationGetRelid(relation),
RelationGetRelid(relation),
isshared, true);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
* initialize the relation lock manager information
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
*/
RelationInitLockInfo(relation); /* see lmgr.c */
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
/*
* initialize physical addressing information for the relation
*/
RelationInitPhysicalAddr(relation);
New file naming. Database OID is used as "tablespace" id and relation OID is used as file node on creation but may be changed later if required. Regression Tests Approved (c) -:)))
2000-10-16 16:52:28 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* initialize the rel-has-index flag, using hardwired knowledge
New relcache hash table with RelFileNode as key to be used from bufmgr - it would be nice to have separate hash in smgr for node <--> fd mappings, but for the moment it's easy to add new hash to relcache. Fixed small bug in xlog.c:ReadRecord.
2000-10-23 06:10:24 +02:00
*/
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
if (IsBootstrapProcessingMode())
{
/* In bootstrap mode, we have no indexes */
relation->rd_rel->relhasindex = false;
}
else
Clean up syscache so that recursive invocation is safe, and remove error message about recursive use of a syscache. Also remove most of the specialized indexscan routines in indexing.c --- it turns out that catcache.c is perfectly able to perform the indexscan for itself, in fact has already looked up all the information needed to do so! This should be faster as well as needing far less boilerplate code.
2000-11-10 01:33:12 +01:00
{
Support toasting of shared system relations, and provide toast tables for pg_database, pg_shadow, pg_group, all of which now have potentially-long fields. Along the way, get rid of SharedSystemRelationNames list: shared rels are now identified in their include/pg_catalog/*.h files by a BKI_SHARED_RELATION macro, while indexes and toast rels inherit sharedness automatically from their parent table. Fix some bugs with failure to detoast pg_group.grolist during ALTER GROUP.
2002-04-27 23:24:34 +02:00
/* Otherwise, all the rels formrdesc is used for have indexes */
relation->rd_rel->relhasindex = true;
Clean up syscache so that recursive invocation is safe, and remove error message about recursive use of a syscache. Also remove most of the specialized indexscan routines in indexing.c --- it turns out that catcache.c is perfectly able to perform the indexscan for itself, in fact has already looked up all the information needed to do so! This should be faster as well as needing far less boilerplate code.
2000-11-10 01:33:12 +01:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* add new reldesc to relcache
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Clean up syscache so that recursive invocation is safe, and remove error message about recursive use of a syscache. Also remove most of the specialized indexscan routines in indexing.c --- it turns out that catcache.c is perfectly able to perform the indexscan for itself, in fact has already looked up all the information needed to do so! This should be faster as well as needing far less boilerplate code.
2000-11-10 01:33:12 +01:00
RelationCacheInsert(relation);
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
/* It's fully valid */
relation->rd_isvalid = true;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
/* ----------------------------------------------------------------
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* Relation Descriptor Lookup Interface
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
* ----------------------------------------------------------------
*/
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
* RelationIdGetRelation
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
* Lookup a reldesc by OID; make one if not already in cache.
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
*
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
* Returns NULL if no pg_class row could be found for the given relid
* (suggesting we are trying to access a just-deleted relation).
* Any other error is reported via elog.
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
*
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
* NB: caller should already have at least AccessShareLock on the
* relation ID, else there are nasty race conditions.
*
* NB: relation ref count is incremented, or set to 1 if new entry.
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
* Caller should eventually decrement count. (Usually,
* that happens by calling RelationClose().)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Relation
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
RelationIdGetRelation(Oid relationId)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Relation rd;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
/*
* first try to find reldesc in the cache
*/
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationIdCacheLookup(relationId, rd);
if (RelationIsValid(rd))
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
{
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationIncrementReferenceCount(rd);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/* revalidate cache entry if necessary */
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
if (!rd->rd_isvalid)
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
{
/*
* Indexes only have a limited number of possible schema changes,
* and we don't want to use the full-blown procedure because it's
* a headache for indexes that reload itself depends on.
*/
if (rd->rd_rel->relkind == RELKIND_INDEX)
RelationReloadIndexInfo(rd);
else
RelationClearRelation(rd, true);
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
return rd;
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* no reldesc in the cache, so have RelationBuildDesc() build one and add
* it.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
rd = RelationBuildDesc(relationId, true);
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
if (RelationIsValid(rd))
RelationIncrementReferenceCount(rd);
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
return rd;
}
/* ----------------------------------------------------------------
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* cache invalidation support routines
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
* ----------------------------------------------------------------
*/
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
/*
* RelationIncrementReferenceCount
* Increments relation reference count.
*
* Note: bootstrap mode has its own weird ideas about relation refcount
* behavior; we ought to fix it someday, but for now, just disable
* reference count ownership tracking in bootstrap mode.
*/
void
RelationIncrementReferenceCount(Relation rel)
{
ResourceOwnerEnlargeRelationRefs(CurrentResourceOwner);
rel->rd_refcnt += 1;
if (!IsBootstrapProcessingMode())
ResourceOwnerRememberRelationRef(CurrentResourceOwner, rel);
}
/*
* RelationDecrementReferenceCount
* Decrements relation reference count.
*/
void
RelationDecrementReferenceCount(Relation rel)
{
Assert(rel->rd_refcnt > 0);
rel->rd_refcnt -= 1;
if (!IsBootstrapProcessingMode())
ResourceOwnerForgetRelationRef(CurrentResourceOwner, rel);
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
* RelationClose - close an open relation
*
Add some debugging support code (ifdef'd out in normal use).
2001-01-08 19:34:44 +01:00
* Actually, we just decrement the refcount.
*
* NOTE: if compiled with -DRELCACHE_FORCE_RELEASE then relcache entries
* will be freed as soon as their refcount goes to zero. In combination
* with aset.c's CLOBBER_FREED_MEMORY option, this provides a good test
* to catch references to already-released relcache entries. It slows
* things down quite a bit, however.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
void
RelationClose(Relation relation)
{
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* Note: no locking manipulations needed */
RelationDecrementReferenceCount(relation);
Add some debugging support code (ifdef'd out in normal use).
2001-01-08 19:34:44 +01:00
#ifdef RELCACHE_FORCE_RELEASE
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
if (RelationHasReferenceCountZero(relation) &&
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
relation->rd_createSubid == InvalidSubTransactionId &&
relation->rd_newRelfilenodeSubid == InvalidSubTransactionId)
Add some debugging support code (ifdef'd out in normal use).
2001-01-08 19:34:44 +01:00
RelationClearRelation(relation, false);
#endif
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
* RelationReloadIndexInfo - reload minimal information for an open index
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
* This function is used only for indexes. A relcache inval on an index
* can mean that its pg_class or pg_index row changed. There are only
* very limited changes that are allowed to an existing index's schema,
* so we can update the relcache entry without a complete rebuild; which
* is fortunate because we can't rebuild an index entry that is "nailed"
* and/or in active use. We support full replacement of the pg_class row,
* as well as updates of a few simple fields of the pg_index row.
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
* We can't necessarily reread the catalog rows right away; we might be
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
* in a failed transaction when we receive the SI notification. If so,
* RelationClearRelation just marks the entry as invalid by setting
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
* rd_isvalid to false. This routine is called to fix the entry when it
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
* is next needed.
Fix LOAD_CRIT_INDEX() macro to take out AccessShareLock on the system index it is trying to build a relcache entry for. This is an oversight in my 8.2 patch that tried to ensure we always took a lock on a relation before trying to build its relcache entry. The implication is that if someone committed a reindex of a critical system index at about the same time that some other backend were starting up without a valid pg_internal.init file, the second one might PANIC due to not seeing any valid version of the index's pg_class row. Improbable case, but definitely not impossible.
2008-04-16 20:23:04 +02:00
*
* We assume that at the time we are called, we have at least AccessShareLock
* on the target index. (Note: in the calls from RelationClearRelation,
* this is legitimate because we know the rel has positive refcount.)
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
*/
static void
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
RelationReloadIndexInfo(Relation relation)
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
{
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
bool indexOK;
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
HeapTuple pg_class_tuple;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Form_pg_class relp;
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
/* Should be called only for invalidated indexes */
Assert(relation->rd_rel->relkind == RELKIND_INDEX &&
!relation->rd_isvalid);
/* Should be closed at smgr level */
Assert(relation->rd_smgr == NULL);
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
Fix up rickety handling of relation-truncation interlocks. Move rd_targblock, rd_fsm_nblocks, and rd_vm_nblocks from relcache to the smgr relation entries, so that they will get reset to InvalidBlockNumber whenever an smgr-level flush happens. Because we now send smgr invalidation messages immediately (not at end of transaction) when a relation truncation occurs, this ensures that other backends will reset their values before they next access the relation. We no longer need the unreliable assumption that a VACUUM that's doing a truncation will hold its AccessExclusive lock until commit --- in fact, we can intentionally release that lock as soon as we've completed the truncation. This patch therefore reverts (most of) Alvaro's patch of 2009-11-10, as well as my marginal hacking on it yesterday. We can also get rid of assorted no-longer-needed relcache flushes, which are far more expensive than an smgr flush because they kill a lot more state. In passing this patch fixes smgr_redo's failure to perform visibility-map truncation, and cleans up some rather dubious assumptions in freespace.c and visibilitymap.c about when rd_fsm_nblocks and rd_vm_nblocks can be out of date.
2010-02-09 22:43:30 +01:00
/* Must free any AM cached data upon relcache flush */
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (relation->rd_amcache)
pfree(relation->rd_amcache);
relation->rd_amcache = NULL;
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* If it's a shared index, we might be called before backend startup has
* finished selecting a database, in which case we have no way to read
* pg_class yet. However, a shared index can never have any significant
* schema updates, so it's okay to ignore the invalidation signal. Just
* mark it valid and return without doing anything more.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*/
if (relation->rd_rel->relisshared && !criticalRelcachesBuilt)
{
relation->rd_isvalid = true;
return;
}
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
/*
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
* Read the pg_class row
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* Don't try to use an indexscan of pg_class_oid_index to reload the info
* for pg_class_oid_index ...
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*/
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
indexOK = (RelationGetRelid(relation) != ClassOidIndexId);
pg_class_tuple = ScanPgRelation(RelationGetRelid(relation), indexOK);
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
if (!HeapTupleIsValid(pg_class_tuple))
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
elog(ERROR, "could not find pg_class tuple for index %u",
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
RelationGetRelid(relation));
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
relp = (Form_pg_class) GETSTRUCT(pg_class_tuple);
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
memcpy(relation->rd_rel, relp, CLASS_TUPLE_SIZE);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
/* Reload reloptions in case they changed */
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
if (relation->rd_options)
pfree(relation->rd_options);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
RelationParseRelOptions(relation, pg_class_tuple);
/* done with pg_class tuple */
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
heap_freetuple(pg_class_tuple);
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
/* We must recalculate physical address in case it changed */
RelationInitPhysicalAddr(relation);
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
/*
* For a non-system index, there are fields of the pg_index row that are
* allowed to change, so re-read that row and update the relcache entry.
* Most of the info derived from pg_index (such as support function lookup
* info) cannot change, and indeed the whole point of this routine is to
* update the relcache entry without clobbering that data; so wholesale
* replacement is not appropriate.
*/
if (!IsSystemRelation(relation))
{
HeapTuple tuple;
Form_pg_index index;
Wrap calls to SearchSysCache and related functions using macros. The purpose of this change is to eliminate the need for every caller of SearchSysCache, SearchSysCacheCopy, SearchSysCacheExists, GetSysCacheOid, and SearchSysCacheList to know the maximum number of allowable keys for a syscache entry (currently 4). This will make it far easier to increase the maximum number of keys in a future release should we choose to do so, and it makes the code shorter, too. Design and review by Tom Lane.
2010-02-14 19:42:19 +01:00
tuple = SearchSysCache1(INDEXRELID,
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
ObjectIdGetDatum(RelationGetRelid(relation)));
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
if (!HeapTupleIsValid(tuple))
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
elog(ERROR, "cache lookup failed for index %u",
RelationGetRelid(relation));
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
index = (Form_pg_index) GETSTRUCT(tuple);
relation->rd_index->indisvalid = index->indisvalid;
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
relation->rd_index->indcheckxmin = index->indcheckxmin;
relation->rd_index->indisready = index->indisready;
HeapTupleHeaderSetXmin(relation->rd_indextuple->t_data,
HeapTupleHeaderGetXmin(tuple->t_data));
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
ReleaseSysCache(tuple);
}
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
/* Okay, now it's valid again */
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
relation->rd_isvalid = true;
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
}
Add some debugging support code (ifdef'd out in normal use).
2001-01-08 19:34:44 +01:00
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/*
* RelationDestroyRelation
*
* Physically delete a relation cache entry and all subsidiary data.
* Caller must already have unhooked the entry from the hash table.
*/
static void
RelationDestroyRelation(Relation relation)
{
Assert(RelationHasReferenceCountZero(relation));
/*
* Make sure smgr and lower levels close the relation's files, if they
* weren't closed already. (This was probably done by caller, but let's
* just be real sure.)
*/
RelationCloseSmgr(relation);
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Free all the subsidiary data structures of the relcache entry, then the
* entry itself.
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
*/
if (relation->rd_rel)
pfree(relation->rd_rel);
/* can't use DecrTupleDescRefCount here */
Assert(relation->rd_att->tdrefcount > 0);
if (--relation->rd_att->tdrefcount == 0)
FreeTupleDesc(relation->rd_att);
list_free(relation->rd_indexlist);
bms_free(relation->rd_indexattr);
FreeTriggerDesc(relation->trigdesc);
if (relation->rd_options)
pfree(relation->rd_options);
if (relation->rd_indextuple)
pfree(relation->rd_indextuple);
if (relation->rd_am)
pfree(relation->rd_am);
if (relation->rd_indexcxt)
MemoryContextDelete(relation->rd_indexcxt);
if (relation->rd_rulescxt)
MemoryContextDelete(relation->rd_rulescxt);
pfree(relation);
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
* RelationClearRelation
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
* Physically blow away a relation cache entry, or reset it and rebuild
* it from scratch (that is, from catalog entries). The latter path is
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
* used when we are notified of a change to an open relation (one with
* refcount > 0).
Fix LOAD_CRIT_INDEX() macro to take out AccessShareLock on the system index it is trying to build a relcache entry for. This is an oversight in my 8.2 patch that tried to ensure we always took a lock on a relation before trying to build its relcache entry. The implication is that if someone committed a reindex of a critical system index at about the same time that some other backend were starting up without a valid pg_internal.init file, the second one might PANIC due to not seeing any valid version of the index's pg_class row. Improbable case, but definitely not impossible.
2008-04-16 20:23:04 +02:00
*
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
* NB: when rebuilding, we'd better hold some lock on the relation,
* else the catalog data we need to read could be changing under us.
pgindent run for 9.0, second run
2010-07-06 21:19:02 +02:00
* Also, a rel to be rebuilt had better have refcnt > 0. This is because
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
* an sinval reset could happen while we're accessing the catalogs, and
* the rel would get blown away underneath us by RelationCacheInvalidate
* if it has zero refcnt.
*
* The "rebuild" parameter is redundant in current usage because it has
* to match the relation's refcnt status, but we keep it as a crosscheck
* that we're doing what the caller expects.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Make functions static where possible, enclose unused functions in #ifdef NOT_USED.
1997-08-19 23:40:56 +02:00
static void
Improve relcache.c error reporting for the next guy who has to debug this thing.
2002-07-15 03:57:51 +02:00
RelationClearRelation(Relation relation, bool rebuild)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
Oid old_reltype = relation->rd_rel->reltype;
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
/*
pgindent run for 9.0, second run
2010-07-06 21:19:02 +02:00
* As per notes above, a rel to be rebuilt MUST have refcnt > 0; while of
* course it would be a bad idea to blow away one with nonzero refcnt.
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
*/
Assert(rebuild ?
!RelationHasReferenceCountZero(relation) :
RelationHasReferenceCountZero(relation));
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
/*
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
* Make sure smgr and lower levels close the relation's files, if they
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* weren't closed already. If the relation is not getting deleted, the
* next smgr access should reopen the files automatically. This ensures
* that the low-level file access state is updated after, say, a vacuum
* truncation.
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
*/
Phase 1 of fix for 'SMgrRelation hashtable corrupted' problem. This is the minimum required fix. I want to look next at taking advantage of it by simplifying the message semantics in the shared inval message queue, but that part can be held over for 8.1 if it turns out too ugly.
2005-01-10 21:02:24 +01:00
RelationCloseSmgr(relation);
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Never, never ever blow away a nailed-in system relation, because we'd
Fix up rickety handling of relation-truncation interlocks. Move rd_targblock, rd_fsm_nblocks, and rd_vm_nblocks from relcache to the smgr relation entries, so that they will get reset to InvalidBlockNumber whenever an smgr-level flush happens. Because we now send smgr invalidation messages immediately (not at end of transaction) when a relation truncation occurs, this ensures that other backends will reset their values before they next access the relation. We no longer need the unreliable assumption that a VACUUM that's doing a truncation will hold its AccessExclusive lock until commit --- in fact, we can intentionally release that lock as soon as we've completed the truncation. This patch therefore reverts (most of) Alvaro's patch of 2009-11-10, as well as my marginal hacking on it yesterday. We can also get rid of assorted no-longer-needed relcache flushes, which are far more expensive than an smgr flush because they kill a lot more state. In passing this patch fixes smgr_redo's failure to perform visibility-map truncation, and cleans up some rather dubious assumptions in freespace.c and visibilitymap.c about when rd_fsm_nblocks and rd_vm_nblocks can be out of date.
2010-02-09 22:43:30 +01:00
* be unable to recover. However, we must redo RelationInitPhysicalAddr
* in case it is a mapped relation whose mapping changed.
Get rid of rd_nblocks field in relcache entries. Turns out this was costing us lots more to maintain than it was worth. On shared tables it was of exactly zero benefit because we couldn't trust it to be up to date. On temp tables it sometimes saved an lseek, but not often enough to be worth getting excited about. And the real problem was that we forced an lseek on every relcache flush in order to update the field. So all in all it seems best to lose the complexity.
2004-05-08 21:09:25 +02:00
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* If it's a nailed index, then we need to re-read the pg_class row to see
* if its relfilenode changed. We can't necessarily do that here, because
* we might be in a failed transaction. We assume it's okay to do it if
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* there are open references to the relcache entry (cf notes for
* AtEOXact_RelationCache). Otherwise just mark the entry as possibly
* invalid, and it'll be fixed when next opened.
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
*/
if (relation->rd_isnailed)
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
{
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
RelationInitPhysicalAddr(relation);
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
if (relation->rd_rel->relkind == RELKIND_INDEX)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
relation->rd_isvalid = false; /* needs to be revalidated */
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
if (relation->rd_refcnt > 1)
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
RelationReloadIndexInfo(relation);
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
return;
REINDEX under WAL.
2000-12-08 07:17:58 +01:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
/*
* Even non-system indexes should not be blown away if they are open and
* have valid index support information. This avoids problems with active
* use of the index support information. As with nailed indexes, we
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* re-read the pg_class row to handle possible physical relocation of the
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
* index, and we check for pg_index updates too.
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
*/
if (relation->rd_rel->relkind == RELKIND_INDEX &&
relation->rd_refcnt > 0 &&
relation->rd_indexcxt != NULL)
{
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
relation->rd_isvalid = false; /* needs to be revalidated */
Fix things so that when CREATE INDEX CONCURRENTLY sets pg_index.indisvalid true at the very end of its processing, the update is broadcast via a shared-cache-inval message for the index; without this, existing backends that already have relcache entries for the index might never see it become valid. Also, force a relcache inval on the index's parent table at the same time, so that any cached plans for that table are re-planned; this ensures that the newly valid index will be used if appropriate. Aside from making C.I.C. behave more reasonably, this is necessary infrastructure for some aspects of the HOT patch. Pavan Deolasee, with a little further stuff from me.
2007-05-02 23:08:46 +02:00
RelationReloadIndexInfo(relation);
Avoid crashing if relcache flush occurs while trying to load data into an index's support-function cache (in index_getprocinfo). Since none of that data can change for an index that's in active use, it seems sufficient to treat all open indexes the same way we were treating "nailed" system indexes --- that is, just re-read the pg_class row and leave the rest of the relcache entry strictly alone. The pg_class re-read might not be strictly necessary either, but since the reltablespace and relfilenode can change in normal operation it seems safest to do it. (We don't support changing any of the other info about an index at all, at the moment.) Back-patch as far as 8.0. It might be possible to adapt the patch to 7.4, but it would take more work than I care to expend for such a low-probability problem. 7.3 is out of luck for sure.
2006-01-19 21:28:43 +01:00
return;
}
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/* Mark it invalid until we've finished rebuild */
relation->rd_isvalid = false;
Repair a bunch of problems in md.c. This builds on Hiroshi's insight that RelationFlushRelation ought to invoke smgrclose, and that the way to make that work is to ensure that mdclose doesn't fail if the relation is already closed (or unlinked, if we are looking at a DROP TABLE). While I was testing that, I was able to identify several problems that we had with multiple-segment relations. The system is now able to do initdb and pass the regression tests with a very small segment size (I had it set to 64Kb per segment for testing). I don't believe that ever worked before. File descriptor leaks seem to be gone too. I have partially addressed the concerns we had about mdtruncate(), too. On a Win32 or NFS filesystem it is not possible to unlink a file that another backend is holding open, so what md.c now does is to truncate unwanted files to zero length before trying to unlink them. The other backends will be forced to close their open files by relation cache invalidation --- but I think it would take considerable work to make that happen before vacuum truncates the relation rather than after. Leaving zero-length files lying around seems a usable compromise.
1999-09-02 04:57:50 +02:00
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
/*
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
* If we're really done with the relcache entry, blow it away. But if
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* someone is still using it, reconstruct the whole deal without moving
* the physical RelationData record (so that the someone's pointer is
* still valid).
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
*/
Improve relcache.c error reporting for the next guy who has to debug this thing.
2002-07-15 03:57:51 +02:00
if (!rebuild)
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
{
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/* Flush any rowtype cache entry */
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
flush_rowtype_cache(old_reltype);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/* Remove it from the hash table */
RelationCacheDelete(relation);
/* And release storage */
RelationDestroyRelation(relation);
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
}
else
{
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Our strategy for rebuilding an open relcache entry is to build a
* new entry from scratch, swap its contents with the old entry, and
* finally delete the new entry (along with any infrastructure swapped
* over from the old entry). This is to avoid trouble in case an
* error causes us to lose control partway through. The old entry
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* will still be marked !rd_isvalid, so we'll try to rebuild it again
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* on next access. Meanwhile it's not any less valid than it was
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
* before, so any code that might expect to continue accessing it
* isn't hurt by the rebuild failure. (Consider for example a
* subtransaction that ALTERs a table and then gets cancelled partway
* through the cache entry rebuild. The outer transaction should
* still see the not-modified cache entry as valid.) The worst
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* consequence of an error is leaking the necessarily-unreferenced new
* entry, and this shouldn't happen often enough for that to be a big
* problem.
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
*
Restructure CLUSTER/newstyle VACUUM FULL/ALTER TABLE support so that swapping of old and new toast tables can be done either at the logical level (by swapping the heaps' reltoastrelid links) or at the physical level (by swapping the relfilenodes of the toast tables and their indexes). This is necessary infrastructure for upcoming changes to support CLUSTER/VAC FULL on shared system catalogs, where we cannot change reltoastrelid. The physical swap saves a few catalog updates too. We unfortunately have to keep the logical-level swap logic because in some cases we will be adding or deleting a toast table, so there's no possibility of a physical swap. However, that only happens as a consequence of schema changes in the table, which we do not need to support for system catalogs, so such cases aren't an obstacle for that. In passing, refactor the cluster support functions a little bit to eliminate unnecessarily-duplicated code; and fix the problem that while CLUSTER had been taught to rename the final toast table at need, ALTER TABLE had not.
2010-02-04 01:09:14 +01:00
* When rebuilding an open relcache entry, we must preserve ref count,
* rd_createSubid/rd_newRelfilenodeSubid, and rd_toastoid state. Also
* attempt to preserve the pg_class entry (rd_rel), tupledesc, and
* rewrite-rule substructures in place, because various places assume
* that these structures won't move while they are working with an
* open relcache entry. (Note: the refcount mechanism for tupledescs
* might someday allow us to remove this hack for the tupledesc.)
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* Note that this process does not touch CurrentResourceOwner; which
* is good because whatever ref counts the entry may have do not
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
* necessarily belong to that resource owner.
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
*/
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
Relation newrel;
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
Oid save_relid = RelationGetRelid(relation);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
bool keep_tupdesc;
bool keep_rules;
/* Build temporary entry, but don't link it into hashtable */
newrel = RelationBuildDesc(save_relid, false);
if (newrel == NULL)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
/* Should only get here if relation was deleted */
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
flush_rowtype_cache(old_reltype);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
RelationCacheDelete(relation);
RelationDestroyRelation(relation);
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
elog(ERROR, "relation %u deleted while still in use", save_relid);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
keep_tupdesc = equalTupleDescs(relation->rd_att, newrel->rd_att);
keep_rules = equalRuleLocks(relation->rd_rules, newrel->rd_rules);
if (!keep_tupdesc)
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
flush_rowtype_cache(old_reltype);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/*
* Perform swapping of the relcache entry contents. Within this
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* process the old entry is momentarily invalid, so there *must* be no
* possibility of CHECK_FOR_INTERRUPTS within this sequence. Do it in
* all-in-line code for safety.
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Since the vast majority of fields should be swapped, our method is
* to swap the whole structures and then re-swap those few fields we
* didn't want swapped.
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
*/
#define SWAPFIELD(fldtype, fldname) \
do { \
fldtype _tmp = newrel->fldname; \
newrel->fldname = relation->fldname; \
relation->fldname = _tmp; \
} while (0)
/* swap all Relation struct fields */
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
{
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
RelationData tmpstruct;
memcpy(&tmpstruct, newrel, sizeof(RelationData));
memcpy(newrel, relation, sizeof(RelationData));
memcpy(relation, &tmpstruct, sizeof(RelationData));
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
}
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/* rd_smgr must not be swapped, due to back-links from smgr level */
SWAPFIELD(SMgrRelation, rd_smgr);
/* rd_refcnt must be preserved */
SWAPFIELD(int, rd_refcnt);
/* isnailed shouldn't change */
Assert(newrel->rd_isnailed == relation->rd_isnailed);
/* creation sub-XIDs must be preserved */
SWAPFIELD(SubTransactionId, rd_createSubid);
SWAPFIELD(SubTransactionId, rd_newRelfilenodeSubid);
/* un-swap rd_rel pointers, swap contents instead */
SWAPFIELD(Form_pg_class, rd_rel);
/* ... but actually, we don't have to update newrel->rd_rel */
memcpy(relation->rd_rel, newrel->rd_rel, CLASS_TUPLE_SIZE);
/* preserve old tupledesc and rules if no logical change */
if (keep_tupdesc)
SWAPFIELD(TupleDesc, rd_att);
if (keep_rules)
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
{
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
SWAPFIELD(RuleLock *, rd_rules);
SWAPFIELD(MemoryContext, rd_rulescxt);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
}
Restructure CLUSTER/newstyle VACUUM FULL/ALTER TABLE support so that swapping of old and new toast tables can be done either at the logical level (by swapping the heaps' reltoastrelid links) or at the physical level (by swapping the relfilenodes of the toast tables and their indexes). This is necessary infrastructure for upcoming changes to support CLUSTER/VAC FULL on shared system catalogs, where we cannot change reltoastrelid. The physical swap saves a few catalog updates too. We unfortunately have to keep the logical-level swap logic because in some cases we will be adding or deleting a toast table, so there's no possibility of a physical swap. However, that only happens as a consequence of schema changes in the table, which we do not need to support for system catalogs, so such cases aren't an obstacle for that. In passing, refactor the cluster support functions a little bit to eliminate unnecessarily-duplicated code; and fix the problem that while CLUSTER had been taught to rename the final toast table at need, ALTER TABLE had not.
2010-02-04 01:09:14 +01:00
/* toast OID override must be preserved */
SWAPFIELD(Oid, rd_toastoid);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
/* pgstat_info must be preserved */
SWAPFIELD(struct PgStat_TableStatus *, pgstat_info);
#undef SWAPFIELD
/* And now we can throw away the temporary entry */
RelationDestroyRelation(newrel);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
* RelationFlushRelation
*
* Rebuild the relation if it is open (refcount > 0), else blow it away.
*/
static void
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
RelationFlushRelation(Relation relation)
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
{
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
if (relation->rd_createSubid != InvalidSubTransactionId ||
relation->rd_newRelfilenodeSubid != InvalidSubTransactionId)
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
{
/*
Code review of CLUSTER patch. Clean up problems with relcache getting confused, toasted data getting lost, etc.
2002-08-11 23:17:35 +02:00
* New relcache entries are always rebuilt, not flushed; else we'd
* forget the "new" status of the relation, which is a useful
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
* optimization to have. Ditto for the new-relfilenode status.
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
*
pgindent run for 9.0, second run
2010-07-06 21:19:02 +02:00
* The rel could have zero refcnt here, so temporarily increment the
* refcnt to ensure it's safe to rebuild it. We can assume that the
* current transaction has some lock on the rel already.
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
*/
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
RelationIncrementReferenceCount(relation);
RelationClearRelation(relation, true);
RelationDecrementReferenceCount(relation);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
}
else
{
/*
Code review of CLUSTER patch. Clean up problems with relcache getting confused, toasted data getting lost, etc.
2002-08-11 23:17:35 +02:00
* Pre-existing rels can be dropped from the relcache if not open.
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
*/
pgindent run for 9.0, second run
2010-07-06 21:19:02 +02:00
bool rebuild = !RelationHasReferenceCountZero(relation);
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
Fix a problem introduced by my patch of 2010-01-12 that revised the way relcache reload works. In the patched code, a relcache entry in process of being rebuilt doesn't get unhooked from the relcache hash table; which means that if a cache flush occurs due to sinval queue overrun while we're rebuilding it, the entry could get blown away by RelationCacheInvalidate, resulting in crash or misbehavior. Fix by ensuring that an entry being rebuilt has positive refcount, so it won't be seen as a target for removal if a cache flush occurs. (This will mean that the entry gets rebuilt twice in such a scenario, but that's okay.) It appears that the problem can only arise within a transaction that has previously reassigned the relfilenode of a pre-existing table, via TRUNCATE or a similar operation. Per bug #5412 from Rusty Conover. Back-patch to 8.2, same as the patch that introduced the problem. I think that the failure can't actually occur in 8.2, since it lacks the rd_newRelfilenodeSubid optimization, but let's make it work like the later branches anyway. Patch by Heikki, slightly editorialized on by me.
2010-04-14 23:31:11 +02:00
RelationClearRelation(relation, rebuild);
}
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
* RelationForgetRelation - unconditionally remove a relcache entry
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
*
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
* External interface for destroying a relcache entry when we
* drop the relation.
New func RelationForgetRelation(); * RelationFlushRelation + if the relation is local then get rid of * the relation descriptor from the newly created relation list.
1997-06-04 10:56:51 +02:00
*/
void
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationForgetRelation(Oid rid)
New func RelationForgetRelation(); * RelationFlushRelation + if the relation is local then get rid of * the relation descriptor from the newly created relation list.
1997-06-04 10:56:51 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Relation relation;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationIdCacheLookup(rid, relation);
Make RelationForgetRelation error out if the relcache entry has nonzero reference count. This avoids leaving dangling pointers around, as in recent bug report against sequences (bug# 671).
2002-05-22 17:57:40 +02:00
if (!PointerIsValid(relation))
return; /* not in cache, nothing to do */
if (!RelationHasReferenceCountZero(relation))
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "relation %u is still open", rid);
Make RelationForgetRelation error out if the relcache entry has nonzero reference count. This avoids leaving dangling pointers around, as in recent bug report against sequences (bug# 671).
2002-05-22 17:57:40 +02:00
/* Unconditionally destroy the relcache entry */
RelationClearRelation(relation, false);
New func RelationForgetRelation(); * RelationFlushRelation + if the relation is local then get rid of * the relation descriptor from the newly created relation list.
1997-06-04 10:56:51 +02:00
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
* RelationCacheInvalidateEntry
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
*
* This routine is invoked for SI cache flush messages.
*
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
* Any relcache entry matching the relid must be flushed. (Note: caller has
* already determined that the relid belongs to our database or is a shared
Phase 1 of fix for 'SMgrRelation hashtable corrupted' problem. This is the minimum required fix. I want to look next at taking advantage of it by simplifying the message semantics in the shared inval message queue, but that part can be held over for 8.1 if it turns out too ugly.
2005-01-10 21:02:24 +01:00
* relation.)
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
*
* We used to skip local relations, on the grounds that they could
* not be targets of cross-backend SI update messages; but it seems
* safer to process them, so that our *own* SI update messages will
* have the same effects during CommandCounterIncrement for both
* local and nonlocal relations.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
void
Phase 1 of fix for 'SMgrRelation hashtable corrupted' problem. This is the minimum required fix. I want to look next at taking advantage of it by simplifying the message semantics in the shared inval message queue, but that part can be held over for 8.1 if it turns out too ugly.
2005-01-10 21:02:24 +01:00
RelationCacheInvalidateEntry(Oid relationId)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
Relation relation;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
RelationIdCacheLookup(relationId, relation);
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
if (PointerIsValid(relation))
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
relcacheInvalsReceived++;
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
RelationFlushRelation(relation);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
/*
* RelationCacheInvalidate
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
* Blow away cached relation descriptors that have zero reference counts,
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
* and rebuild those with positive reference counts. Also reset the smgr
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
* relation cache and re-read relation mapping data.
Modify uses of RelationFlushRelation and RelationCacheInvalidate so that we *always* rebuild, rather than deleting, an invalidated relcache entry that has positive refcount. Otherwise an SI cache overrun leads to dangling Relation pointers all over the place!
2000-01-29 20:51:59 +01:00
*
Fix relcache.c so that local relations (those created during current transaction) are not flushed by shared-cache-inval reset message. SI reset actually works now, for probably the first time in a long time. I was able to run initdb and regression tests with a 16-element SI message array, with a lot of NOTICE: cache state reset messages but no crashes.
1999-09-06 21:33:16 +02:00
* This is currently used only to recover from SI message buffer overflow,
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
* so we do not touch new-in-transaction relations; they cannot be targets
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
* of cross-backend SI updates (and our own updates now go through a
* separate linked list that isn't limited by the SI message buffer size).
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
* Likewise, we need not discard new-relfilenode-in-transaction hints,
* since any invalidation of those would be a local event.
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
*
* We do this in two phases: the first pass deletes deletable items, and
* the second one rebuilds the rebuildable items. This is essential for
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
* safety, because hash_seq_search only copes with concurrent deletion of
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
* the element it is currently visiting. If a second SI overflow were to
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
* occur while we are walking the table, resulting in recursive entry to
* this routine, we could crash because the inner invocation blows away
* the entry next to be visited by the outer scan. But this way is OK,
* because (a) during the first pass we won't process any more SI messages,
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
* so hash_seq_search will complete safely; (b) during the second pass we
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
* only hold onto pointers to nondeletable entries.
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*
* The two-phase approach also makes it easy to ensure that we process
* nailed-in-cache indexes before other nondeletable items, and that we
* process pg_class_oid_index first of all. In scenarios where a nailed
* index has been given a new relfilenode, we have to detect that update
* before the nailed index is used in reloading any other relcache entry.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
void
Fix problems seen in parallel regress tests when SI buffer overruns (causing syscache and relcache flushes). Relcache entry rebuild now preserves original tupledesc, rewrite rules, and triggers if possible, so that pointers to these things remain valid --- if these things change while relcache entry has positive refcount, we elog(ERROR) to avoid later crash. Arrange for xact-local rels to be rebuilt when an SI inval message is seen for them, so that they are updated by CommandCounterIncrement the same as regular rels. (This is useful because of Hiroshi's recent changes to process our own SI messages at CommandCounterIncrement time.) This allows simplification of some routines that previously hacked around the lack of an automatic update. catcache now keeps its own copy of tupledesc for its relation, rather than depending on the relcache's copy; this avoids needing to reinitialize catcache during a cache flush, which saves some cycles and eliminates nasty circularity problems that occur if a cache flush happens while trying to initialize a catcache. Eliminate a number of permanent memory leaks that used to happen during catcache or relcache flush; not least of which was that catcache never freed any cached tuples! (Rule parsetree storage is still leaked, however; will fix that separately.) Nothing done yet about code that uses tuples retrieved by SearchSysCache for longer than is safe.
2000-01-31 05:35:57 +01:00
RelationCacheInvalidate(void)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
HASH_SEQ_STATUS status;
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
RelIdCacheEnt *idhentry;
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
Relation relation;
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
List *rebuildFirstList = NIL;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
List *rebuildList = NIL;
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
ListCell *l;
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
/* Phase 1 */
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
hash_seq_init(&status, RelationIdCache);
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
while ((idhentry = (RelIdCacheEnt *) hash_seq_search(&status)) != NULL)
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
{
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
relation = idhentry->reldesc;
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
/* Must close all smgr references to avoid leaving dangling ptrs */
Phase 1 of fix for 'SMgrRelation hashtable corrupted' problem. This is the minimum required fix. I want to look next at taking advantage of it by simplifying the message semantics in the shared inval message queue, but that part can be held over for 8.1 if it turns out too ugly.
2005-01-10 21:02:24 +01:00
RelationCloseSmgr(relation);
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
/* Ignore new relations, since they are never SI targets */
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
if (relation->rd_createSubid != InvalidSubTransactionId)
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
continue;
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
relcacheInvalsReceived++;
Code review of CLUSTER patch. Clean up problems with relcache getting confused, toasted data getting lost, etc.
2002-08-11 23:17:35 +02:00
if (RelationHasReferenceCountZero(relation))
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
{
/* Delete this entry immediately */
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
Assert(!relation->rd_isnailed);
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
RelationClearRelation(relation, false);
}
else
{
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
/*
* Add this entry to list of stuff to rebuild in second pass.
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* pg_class_oid_index goes on the front of rebuildFirstList, other
* nailed indexes on the back, and everything else into
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
* rebuildList (in no particular order).
*/
if (relation->rd_isnailed &&
relation->rd_rel->relkind == RELKIND_INDEX)
{
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
if (RelationGetRelid(relation) == ClassOidIndexId)
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
rebuildFirstList = lcons(relation, rebuildFirstList);
else
rebuildFirstList = lappend(rebuildFirstList, relation);
}
else
rebuildList = lcons(relation, rebuildList);
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
}
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
}
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Now zap any remaining smgr cache entries. This must happen before we
* start to rebuild entries, since that may involve catalog fetches which
* will re-open catalog files.
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
*/
smgrcloseall();
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/*
* Reload relation mapping data before starting to reconstruct cache.
*/
RelationMapInvalidateAll();
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
/* Phase 2: rebuild the items found to need rebuild in phase 1 */
Fix a tiny memory leak (one List header) in RelationCacheInvalidate(). This is utterly insignificant in normal operation, but it becomes a problem during cache inval stress testing. The original coding in fact had no leak --- the 8.0 List rewrite created the issue. I wonder whether list_concat should pfree the discarded header?
2006-01-19 01:27:08 +01:00
foreach(l, rebuildFirstList)
{
relation = (Relation) lfirst(l);
RelationClearRelation(relation, true);
}
list_free(rebuildFirstList);
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
foreach(l, rebuildList)
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
{
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
relation = (Relation) lfirst(l);
RelationClearRelation(relation, true);
Clean up non-reentrant interface for hash_seq/HashTableWalk, so that starting a new hashtable search no longer clobbers any other search active anywhere in the system. Fix RelationCacheInvalidate() so that it will not crash or go into an infinite loop if invoked recursively, as for example by a second SI Reset message arriving while we are still processing a prior one.
2001-01-02 05:33:24 +01:00
}
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
list_free(rebuildList);
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/*
* RelationCloseSmgrByOid - close a relcache entry's smgr link
*
* Needed in some cases where we are changing a relation's physical mapping.
* The link will be automatically reopened on next use.
*/
void
RelationCloseSmgrByOid(Oid relationId)
{
Relation relation;
RelationIdCacheLookup(relationId, relation);
if (!PointerIsValid(relation))
return; /* not in cache, nothing to do */
RelationCloseSmgr(relation);
}
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
/*
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
* AtEOXact_RelationCache
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
*
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
* Clean up the relcache at main-transaction commit or abort.
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*
* Note: this must be called *before* processing invalidation messages.
* In the case of abort, we don't want to try to rebuild any invalidated
* cache entries (since we can't safely do database accesses). Therefore
* we must reset refcnts before handling pending invalidations.
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
*
* As of PostgreSQL 8.1, relcache refcnts should get released by the
* ResourceOwner mechanism. This routine just does a debugging
* cross-check that no pins remain. However, we also need to do special
* cleanup when the current transaction created any relations or made use
* of forced index lists.
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
*/
void
Nested transactions. There is still much left to do, especially on the performance front, but with feature freeze upon us I think it's time to drive a stake in the ground and say that this will be in 7.5. Alvaro Herrera, with some help from Tom Lane.
2004-07-01 02:52:04 +02:00
AtEOXact_RelationCache(bool isCommit)
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
{
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
HASH_SEQ_STATUS status;
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
RelIdCacheEnt *idhentry;
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
/*
* To speed up transaction exit, we want to avoid scanning the relcache
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* unless there is actually something for this routine to do. Other than
* the debug-only Assert checks, most transactions don't create any work
* for us to do here, so we keep a static flag that gets set if there is
* anything to do. (Currently, this means either a relation is created in
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
* the current xact, or one is given a new relfilenode, or an index list
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* is forced.) For simplicity, the flag remains set till end of top-level
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
* transaction, even though we could clear it at subtransaction end in
* some cases.
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
*/
if (!need_eoxact_work
#ifdef USE_ASSERT_CHECKING
&& !assert_enabled
#endif
)
return;
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
hash_seq_init(&status, RelationIdCache);
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
while ((idhentry = (RelIdCacheEnt *) hash_seq_search(&status)) != NULL)
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
{
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
Relation relation = idhentry->reldesc;
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
/*
* The relcache entry's ref count should be back to its normal
* not-in-a-transaction state: 0 unless it's nailed in cache.
*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* In bootstrap mode, this is NOT true, so don't check it --- the
* bootstrap code expects relations to stay open across start/commit
* transaction calls. (That seems bogus, but it's not worth fixing.)
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
*/
#ifdef USE_ASSERT_CHECKING
if (!IsBootstrapProcessingMode())
{
int expected_refcnt;
expected_refcnt = relation->rd_isnailed ? 1 : 0;
Assert(relation->rd_refcnt == expected_refcnt);
}
#endif
When compiling with --enable-cassert, check for reference count leaks in the relcache. It's rather silly that we have reference count leak checks in bufmgr and in catcache, but not in relcache which will normally have many fewer entries. Chris K-L would have caught at least one bug in his recent DROP patch if he'd had this.
2002-08-03 00:36:05 +02:00
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
/*
* Is it a relation created in the current transaction?
*
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
* During commit, reset the flag to zero, since we are now out of the
* creating transaction. During abort, simply delete the relcache
* entry --- it isn't interesting any longer. (NOTE: if we have
* forgotten the new-ness of a new relation due to a forced cache
* flush, the entry will get deleted anyway by shared-cache-inval
* processing of the aborted pg_class insertion.)
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
*/
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
if (relation->rd_createSubid != InvalidSubTransactionId)
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
{
Nested transactions. There is still much left to do, especially on the performance front, but with feature freeze upon us I think it's time to drive a stake in the ground and say that this will be in 7.5. Alvaro Herrera, with some help from Tom Lane.
2004-07-01 02:52:04 +02:00
if (isCommit)
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
relation->rd_createSubid = InvalidSubTransactionId;
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
else
{
RelationClearRelation(relation, false);
continue;
}
}
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
/*
* Likewise, reset the hint about the relfilenode being new.
*/
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
relation->rd_newRelfilenodeSubid = InvalidSubTransactionId;
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
/*
* Flush any temporary index list.
*/
if (relation->rd_indexvalid == 2)
{
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
list_free(relation->rd_indexlist);
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
relation->rd_indexlist = NIL;
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
relation->rd_oidindex = InvalidOid;
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
relation->rd_indexvalid = 0;
}
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
}
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
/* Once done with the transaction, we can reset need_eoxact_work */
need_eoxact_work = false;
Modify RelationFlushRelation so that if the relcache entry has positive refcount, it is rebuilt from pg_class data. This ensures that relcache entries will track changes made by other backends. Formerly, a shared inval report would just be ignored if it happened to arrive while the relcache entry was in use. Also, fix relcache to reset ref counts to zero during transaction abort. Finally, change LockRelation() so that it checks for shared inval reports after obtaining the lock. In this way, once any kind of lock has been obtained on a rel, we can trust the relcache entry to be up-to-date.
1999-09-04 20:42:15 +02:00
}
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
/*
* AtEOSubXact_RelationCache
*
* Clean up the relcache at sub-transaction commit or abort.
*
* Note: this must be called *before* processing invalidation messages.
*/
void
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
AtEOSubXact_RelationCache(bool isCommit, SubTransactionId mySubid,
SubTransactionId parentSubid)
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
{
HASH_SEQ_STATUS status;
RelIdCacheEnt *idhentry;
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
/*
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
* Skip the relcache scan if nothing to do --- see notes for
* AtEOXact_RelationCache.
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
*/
Modify AtEOXact_CatCache and AtEOXact_RelationCache to assume that the ResourceOwner mechanism already released all reference counts for the cache entries; therefore, we do not need to scan the catcache or relcache at transaction end, unless we want to do it as a debugging crosscheck. Do the crosscheck only in Assert mode. This is the same logic we had previously installed in AtEOXact_Buffers to avoid overhead with large numbers of shared buffers. I thought it'd be a good idea to do it here too, in view of Kari Lavikka's recent report showing a real-world case where AtEOXact_CatCache is taking a significant fraction of runtime.
2005-08-08 21:17:23 +02:00
if (!need_eoxact_work)
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
return;
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
hash_seq_init(&status, RelationIdCache);
while ((idhentry = (RelIdCacheEnt *) hash_seq_search(&status)) != NULL)
{
Relation relation = idhentry->reldesc;
/*
* Is it a relation created in the current subtransaction?
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* During subcommit, mark it as belonging to the parent, instead.
* During subabort, simply delete the relcache entry.
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
*/
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
if (relation->rd_createSubid == mySubid)
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
{
if (isCommit)
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
relation->rd_createSubid = parentSubid;
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
else
{
RelationClearRelation(relation, false);
continue;
}
}
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
/*
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
* Likewise, update or drop any new-relfilenode-in-subtransaction
* hint.
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
*/
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
if (relation->rd_newRelfilenodeSubid == mySubid)
{
if (isCommit)
relation->rd_newRelfilenodeSubid = parentSubid;
else
Fix for COPY-after-truncate feature. Simon Riggs
2007-03-03 21:08:41 +01:00
relation->rd_newRelfilenodeSubid = InvalidSubTransactionId;
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
}
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
/*
* Flush any temporary index list.
*/
if (relation->rd_indexvalid == 2)
{
list_free(relation->rd_indexlist);
relation->rd_indexlist = NIL;
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
relation->rd_oidindex = InvalidOid;
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
relation->rd_indexvalid = 0;
}
}
}
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
* RelationBuildLocalRelation
* Build a relcache entry for an about-to-be-created relation,
* and enter it into the relcache.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
Relation
RelationBuildLocalRelation(const char *relname,
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
Oid relnamespace,
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
TupleDesc tupDesc,
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
Oid relid,
Oid reltablespace,
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
bool shared_relation,
bool mapped_relation)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
Relation rel;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
MemoryContext oldcxt;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
int natts = tupDesc->natts;
int i;
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
bool has_not_null;
First phase of project to use fixed OIDs for all system catalogs and indexes. Extend the macros in include/catalog/*.h to carry the info about hand-assigned OIDs, and adjust the genbki script and bootstrap code to make the relations actually get those OIDs. Remove the small number of RelOid_pg_foo macros that we had in favor of a complete set named like the catname.h and indexing.h macros. Next phase will get rid of internal use of names for looking up catalogs and indexes; but this completes the changes forcing an initdb, so it looks like a good place to commit. Along the way, I made the shared relations (pg_database etc) not be 'bootstrap' relations any more, so as to reduce the number of hardwired entries and simplify changing those relations in future. I'm not sure whether they ever really needed to be handled as bootstrap relations, but it seems to work fine to not do so now.
2005-04-14 03:38:22 +02:00
bool nailit;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Fix ALTER TABLE ADD COLUMN to disallow the same column types that are disallowed by CREATE TABLE (eg, pseudo-types); also disallow these types from being introduced by the range-function syntax. While at it, allow CREATE TABLE to create zero-column tables, per recent pghackers discussion. I am back-patching this into 7.3 since failure to disallow pseudo-types is arguably a security hole.
2002-12-16 19:39:22 +01:00
AssertArg(natts >= 0);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
First phase of project to use fixed OIDs for all system catalogs and indexes. Extend the macros in include/catalog/*.h to carry the info about hand-assigned OIDs, and adjust the genbki script and bootstrap code to make the relations actually get those OIDs. Remove the small number of RelOid_pg_foo macros that we had in favor of a complete set named like the catname.h and indexing.h macros. Next phase will get rid of internal use of names for looking up catalogs and indexes; but this completes the changes forcing an initdb, so it looks like a good place to commit. Along the way, I made the shared relations (pg_database etc) not be 'bootstrap' relations any more, so as to reduce the number of hardwired entries and simplify changing those relations in future. I'm not sure whether they ever really needed to be handled as bootstrap relations, but it seems to work fine to not do so now.
2005-04-14 03:38:22 +02:00
/*
* check for creation of a rel that must be nailed in cache.
*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* XXX this list had better match the relations specially handled in
* RelationCacheInitializePhase2/3.
First phase of project to use fixed OIDs for all system catalogs and indexes. Extend the macros in include/catalog/*.h to carry the info about hand-assigned OIDs, and adjust the genbki script and bootstrap code to make the relations actually get those OIDs. Remove the small number of RelOid_pg_foo macros that we had in favor of a complete set named like the catname.h and indexing.h macros. Next phase will get rid of internal use of names for looking up catalogs and indexes; but this completes the changes forcing an initdb, so it looks like a good place to commit. Along the way, I made the shared relations (pg_database etc) not be 'bootstrap' relations any more, so as to reduce the number of hardwired entries and simplify changing those relations in future. I'm not sure whether they ever really needed to be handled as bootstrap relations, but it seems to work fine to not do so now.
2005-04-14 03:38:22 +02:00
*/
switch (relid)
{
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
case DatabaseRelationId:
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
case AuthIdRelationId:
case AuthMemRelationId:
First phase of project to use fixed OIDs for all system catalogs and indexes. Extend the macros in include/catalog/*.h to carry the info about hand-assigned OIDs, and adjust the genbki script and bootstrap code to make the relations actually get those OIDs. Remove the small number of RelOid_pg_foo macros that we had in favor of a complete set named like the catname.h and indexing.h macros. Next phase will get rid of internal use of names for looking up catalogs and indexes; but this completes the changes forcing an initdb, so it looks like a good place to commit. Along the way, I made the shared relations (pg_database etc) not be 'bootstrap' relations any more, so as to reduce the number of hardwired entries and simplify changing those relations in future. I'm not sure whether they ever really needed to be handled as bootstrap relations, but it seems to work fine to not do so now.
2005-04-14 03:38:22 +02:00
case RelationRelationId:
case AttributeRelationId:
case ProcedureRelationId:
case TypeRelationId:
nailit = true;
break;
default:
nailit = false;
break;
}
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
/*
* check that hardwired list of shared rels matches what's in the
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* bootstrap .bki file. If you get a failure here during initdb, you
* probably need to fix IsSharedRelation() to match whatever you've done
* to the set of shared relations.
Change the relation_open protocol so that we obtain lock on a relation (table or index) before trying to open its relcache entry. This fixes race conditions in which someone else commits a change to the relation's catalog entries while we are in process of doing relcache load. Problems of that ilk have been reported sporadically for years, but it was not really practical to fix until recently --- for instance, the recent addition of WAL-log support for in-place updates helped. Along the way, remove pg_am.amconcurrent: all AMs are now expected to support concurrent update.
2006-07-31 22:09:10 +02:00
*/
if (shared_relation != IsSharedRelation(relid))
elog(ERROR, "shared_relation flag for \"%s\" does not match IsSharedRelation(%u)",
relname, relid);
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/* Shared relations had better be mapped, too */
Assert(mapped_relation || !shared_relation);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
* switch to the cache context to create the relcache entry.
*/
if (!CacheMemoryContext)
CreateCacheMemoryContext();
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
* allocate a new relation descriptor and fill in basic state fields.
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
*/
Add new palloc0 call as merge of palloc and MemSet(0).
2002-11-13 01:39:48 +01:00
rel = (Relation) palloc0(sizeof(RelationData));
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/* make sure relation is marked as having no open file yet */
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
rel->rd_smgr = NULL;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
First phase of project to use fixed OIDs for all system catalogs and indexes. Extend the macros in include/catalog/*.h to carry the info about hand-assigned OIDs, and adjust the genbki script and bootstrap code to make the relations actually get those OIDs. Remove the small number of RelOid_pg_foo macros that we had in favor of a complete set named like the catname.h and indexing.h macros. Next phase will get rid of internal use of names for looking up catalogs and indexes; but this completes the changes forcing an initdb, so it looks like a good place to commit. Along the way, I made the shared relations (pg_database etc) not be 'bootstrap' relations any more, so as to reduce the number of hardwired entries and simplify changing those relations in future. I'm not sure whether they ever really needed to be handled as bootstrap relations, but it seems to work fine to not do so now.
2005-04-14 03:38:22 +02:00
/* mark it nailed if appropriate */
rel->rd_isnailed = nailit;
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
rel->rd_refcnt = nailit ? 1 : 0;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
/* it's being created in this transaction */
Restructure subtransaction handling to reduce resource consumption, as per recent discussions. Invent SubTransactionIds that are managed like CommandIds (ie, counter is reset at start of each top transaction), and use these instead of TransactionIds to keep track of subtransaction status in those modules that need it. This means that a subtransaction does not need an XID unless it actually inserts/modifies rows in the database. Accordingly, don't assign it an XID nor take a lock on the XID until it tries to do that. This saves a lot of overhead for subtransactions that are only used for error recovery (eg plpgsql exceptions). Also, arrange to release a subtransaction's XID lock as soon as the subtransaction exits, in both the commit and abort cases. This avoids holding many unique locks after a long series of subtransactions. The price is some additional overhead in XactLockTableWait, but that seems acceptable. Finally, restructure the state machine in xact.c to have a more orthogonal set of states for subtransactions.
2004-09-16 18:58:44 +02:00
rel->rd_createSubid = GetCurrentSubTransactionId();
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
rel->rd_newRelfilenodeSubid = InvalidSubTransactionId;
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
/* must flag that we have rels created in this transaction */
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
need_eoxact_work = true;
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
Modify the relcache to record the temp status of both local and nonlocal temp relations; this is no more expensive than before, now that we have pg_class.relistemp. Insert tests into bufmgr.c to prevent attempting to fetch pages from nonlocal temp relations. This provides a low-level defense against bugs-of-omission allowing temp pages to be loaded into shared buffers, as in the contrib/pgstattuple problem reported by Stuart Bishop. While at it, tweak a bunch of places to use new relcache tests (instead of expensive probes into pg_namespace) to detect local or nonlocal temp tables.
2009-04-01 00:12:48 +02:00
/* it is temporary if and only if it is in my temp-table namespace */
Arrange to put TOAST tables belonging to temporary tables into special schemas named pg_toast_temp_nnn, alongside the pg_temp_nnn schemas used for the temp tables themselves. This allows low-level code such as the relcache to recognize that these tables are indeed temporary, which enables various optimizations such as not WAL-logging changes and using local rather than shared buffers for access. Aside from obvious performance benefits, this provides a solution to bug #3483, in which other backends unexpectedly held open file references to temporary tables. The scheme preserves the property that TOAST tables are not in any schema that's normally in the search path, so they don't conflict with user table names. initdb forced because of changes in system view definitions.
2007-07-26 00:16:18 +02:00
rel->rd_istemp = isTempOrToastNamespace(relnamespace);
Modify the relcache to record the temp status of both local and nonlocal temp relations; this is no more expensive than before, now that we have pg_class.relistemp. Insert tests into bufmgr.c to prevent attempting to fetch pages from nonlocal temp relations. This provides a low-level defense against bugs-of-omission allowing temp pages to be loaded into shared buffers, as in the contrib/pgstattuple problem reported by Stuart Bishop. While at it, tweak a bunch of places to use new relcache tests (instead of expensive probes into pg_namespace) to detect local or nonlocal temp tables.
2009-04-01 00:12:48 +02:00
rel->rd_islocaltemp = rel->rd_istemp;
Restructure local-buffer handling per recent pghackers discussion. The local buffer manager is no longer used for newly-created relations (unless they are TEMP); a new non-TEMP relation goes through the shared bufmgr and thus will participate normally in checkpoints. But TEMP relations use the local buffer manager throughout their lifespan. Also, operations in TEMP relations are not logged in WAL, thus improving performance. Since it's no longer necessary to fsync relations as they move out of the local buffers into shared buffers, quite a lot of smgr.c/md.c/fd.c code is no longer needed and has been removed: there's no concept of a dirty relation anymore in md.c/fd.c, and we never fsync anything but WAL. Still TODO: improve local buffer management algorithms so that it would be reasonable to increase NLocBuffer.
2002-08-06 04:36:35 +02:00
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
Catcaches can now store negative entries as well as positive ones, to speed up repetitive failed searches; per pghackers discussion in late January. inval.c logic substantially simplified, since we can now treat inserts and deletes alike as far as inval events are concerned. Some repair work needed in heap_create_with_catalog, which turns out to have been doing CommandCounterIncrement at a point where the new relation has non-self-consistent catalog entries. With the new inval code, that resulted in assert failures during a relcache entry rebuild.
2002-03-03 18:47:56 +01:00
* create a new tuple descriptor from the one passed in. We do this
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* partly to copy it into the cache context, and partly because the new
* relation can't have any defaults or constraints yet; they have to be
* added in later steps, because they require additions to multiple system
* catalogs. We can copy attnotnull constraints here, however.
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
*/
Catcaches can now store negative entries as well as positive ones, to speed up repetitive failed searches; per pghackers discussion in late January. inval.c logic substantially simplified, since we can now treat inserts and deletes alike as far as inval events are concerned. Some repair work needed in heap_create_with_catalog, which turns out to have been doing CommandCounterIncrement at a point where the new relation has non-self-consistent catalog entries. With the new inval code, that resulted in assert failures during a relcache entry rebuild.
2002-03-03 18:47:56 +01:00
rel->rd_att = CreateTupleDescCopy(tupDesc);
Fix problems with cached tuple descriptors disappearing while still in use by creating a reference-count mechanism, similar to what we did a long time ago for catcache entries. The back branches have an ugly solution involving lots of extra copies, but this way is more efficient. Reference counting is only applied to tupdescs that are actually in caches --- there seems no need to use it for tupdescs that are generated in the executor, since they'll go away during plan shutdown by virtue of being in the per-query memory context. Neil Conway and Tom Lane
2006-06-16 20:42:24 +02:00
rel->rd_att->tdrefcount = 1; /* mark as refcounted */
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
has_not_null = false;
Catcaches can now store negative entries as well as positive ones, to speed up repetitive failed searches; per pghackers discussion in late January. inval.c logic substantially simplified, since we can now treat inserts and deletes alike as far as inval events are concerned. Some repair work needed in heap_create_with_catalog, which turns out to have been doing CommandCounterIncrement at a point where the new relation has non-self-consistent catalog entries. With the new inval code, that resulted in assert failures during a relcache entry rebuild.
2002-03-03 18:47:56 +01:00
for (i = 0; i < natts; i++)
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
{
Catcaches can now store negative entries as well as positive ones, to speed up repetitive failed searches; per pghackers discussion in late January. inval.c logic substantially simplified, since we can now treat inserts and deletes alike as far as inval events are concerned. Some repair work needed in heap_create_with_catalog, which turns out to have been doing CommandCounterIncrement at a point where the new relation has non-self-consistent catalog entries. With the new inval code, that resulted in assert failures during a relcache entry rebuild.
2002-03-03 18:47:56 +01:00
rel->rd_att->attrs[i]->attnotnull = tupDesc->attrs[i]->attnotnull;
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
has_not_null |= tupDesc->attrs[i]->attnotnull;
}
if (has_not_null)
{
TupleConstr *constr = (TupleConstr *) palloc0(sizeof(TupleConstr));
constr->has_not_null = true;
rel->rd_att->constr = constr;
}
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
* initialize relation tuple form (caller may add/override data later)
*/
Add new palloc0 call as merge of palloc and MemSet(0).
2002-11-13 01:39:48 +01:00
rel->rd_rel = (Form_pg_class) palloc0(CLASS_TUPLE_SIZE);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
namestrcpy(&rel->rd_rel->relname, relname);
rel->rd_rel->relnamespace = relnamespace;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
rel->rd_rel->relkind = RELKIND_UNCATALOGED;
Code review for HeapTupleHeader changes. Add version number to page headers (overlaying low byte of page size) and add HEAP_HASOID bit to t_infomask, per earlier discussion. Simplify scheme for overlaying fields in tuple header (no need for cmax to live in more than one place). Don't try to clear infomask status bits in tqual.c --- not safe to do it there. Don't try to force output table of a SELECT INTO to have OIDs, either. Get rid of unnecessarily complex three-state scheme for TupleDesc.tdhasoids, which has already caused one recent failure. Improve documentation.
2002-09-02 03:05:06 +02:00
rel->rd_rel->relhasoids = rel->rd_att->tdhasoid;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
rel->rd_rel->relnatts = natts;
rel->rd_rel->reltype = InvalidOid;
Arrange for indexes and toast tables to inherit their ownership from the parent table, even if the command that creates them is executed by someone else (such as a superuser or a member of the owning role). Per gripe from Michael Fuhr.
2005-08-26 05:08:15 +02:00
/* needed when bootstrapping: */
rel->rd_rel->relowner = BOOTSTRAP_SUPERUSERID;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Insert relation physical and logical identifiers (OIDs) into the right
* places. Note that the physical ID (relfilenode) is initially the same
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
* as the logical ID (OID); except that for a mapped relation, we set
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* relfilenode to zero and rely on RelationInitPhysicalAddr to consult the
* map.
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
*/
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
rel->rd_rel->relisshared = shared_relation;
Modify the relcache to record the temp status of both local and nonlocal temp relations; this is no more expensive than before, now that we have pg_class.relistemp. Insert tests into bufmgr.c to prevent attempting to fetch pages from nonlocal temp relations. This provides a low-level defense against bugs-of-omission allowing temp pages to be loaded into shared buffers, as in the contrib/pgstattuple problem reported by Stuart Bishop. While at it, tweak a bunch of places to use new relcache tests (instead of expensive probes into pg_namespace) to detect local or nonlocal temp tables.
2009-04-01 00:12:48 +02:00
rel->rd_rel->relistemp = rel->rd_istemp;
Add a "relistemp" boolean column to pg_class, which is true for temporary relations (including a temp table's indexes and toast table/index), and false for normal relations. For ease of checking, this commit just adds the column and fills it correctly --- revising the relation access machinery to use it will come separately.
2009-03-31 19:59:56 +02:00
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
RelationGetRelid(rel) = relid;
for (i = 0; i < natts; i++)
rel->rd_att->attrs[i]->attrelid = relid;
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
rel->rd_rel->reltablespace = reltablespace;
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
if (mapped_relation)
{
rel->rd_rel->relfilenode = InvalidOid;
/* Add it to the active mapping information */
RelationMapUpdateMap(relid, relid, shared_relation, true);
}
else
rel->rd_rel->relfilenode = relid;
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
RelationInitLockInfo(rel); /* see lmgr.c */
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
RelationInitPhysicalAddr(rel);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
* Okay to insert into the relcache hash tables.
*/
RelationCacheInsert(rel);
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
/*
* done building relcache entry.
*/
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
MemoryContextSwitchTo(oldcxt);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
/* It's fully valid */
rel->rd_isvalid = true;
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
/*
* Caller expects us to pin the returned entry.
*/
RelationIncrementReferenceCount(rel);
Further work on connecting the free space map (which is still just a stub) into the rest of the system. Adopt a cleaner approach to preventing deadlock in concurrent heap_updates: allow RelationGetBufferForTuple to select any page of the rel, and put the onus on it to lock both buffers in a consistent order. Remove no-longer-needed isExtend hack from API of ReleaseAndReadBuffer.
2001-06-29 23:08:25 +02:00
return rel;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
/*
* RelationSetNewRelfilenode
*
* Assign a new relfilenode (physical file name) to the relation.
*
* This allows a full rewrite of the relation to be done with transactional
* safety (since the filenode assignment can be rolled back). Note however
* that there is no simple way to access the relation's old data for the
* remainder of the current transaction. This limits the usefulness to cases
* such as TRUNCATE or rebuilding an index from scratch.
*
* Caller must already hold exclusive lock on the relation.
*
* The relation is marked with relfrozenxid = freezeXid (InvalidTransactionId
* must be passed for indexes). This should be a lower bound on the XIDs
* that will be put into the new relation contents.
*/
void
RelationSetNewRelfilenode(Relation relation, TransactionId freezeXid)
{
Oid newrelfilenode;
RelFileNode newrnode;
Relation pg_class;
HeapTuple tuple;
Form_pg_class classform;
/* Indexes must have Invalid frozenxid; other relations must not */
Assert((relation->rd_rel->relkind == RELKIND_INDEX &&
freezeXid == InvalidTransactionId) ||
TransactionIdIsNormal(freezeXid));
/* Allocate a new relfilenode */
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
newrelfilenode = GetNewRelFileNode(relation->rd_rel->reltablespace, NULL);
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
/*
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
* Get a writable copy of the pg_class tuple for the given relation.
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
*/
pg_class = heap_open(RelationRelationId, RowExclusiveLock);
Wrap calls to SearchSysCache and related functions using macros. The purpose of this change is to eliminate the need for every caller of SearchSysCache, SearchSysCacheCopy, SearchSysCacheExists, GetSysCacheOid, and SearchSysCacheList to know the maximum number of allowable keys for a syscache entry (currently 4). This will make it far easier to increase the maximum number of keys in a future release should we choose to do so, and it makes the code shorter, too. Design and review by Tom Lane.
2010-02-14 19:42:19 +01:00
tuple = SearchSysCacheCopy1(RELOID,
ObjectIdGetDatum(RelationGetRelid(relation)));
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
if (!HeapTupleIsValid(tuple))
elog(ERROR, "could not find tuple for relation %u",
RelationGetRelid(relation));
classform = (Form_pg_class) GETSTRUCT(tuple);
/*
* Create storage for the main fork of the new relfilenode.
*
* NOTE: any conflict in relfilenode value will be caught here, if
* GetNewRelFileNode messes up for any reason.
*/
newrnode = relation->rd_node;
newrnode.relNode = newrelfilenode;
RelationCreateStorage(newrnode, relation->rd_istemp);
smgrclosenode(newrnode);
/*
* Schedule unlinking of the old storage at transaction commit.
*/
RelationDropStorage(relation);
/*
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
* Now update the pg_class row. However, if we're dealing with a mapped
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* index, pg_class.relfilenode doesn't change; instead we have to send the
* update to the relation mapper.
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
*/
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
if (RelationIsMapped(relation))
RelationMapUpdateMap(RelationGetRelid(relation),
newrelfilenode,
relation->rd_rel->relisshared,
false);
else
classform->relfilenode = newrelfilenode;
/* These changes are safe even for a mapped relation */
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
classform->relpages = 0; /* it's empty until further notice */
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
classform->reltuples = 0;
classform->relfrozenxid = freezeXid;
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
simple_heap_update(pg_class, &tuple->t_self, tuple);
CatalogUpdateIndexes(pg_class, tuple);
heap_freetuple(tuple);
heap_close(pg_class, RowExclusiveLock);
/*
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
* Make the pg_class row change visible, as well as the relation map
* change if any. This will cause the relcache entry to get updated, too.
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
*/
CommandCounterIncrement();
/*
* Mark the rel as having been given a new relfilenode in the current
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* (sub) transaction. This is a hint that can be used to optimize later
* operations on the rel in the same transaction.
Assorted cleanups in preparation for using a map file to support altering the relfilenode of currently-not-relocatable system catalogs. 1. Get rid of inval.c's dependency on relfilenode, by not having it emit smgr invalidations as a result of relcache flushes. Instead, smgr sinval messages are sent directly from smgr.c when an actual relation delete or truncate is done. This makes considerably more structural sense and allows elimination of a large number of useless smgr inval messages that were formerly sent even in cases where nothing was changing at the physical-relation level. Note that this reintroduces the concept of nontransactional inval messages, but that's okay --- because the messages are sent by smgr.c, they will be sent in Hot Standby slaves, just from a lower logical level than before. 2. Move setNewRelfilenode out of catalog/index.c, where it never logically belonged, into relcache.c; which is a somewhat debatable choice as well but better than before. (I considered catalog/storage.c, but that seemed too low level.) Rename to RelationSetNewRelfilenode. 3. Cosmetic cleanups of some other relfilenode manipulations.
2010-02-03 02:14:17 +01:00
*/
relation->rd_newRelfilenodeSubid = GetCurrentSubTransactionId();
/* ... and now we have eoxact cleanup work to do */
need_eoxact_work = true;
}
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
* RelationCacheInitialize
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* This initializes the relation descriptor cache. At the time
* that this is invoked, we can't do database access yet (mainly
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
* because the transaction subsystem is not up); all we are doing
* is making an empty cache hashtable. This must be done before
* starting the initialization transaction, because otherwise
* AtEOXact_RelationCache would crash if that transaction aborts
* before we can get the relcache set up.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
#define INITRELCACHESIZE 400
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
void
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
RelationCacheInitialize(void)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
HASHCTL ctl;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Remove a couple of unnecessary calls of CreateCacheMemoryContext. These probably got there via blind copy-and-paste from one of the legitimate callers, so rearrange and comment that code a bit to make it clearer that this isn't a necessary prerequisite to hash_create. Per observation from Robert Haas.
2009-12-27 19:55:52 +01:00
* make sure cache memory context exists
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
First phase of memory management rewrite (see backend/utils/mmgr/README for details). It doesn't really do that much yet, since there are no short-term memory contexts in the executor, but the infrastructure is in place and long-term contexts are handled reasonably. A few long- standing bugs have been fixed, such as 'VACUUM; anything' in a single query string crashing. Also, out-of-memory is now considered a recoverable ERROR, not FATAL. Eliminate a large amount of crufty, now-dead code in and around memory management. Fix problem with holding off SIGTRAP, SIGSEGV, etc in postmaster and backend startup.
2000-06-28 05:33:33 +02:00
if (!CacheMemoryContext)
CreateCacheMemoryContext();
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Further minor simplification of relcache startup: don't need a static needNewCacheFile flag anymore, it can just be local in RelationCacheInitializePhase2.
2006-05-06 17:51:07 +02:00
* create hashtable that indexes the relcache
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
Another round of cleanups for dynahash.c (maybe it's finally clean of portability issues). Caller-visible data structures are now allocated on MAXALIGN boundaries, allowing safe use of datatypes wider than 'long'. Rejigger hash_create API so that caller specifies size of key and total size of entry, not size of key and size of rest of entry. This simplifies life considerably since each number is just a sizeof(), and padding issues etc. are taken care of automatically.
2001-10-01 07:36:17 +02:00
MemSet(&ctl, 0, sizeof(ctl));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
ctl.keysize = sizeof(Oid);
Another round of cleanups for dynahash.c (maybe it's finally clean of portability issues). Caller-visible data structures are now allocated on MAXALIGN boundaries, allowing safe use of datatypes wider than 'long'. Rejigger hash_create API so that caller specifies size of key and total size of entry, not size of key and size of rest of entry. This simplifies life considerably since each number is just a sizeof(), and padding issues etc. are taken care of automatically.
2001-10-01 07:36:17 +02:00
ctl.entrysize = sizeof(RelIdCacheEnt);
Marginal hack to use a specialized hash function for dynahash hashtables whose keys are OIDs. The only one that looks particularly performance critical is the relcache hashtable, but as long as we've got the function we may as well use it wherever it's applicable.
2005-04-14 22:32:43 +02:00
ctl.hash = oid_hash;
Further cleanup of dynahash.c API, in pursuit of portability and readability. Bizarre '(long *) TRUE' return convention is gone, in favor of just raising an error internally in dynahash.c when we detect hashtable corruption. HashTableWalk is gone, in favor of using hash_seq_search directly, since it had no hope of working with non-LONGALIGNable datatypes. Simplify some other code that was made undesirably grotty by promixity to HashTableWalk.
2001-10-05 19:28:13 +02:00
RelationIdCache = hash_create("Relcache by OID", INITRELCACHESIZE,
&ctl, HASH_ELEM | HASH_FUNCTION);
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/*
* relation mapper needs initialized too
*/
RelationMapInitialize();
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
}
/*
* RelationCacheInitializePhase2
*
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* This is called to prepare for access to shared catalogs during startup.
* We must at least set up nailed reldescs for pg_database, pg_authid,
* and pg_auth_members. Ideally we'd like to have reldescs for their
* indexes, too. We attempt to load this information from the shared
* relcache init file. If that's missing or broken, just make phony
* entries for the catalogs themselves. RelationCacheInitializePhase3
* will clean up as needed.
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
*/
void
RelationCacheInitializePhase2(void)
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
{
MemoryContext oldcxt;
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/*
* relation mapper needs initialized too
*/
RelationMapInitializePhase2();
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
pgindent run for 9.0, second run
2010-07-06 21:19:02 +02:00
* In bootstrap mode, the shared catalogs aren't there yet anyway, so do
* nothing.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*/
if (IsBootstrapProcessingMode())
return;
/*
* switch to cache memory context
*/
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Try to load the shared relcache cache file. If unsuccessful, bootstrap
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* the cache with pre-made descriptors for the critical shared catalogs.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*/
if (!load_relcache_init_file(true))
{
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
formrdesc("pg_database", DatabaseRelation_Rowtype_Id, true,
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
true, Natts_pg_database, Desc_pg_database);
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
formrdesc("pg_authid", AuthIdRelation_Rowtype_Id, true,
true, Natts_pg_authid, Desc_pg_authid);
formrdesc("pg_auth_members", AuthMemRelation_Rowtype_Id, true,
false, Natts_pg_auth_members, Desc_pg_auth_members);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
#define NUM_CRITICAL_SHARED_RELS 3 /* fix if you change list above */
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
}
MemoryContextSwitchTo(oldcxt);
}
/*
* RelationCacheInitializePhase3
*
* This is called as soon as the catcache and transaction system
* are functional and we have determined MyDatabaseId. At this point
* we can actually read data from the database's system catalogs.
* We first try to read pre-computed relcache entries from the local
* relcache init file. If that's missing or broken, make phony entries
* for the minimum set of nailed-in-cache relations. Then (unless
* bootstrapping) make sure we have entries for the critical system
* indexes. Once we've done all this, we have enough infrastructure to
* open any system catalog or use any catcache. The last step is to
* rewrite the cache files if needed.
*/
void
RelationCacheInitializePhase3(void)
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
{
HASH_SEQ_STATUS status;
RelIdCacheEnt *idhentry;
MemoryContext oldcxt;
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
bool needNewCacheFile = !criticalSharedRelcachesBuilt;
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
Create a "relation mapping" infrastructure to support changing the relfilenodes of shared or nailed system catalogs. This has two key benefits: * The new CLUSTER-based VACUUM FULL can be applied safely to all catalogs. * We no longer have to use an unsafe reindex-in-place approach for reindexing shared catalogs. CLUSTER on nailed catalogs now works too, although I left it disabled on shared catalogs because the resulting pg_index.indisclustered update would only be visible in one database. Since reindexing shared system catalogs is now fully transactional and crash-safe, the former special cases in REINDEX behavior have been removed; shared catalogs are treated the same as non-shared. This commit does not do anything about the recently-discussed problem of deadlocks between VACUUM FULL/CLUSTER on a system catalog and other concurrent queries; will address that in a separate patch. As a stopgap, parallel_schedule has been tweaked to run vacuum.sql by itself, to avoid such failures during the regression tests.
2010-02-07 21:48:13 +01:00
/*
* relation mapper needs initialized too
*/
RelationMapInitializePhase3();
Change /*---- commants to /* where appropriate. pgindent will tighten up the comments later.
2001-02-22 19:39:20 +01:00
/*
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
* switch to cache memory context
*/
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Try to load the local relcache cache file. If unsuccessful, bootstrap
* the cache with pre-made descriptors for the critical "nailed-in" system
* catalogs.
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (IsBootstrapProcessingMode() ||
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
!load_relcache_init_file(false))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
Further minor simplification of relcache startup: don't need a static needNewCacheFile flag anymore, it can just be local in RelationCacheInitializePhase2.
2006-05-06 17:51:07 +02:00
needNewCacheFile = true;
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
formrdesc("pg_class", RelationRelation_Rowtype_Id, false,
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
true, Natts_pg_class, Desc_pg_class);
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
formrdesc("pg_attribute", AttributeRelation_Rowtype_Id, false,
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
false, Natts_pg_attribute, Desc_pg_attribute);
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
formrdesc("pg_proc", ProcedureRelation_Rowtype_Id, false,
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
true, Natts_pg_proc, Desc_pg_proc);
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
formrdesc("pg_type", TypeRelation_Rowtype_Id, false,
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
true, Natts_pg_type, Desc_pg_type);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
#define NUM_CRITICAL_LOCAL_RELS 4 /* fix if you change list above */
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
MemoryContextSwitchTo(oldcxt);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
/* In bootstrap mode, the faked-up formrdesc info is all we'll have */
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (IsBootstrapProcessingMode())
return;
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* If we didn't get the critical system indexes loaded into relcache, do
Found the problem with my operator-family changes: by fetching from pg_opclass during LookupOpclassInfo(), I'd turned pg_opclass_oid_index into a critical system index. However the problem could only manifest during a backend's first attempt to load opclass data, and then only if it had successfully loaded pg_internal.init and subsequently received a relcache flush; which made it impossible to reproduce in sequential tests and darn hard even in parallel tests. Memo to self: when exercising cache flush scenarios, must disable LookupOpclassInfo's internal cache too.
2006-12-31 21:32:04 +01:00
* so now. These are critical because the catcache and/or opclass cache
* depend on them for fetches done during relcache load. Thus, we have an
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* infinite-recursion problem. We can break the recursion by doing
* heapscans instead of indexscans at certain key spots. To avoid hobbling
* performance, we only want to do that until we have the critical indexes
* loaded into relcache. Thus, the flag criticalRelcachesBuilt is used to
* decide whether to do heapscan or indexscan at the key spots, and we set
* it true after we've loaded the critical indexes.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* The critical indexes are marked as "nailed in cache", partly to make it
* easy for load_relcache_init_file to count them, but mainly because we
* cannot flush and rebuild them once we've set criticalRelcachesBuilt to
* true. (NOTE: perhaps it would be possible to reload them by
* temporarily setting criticalRelcachesBuilt to false again. For now,
* though, we just nail 'em in.)
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
*
* RewriteRelRulenameIndexId and TriggerRelidNameIndexId are not critical
* in the same way as the others, because the critical catalogs don't
* (currently) have any rules or triggers, and so these indexes can be
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
* rebuilt without inducing recursion. However they are used during
Avoid leaking memory while reading toasted entries from pg_rewrite, and nail a couple more system indexes into cache. This doesn't make any difference in normal system operation, but when forcing constant cache resets it's difficult to get through the rules regression test without these changes.
2006-01-08 21:04:41 +01:00
* relcache load when a rel does have rules or triggers, so we choose to
* nail them for performance reasons.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
pgindent run.
2002-09-04 22:31:48 +02:00
if (!criticalRelcachesBuilt)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
load_critical_index(ClassOidIndexId,
RelationRelationId);
load_critical_index(AttributeRelidNumIndexId,
AttributeRelationId);
load_critical_index(IndexRelidIndexId,
IndexRelationId);
load_critical_index(OpclassOidIndexId,
OperatorClassRelationId);
load_critical_index(AccessMethodStrategyIndexId,
AccessMethodOperatorRelationId);
load_critical_index(AccessMethodProcedureIndexId,
AccessMethodProcedureRelationId);
load_critical_index(OperatorOidIndexId,
OperatorRelationId);
load_critical_index(RewriteRelRulenameIndexId,
RewriteRelationId);
load_critical_index(TriggerRelidNameIndexId,
TriggerRelationId);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
#define NUM_CRITICAL_LOCAL_INDEXES 9 /* fix if you change list above */
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
criticalRelcachesBuilt = true;
}
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
* Process critical shared indexes too.
*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* DatabaseNameIndexId isn't critical for relcache loading, but rather for
* initial lookup of MyDatabaseId, without which we'll never find any
* non-shared catalogs at all. Autovacuum calls InitPostgres with a
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
* database OID, so it instead depends on DatabaseOidIndexId. We also
* need to nail up some indexes on pg_authid and pg_auth_members for use
* during client authentication.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*/
if (!criticalSharedRelcachesBuilt)
{
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
load_critical_index(DatabaseNameIndexId,
DatabaseRelationId);
load_critical_index(DatabaseOidIndexId,
DatabaseRelationId);
Arrange for client authentication to occur before we select a specific database to connect to. This is necessary for the walsender code to work properly (it was previously using an untenable assumption that template1 would always be available to connect to). This also gets rid of a small security shortcoming that was introduced in the original patch to eliminate the flat authentication files: before, you could find out whether or not the requested database existed even if you couldn't pass the authentication checks. The changes needed to support this are mainly just to treat pg_authid and pg_auth_members as nailed relations, so that we can read them without having to be able to locate real pg_class entries for them. This mechanism was already debugged for pg_database, but we hadn't recognized the value of applying it to those catalogs too. Since the current code doesn't have support for accessing toast tables before we've brought up all of the relcache, remove pg_authid's toast table to ensure that no one can store an out-of-line toasted value of rolpassword. The case seems quite unlikely to occur in practice, and was effectively unsupported anyway in the old "flatfiles" implementation. Update genbki.pl to actually implement the same rules as bootstrap.c does for not-nullability of catalog columns. The previous coding was a bit cheesy but worked all right for the previous set of bootstrap catalogs. It does not work for pg_authid, where rolvaliduntil needs to be nullable. Initdb forced due to minor catalog changes (mainly the toast table removal).
2010-04-21 01:48:47 +02:00
load_critical_index(AuthIdRolnameIndexId,
AuthIdRelationId);
load_critical_index(AuthIdOidIndexId,
AuthIdRelationId);
load_critical_index(AuthMemMemRoleIndexId,
AuthMemRelationId);
#define NUM_CRITICAL_SHARED_INDEXES 5 /* fix if you change list above */
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
criticalSharedRelcachesBuilt = true;
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Now, scan all the relcache entries and update anything that might be
* wrong in the results from formrdesc or the relcache cache file. If we
* faked up relcache entries using formrdesc, then read the real pg_class
* rows and replace the fake entries with them. Also, if any of the
* relcache entries have rules or triggers, load that info the hard way
* since it isn't recorded in the cache file.
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Whenever we access the catalogs to read data, there is a possibility of
* a shared-inval cache flush causing relcache entries to be removed.
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
* Since hash_seq_search only guarantees to still work after the *current*
* entry is removed, it's unsafe to continue the hashtable scan afterward.
* We handle this by restarting the scan from scratch after each access.
* This is theoretically O(N^2), but the number of entries that actually
* need to be fixed is small enough that it doesn't matter.
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
*/
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
hash_seq_init(&status, RelationIdCache);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
while ((idhentry = (RelIdCacheEnt *) hash_seq_search(&status)) != NULL)
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
{
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
Relation relation = idhentry->reldesc;
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
bool restart = false;
/*
* Make sure *this* entry doesn't get flushed while we work with it.
*/
RelationIncrementReferenceCount(relation);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
init_irels() is changed to be called in RelationCacheInitializePhase2() so that transactional control could guarantee the consistency.
2001-01-06 02:48:59 +01:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* If it's a faked-up entry, read the real pg_class tuple.
init_irels() is changed to be called in RelationCacheInitializePhase2() so that transactional control could guarantee the consistency.
2001-01-06 02:48:59 +01:00
*/
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
if (relation->rd_rel->relowner == InvalidOid)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
HeapTuple htup;
Form_pg_class relp;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
Wrap calls to SearchSysCache and related functions using macros. The purpose of this change is to eliminate the need for every caller of SearchSysCache, SearchSysCacheCopy, SearchSysCacheExists, GetSysCacheOid, and SearchSysCacheList to know the maximum number of allowable keys for a syscache entry (currently 4). This will make it far easier to increase the maximum number of keys in a future release should we choose to do so, and it makes the code shorter, too. Design and review by Tom Lane.
2010-02-14 19:42:19 +01:00
htup = SearchSysCache1(RELOID,
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
ObjectIdGetDatum(RelationGetRelid(relation)));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (!HeapTupleIsValid(htup))
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(FATAL, "cache lookup failed for relation %u",
RelationGetRelid(relation));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
relp = (Form_pg_class) GETSTRUCT(htup);
pgindent run.
2002-09-04 22:31:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* Copy tuple to relation->rd_rel. (See notes in
* AllocateRelationDesc())
*/
memcpy((char *) relation->rd_rel, (char *) relp, CLASS_TUPLE_SIZE);
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
/* Update rd_options while we have the tuple */
if (relation->rd_options)
pfree(relation->rd_options);
RelationParseRelOptions(relation, htup);
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
/*
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
* Check the values in rd_att were set up correctly. (We cannot
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* just copy them over now: formrdesc must have set up the rd_att
* data correctly to start with, because it may already have been
* copied into one or more catcache entries.)
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
*/
Revert my ill-considered change that made formrdesc not insert the correct relation rowtype OID into the relcache entries it builds. This ensures that catcache copies of the relation tupdescs will be fully correct. While the deficiency doesn't seem to have any effect in the current sources, we have been bitten by not-quite-right catcache tupdescs before, so it seems like a good idea to maintain the rule that they should be right.
2009-09-27 01:08:22 +02:00
Assert(relation->rd_att->tdtypeid == relp->reltype);
Assert(relation->rd_att->tdtypmod == -1);
Assert(relation->rd_att->tdhasoid == relp->relhasoids);
init_irels() is changed to be called in RelationCacheInitializePhase2() so that transactional control could guarantee the consistency.
2001-01-06 02:48:59 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
ReleaseSysCache(htup);
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
/* relowner had better be OK now, else we'll loop forever */
if (relation->rd_rel->relowner == InvalidOid)
elog(ERROR, "invalid relowner in pg_class entry for \"%s\"",
RelationGetRelationName(relation));
restart = true;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
/*
* Fix data that isn't saved in relcache cache file.
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
*
* relhasrules or relhastriggers could possibly be wrong or out of
* date. If we don't actually find any rules or triggers, clear the
* local copy of the flag so that we don't get into an infinite loop
* here. We don't make any attempt to fix the pg_class entry, though.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
if (relation->rd_rel->relhasrules && relation->rd_rules == NULL)
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
RelationBuildRuleLock(relation);
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
if (relation->rd_rules == NULL)
relation->rd_rel->relhasrules = false;
restart = true;
}
Replace pg_class.reltriggers with relhastriggers, which is just a boolean hint ("there might be triggers") rather than an exact count. This is necessary catalog infrastructure for the upcoming patch to reduce the strength of locking needed for trigger addition/removal. Split out and committed separately for ease of reviewing/testing. In passing, also get rid of the unused pg_class columns relukeys, relfkeys, and relrefs, which haven't been maintained in many years and now have no chance of ever being maintained (because of wishing to avoid locking). Simon Riggs
2008-11-09 22:24:33 +01:00
if (relation->rd_rel->relhastriggers && relation->trigdesc == NULL)
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
RelationBuildTriggers(relation);
Fix RelationCacheInitializePhase2 (Phase3, in HEAD) to cope with the possibility of shared-inval messages causing a relcache flush while it tries to fill in missing data in preloaded relcache entries. There are actually two distinct failure modes here: 1. The flush could delete the next-to-be-processed cache entry, causing the subsequent hash_seq_search calls to go off into the weeds. This is the problem reported by Michael Brown, and I believe it also accounts for bug #5074. The simplest fix is to restart the hashtable scan after we've read any new data from the catalogs. It appears that pre-8.4 branches have not suffered from this failure, because by chance there were no other catalogs sharing the same hash chains with the catalogs that RelationCacheInitializePhase2 had work to do for. However that's obviously pretty fragile, and it seems possible that derivative versions with additional system catalogs might be vulnerable, so I'm back-patching this part of the fix anyway. 2. The flush could delete the *current* cache entry, in which case the pointer to the newly-loaded data would end up being stored into an already-deleted Relation struct. As long as it was still deleted, the only consequence would be some leaked space in CacheMemoryContext. But it seems possible that the Relation struct could already have been recycled, in which case this represents a hard-to-reproduce clobber of cached data structures, with unforeseeable consequences. The fix here is to pin the entry while we work on it. In passing, also change RelationCacheInitializePhase2 to Assert that formrdesc() set up the relation's cached TupleDesc (rd_att) with the correct type OID and hasoids values. This is more appropriate than silently updating the values, because the original tupdesc might already have been copied into the catcache. However this part of the patch is not in HEAD because it fails due to some questionable recent changes in formrdesc :-(. That will be cleaned up in a subsequent patch.
2009-09-26 20:24:49 +02:00
if (relation->trigdesc == NULL)
relation->rd_rel->relhastriggers = false;
restart = true;
}
/* Release hold on the relation */
RelationDecrementReferenceCount(relation);
/* Now, restart the hashtable scan if needed */
if (restart)
{
hash_seq_term(&status);
hash_seq_init(&status, RelationIdCache);
}
Toast all the system-table columns that seem to need it. It turns out that giving pg_proc a toast table required solving the same problems we'd have to solve for pg_class --- pg_proc is one of the relations that gets bootstrapped in relcache.c. Solution is to go back at the end of initialization and read in the *real* pg_class row to replace the phony entry created by formrdesc(). This should work as long as there's no need to touch any toasted values during initialization, which seems a reasonable assumption. Although I did not add a toast-table for every single system table with a varlena attribute, I believe that it would work to just do ALTER TABLE pg_class CREATE TOAST TABLE. So anyone who's really intent on having several thousand ACL entries for a rel could do it. NOTE: I didn't force initdb, but you must do one to see the effects of this patch.
2000-08-06 06:40:08 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
/*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* Lastly, write out new relcache cache files if needed. We don't bother
* to distinguish cases where only one of the two needs an update.
Simplify relcache startup sequence. With the new design of InitPostgres it's not necessary to have three separate calls anymore. This patch also fixes things so we don't try to read pg_internal.init until after we've obtained lock on the target database; which was fairly harmless, but it's certainly cleaner this way.
2006-05-04 20:51:36 +02:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (needNewCacheFile)
{
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Force all the catcaches to finish initializing and thereby open the
* catalogs and indexes they use. This will preload the relcache with
* entries for all the most important system catalogs and indexes, so
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* that the init files will be most useful for future backends.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
InitCatalogCachePhase2();
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/* reset initFileRelationIds list; we'll fill it during write */
initFileRelationIds = NIL;
/* now write the files */
write_relcache_init_file(true);
write_relcache_init_file(false);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
}
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
* Load one critical system index into the relcache
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
*
* indexoid is the OID of the target index, heapoid is the OID of the catalog
* it belongs to.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*/
static void
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
load_critical_index(Oid indexoid, Oid heapoid)
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
{
Relation ird;
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
/*
* We must lock the underlying catalog before locking the index to avoid
* deadlock, since RelationBuildDesc might well need to read the catalog,
* and if anyone else is exclusive-locking this catalog and index they'll
* be doing it in that order.
*/
LockRelationOid(heapoid, AccessShareLock);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
LockRelationOid(indexoid, AccessShareLock);
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
ird = RelationBuildDesc(indexoid, true);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (ird == NULL)
elog(PANIC, "could not open critical system index %u", indexoid);
ird->rd_isnailed = true;
ird->rd_refcnt = 1;
UnlockRelationOid(indexoid, AccessShareLock);
When loading critical system indexes into the relcache, ensure we lock the underlying catalog not only the index itself. Otherwise, if the cache load process touches the catalog (which will happen for many though not all of these indexes), we are locking index before parent table, which can result in a deadlock against processes that are trying to lock them in the normal order. Per today's failure on buildfarm member gothic_moth; it's surprising the problem hadn't been identified before. Back-patch to 8.2. Earlier releases didn't have the issue because they didn't try to lock these indexes during load (instead assuming that they couldn't change schema at all during multiuser operation).
2010-01-14 00:07:08 +01:00
UnlockRelationOid(heapoid, AccessShareLock);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
}
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
/*
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
* GetPgClassDescriptor -- get a predefined tuple descriptor for pg_class
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
* GetPgIndexDescriptor -- get a predefined tuple descriptor for pg_index
*
* We need this kluge because we have to be able to access non-fixed-width
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
* fields of pg_class and pg_index before we have the standard catalog caches
* available. We use predefined data that's set up in just the same way as
* the bootstrapped reldescs used by formrdesc(). The resulting tupdesc is
* not 100% kosher: it does not have the correct rowtype OID in tdtypeid, nor
* does it have a TupleConstr field. But it's good enough for the purpose of
* extracting fields.
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
*/
static TupleDesc
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
BuildHardcodedDescriptor(int natts, const FormData_pg_attribute *attrs,
bool hasoids)
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
{
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
TupleDesc result;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
MemoryContext oldcxt;
int i;
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
result = CreateTemplateTupleDesc(natts, hasoids);
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
result->tdtypeid = RECORDOID; /* not right, but we don't care */
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
result->tdtypmod = -1;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
for (i = 0; i < natts; i++)
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
{
Support column-level privileges, as required by SQL standard. Stephen Frost, with help from KaiGai Kohei and others
2009-01-22 21:16:10 +01:00
memcpy(result->attrs[i], &attrs[i], ATTRIBUTE_FIXED_PART_SIZE);
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
/* make sure attcacheoff is valid */
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
result->attrs[i]->attcacheoff = -1;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
}
/* initialize first attribute's attcacheoff, cf RelationBuildTupleDesc */
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
result->attrs[0]->attcacheoff = 0;
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
/* Note: we don't bother to set up a TupleConstr entry */
MemoryContextSwitchTo(oldcxt);
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
return result;
}
static TupleDesc
GetPgClassDescriptor(void)
{
static TupleDesc pgclassdesc = NULL;
/* Already done? */
if (pgclassdesc == NULL)
pgclassdesc = BuildHardcodedDescriptor(Natts_pg_class,
Desc_pg_class,
true);
return pgclassdesc;
}
static TupleDesc
GetPgIndexDescriptor(void)
{
static TupleDesc pgindexdesc = NULL;
/* Already done? */
if (pgindexdesc == NULL)
pgindexdesc = BuildHardcodedDescriptor(Natts_pg_index,
Desc_pg_index,
false);
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
return pgindexdesc;
}
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
* Load any default attribute value definitions for the relation.
*/
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
static void
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
AttrDefaultFetch(Relation relation)
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
AttrDefault *attrdef = relation->rd_att->constr->defval;
int ndef = relation->rd_att->constr->num_defval;
Relation adrel;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
SysScanDesc adscan;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
ScanKeyData skey;
Implement reindex command
2000-02-18 10:30:20 +01:00
HeapTuple htup;
Update textin() and textout() to new fmgr style. This is just phase one of updating the whole text datatype, but there are so dang many calls of these two routines that it seems worth a separate commit.
2000-07-06 01:12:09 +02:00
Datum val;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
bool isnull;
int found;
int i;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey,
Anum_pg_attrdef_adrelid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(RelationGetRelid(relation)));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
adrel = heap_open(AttrDefaultRelationId, AccessShareLock);
adscan = systable_beginscan(adrel, AttrDefaultIndexId, true,
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
SnapshotNow, 1, &skey);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
found = 0;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
while (HeapTupleIsValid(htup = systable_getnext(adscan)))
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Form_pg_attrdef adform = (Form_pg_attrdef) GETSTRUCT(htup);
OK, folks, here is the pgindent output.
1998-09-01 06:40:42 +02:00
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
for (i = 0; i < ndef; i++)
{
if (adform->adnum != attrdef[i].adnum)
continue;
Reimplement parsing and storage of default expressions and constraint expressions in CREATE TABLE. There is no longer an emasculated expression syntax for these things; it's full a_expr for constraints, and b_expr for defaults (unfortunately the fact that NOT NULL is a part of the column constraint syntax causes a shift/reduce conflict if you try a_expr. Oh well --- at least parenthesized boolean expressions work now). Also, stored expression for a column default is not pre-coerced to the column type; we rely on transformInsertStatement to do that when the default is actually used. This means "f1 datetime default 'now'" behaves the way people usually expect it to. BTW, all the support code is now there to implement ALTER TABLE ADD CONSTRAINT and ALTER TABLE ADD COLUMN with a default value. I didn't actually teach ALTER TABLE to call it, but it wouldn't be much work.
1999-10-04 01:55:40 +02:00
if (attrdef[i].adbin != NULL)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(WARNING, "multiple attrdef records found for attr %s of rel %s",
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
NameStr(relation->rd_att->attrs[adform->adnum - 1]->attname),
New NameStr macro to convert Name to Str. No need for var.data anymore. Fewer calls to nameout. Better use of RelationGetRelationName.
1999-11-08 00:08:36 +01:00
RelationGetRelationName(relation));
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
else
found++;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Update textin() and textout() to new fmgr style. This is just phase one of updating the whole text datatype, but there are so dang many calls of these two routines that it seems worth a separate commit.
2000-07-06 01:12:09 +02:00
val = fastgetattr(htup,
Anum_pg_attrdef_adbin,
adrel->rd_att, &isnull);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
if (isnull)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(WARNING, "null adbin for attr %s of rel %s",
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
NameStr(relation->rd_att->attrs[adform->adnum - 1]->attname),
New NameStr macro to convert Name to Str. No need for var.data anymore. Fewer calls to nameout. Better use of RelationGetRelationName.
1999-11-08 00:08:36 +01:00
RelationGetRelationName(relation));
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
else
attrdef[i].adbin = MemoryContextStrdup(CacheMemoryContext,
8.4 pgindent run, with new combined Linux/FreeBSD/MinGW typedef list provided by Andrew.
2009-06-11 16:49:15 +02:00
TextDatumGetCString(val));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
break;
}
OK, folks, here is the pgindent output.
1998-09-01 06:40:42 +02:00
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
if (i >= ndef)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(WARNING, "unexpected attrdef record found for attr %d of rel %s",
adform->adnum, RelationGetRelationName(relation));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
systable_endscan(adscan);
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
heap_close(adrel, AccessShareLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (found != ndef)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(WARNING, "%d attrdef record(s) missing for rel %s",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
ndef - found, RelationGetRelationName(relation));
Read info for DEFAULT from pg_attrdef.
1997-08-21 03:36:09 +02:00
}
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/*
* Load any check constraints for the relation.
*/
Fetch information about DEFAULT/CHECK while openning a relation.
1997-08-22 05:35:44 +02:00
static void
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
CheckConstraintFetch(Relation relation)
Fetch information about DEFAULT/CHECK while openning a relation.
1997-08-22 05:35:44 +02:00
{
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
ConstrCheck *check = relation->rd_att->constr->check;
int ncheck = relation->rd_att->constr->num_check;
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
Relation conrel;
SysScanDesc conscan;
ScanKeyData skey[1];
Implement reindex command
2000-02-18 10:30:20 +01:00
HeapTuple htup;
Update textin() and textout() to new fmgr style. This is just phase one of updating the whole text datatype, but there are so dang many calls of these two routines that it seems worth a separate commit.
2000-07-06 01:12:09 +02:00
Datum val;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
bool isnull;
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
int found = 0;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey[0],
Anum_pg_constraint_conrelid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(RelationGetRelid(relation)));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
conrel = heap_open(ConstraintRelationId, AccessShareLock);
conscan = systable_beginscan(conrel, ConstraintRelidIndexId, true,
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
SnapshotNow, 1, skey);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
while (HeapTupleIsValid(htup = systable_getnext(conscan)))
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
Form_pg_constraint conform = (Form_pg_constraint) GETSTRUCT(htup);
/* We want check constraints only */
if (conform->contype != CONSTRAINT_CHECK)
continue;
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
if (found >= ncheck)
elog(ERROR, "unexpected constraint record found for rel %s",
New NameStr macro to convert Name to Str. No need for var.data anymore. Fewer calls to nameout. Better use of RelationGetRelationName.
1999-11-08 00:08:36 +01:00
RelationGetRelationName(relation));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
check[found].ccname = MemoryContextStrdup(CacheMemoryContext,
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
NameStr(conform->conname));
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
/* Grab and test conbin is actually set */
Update textin() and textout() to new fmgr style. This is just phase one of updating the whole text datatype, but there are so dang many calls of these two routines that it seems worth a separate commit.
2000-07-06 01:12:09 +02:00
val = fastgetattr(htup,
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
Anum_pg_constraint_conbin,
conrel->rd_att, &isnull);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
if (isnull)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "null conbin for rel %s",
New NameStr macro to convert Name to Str. No need for var.data anymore. Fewer calls to nameout. Better use of RelationGetRelationName.
1999-11-08 00:08:36 +01:00
RelationGetRelationName(relation));
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
Use a private memory context to store rule information in each relcache entry that has rules. This allows us to release the rule parsetrees on relcache flush without needing a working freeObject() routine. Formerly, the rule trees were leaked permanently at relcache flush. Also, clean up handling of rule creation and deletion --- there was not sufficient locking of the relation being modified, and there was no reliable notification of other backends that a relcache reload was needed. Also, clean up relcache.c code so that scans of system tables needed to load a relcache entry are done in the caller's memory context, not in CacheMemoryContext. This prevents any un-pfreed memory from those scans from becoming a permanent memory leak.
2000-06-30 09:04:23 +02:00
check[found].ccbin = MemoryContextStrdup(CacheMemoryContext,
Simplify and standardize conversions between TEXT datums and ordinary C strings. This patch introduces four support functions cstring_to_text, cstring_to_text_with_len, text_to_cstring, and text_to_cstring_buffer, and two macros CStringGetTextDatum and TextDatumGetCString. A number of existing macros that provided variants on these themes were removed. Most of the places that need to make such conversions now require just one function or macro call, in place of the multiple notational layers that used to be needed. There are no longer any direct calls of textout or textin, and we got most of the places that were using handmade conversions via memcpy (there may be a few still lurking, though). This commit doesn't make any serious effort to eliminate transient memory leaks caused by detoasting toasted text objects before they reach text_to_cstring. We changed PG_GETARG_TEXT_P to PG_GETARG_TEXT_PP in a few places where it was easy, but much more could be done. Brendan Jurd and Tom Lane
2008-03-25 23:42:46 +01:00
TextDatumGetCString(val));
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
found++;
}
Second phase of committing Rod Taylor's pg_depend/pg_constraint patch. pg_relcheck is gone; CHECK, UNIQUE, PRIMARY KEY, and FOREIGN KEY constraints all have real live entries in pg_constraint. pg_depend exists, and RESTRICT/CASCADE options work on most kinds of DROP; however, pg_depend is not yet very well populated with dependencies. (Most of the ones that are present at this point just replace formerly hardwired associations, such as the implicit drop of a relation's pg_type entry when the relation is dropped.) Need to add more logic to create dependency entries, improve pg_dump to dump constraints in place of indexes and triggers, and add some regression tests.
2002-07-12 20:43:19 +02:00
systable_endscan(conscan);
heap_close(conrel, AccessShareLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (found != ncheck)
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
elog(ERROR, "%d constraint record(s) missing for rel %s",
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
ncheck - found, RelationGetRelationName(relation));
Fetch information about DEFAULT/CHECK while openning a relation.
1997-08-22 05:35:44 +02:00
}
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
/*
* RelationGetIndexList -- get a list of OIDs of indexes on this relation
*
* The index list is created only if someone requests it. We scan pg_index
* to find relevant indexes, and add the list to the relcache entry so that
* we won't have to compute it again. Note that shared cache inval of a
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
* relcache entry will delete the old list and set rd_indexvalid to 0,
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
* so that we must recompute the index list on next request. This handles
* creation or deletion of an index.
*
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
* The returned list is guaranteed to be sorted in order by OID. This is
* needed by the executor, since for index types that we obtain exclusive
* locks on when updating the index, all backends must lock the indexes in
* the same order or we will get deadlocks (see ExecOpenIndices()). Any
* consistent ordering would do, but ordering by OID is easy.
*
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
* Since shared cache inval causes the relcache's copy of the list to go away,
* we return a copy of the list palloc'd in the caller's context. The caller
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* may list_free() the returned list after scanning it. This is necessary
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
* since the caller will typically be doing syscache lookups on the relevant
* indexes, and syscache lookup could cause SI messages to be processed!
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
*
* We also update rd_oidindex, which this module treats as effectively part
* of the index list. rd_oidindex is valid when rd_indexvalid isn't zero;
* it is the pg_class OID of a unique index on OID when the relation has one,
* and InvalidOid if there is no such index.
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
*/
List *
RelationGetIndexList(Relation relation)
{
Relation indrel;
pgindent run.
2002-09-04 22:31:48 +02:00
SysScanDesc indscan;
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
ScanKeyData skey;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
HeapTuple htup;
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
List *result;
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
Oid oidIndex;
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
MemoryContext oldcxt;
/* Quick exit if we already computed the list. */
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
if (relation->rd_indexvalid != 0)
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
return list_copy(relation->rd_indexlist);
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We build the list we intend to return (in the caller's context) while
* doing the scan. After successfully completing the scan, we copy that
* list into the relcache entry. This avoids cache-context memory leakage
* if we get some sort of error partway through.
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
*/
result = NIL;
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
oidIndex = InvalidOid;
pgindent run. Make it all clean.
2001-03-22 05:01:46 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* Prepare to scan pg_index for entries having indrelid = this rel. */
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
ScanKeyInit(&skey,
Anum_pg_index_indrelid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(RelationGetRelid(relation)));
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
Completion of project to use fixed OIDs for all system catalogs and indexes. Replace all heap_openr and index_openr calls by heap_open and index_open. Remove runtime lookups of catalog OID numbers in various places. Remove relcache's support for looking up system catalogs by name. Bulky but mostly very boring patch ...
2005-04-14 22:03:27 +02:00
indrel = heap_open(IndexRelationId, AccessShareLock);
indscan = systable_beginscan(indrel, IndexIndrelidIndexId, true,
Cross-data-type comparisons are now indexable by btrees, pursuant to my pghackers proposal of 8-Nov. All the existing cross-type comparison operators (int2/int4/int8 and float4/float8) have appropriate support. The original proposal of storing the right-hand-side datatype as part of the primary key for pg_amop and pg_amproc got modified a bit in the event; it is easier to store zero as the 'default' case and only store a nonzero when the operator is actually cross-type. Along the way, remove the long-since-defunct bigbox_ops operator class.
2003-11-12 22:15:59 +01:00
SnapshotNow, 1, &skey);
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
while (HeapTupleIsValid(htup = systable_getnext(indscan)))
{
Form_pg_index index = (Form_pg_index) GETSTRUCT(htup);
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
/* Add index's OID to result list in the proper order */
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
result = insert_ordered_oid(result, index->indexrelid);
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
/* Check to see if it is a unique, non-partial btree index on OID */
if (index->indnatts == 1 &&
Support deferrable uniqueness constraints. The current implementation fires an AFTER ROW trigger for each tuple that looks like it might be non-unique according to the index contents at the time of insertion. This works well as long as there aren't many conflicts, but won't scale to massive unique-key reassignments. Improving that case is a TODO item. Dean Rasheed
2009-07-29 22:56:21 +02:00
index->indisunique && index->indimmediate &&
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
index->indkey.values[0] == ObjectIdAttributeNumber &&
index->indclass.values[0] == OID_BTREE_OPS_OID &&
heap_attisnull(htup, Anum_pg_index_indpred))
oidIndex = index->indexrelid;
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
systable_endscan(indscan);
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
heap_close(indrel, AccessShareLock);
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
/* Now save a copy of the completed list in the relcache entry. */
First phase of memory management rewrite (see backend/utils/mmgr/README for details). It doesn't really do that much yet, since there are no short-term memory contexts in the executor, but the infrastructure is in place and long-term contexts are handled reasonably. A few long- standing bugs have been fixed, such as 'VACUUM; anything' in a single query string crashing. Also, out-of-memory is now considered a recoverable ERROR, not FATAL. Eliminate a large amount of crufty, now-dead code in and around memory management. Fix problem with holding off SIGTRAP, SIGSEGV, etc in postmaster and backend startup.
2000-06-28 05:33:33 +02:00
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
relation->rd_indexlist = list_copy(result);
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
relation->rd_oidindex = oidIndex;
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
relation->rd_indexvalid = 1;
Fix performance problems with pg_index lookups (see, for example, discussion of 5/19/00). pg_index is now searched for indexes of a relation using an indexscan. Moreover, this is done once and cached in the relcache entry for the relation, in the form of a list of OIDs for the indexes. This list is used by the parser and executor to drive lookups in the pg_index syscache when they want to know the properties of the indexes. Net result: index information will be fully cached for repetitive operations such as inserts.
2000-06-17 23:49:04 +02:00
MemoryContextSwitchTo(oldcxt);
return result;
}
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
/*
* insert_ordered_oid
* Insert a new Oid into a sorted list of Oids, preserving ordering
*
* Building the ordered list this way is O(N^2), but with a pretty small
* constant, so for the number of entries we expect it will probably be
* faster than trying to apply qsort(). Most tables don't have very many
* indexes...
*/
static List *
insert_ordered_oid(List *list, Oid datum)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
ListCell *prev;
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
/* Does the datum belong at the front? */
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
if (list == NIL || datum < linitial_oid(list))
return lcons_oid(datum, list);
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
/* No, so find the entry it belongs after */
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
prev = list_head(list);
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
for (;;)
{
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
ListCell *curr = lnext(prev);
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
if (curr == NULL || datum < lfirst_oid(curr))
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
break; /* it belongs after 'prev', before 'curr' */
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
prev = curr;
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
}
Reimplement the linked list data structure used throughout the backend. In the past, we used a 'Lispy' linked list implementation: a "list" was merely a pointer to the head node of the list. The problem with that design is that it makes lappend() and length() linear time. This patch fixes that problem (and others) by maintaining a count of the list length and a pointer to the tail node along with each head node pointer. A "list" is now a pointer to a structure containing some meta-data about the list; the head and tail pointers in that structure refer to ListCell structures that maintain the actual linked list of nodes. The function names of the list API have also been changed to, I hope, be more logically consistent. By default, the old function names are still available; they will be disabled-by-default once the rest of the tree has been updated to use the new API names.
2004-05-26 06:41:50 +02:00
/* Insert datum into list after 'prev' */
lappend_cell_oid(list, prev, datum);
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
return list;
}
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
/*
* RelationSetIndexList -- externally force the index list contents
*
* This is used to temporarily override what we think the set of valid
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
* indexes is (including the presence or absence of an OID index).
* The forcing will be valid only until transaction commit or abort.
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*
* This should only be applied to nailed relations, because in a non-nailed
* relation the hacked index list could be lost at any time due to SI
* messages. In practice it is only used on pg_class (see REINDEX).
*
* It is up to the caller to make sure the given list is correctly ordered.
Fix corner-case bug introduced with HOT: if REINDEX TABLE pg_class (or a REINDEX DATABASE including same) is done before a session has done any other update on pg_class, the pg_class relcache entry was left with an incorrect setting of rd_indexattr, because the indexed-attributes set would be first demanded at a time when we'd forced a partial list of indexes into the pg_class entry, and it would remain cached after that. This could result in incorrect decisions about HOT-update safety later in the same session. In practice, since only pg_class_relname_nsp_index would be missed out, only ALTER TABLE RENAME and ALTER TABLE SET SCHEMA could trigger a problem. Per report and test case from Ondrej Jirman.
2008-08-10 21:02:33 +02:00
*
* We deliberately do not change rd_indexattr here: even when operating
* with a temporary partial index list, HOT-update decisions must be made
* correctly with respect to the full index set. It is up to the caller
* to ensure that a correct rd_indexattr set has been cached before first
* calling RelationSetIndexList; else a subsequent inquiry might cause a
* wrong rd_indexattr set to get computed and cached.
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
*/
void
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
RelationSetIndexList(Relation relation, List *indexIds, Oid oidIndex)
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
{
MemoryContext oldcxt;
Fix relcache to account properly for subtransaction status of 'new' relcache entries. Also, change TransactionIdIsCurrentTransactionId() so that if consulted during transaction abort, it will not say that the aborted xact is still current. (It would be better to ensure that it's never called at all during abort, but I'm not sure we can easily guarantee that.) In combination, these fix a crash we have seen occasionally during parallel regression tests of 8.0.
2004-08-28 22:31:44 +02:00
Assert(relation->rd_isnailed);
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
/* Copy the list into the cache context (could fail for lack of mem) */
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
indexIds = list_copy(indexIds);
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
MemoryContextSwitchTo(oldcxt);
/* Okay to replace old list */
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
list_free(relation->rd_indexlist);
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
relation->rd_indexlist = indexIds;
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
relation->rd_oidindex = oidIndex;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
relation->rd_indexvalid = 2; /* mark list as forced */
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
/* must flag that we have a forced index list */
Teach CLUSTER to skip writing WAL if not needed (ie, not using archiving) --- Simon. Also, code review and cleanup for the previous COPY-no-WAL patches --- Tom.
2007-03-29 02:15:39 +02:00
need_eoxact_work = true;
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
}
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
/*
* RelationGetOidIndex -- get the pg_class OID of the relation's OID index
*
* Returns InvalidOid if there is no such index.
*/
Oid
RelationGetOidIndex(Relation relation)
{
List *ilist;
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* If relation doesn't have OIDs at all, caller is probably confused. (We
* could just silently return InvalidOid, but it seems better to throw an
* assertion.)
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
*/
Assert(relation->rd_rel->relhasoids);
if (relation->rd_indexvalid == 0)
{
/* RelationGetIndexList does the heavy lifting. */
ilist = RelationGetIndexList(relation);
list_free(ilist);
Assert(relation->rd_indexvalid != 0);
}
return relation->rd_oidindex;
}
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/*
* RelationGetIndexExpressions -- get the index expressions for an index
*
* We cache the result of transforming pg_index.indexprs into a node tree.
* If the rel is not an index or has no expressional columns, we return NIL.
* Otherwise, the returned tree is copied into the caller's memory context.
* (We don't want to return a pointer to the relcache copy, since it could
* disappear due to relcache invalidation.)
*/
List *
RelationGetIndexExpressions(Relation relation)
{
List *result;
Datum exprsDatum;
bool isnull;
char *exprsString;
MemoryContext oldcxt;
/* Quick exit if we already computed the result. */
if (relation->rd_indexprs)
return (List *) copyObject(relation->rd_indexprs);
/* Quick exit if there is nothing to do. */
if (relation->rd_indextuple == NULL ||
heap_attisnull(relation->rd_indextuple, Anum_pg_index_indexprs))
return NIL;
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We build the tree we intend to return in the caller's context. After
* successfully completing the work, we copy it into the relcache entry.
* This avoids problems if we get some sort of error partway through.
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
*/
exprsDatum = heap_getattr(relation->rd_indextuple,
Anum_pg_index_indexprs,
GetPgIndexDescriptor(),
&isnull);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
Assert(!isnull);
Simplify and standardize conversions between TEXT datums and ordinary C strings. This patch introduces four support functions cstring_to_text, cstring_to_text_with_len, text_to_cstring, and text_to_cstring_buffer, and two macros CStringGetTextDatum and TextDatumGetCString. A number of existing macros that provided variants on these themes were removed. Most of the places that need to make such conversions now require just one function or macro call, in place of the multiple notational layers that used to be needed. There are no longer any direct calls of textout or textin, and we got most of the places that were using handmade conversions via memcpy (there may be a few still lurking, though). This commit doesn't make any serious effort to eliminate transient memory leaks caused by detoasting toasted text objects before they reach text_to_cstring. We changed PG_GETARG_TEXT_P to PG_GETARG_TEXT_PP in a few places where it was easy, but much more could be done. Brendan Jurd and Tom Lane
2008-03-25 23:42:46 +01:00
exprsString = TextDatumGetCString(exprsDatum);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
result = (List *) stringToNode(exprsString);
pfree(exprsString);
/*
Rethink the order of expression preprocessing: eval_const_expressions really ought to run before canonicalize_qual, because it can now produce forms that canonicalize_qual knows how to improve (eg, NOT clauses). Also, because eval_const_expressions already knows about flattening nested ANDs and ORs into N-argument form, the initial flatten_andors pass in canonicalize_qual is now completely redundant and can be removed. This doesn't save a whole lot of code, but the time and palloc traffic eliminated is a useful gain on large expression trees.
2005-03-28 02:58:26 +02:00
* Run the expressions through eval_const_expressions. This is not just an
* optimization, but is necessary, because the planner will be comparing
* them to similarly-processed qual clauses, and may fail to detect valid
* matches without this. We don't bother with canonicalize_qual, however.
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
*/
Fix an oversight I made in a cleanup patch over a year ago: eval_const_expressions needs to be passed the PlannerInfo ("root") structure, because in some cases we want it to substitute values for Param nodes. (So "constant" is not so constant as all that ...) This mistake partially disabled optimization of unnamed extended-Query statements in 8.3: in particular the LIKE-to-indexscan optimization would never be applied if the LIKE pattern was passed as a parameter, and constraint exclusion depending on a parameter value didn't work either.
2008-04-01 02:48:33 +02:00
result = (List *) eval_const_expressions(NULL, (Node *) result);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
Tweak planner so that index expressions and predicates are matched to queries without regard to whether coercions are stated explicitly or implicitly. Per suggestion from Stephan Szabo.
2004-03-15 00:41:27 +01:00
/*
* Also mark any coercion format fields as "don't care", so that the
* planner can match to both explicit and implicit coercions.
*/
set_coercionform_dontcare((Node *) result);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/* May as well fix opfuncids too */
fix_opfuncids((Node *) result);
/* Now save a copy of the completed tree in the relcache entry. */
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
oldcxt = MemoryContextSwitchTo(relation->rd_indexcxt);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
relation->rd_indexprs = (List *) copyObject(result);
MemoryContextSwitchTo(oldcxt);
return result;
}
/*
* RelationGetIndexPredicate -- get the index predicate for an index
*
Clean up the usage of canonicalize_qual(): in particular, be consistent about whether it is applied before or after eval_const_expressions(). I believe there were some corner cases where the system would fail to recognize that a partial index is applicable because of the previous inconsistency. Store normal rather than 'implicit AND' representations of constraints and index predicates in the catalogs. initdb forced due to representation change of constraints/predicates.
2003-12-28 22:57:37 +01:00
* We cache the result of transforming pg_index.indpred into an implicit-AND
* node tree (suitable for ExecQual).
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
* If the rel is not an index or has no predicate, we return NIL.
* Otherwise, the returned tree is copied into the caller's memory context.
* (We don't want to return a pointer to the relcache copy, since it could
* disappear due to relcache invalidation.)
*/
List *
RelationGetIndexPredicate(Relation relation)
{
List *result;
Datum predDatum;
bool isnull;
char *predString;
MemoryContext oldcxt;
/* Quick exit if we already computed the result. */
if (relation->rd_indpred)
return (List *) copyObject(relation->rd_indpred);
/* Quick exit if there is nothing to do. */
if (relation->rd_indextuple == NULL ||
heap_attisnull(relation->rd_indextuple, Anum_pg_index_indpred))
return NIL;
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We build the tree we intend to return in the caller's context. After
* successfully completing the work, we copy it into the relcache entry.
* This avoids problems if we get some sort of error partway through.
Convert oidvector and int2vector into variable-length arrays. This change saves a great deal of space in pg_proc and its primary index, and it eliminates the former requirement that INDEX_MAX_KEYS and FUNC_MAX_ARGS have the same value. INDEX_MAX_KEYS is still embedded in the on-disk representation (because it affects index tuple header size), but FUNC_MAX_ARGS is not. I believe it would now be possible to increase FUNC_MAX_ARGS at little cost, but haven't experimented yet. There are still a lot of vestigial references to FUNC_MAX_ARGS, which I will clean up in a separate pass. However, getting rid of it altogether would require changing the FunctionCallInfoData struct, and I'm not sure I want to buy into that.
2005-03-29 02:17:27 +02:00
*/
predDatum = heap_getattr(relation->rd_indextuple,
Anum_pg_index_indpred,
GetPgIndexDescriptor(),
&isnull);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
Assert(!isnull);
Simplify and standardize conversions between TEXT datums and ordinary C strings. This patch introduces four support functions cstring_to_text, cstring_to_text_with_len, text_to_cstring, and text_to_cstring_buffer, and two macros CStringGetTextDatum and TextDatumGetCString. A number of existing macros that provided variants on these themes were removed. Most of the places that need to make such conversions now require just one function or macro call, in place of the multiple notational layers that used to be needed. There are no longer any direct calls of textout or textin, and we got most of the places that were using handmade conversions via memcpy (there may be a few still lurking, though). This commit doesn't make any serious effort to eliminate transient memory leaks caused by detoasting toasted text objects before they reach text_to_cstring. We changed PG_GETARG_TEXT_P to PG_GETARG_TEXT_PP in a few places where it was easy, but much more could be done. Brendan Jurd and Tom Lane
2008-03-25 23:42:46 +01:00
predString = TextDatumGetCString(predDatum);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
result = (List *) stringToNode(predString);
pfree(predString);
/*
Rethink the order of expression preprocessing: eval_const_expressions really ought to run before canonicalize_qual, because it can now produce forms that canonicalize_qual knows how to improve (eg, NOT clauses). Also, because eval_const_expressions already knows about flattening nested ANDs and ORs into N-argument form, the initial flatten_andors pass in canonicalize_qual is now completely redundant and can be removed. This doesn't save a whole lot of code, but the time and palloc traffic eliminated is a useful gain on large expression trees.
2005-03-28 02:58:26 +02:00
* Run the expression through const-simplification and canonicalization.
* This is not just an optimization, but is necessary, because the planner
* will be comparing it to similarly-processed qual clauses, and may fail
* to detect valid matches without this. This must match the processing
* done to qual clauses in preprocess_expression()! (We can skip the
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* stuff involving subqueries, however, since we don't allow any in index
* predicates.)
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
*/
Fix an oversight I made in a cleanup patch over a year ago: eval_const_expressions needs to be passed the PlannerInfo ("root") structure, because in some cases we want it to substitute values for Param nodes. (So "constant" is not so constant as all that ...) This mistake partially disabled optimization of unnamed extended-Query statements in 8.3: in particular the LIKE-to-indexscan optimization would never be applied if the LIKE pattern was passed as a parameter, and constraint exclusion depending on a parameter value didn't work either.
2008-04-01 02:48:33 +02:00
result = (List *) eval_const_expressions(NULL, (Node *) result);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
Rethink the order of expression preprocessing: eval_const_expressions really ought to run before canonicalize_qual, because it can now produce forms that canonicalize_qual knows how to improve (eg, NOT clauses). Also, because eval_const_expressions already knows about flattening nested ANDs and ORs into N-argument form, the initial flatten_andors pass in canonicalize_qual is now completely redundant and can be removed. This doesn't save a whole lot of code, but the time and palloc traffic eliminated is a useful gain on large expression trees.
2005-03-28 02:58:26 +02:00
result = (List *) canonicalize_qual((Expr *) result);
Tweak planner so that index expressions and predicates are matched to queries without regard to whether coercions are stated explicitly or implicitly. Per suggestion from Stephan Szabo.
2004-03-15 00:41:27 +01:00
/*
* Also mark any coercion format fields as "don't care", so that the
* planner can match to both explicit and implicit coercions.
*/
set_coercionform_dontcare((Node *) result);
Clean up the usage of canonicalize_qual(): in particular, be consistent about whether it is applied before or after eval_const_expressions(). I believe there were some corner cases where the system would fail to recognize that a partial index is applicable because of the previous inconsistency. Store normal rather than 'implicit AND' representations of constraints and index predicates in the catalogs. initdb forced due to representation change of constraints/predicates.
2003-12-28 22:57:37 +01:00
/* Also convert to implicit-AND format */
result = make_ands_implicit((Expr *) result);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/* May as well fix opfuncids too */
fix_opfuncids((Node *) result);
/* Now save a copy of the completed tree in the relcache entry. */
Fix relcache reload mechanism to be more robust in the face of errors occurring during a reload, such as query-cancel. Instead of zeroing out an existing relcache entry and rebuilding it in place, build a new relcache entry, then swap its contents with the old one, then free the new entry. This avoids problems with code believing that a previously obtained pointer to a cache entry must still reference a valid entry, as seen in recent failures on buildfarm member jaguar. (jaguar is using CLOBBER_CACHE_ALWAYS which raises the probability of failure substantially, but the problem could occur in the field without that.) The previous design was okay when it was made, but subtransactions and the ResourceOwner mechanism make it unsafe now. Also, make more use of the already existing rd_isvalid flag, so that we remember that the entry requires rebuilding even if the first attempt fails. Back-patch as far as 8.2. Prior versions have enough issues around relcache reload anyway (due to inadequate locking) that fixing this one doesn't seem worthwhile.
2010-01-12 19:12:18 +01:00
oldcxt = MemoryContextSwitchTo(relation->rd_indexcxt);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
relation->rd_indpred = (List *) copyObject(result);
MemoryContextSwitchTo(oldcxt);
return result;
}
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
/*
* RelationGetIndexAttrBitmap -- get a bitmap of index attribute numbers
*
* The result has a bit set for each attribute used anywhere in the index
* definitions of all the indexes on this relation. (This includes not only
* simple index keys, but attributes used in expressions and partial-index
* predicates.)
*
* Attribute numbers are offset by FirstLowInvalidHeapAttributeNumber so that
* we can include system attributes (e.g., OID) in the bitmap representation.
*
* The returned result is palloc'd in the caller's memory context and should
* be bms_free'd when not needed anymore.
*/
Bitmapset *
RelationGetIndexAttrBitmap(Relation relation)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
Bitmapset *indexattrs;
List *indexoidlist;
ListCell *l;
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
MemoryContext oldcxt;
/* Quick exit if we already computed the result. */
if (relation->rd_indexattr != NULL)
return bms_copy(relation->rd_indexattr);
/* Fast path if definitely no indexes */
if (!RelationGetForm(relation)->relhasindex)
return NULL;
/*
* Get cached list of index OIDs
*/
indexoidlist = RelationGetIndexList(relation);
/* Fall out if no indexes (but relhasindex was set) */
if (indexoidlist == NIL)
return NULL;
/*
* For each index, add referenced attributes to indexattrs.
*/
indexattrs = NULL;
foreach(l, indexoidlist)
{
Oid indexOid = lfirst_oid(l);
Relation indexDesc;
IndexInfo *indexInfo;
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int i;
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
indexDesc = index_open(indexOid, AccessShareLock);
/* Extract index key information from the index's pg_index row */
indexInfo = BuildIndexInfo(indexDesc);
/* Collect simple attribute references */
for (i = 0; i < indexInfo->ii_NumIndexAttrs; i++)
{
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
int attrnum = indexInfo->ii_KeyAttrNumbers[i];
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
if (attrnum != 0)
indexattrs = bms_add_member(indexattrs,
pgindent run for 8.3.
2007-11-15 22:14:46 +01:00
attrnum - FirstLowInvalidHeapAttributeNumber);
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
}
/* Collect all attributes used in expressions, too */
pull_varattnos((Node *) indexInfo->ii_Expressions, &indexattrs);
/* Collect all attributes in the index predicate, too */
pull_varattnos((Node *) indexInfo->ii_Predicate, &indexattrs);
index_close(indexDesc, AccessShareLock);
}
list_free(indexoidlist);
/* Now save a copy of the bitmap in the relcache entry. */
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
relation->rd_indexattr = bms_copy(indexattrs);
MemoryContextSwitchTo(oldcxt);
/* We return our original working copy for caller to play with */
return indexattrs;
}
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
/*
* RelationGetExclusionInfo -- get info about index's exclusion constraint
*
* This should be called only for an index that is known to have an
* associated exclusion constraint. It returns arrays (palloc'd in caller's
* context) of the exclusion operator OIDs, their underlying functions'
* OIDs, and their strategy numbers in the index's opclasses. We cache
* all this information since it requires a fair amount of work to get.
*/
void
RelationGetExclusionInfo(Relation indexRelation,
Oid **operators,
Oid **procs,
uint16 **strategies)
{
int ncols = indexRelation->rd_rel->relnatts;
Oid *ops;
Oid *funcs;
uint16 *strats;
Relation conrel;
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
SysScanDesc conscan;
ScanKeyData skey[1];
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
HeapTuple htup;
bool found;
MemoryContext oldcxt;
int i;
/* Allocate result space in caller context */
*operators = ops = (Oid *) palloc(sizeof(Oid) * ncols);
*procs = funcs = (Oid *) palloc(sizeof(Oid) * ncols);
*strategies = strats = (uint16 *) palloc(sizeof(uint16) * ncols);
/* Quick exit if we have the data cached already */
if (indexRelation->rd_exclstrats != NULL)
{
memcpy(ops, indexRelation->rd_exclops, sizeof(Oid) * ncols);
memcpy(funcs, indexRelation->rd_exclprocs, sizeof(Oid) * ncols);
memcpy(strats, indexRelation->rd_exclstrats, sizeof(uint16) * ncols);
return;
}
/*
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* Search pg_constraint for the constraint associated with the index. To
* make this not too painfully slow, we use the index on conrelid; that
* will hold the parent relation's OID not the index's own OID.
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
*/
ScanKeyInit(&skey[0],
Anum_pg_constraint_conrelid,
BTEqualStrategyNumber, F_OIDEQ,
ObjectIdGetDatum(indexRelation->rd_index->indrelid));
conrel = heap_open(ConstraintRelationId, AccessShareLock);
conscan = systable_beginscan(conrel, ConstraintRelidIndexId, true,
SnapshotNow, 1, skey);
found = false;
while (HeapTupleIsValid(htup = systable_getnext(conscan)))
{
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
Form_pg_constraint conform = (Form_pg_constraint) GETSTRUCT(htup);
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
Datum val;
bool isnull;
ArrayType *arr;
int nelem;
/* We want the exclusion constraint owning the index */
if (conform->contype != CONSTRAINT_EXCLUSION ||
conform->conindid != RelationGetRelid(indexRelation))
continue;
/* There should be only one */
if (found)
elog(ERROR, "unexpected exclusion constraint record found for rel %s",
RelationGetRelationName(indexRelation));
found = true;
/* Extract the operator OIDS from conexclop */
val = fastgetattr(htup,
Anum_pg_constraint_conexclop,
conrel->rd_att, &isnull);
if (isnull)
elog(ERROR, "null conexclop for rel %s",
RelationGetRelationName(indexRelation));
arr = DatumGetArrayTypeP(val); /* ensure not toasted */
nelem = ARR_DIMS(arr)[0];
if (ARR_NDIM(arr) != 1 ||
nelem != ncols ||
ARR_HASNULL(arr) ||
ARR_ELEMTYPE(arr) != OIDOID)
elog(ERROR, "conexclop is not a 1-D Oid array");
memcpy(ops, ARR_DATA_PTR(arr), sizeof(Oid) * ncols);
}
systable_endscan(conscan);
heap_close(conrel, AccessShareLock);
if (!found)
elog(ERROR, "exclusion constraint record missing for rel %s",
RelationGetRelationName(indexRelation));
/* We need the func OIDs and strategy numbers too */
for (i = 0; i < ncols; i++)
{
funcs[i] = get_opcode(ops[i]);
strats[i] = get_op_opfamily_strategy(ops[i],
indexRelation->rd_opfamily[i]);
/* shouldn't fail, since it was checked at index creation */
if (strats[i] == InvalidStrategy)
elog(ERROR, "could not find strategy for operator %u in family %u",
ops[i], indexRelation->rd_opfamily[i]);
}
/* Save a copy of the results in the relcache entry. */
oldcxt = MemoryContextSwitchTo(indexRelation->rd_indexcxt);
indexRelation->rd_exclops = (Oid *) palloc(sizeof(Oid) * ncols);
indexRelation->rd_exclprocs = (Oid *) palloc(sizeof(Oid) * ncols);
indexRelation->rd_exclstrats = (uint16 *) palloc(sizeof(uint16) * ncols);
memcpy(indexRelation->rd_exclops, ops, sizeof(Oid) * ncols);
memcpy(indexRelation->rd_exclprocs, funcs, sizeof(Oid) * ncols);
memcpy(indexRelation->rd_exclstrats, strats, sizeof(uint16) * ncols);
MemoryContextSwitchTo(oldcxt);
}
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* load_relcache_init_file, write_relcache_init_file
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* In late 1992, we started regularly having databases with more than
* a thousand classes in them. With this number of classes, it became
* critical to do indexed lookups on the system catalogs.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* Bootstrapping these lookups is very hard. We want to be able to
* use an index on pg_attribute, for example, but in order to do so,
* we must have read pg_attribute for the attributes in the index,
* which implies that we need to use the index.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* In order to get around the problem, we do the following:
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* + When the database system is initialized (at initdb time), we
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* don't use indexes. We do sequential scans.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* + When the backend is started up in normal mode, we load an image
* of the appropriate relation descriptors, in internal format,
* from an initialization file in the data/base/... directory.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* + If the initialization file isn't there, then we create the
Arrange for VACUUM to delete the init file that relcache.c uses to save a little bit of backend startup time. This way, the first backend started after a VACUUM will rebuild the init file with up-to-date statistics for the critical system indexes.
1999-05-01 21:09:46 +02:00
* relation descriptors using sequential scans and write 'em to
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
* the initialization file for use by subsequent backends.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*
Stamp HEAD as 9.0devel, and update various places that were referring to 8.5 (hope I got 'em all). Per discussion, this release will be 9.0 not 8.5.
2010-02-17 05:19:41 +01:00
* As of Postgres 9.0, there is one local initialization file in each
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* database, plus one shared initialization file for shared catalogs.
*
* We could dispense with the initialization files and just build the
Arrange for VACUUM to delete the init file that relcache.c uses to save a little bit of backend startup time. This way, the first backend started after a VACUUM will rebuild the init file with up-to-date statistics for the critical system indexes.
1999-05-01 21:09:46 +02:00
* critical reldescs the hard way on every backend startup, but that
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* slows down backend startup noticeably.
*
* We can in fact go further, and save more relcache entries than
* just the ones that are absolutely critical; this allows us to speed
* up backend startup by not having to build such entries the hard way.
* Presently, all the catalog and index entries that are referred to
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* by catcaches are stored in the initialization files.
Arrange for VACUUM to delete the init file that relcache.c uses to save a little bit of backend startup time. This way, the first backend started after a VACUUM will rebuild the init file with up-to-date statistics for the critical system indexes.
1999-05-01 21:09:46 +02:00
*
Fix obsolete comment.
2002-08-04 20:12:15 +02:00
* The same mechanism that detects when catcache and relcache entries
* need to be invalidated (due to catalog updates) also arranges to
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* unlink the initialization files when the contents may be out of date.
* The files will then be rebuilt during the next backend startup.
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* load_relcache_init_file -- attempt to load cache from the shared
* or local cache init file
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* If successful, return TRUE and set criticalRelcachesBuilt or
* criticalSharedRelcachesBuilt to true.
Further minor simplification of relcache startup: don't need a static needNewCacheFile flag anymore, it can just be local in RelationCacheInitializePhase2.
2006-05-06 17:51:07 +02:00
* If not successful, return FALSE.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
* NOTE: we assume we are already switched into CacheMemoryContext.
*/
static bool
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
load_relcache_init_file(bool shared)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
FILE *fp;
char initfilename[MAXPGPATH];
Relation *rels;
int relno,
num_rels,
max_rels,
nailed_rels,
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
nailed_indexes,
magic;
Another PGINDENT run that changes variable indenting and case label indenting. Also static variable indenting.
1997-09-08 04:41:22 +02:00
int i;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (shared)
snprintf(initfilename, sizeof(initfilename), "global/%s",
RELCACHE_INIT_FILENAME);
else
snprintf(initfilename, sizeof(initfilename), "%s/%s",
DatabasePath, RELCACHE_INIT_FILENAME);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
fp = AllocateFile(initfilename, PG_BINARY_R);
if (fp == NULL)
return false;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* Read the index relcache entries from the file. Note we will not enter
* any of them into the cache if the read fails partway through; this
* helps to guard against broken init files.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
max_rels = 100;
rels = (Relation *) palloc(max_rels * sizeof(Relation));
num_rels = 0;
nailed_rels = nailed_indexes = 0;
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
/* check for correct magic number (compatible version) */
if (fread(&magic, 1, sizeof(magic), fp) != sizeof(magic))
goto read_failed;
if (magic != RELCACHE_INIT_FILEMAGIC)
goto read_failed;
pgindent run.
2002-09-04 22:31:48 +02:00
for (relno = 0;; relno++)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Size len;
size_t nread;
Relation rel;
Form_pg_class relform;
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
bool has_not_null;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* first read the relation descriptor length */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
nread = fread(&len, 1, sizeof(len), fp);
if (nread != sizeof(len))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
{
if (nread == 0)
break; /* end of file */
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
/* safety check for incompatible relcache layout */
if (len != sizeof(RelationData))
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* allocate another relcache header */
if (num_rels >= max_rels)
{
max_rels *= 2;
rels = (Relation *) repalloc(rels, max_rels * sizeof(Relation));
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
rel = rels[num_rels++] = (Relation) palloc(len);
New file naming. Database OID is used as "tablespace" id and relation OID is used as file node on creation but may be changed later if required. Regression Tests Approved (c) -:)))
2000-10-16 16:52:28 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* then, read the Relation structure */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(rel, 1, len, fp) != len)
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* next read the relation tuple form */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
relform = (Form_pg_class) palloc(len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(relform, 1, len, fp) != len)
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
rel->rd_rel = relform;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* initialize attribute tuple forms */
Code review for HeapTupleHeader changes. Add version number to page headers (overlaying low byte of page size) and add HEAP_HASOID bit to t_infomask, per earlier discussion. Simplify scheme for overlaying fields in tuple header (no need for cmax to live in more than one place). Don't try to clear infomask status bits in tqual.c --- not safe to do it there. Don't try to force output table of a SELECT INTO to have OIDs, either. Get rid of unnecessarily complex three-state scheme for TupleDesc.tdhasoids, which has already caused one recent failure. Improve documentation.
2002-09-02 03:05:06 +02:00
rel->rd_att = CreateTemplateTupleDesc(relform->relnatts,
relform->relhasoids);
Fix problems with cached tuple descriptors disappearing while still in use by creating a reference-count mechanism, similar to what we did a long time ago for catcache entries. The back branches have an ugly solution involving lots of extra copies, but this way is more efficient. Reference counting is only applied to tupdescs that are actually in caches --- there seems no need to use it for tupdescs that are generated in the executor, since they'll go away during plan shutdown by virtue of being in the per-query memory context. Neil Conway and Tom Lane
2006-06-16 20:42:24 +02:00
rel->rd_att->tdrefcount = 1; /* mark as refcounted */
Replace TupleTableSlot convention for whole-row variables and function results with tuples as ordinary varlena Datums. This commit does not in itself do much for us, except eliminate the horrid memory leak associated with evaluation of whole-row variables. However, it lays the groundwork for allowing composite types as table columns, and perhaps some other useful features as well. Per my proposal of a few days ago.
2004-04-01 23:28:47 +02:00
rel->rd_att->tdtypeid = relform->reltype;
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
rel->rd_att->tdtypmod = -1; /* unnecessary, but... */
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* next read all the attribute tuple form data entries */
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
has_not_null = false;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
for (i = 0; i < relform->relnatts; i++)
{
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
Support column-level privileges, as required by SQL standard. Stephen Frost, with help from KaiGai Kohei and others
2009-01-22 21:16:10 +01:00
if (len != ATTRIBUTE_FIXED_PART_SIZE)
Adjust creation/destruction of TupleDesc data structure to reduce the number of palloc calls. This has a salutory impact on plpgsql operations with record variables (which create and destroy tupdescs constantly) and probably helps a bit in some other cases too.
2005-03-07 05:42:17 +01:00
goto read_failed;
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(rel->rd_att->attrs[i], 1, len, fp) != len)
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
goto read_failed;
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
has_not_null |= rel->rd_att->attrs[i]->attnotnull;
}
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
/* next read the access method specific field */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
goto read_failed;
if (len > 0)
{
rel->rd_options = palloc(len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(rel->rd_options, 1, len, fp) != len)
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
goto read_failed;
Replace direct assignments to VARATT_SIZEP(x) with SET_VARSIZE(x, len). Get rid of VARATT_SIZE and VARATT_DATA, which were simply redundant with VARSIZE and VARDATA, and as a consequence almost no code was using the longer names. Rename the length fields of struct varlena and various derived structures to catch anyplace that was accessing them directly; and clean up various places so caught. In itself this patch doesn't change any behavior at all, but it is necessary infrastructure if we hope to play any games with the representation of varlena headers. Greg Stark and Tom Lane
2007-02-28 00:48:10 +01:00
if (len != VARSIZE(rel->rd_options))
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
goto read_failed; /* sanity check */
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
}
else
{
rel->rd_options = NULL;
}
Add relcache support for system catalogs with attnotnull columns. The code was not making TupleConstr structs for such catalogs in several places; with the consequence that the not-null constraint wasn't actually enforced. With this change, INSERT INTO pg_proc VALUES('sdf'); generates a 'Fail to add null value' error instead of a core dump.
2002-11-15 18:18:49 +01:00
/* mark not-null status */
if (has_not_null)
{
TupleConstr *constr = (TupleConstr *) palloc0(sizeof(TupleConstr));
constr->has_not_null = true;
rel->rd_att->constr = constr;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* If it's an index, there's more to do */
if (rel->rd_rel->relkind == RELKIND_INDEX)
{
Form_pg_am am;
MemoryContext indexcxt;
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Oid *opfamily;
Oid *opcintype;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Oid *operator;
RegProcedure *support;
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
int nsupport;
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
int16 *indoption;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* Count nailed indexes to ensure we have 'em all */
if (rel->rd_isnailed)
nailed_indexes++;
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/* next, read the pg_index tuple */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
rel->rd_indextuple = (HeapTuple) palloc(len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(rel->rd_indextuple, 1, len, fp) != len)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/* Fix up internal pointers in the tuple -- see heap_copytuple */
rel->rd_indextuple->t_data = (HeapTupleHeader) ((char *) rel->rd_indextuple + HEAPTUPLESIZE);
rel->rd_index = (Form_pg_index) GETSTRUCT(rel->rd_indextuple);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* next, read the access method tuple form */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
am = (Form_pg_am) palloc(len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(am, 1, len, fp) != len)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
rel->rd_am = am;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* prepare index info context --- parameters should match
* RelationInitIndexAccessInfo
*/
indexcxt = AllocSetContextCreate(CacheMemoryContext,
RelationGetRelationName(rel),
Tweak default memory context allocation policy so that a context is not given any malloc block until something is first allocated in it; but thereafter, MemoryContextReset won't release that first malloc block. This preserves the quick-reset property of the original policy, without forcing 8K to be allocated to every context whether any of it is ever used or not. Also, remove some more no-longer-needed explicit freeing during ExecEndPlan.
2002-12-15 22:01:34 +01:00
ALLOCSET_SMALL_MINSIZE,
ALLOCSET_SMALL_INITSIZE,
ALLOCSET_SMALL_MAXSIZE);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
rel->rd_indexcxt = indexcxt;
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
/* next, read the vector of opfamily OIDs */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
goto read_failed;
opfamily = (Oid *) MemoryContextAlloc(indexcxt, len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(opfamily, 1, len, fp) != len)
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
goto read_failed;
rel->rd_opfamily = opfamily;
/* next, read the vector of opcintype OIDs */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
goto read_failed;
opcintype = (Oid *) MemoryContextAlloc(indexcxt, len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(opcintype, 1, len, fp) != len)
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
goto read_failed;
rel->rd_opcintype = opcintype;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* next, read the vector of operator OIDs */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
operator = (Oid *) MemoryContextAlloc(indexcxt, len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(operator, 1, len, fp) != len)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
rel->rd_operator = operator;
Rearrange fmgr.c and relcache so that it's possible to keep FmgrInfo lookup info in the relcache for index access method support functions. This makes a huge difference for dynamically loaded support functions, and should save a few cycles even for built-in ones. Also tweak dfmgr.c so that load_external_function is called only once, not twice, when doing fmgr_info for a dynamically loaded function. All per performance gripe from Teodor Sigaev, 5-Oct-01.
2001-10-07 01:21:45 +02:00
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
/* next, read the vector of support procedures */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
support = (RegProcedure *) MemoryContextAlloc(indexcxt, len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(support, 1, len, fp) != len)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
goto read_failed;
rel->rd_support = support;
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
/* finally, read the vector of indoption values */
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(&len, 1, sizeof(len), fp) != sizeof(len))
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
goto read_failed;
indoption = (int16 *) MemoryContextAlloc(indexcxt, len);
Remove some useless assignments of the result of fread(). Quiets warnings from clang static checker, and makes the code more readable anyway IMO.
2009-08-30 19:18:52 +02:00
if (fread(indoption, 1, len, fp) != len)
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
goto read_failed;
rel->rd_indoption = indoption;
Arrange to cache fmgr lookup information for an index's access method routines in the index's relcache entry, instead of doing a fresh fmgr_info on every index access. We were already doing this for the index's opclass support functions; not sure why we didn't think to do it for the AM functions too. This supersedes the former method of caching (only) amgettuple in indexscan scan descriptors; it's an improvement because the function lookup can be amortized across multiple statements instead of being repeated for each statement. Even though lookup for builtin functions is pretty cheap, this seems to drop a percent or two off some simple benchmarks.
2005-05-28 01:31:21 +02:00
/* set up zeroed fmgr-info vectors */
rel->rd_aminfo = (RelationAmInfo *)
MemoryContextAllocZero(indexcxt, sizeof(RelationAmInfo));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
nsupport = relform->relnatts * am->amsupport;
rel->rd_supportinfo = (FmgrInfo *)
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
MemoryContextAllocZero(indexcxt, nsupport * sizeof(FmgrInfo));
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
else
{
/* Count nailed rels to ensure we have 'em all */
if (rel->rd_isnailed)
nailed_rels++;
Assert(rel->rd_index == NULL);
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
Assert(rel->rd_indextuple == NULL);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Assert(rel->rd_am == NULL);
Assert(rel->rd_indexcxt == NULL);
Arrange to cache fmgr lookup information for an index's access method routines in the index's relcache entry, instead of doing a fresh fmgr_info on every index access. We were already doing this for the index's opclass support functions; not sure why we didn't think to do it for the AM functions too. This supersedes the former method of caching (only) amgettuple in indexscan scan descriptors; it's an improvement because the function lookup can be amortized across multiple statements instead of being repeated for each statement. Even though lookup for builtin functions is pretty cheap, this seems to drop a percent or two off some simple benchmarks.
2005-05-28 01:31:21 +02:00
Assert(rel->rd_aminfo == NULL);
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
Assert(rel->rd_opfamily == NULL);
Assert(rel->rd_opcintype == NULL);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Assert(rel->rd_operator == NULL);
Assert(rel->rd_support == NULL);
Assert(rel->rd_supportinfo == NULL);
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
Assert(rel->rd_indoption == NULL);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
/*
* Rules and triggers are not saved (mainly because the internal
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* format is complex and subject to change). They must be rebuilt if
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* needed by RelationCacheInitializePhase3. This is not expected to
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* be a big performance hit since few system catalogs have such. Ditto
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
* for index expressions, predicates, and exclusion info.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
rel->rd_rules = NULL;
rel->rd_rulescxt = NULL;
rel->trigdesc = NULL;
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
rel->rd_indexprs = NIL;
rel->rd_indpred = NIL;
Add exclusion constraints, which generalize the concept of uniqueness to support any indexable commutative operator, not just equality. Two rows violate the exclusion constraint if "row1.col OP row2.col" is TRUE for each of the columns in the constraint. Jeff Davis, reviewed by Robert Haas
2009-12-07 06:22:23 +01:00
rel->rd_exclops = NULL;
rel->rd_exclprocs = NULL;
rel->rd_exclstrats = NULL;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* Reset transient-state fields in the relcache entry
*/
Restructure smgr API as per recent proposal. smgr no longer depends on the relcache, and so the notion of 'blind write' is gone. This should improve efficiency in bgwriter and background checkpoint processes. Internal restructuring in md.c to remove the not-very-useful array of MdfdVec objects --- might as well just use pointers. Also remove the long-dead 'persistent main memory' storage manager (mm.c), since it seems quite unlikely to ever get resurrected.
2004-02-10 02:55:27 +01:00
rel->rd_smgr = NULL;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
if (rel->rd_isnailed)
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
rel->rd_refcnt = 1;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
else
Invent ResourceOwner mechanism as per my recent proposal, and use it to keep track of portal-related resources separately from transaction-related resources. This allows cursors to work in a somewhat sane fashion with nested transactions. For now, cursor behavior is non-subtransactional, that is a cursor's state does not roll back if you abort a subtransaction that fetched from the cursor. We might want to change that later.
2004-07-17 05:32:14 +02:00
rel->rd_refcnt = 0;
Repair some REINDEX problems per recent discussions. The relcache is now able to cope with assigning new relfilenode values to nailed-in-cache indexes, so they can be reindexed using the fully crash-safe method. This leaves only shared system indexes as special cases. Remove the 'index deactivation' code, since it provides no useful protection in the shared- index case. Require reindexing of shared indexes to be done in standalone mode, but remove other restrictions on REINDEX. -P (IgnoreSystemIndexes) now prevents using indexes for lookups, but does not disable index updates. It is therefore safe to allow from PGOPTIONS. Upshot: reindexing system catalogs can be done without a standalone backend for all cases except shared catalogs.
2003-09-24 20:54:02 +02:00
rel->rd_indexvalid = 0;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
rel->rd_indexlist = NIL;
HOT updates. When we update a tuple without changing any of its indexed columns, and the new version can be stored on the same heap page, we no longer generate extra index entries for the new version. Instead, index searches follow the HOT-chain links to ensure they find the correct tuple version. In addition, this patch introduces the ability to "prune" dead tuples on a per-page basis, without having to do a complete VACUUM pass to recover space. VACUUM is still needed to clean up dead index entries, however. Pavan Deolasee, with help from a bunch of other people.
2007-09-20 19:56:33 +02:00
rel->rd_indexattr = NULL;
Solve the problem of OID collisions by probing for duplicate OIDs whenever we generate a new OID. This prevents occasional duplicate-OID errors that can otherwise occur once the OID counter has wrapped around. Duplicate relfilenode values are also checked for when creating new physical files. Per my recent proposal.
2005-08-12 03:36:05 +02:00
rel->rd_oidindex = InvalidOid;
Avoid scanning the relcache during AtEOSubXact_RelationCache when there is nothing to do, which is most of the time. This is another simple improvement to cut subtransaction entry/exit overhead.
2004-11-20 21:19:52 +01:00
rel->rd_createSubid = InvalidSubTransactionId;
Prevent WAL logging when COPY is done in the same transation that created it. Simon Riggs
2007-01-25 03:17:26 +01:00
rel->rd_newRelfilenodeSubid = InvalidSubTransactionId;
Arrange to cache btree metapage data in the relcache entry for the index, thereby saving a visit to the metapage in most index searches/updates. This wouldn't actually save any I/O (since in the old regime the metapage generally stayed in cache anyway), but it does provide a useful decrease in bufmgr traffic in high-contention scenarios. Per my recent proposal.
2006-04-26 00:46:05 +02:00
rel->rd_amcache = NULL;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
MemSet(&rel->pgstat_info, 0, sizeof(rel->pgstat_info));
Mega-commit to make heap_open/heap_openr/heap_close take an additional argument specifying the kind of lock to acquire/release (or 'NoLock' to do no lock processing). Ensure that all relations are locked with some appropriate lock level before being examined --- this ensures that relevant shared-inval messages have been processed and should prevent problems caused by concurrent VACUUM. Fix several bugs having to do with mismatched increment/decrement of relation ref count and mismatched heap_open/close (which amounts to the same thing). A bogus ref count on a relation doesn't matter much *unless* a SI Inval message happens to arrive at the wrong time, which is probably why we got away with this sloppiness for so long. Repair missing grab of AccessExclusiveLock in DROP TABLE, ALTER/RENAME TABLE, etc, as noted by Hiroshi. Recommend 'make clean all' after pulling this update; I modified the Relation struct layout slightly. Will post further discussion to pghackers list shortly.
1999-09-18 21:08:25 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
* Recompute lock and physical addressing info. This is needed in
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* case the pg_internal.init file was copied from some other database
* by CREATE DATABASE.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
RelationInitLockInfo(rel);
Tablespaces. Alternate database locations are dead, long live tablespaces. There are various things left to do: contrib dbsize and oid2name modules need work, and so does the documentation. Also someone should think about COMMENT ON TABLESPACE and maybe RENAME TABLESPACE. Also initlocation is dead, it just doesn't know it yet. Gavin Sherry and Tom Lane.
2004-06-18 08:14:31 +02:00
RelationInitPhysicalAddr(rel);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* We reached the end of the init file without apparent problem. Did we
* get the right number of nailed items? (This is a useful crosscheck in
* case the set of critical rels or indexes changes.)
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (shared)
{
if (nailed_rels != NUM_CRITICAL_SHARED_RELS ||
nailed_indexes != NUM_CRITICAL_SHARED_INDEXES)
goto read_failed;
}
else
{
if (nailed_rels != NUM_CRITICAL_LOCAL_RELS ||
nailed_indexes != NUM_CRITICAL_LOCAL_INDEXES)
goto read_failed;
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* OK, all appears well.
*
* Now insert all the new relcache entries into the cache.
*/
for (relno = 0; relno < num_rels; relno++)
{
RelationCacheInsert(rels[relno]);
/* also make a list of their OIDs, for RelationIdIsInInitFile */
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (!shared)
initFileRelationIds = lcons_oid(RelationGetRelid(rels[relno]),
initFileRelationIds);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
pfree(rels);
FreeFile(fp);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (shared)
criticalSharedRelcachesBuilt = true;
else
criticalRelcachesBuilt = true;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
return true;
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* init file is broken, so do it the hard way. We don't bother trying to
* free the clutter we just allocated; it's not in the relcache so it
* won't hurt.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
read_failed:
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
pfree(rels);
FreeFile(fp);
return false;
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* Write out a new initialization file with the current contents
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* of the relcache (either shared rels or local rels, as indicated).
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
Make functions static where possible, enclose unused functions in #ifdef NOT_USED.
1997-08-19 23:40:56 +02:00
static void
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
write_relcache_init_file(bool shared)
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
{
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
FILE *fp;
Fix low-probability bug in relcache startup: write_irels wrote the pg_internal.init file in-place, which meant that if another backend started at about the same time, it might read the incomplete file. init_irels tries to guard against that, but I have now seen a crash due to reading bad data from a partly-written file. (This may indicate a kernel bug on my platform? Not sure.) Anyway, clearly the safest course is to write the new pg_internal.init file under a unique temporary filename, and rename it into place only after it's all written.
2000-03-31 21:39:22 +02:00
char tempfilename[MAXPGPATH];
char finalfilename[MAXPGPATH];
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
int magic;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
HASH_SEQ_STATUS status;
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
RelIdCacheEnt *idhentry;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
MemoryContext oldcxt;
int i;
Fix low-probability bug in relcache startup: write_irels wrote the pg_internal.init file in-place, which meant that if another backend started at about the same time, it might read the incomplete file. init_irels tries to guard against that, but I have now seen a crash due to reading bad data from a partly-written file. (This may indicate a kernel bug on my platform? Not sure.) Anyway, clearly the safest course is to write the new pg_internal.init file under a unique temporary filename, and rename it into place only after it's all written.
2000-03-31 21:39:22 +02:00
/*
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
* We must write a temporary file and rename it into place. Otherwise,
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* another backend starting at about the same time might crash trying to
* read the partially-complete file.
Fix low-probability bug in relcache startup: write_irels wrote the pg_internal.init file in-place, which meant that if another backend started at about the same time, it might read the incomplete file. init_irels tries to guard against that, but I have now seen a crash due to reading bad data from a partly-written file. (This may indicate a kernel bug on my platform? Not sure.) Anyway, clearly the safest course is to write the new pg_internal.init file under a unique temporary filename, and rename it into place only after it's all written.
2000-03-31 21:39:22 +02:00
*/
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (shared)
{
snprintf(tempfilename, sizeof(tempfilename), "global/%s.%d",
RELCACHE_INIT_FILENAME, MyProcPid);
snprintf(finalfilename, sizeof(finalfilename), "global/%s",
RELCACHE_INIT_FILENAME);
}
else
{
snprintf(tempfilename, sizeof(tempfilename), "%s/%s.%d",
DatabasePath, RELCACHE_INIT_FILENAME, MyProcPid);
snprintf(finalfilename, sizeof(finalfilename), "%s/%s",
DatabasePath, RELCACHE_INIT_FILENAME);
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
unlink(tempfilename); /* in case it exists w/wrong permissions */
fp = AllocateFile(tempfilename, PG_BINARY_W);
if (fp == NULL)
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
{
/*
* We used to consider this a fatal error, but we might as well
* continue with backend startup ...
*/
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
ereport(WARNING,
(errcode_for_file_access(),
Message editing: remove gratuitous variations in message wording, standardize terms, add some clarifications, fix some untranslatable attempts at dynamic message building.
2003-09-25 08:58:07 +02:00
errmsg("could not create relation-cache initialization file \"%s\": %m",
Error message editing in backend/utils (except /adt).
2003-07-25 22:18:01 +02:00
tempfilename),
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
errdetail("Continuing anyway, but there's something wrong.")));
Modify index-opening code to guarantee that the indexes of a relation are opened in a consistent order by different backends (I ordered them by index OID because that's easy, but any other consistent order would do as well). This avoids potential deadlock for index types that we acquire exclusive locks on ... ie, rtree.
2000-06-20 01:40:48 +02:00
return;
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
/*
Pgindent run for 8.0.
2004-08-29 07:07:03 +02:00
* Write a magic number to serve as a file version identifier. We can
Add operator strategy and comparison-value datatype fields to ScanKey. Remove the 'strategy map' code, which was a large amount of mechanism that no longer had any use except reverse-mapping from procedure OID to strategy number. Passing the strategy number to the index AM in the first place is simpler and faster. This is a preliminary step in planned support for cross-datatype index operations. I'm committing it now since the ScanKeyEntryInitialize() API change touches quite a lot of files, and I want to commit those changes before the tree drifts under me.
2003-11-09 22:30:38 +01:00
* change the magic number whenever the relcache layout changes.
*/
magic = RELCACHE_INIT_FILEMAGIC;
if (fwrite(&magic, 1, sizeof(magic), fp) != sizeof(magic))
elog(FATAL, "could not write init file");
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* Write all the appropriate reldescs (in no particular order).
Implement reindex command
2000-02-18 10:30:20 +01:00
*/
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
hash_seq_init(&status, RelationIdCache);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
while ((idhentry = (RelIdCacheEnt *) hash_seq_search(&status)) != NULL)
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
{
pg_class has a relnamespace column. You can create and access tables in schemas other than the system namespace; however, there's no search path yet, and not all operations work yet on tables outside the system namespace.
2002-03-26 20:17:02 +01:00
Relation rel = idhentry->reldesc;
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Form_pg_class relform = rel->rd_rel;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/* ignore if not correct group */
if (relform->relisshared != shared)
continue;
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
/* first write the relcache entry proper */
write_item(rel, sizeof(RelationData), fp);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* next write the relation tuple form */
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
write_item(relform, CLASS_TUPLE_SIZE, fp);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
/* next, do all the attribute tuple form data entries */
for (i = 0; i < relform->relnatts; i++)
{
Support column-level privileges, as required by SQL standard. Stephen Frost, with help from KaiGai Kohei and others
2009-01-22 21:16:10 +01:00
write_item(rel->rd_att->attrs[i], ATTRIBUTE_FIXED_PART_SIZE, fp);
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
}
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
/* next, do the access method specific field */
write_item(rel->rd_options,
Replace direct assignments to VARATT_SIZEP(x) with SET_VARSIZE(x, len). Get rid of VARATT_SIZE and VARATT_DATA, which were simply redundant with VARSIZE and VARDATA, and as a consequence almost no code was using the longer names. Rename the length fields of struct varlena and various derived structures to catch anyplace that was accessing them directly; and clean up various places so caught. In itself this patch doesn't change any behavior at all, but it is necessary infrastructure if we hope to play any games with the representation of varlena headers. Greg Stark and Tom Lane
2007-02-28 00:48:10 +01:00
(rel->rd_options ? VARSIZE(rel->rd_options) : 0),
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
fp);
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* If it's an index, there's more to do */
if (rel->rd_rel->relkind == RELKIND_INDEX)
{
Form_pg_am am = rel->rd_am;
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Replace functional-index facility with expressional indexes. Any column of an index can now be a computed expression instead of a simple variable. Restrictions on expressions are the same as for predicates (only immutable functions, no sub-selects). This fixes problems recently introduced with inlining SQL functions, because the inlining transformation is applied to both expression trees so the planner can still match them up. Along the way, improve efficiency of handling index predicates (both predicates and index expressions are now cached by the relcache) and fix 7.3 oversight that didn't record dependencies of predicate expressions.
2003-05-28 18:04:02 +02:00
/* write the pg_index tuple */
/* we assume this was created by heap_copytuple! */
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
write_item(rel->rd_indextuple,
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
HEAPTUPLESIZE + rel->rd_indextuple->t_len,
fp);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* next, write the access method tuple form */
Add FILLFACTOR to CREATE INDEX. ITAGAKI Takahiro
2006-07-02 04:23:23 +02:00
write_item(am, sizeof(FormData_pg_am), fp);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Restructure operator classes to allow improved handling of cross-data-type cases. Operator classes now exist within "operator families". While most families are equivalent to a single class, related classes can be grouped into one family to represent the fact that they are semantically compatible. Cross-type operators are now naturally adjunct parts of a family, without having to wedge them into a particular opclass as we had done originally. This commit restructures the catalogs and cleans up enough of the fallout so that everything still works at least as well as before, but most of the work needed to actually improve the planner's behavior will come later. Also, there are not yet CREATE/DROP/ALTER OPERATOR FAMILY commands; the only way to create a new family right now is to allow CREATE OPERATOR CLASS to make one by default. I owe some more documentation work, too. But that can all be done in smaller pieces once this infrastructure is in place.
2006-12-23 01:43:13 +01:00
/* next, write the vector of opfamily OIDs */
write_item(rel->rd_opfamily,
relform->relnatts * sizeof(Oid),
fp);
/* next, write the vector of opcintype OIDs */
write_item(rel->rd_opcintype,
relform->relnatts * sizeof(Oid),
fp);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* next, write the vector of operator OIDs */
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
write_item(rel->rd_operator,
relform->relnatts * (am->amstrategies * sizeof(Oid)),
fp);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
/* next, write the vector of support procedures */
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
write_item(rel->rd_support,
pgindent run for 8.2.
2006-10-04 02:30:14 +02:00
relform->relnatts * (am->amsupport * sizeof(RegProcedure)),
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
fp);
Support ORDER BY ... NULLS FIRST/LAST, and add ASC/DESC/NULLS FIRST/NULLS LAST per-column options for btree indexes. The planner's support for this is still pretty rudimentary; it does not yet know how to plan mergejoins with nondefault ordering options. The documentation is pretty rudimentary, too. I'll work on improving that stuff later. Note incompatible change from prior behavior: ORDER BY ... USING will now be rejected if the operator is not a less-than or greater-than member of some btree opclass. This prevents less-than-sane behavior if an operator that doesn't actually define a proper sort ordering is selected.
2007-01-09 03:14:16 +01:00
/* finally, write the vector of indoption values */
write_item(rel->rd_indoption,
relform->relnatts * sizeof(int16),
fp);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/* also make a list of their OIDs, for RelationIdIsInInitFile */
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
if (!shared)
{
oldcxt = MemoryContextSwitchTo(CacheMemoryContext);
initFileRelationIds = lcons_oid(RelationGetRelid(rel),
initFileRelationIds);
MemoryContextSwitchTo(oldcxt);
}
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Massive commit to run PGINDENT on all *.c and *.h files.
1997-09-07 07:04:48 +02:00
Ensure that close() and fclose() are checked for errors, at least in cases involving writes. Per recent discussion about the possibility of close-time failures on some filesystems. There is a TODO item for this, too.
2004-01-26 23:35:32 +01:00
if (FreeFile(fp))
elog(FATAL, "could not write init file");
Fix low-probability bug in relcache startup: write_irels wrote the pg_internal.init file in-place, which meant that if another backend started at about the same time, it might read the incomplete file. init_irels tries to guard against that, but I have now seen a crash due to reading bad data from a partly-written file. (This may indicate a kernel bug on my platform? Not sure.) Anyway, clearly the safest course is to write the new pg_internal.init file under a unique temporary filename, and rename it into place only after it's all written.
2000-03-31 21:39:22 +02:00
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* Now we have to check whether the data we've so painstakingly
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* accumulated is already obsolete due to someone else's just-committed
* catalog changes. If so, we just delete the temp file and leave it to
* the next backend to try again. (Our own relcache entries will be
* updated by SI message processing, but we can't be sure whether what we
* wrote out was up-to-date.)
Fix init_irels to close the pg_internal.init file before returning. This saves one open file descriptor per backend, and avoids an annoying NOTICE on Cygwin (which has trouble deleting open files). Bug appears to date back to original coding of init_irels, circa 1992.
2002-01-16 18:34:42 +01:00
*
pgindent run.
2002-09-04 22:31:48 +02:00
* This mustn't run concurrently with RelationCacheInitFileInvalidate, so
* grab a serialization lock for the duration.
Ye-old pgindent run. Same 4-space tabs.
2000-04-12 19:17:23 +02:00
*/
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
LWLockAcquire(RelCacheInitLock, LW_EXCLUSIVE);
/* Make sure we have seen all incoming SI messages */
AcceptInvalidationMessages();
/*
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* If we have received any SI relcache invals since backend start, assume
* we may have written out-of-date data.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
if (relcacheInvalsReceived == 0L)
If we fail to rename pg_internal.init into place, delete the useless temporary file. This seems to be a known failure mode under Cygwin, so we might as well expend the extra line of code to be tidy.
2002-01-15 23:33:20 +01:00
{
/*
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* OK, rename the temp file to its final name, deleting any
* previously-existing init file.
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
*
Re-run pgindent, fixing a problem where comment lines after a blank comment line where output as too long, and update typedefs for /lib directory. Also fix case where identifiers were used as variable names in the backend, but as typedefs in ecpg (favor the backend for indenting). Backpatch to 8.1.X.
2005-11-22 19:17:34 +01:00
* Note: a failure here is possible under Cygwin, if some other
* backend is holding open an unlinked-but-not-yet-gone init file. So
* treat this as a noncritical failure; just remove the useless temp
* file on failure.
If we fail to rename pg_internal.init into place, delete the useless temporary file. This seems to be a known failure mode under Cygwin, so we might as well expend the extra line of code to be tidy.
2002-01-15 23:33:20 +01:00
*/
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
if (rename(tempfilename, finalfilename) < 0)
unlink(tempfilename);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
else
{
/* Delete the already-obsolete temp file */
If we fail to rename pg_internal.init into place, delete the useless temporary file. This seems to be a known failure mode under Cygwin, so we might as well expend the extra line of code to be tidy.
2002-01-15 23:33:20 +01:00
unlink(tempfilename);
}
Upgrade formrdesc() so that it can correctly initialize the tupledesc (rd_att) field of a nailed-in-cache relcache entry. This fixes the bug reported by Alvaro 8-Dec-2004; I believe it probably also explains Grant Finnemore's report of 10-Sep-2004. In an unrelated change in the same file, put back 7.4's response to failure to rename() the relcache init file, ie, unlink the useless temp file. I did not put back the warning message, since there might actually be some reason not to have that.
2004-12-12 06:07:50 +01:00
LWLockRelease(RelCacheInitLock);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
Code review for FILLFACTOR patch. Change WITH grammar as per earlier discussion (including making def_arg allow reserved words), add missed opt_definition for UNIQUE case. Put the reloptions support code in a less random place (I chose to make a new file access/common/reloptions.c). Eliminate header inclusion creep. Make the index options functions safely user-callable (seems like client apps might like to be able to test validity of options before trying to make an index). Reduce overhead for normal case with no options by allowing rd_options to be NULL. Fix some unmaintainably klugy code, including getting rid of Natts_pg_class_fixed at long last. Some stylistic cleanup too, and pay attention to keeping comments in sync with code. Documentation still needs work, though I did fix the omissions in catalogs.sgml and indexam.sgml.
2006-07-04 00:45:41 +02:00
/* write a chunk of data preceded by its length */
static void
write_item(const void *data, Size len, FILE *fp)
{
if (fwrite(&len, 1, sizeof(len), fp) != sizeof(len))
elog(FATAL, "could not write init file");
if (fwrite(data, 1, len, fp) != len)
elog(FATAL, "could not write init file");
}
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
/*
* Detect whether a given relation (identified by OID) is one of the ones
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* we store in the local relcache init file.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
* Note that we effectively assume that all backends running in a database
* would choose to store the same set of relations in the init file;
* otherwise there are cases where we'd fail to detect the need for an init
* file invalidation. This does not seem likely to be a problem in practice.
*/
bool
RelationIdIsInInitFile(Oid relationId)
{
Use the new List API function names throughout the backend, and disable the list compatibility API by default. While doing this, I decided to keep the llast() macro around and introduce llast_int() and llast_oid() variants.
2004-05-31 01:40:41 +02:00
return list_member_oid(initFileRelationIds, relationId);
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
}
/*
* Invalidate (remove) the init file during commit of a transaction that
* changed one or more of the relation cache entries that are kept in the
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* local init file.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*
* We actually need to remove the init file twice: once just before sending
* the SI messages that include relcache inval for such relations, and once
* just after sending them. The unlink before ensures that a backend that's
* currently starting cannot read the now-obsolete init file and then miss
* the SI messages that will force it to update its relcache entries. (This
Simplify lock manager data structures by making a clear separation between the data defining the semantics of a lock method (ie, conflict resolution table and ancillary data, which is all constant) and the hash tables storing the current state. The only thing we give up by this is the ability to use separate hashtables for different lock methods, but there is no need for that anyway. Put some extra fields into the LockMethod definition structs to clean up some other uglinesses, like hard-wired tests for DEFAULT_LOCKMETHOD and USER_LOCKMETHOD. This commit doesn't do anything about the performance issues we were discussing, but it clears away some of the underbrush that's in the way of fixing that.
2005-12-09 02:22:04 +01:00
* works because the backend startup sequence gets into the PGPROC array before
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
* trying to load the init file.) The unlink after is to synchronize with a
* backend that may currently be trying to write an init file based on data
* that we've just rendered invalid. Such a backend will see the SI messages,
* but we can't leave the init file sitting around to fool later backends.
*
* Ignore any failure to unlink the file, since it might not be there if
* no backend has been started since the last removal.
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
*
* Notice this deals only with the local init file, not the shared init file.
* The reason is that there can never be a "significant" change to the
* relcache entry of a shared relation; the most that could happen is
* updates of noncritical fields such as relpages/reltuples. So, while
* it's worth updating the shared init file from time to time, it can never
* be invalid enough to make it necessary to remove it.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
void
RelationCacheInitFileInvalidate(bool beforeSend)
{
char initfilename[MAXPGPATH];
snprintf(initfilename, sizeof(initfilename), "%s/%s",
DatabasePath, RELCACHE_INIT_FILENAME);
if (beforeSend)
{
/* no interlock needed here */
unlink(initfilename);
}
else
{
/*
pgindent run.
2002-09-04 22:31:48 +02:00
* We need to interlock this against write_relcache_init_file, to
* guard against possibility that someone renames a new-but-
Standard pgindent run for 8.1.
2005-10-15 04:49:52 +02:00
* already-obsolete init file into place just after we unlink. With
* the interlock, it's certain that write_relcache_init_file will
* notice our SI inval message before renaming into place, or else
* that we will execute second and successfully unlink the file.
A bunch of changes aimed at reducing backend startup time... Improve 'pg_internal.init' relcache entry preload mechanism so that it is safe to use for all system catalogs, and arrange to preload a realistic set of system-catalog entries instead of only the three nailed-in-cache indexes that were formerly loaded this way. Fix mechanism for deleting out-of-date pg_internal.init files: this must be synchronized with transaction commit, not just done at random times within transactions. Drive it off relcache invalidation mechanism so that no special-case tests are needed. Cache additional information in relcache entries for indexes (their pg_index tuples and index-operator OIDs) to eliminate repeated lookups. Also cache index opclass info at the per-opclass level to avoid repeated lookups during relcache load. Generalize 'systable scan' utilities originally developed by Hiroshi, move them into genam.c, use in a number of places where there was formerly ugly code for choosing either heap or index scan. In particular this allows simplification of the logic that prevents infinite recursion between syscache and relcache during startup: we can easily switch to heapscans in relcache.c when and where needed to avoid recursion, so IndexScanOK becomes simpler and does not need any expensive initialization. Eliminate useless opening of a heapscan data structure while doing an indexscan (this saves an mdnblocks call and thus at least one kernel call).
2002-02-19 21:11:20 +01:00
*/
LWLockAcquire(RelCacheInitLock, LW_EXCLUSIVE);
unlink(initfilename);
LWLockRelease(RelCacheInitLock);
}
Postgres95 1.01 Distribution - Virgin Sources
1996-07-09 08:22:35 +02:00
}
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
/*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* Remove the init files during postmaster startup.
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
*
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* We used to keep the init files across restarts, but that is unsafe in PITR
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
* scenarios, and even in simple crash-recovery cases there are windows for
pgindent run for 9.0
2010-02-26 03:01:40 +01:00
* the init files to become out-of-sync with the database. So now we just
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
* remove them during startup and expect the first backend launch to rebuild
* them. Of course, this has to happen in each database of the cluster.
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
*/
void
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
RelationCacheInitFileRemove(void)
{
const char *tblspcdir = "pg_tblspc";
DIR *dir;
struct dirent *de;
char path[MAXPGPATH];
/*
* We zap the shared cache file too. In theory it can't get out of sync
* enough to be a problem, but in data-corruption cases, who knows ...
*/
snprintf(path, sizeof(path), "global/%s",
RELCACHE_INIT_FILENAME);
unlink_initfile(path);
/* Scan everything in the default tablespace */
RelationCacheInitFileRemoveInDir("base");
/* Scan the tablespace link directory to find non-default tablespaces */
dir = AllocateDir(tblspcdir);
if (dir == NULL)
{
elog(LOG, "could not open tablespace link directory \"%s\": %m",
tblspcdir);
return;
}
while ((de = ReadDir(dir, tblspcdir)) != NULL)
{
if (strspn(de->d_name, "0123456789") == strlen(de->d_name))
{
/* Scan the tablespace dir for per-database dirs */
Please tablespace directories in their own subdirectory so pg_migrator can upgrade clusters without renaming the tablespace directories. New directory structure format is, e.g.: $PGDATA/pg_tblspc/20981/PG_8.5_201001061/719849/83292814
2010-01-12 03:42:52 +01:00
snprintf(path, sizeof(path), "%s/%s/%s",
tblspcdir, de->d_name, TABLESPACE_VERSION_DIRECTORY);
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
RelationCacheInitFileRemoveInDir(path);
}
}
FreeDir(dir);
}
/* Process one per-tablespace directory for RelationCacheInitFileRemove */
static void
RelationCacheInitFileRemoveInDir(const char *tblspcpath)
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
{
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
DIR *dir;
struct dirent *de;
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
char initfilename[MAXPGPATH];
Allow backends to start up without use of the flat-file copy of pg_database. To make this work in the base case, pg_database now has a nailed-in-cache relation descriptor that is initialized using hardwired knowledge in relcache.c. This means pg_database is added to the set of relations that need to have a Schema_pg_xxx macro maintained in pg_attribute.h. When this path is taken, we'll have to do a seqscan of pg_database to find the row we need. In the normal case, we are able to do an indexscan to find the database's row by name. This is made possible by storing a global relcache init file that describes only the shared catalogs and their indexes (and therefore is usable by all backends in any database). A new backend loads this cache file, finds its database OID after an indexscan on pg_database, and then loads the local relcache init file for that database. This change should effectively eliminate number of databases as a factor in backend startup time, even with large numbers of databases. However, the real reason for doing it is as a first step towards getting rid of the flat files altogether. There are still several other sub-projects to be tackled before that can happen.
2009-08-12 22:53:31 +02:00
/* Scan the tablespace directory to find per-database directories */
dir = AllocateDir(tblspcpath);
if (dir == NULL)
{
elog(LOG, "could not open tablespace directory \"%s\": %m",
tblspcpath);
return;
}
while ((de = ReadDir(dir, tblspcpath)) != NULL)
{
if (strspn(de->d_name, "0123456789") == strlen(de->d_name))
{
/* Try to remove the init file in each database */
snprintf(initfilename, sizeof(initfilename), "%s/%s/%s",
tblspcpath, de->d_name, RELCACHE_INIT_FILENAME);
unlink_initfile(initfilename);
}
}
FreeDir(dir);
}
static void
unlink_initfile(const char *initfilename)
{
if (unlink(initfilename) < 0)
{
/* It might not be there, but log any error other than ENOENT */
if (errno != ENOENT)
elog(LOG, "could not remove cache file \"%s\": %m", initfilename);
}
Fix recently-identified PITR recovery hazard: the base backup could contain stale relcache init files (pg_internal.init), and there is no mechanism for updating them during WAL replay. Easiest solution is just to delete the init files at conclusion of startup, and let the first backend started in each database take care of rebuilding the init file. Simon Riggs and Tom Lane. Back-patched to 8.1. Arguably this should be fixed in 8.0 too, but it would require significantly more code since 8.0 has no handy startup-time scan of pg_database to piggyback on. Manual solution of the problem is possible in 8.0 (just delete the pg_internal.init files before starting WAL replay), so that may be a sufficient answer.
2006-11-06 00:40:31 +01:00
}
Reference in New Issue Copy Permalink