The error message about columns in the primary key not including all of
the partition key was unclear; reword it.
Backpatch all the way to pg11, where it appeared.
Reported-by: Nagaraj Raj <nagaraj.sf@yahoo.com>
Discussion: https://postgr.es/m/64062533.78364.1601415362244@mail.yahoo.com
Previously, a conversion such as
to_date('-44-02-01','YYYY-MM-DD')
would result in '0045-02-01 BC', as the code attempted to interpret
the negative year as BC, but failed to apply the correction needed
for our internal handling of BC years. Fix the off-by-one problem.
Also, arrange for the combination of a negative year and an
explicit "BC" marker to cancel out and produce AD. This is how
the negative-century case works, so it seems sane to do likewise.
Continue to read "year 0000" as 1 BC. Oracle would throw an error,
but we've accepted that case for a long time so I'm hesitant to
change it in a back-patch.
Per bug #16419 from Saeed Hubaishan. Back-patch to all supported
branches.
Dar Alathar-Yemen and Tom Lane
Discussion: https://postgr.es/m/16419-d8d9db0a7553f01b@postgresql.org
Previously we threw an error. But make_date already allowed the case,
so it is inconsistent as well as unhelpful for make_timestamp not to.
Both functions continue to reject year zero.
Code and test fixes by Peter Eisentraut, doc changes by me
Discussion: https://postgr.es/m/13c0992c-f15a-a0ca-d839-91d3efd965d9@2ndquadrant.com
The SQL standard doesn't require jsonpath .datetime() method to support the
ISO 8601 format. But our to_json[b]() functions convert timestamps to text in
the ISO 8601 format in the sake of compatibility with javascript. So, we add
support of the ISO 8601 to the jsonpath .datetime() in the sake compatibility
with to_json[b]().
The standard mode of datetime parsing currently supports just template patterns
and separators in the format string. In order to implement ISO 8601, we have to
add support of the format string double quotes to the standard parsing mode.
Discussion: https://postgr.es/m/94321be0-cc96-1a81-b6df-796f437f7c66%40postgrespro.ru
Author: Nikita Glukhov, revised by me
Backpatch-through: 13
bffe1bd684 has introduced jsonpath .datetime() method, but default formats
for time and timestamp contain excess space between time and timezone. This
commit removes this excess space making behavior of .datetime() method
standard-compliant.
Discussion: https://postgr.es/m/94321be0-cc96-1a81-b6df-796f437f7c66%40postgrespro.ru
Author: Nikita Glukhov
Backpatch-through: 13
Previously the standby server didn't archive timeline history files
streamed from the primary even when archive_mode is set to "always",
while it archives the streamed WAL files. This could cause the PITR to
fail because there was no required timeline history file in the archive.
The cause of this issue was that walreceiver didn't mark those files as
ready for archiving.
This commit makes walreceiver mark those streamed timeline history
files as ready for archiving if archive_mode=always. Then the archiver
process archives the marked timeline history files.
Back-patch to all supported versions.
Reported-by: Grigory Smolkin
Author: Grigory Smolkin, Fujii Masao
Reviewed-by: David Zhang, Anastasia Lubennikova
Discussion: https://postgr.es/m/54b059d4-2b48-13a4-6f43-95a087c92367@postgrespro.ru
This addresses a couple of issues with the so-said subject:
- Report the correct parent relation with the index actually being
rebuilt or validated. Previously, the command status remained set to
the last index created for the progress of the index build and
validation, which would be incorrect when working on a table that has
more than one index.
- Use the correct phase when waiting before the drop of the old
indexes. Previously, this was reported with the same status as when
waiting before the old indexes are marked as dead.
Author: Matthias van de Meent, Michael Paquier
Discussion: https://postgr.es/m/CAEze2WhqFgcwe1_tv=sFYhLWV2AdpfukumotJ6JNcAOQs3jufg@mail.gmail.com
Backpatch-through: 12
We have a dozen or so places that need to iterate over all but the
first cell of a List. Prior to v13 this was typically written as
for_each_cell(lc, lnext(list_head(list)))
Commit 1cff1b95a changed these to
for_each_cell(lc, list, list_second_cell(list))
This patch introduces a new macro for_each_from() which expresses
the start point as a list index, allowing these to be written as
for_each_from(lc, list, 1)
This is marginally more efficient, since ForEachState.i can be
initialized directly instead of backing into it from a ListCell
address. It also seems clearer and less typo-prone.
Some of the remaining uses of for_each_cell() look like they could
profitably be changed to for_each_from(), but here I confined myself
to changing uses of list_second_cell().
Also, fix for_each_cell_setup() and for_both_cell_setup() to
const-ify their arguments; that's a simple oversight in 1cff1b95a.
Back-patch into v13, on the grounds that (1) the const-ification
is a minor bug fix, and (2) it's better for back-patching purposes
if we only have two ways to write these loops rather than three.
In HEAD, also remove list_third_cell() and list_fourth_cell(),
which were also introduced in 1cff1b95a, and are unused as of
cc99baa43. It seems unlikely that any third-party code would
have started to use them already; anyone who has can be directed
to list_nth_cell instead.
Discussion: https://postgr.es/m/CAApHDvpo1zj9KhEpU2cCRZfSM3Q6XGdhzuAS2v79PH7WJBkYVA@mail.gmail.com
transformPartitionBoundValue went out of its way to do the wrong
thing: there is no reason to complain about a non-matching COLLATE
clause in a partition boundary expression. We're coercing the
bound expression to the target column type as though by an
implicit assignment, and the rules for implicit assignment say
that collations can be implicitly converted.
What we *do* need to do, and the code is not doing, is apply
assign_expr_collations() to the bound expression. While this is
merely a definition disagreement, that is a bug that needs to be
back-patched, so I'll commit it separately.
Discussion: https://postgr.es/m/CAJV4CdrZ5mKuaEsRSbLf2URQ3h6iMtKD=hik8MaF5WwdmC9uZw@mail.gmail.com
SQL operations such as CURRENT_DATE, CURRENT_TIME, LOCALTIME, and
conversion of "now" in a datetime input string have to obtain the
transaction start timestamp ("now()") as a broken-down struct pg_tm.
This is a remarkably expensive conversion, and since now() does not
change intra-transaction, it doesn't really need to be done more than
once per transaction. Introducing a simple cache provides visible
speedups in queries that compute these values many times, for example
insertion of many rows that use a default value of CURRENT_DATE.
Peter Smith, with a bit of kibitzing by me
Discussion: https://postgr.es/m/CAHut+Pu89TWjq530V2gY5O6SWi=OEJMQ_VHMt8bdZB_9JFna5A@mail.gmail.com
When commit bd3daddaf introduced AlternativeSubPlans, I had some
ambitions towards allowing the choice of subplan to change during
execution. That has not happened, or even been thought about, in the
ensuing twelve years; so it seems like a failed experiment. So let's
rip that out and resolve the choice of subplan at the end of planning
(in setrefs.c) rather than during executor startup. This has a number
of positive benefits:
* Removal of a few hundred lines of executor code, since
AlternativeSubPlans need no longer be supported there.
* Removal of executor-startup overhead (particularly, initialization
of subplans that won't be used).
* Removal of incidental costs of having a larger plan tree, such as
tree-scanning and copying costs in the plancache; not to mention
setrefs.c's own costs of processing the discarded subplans.
* EXPLAIN no longer has to print a weird (and undocumented)
representation of an AlternativeSubPlan choice; it sees only the
subplan actually used. This should mean less confusion for users.
* Since setrefs.c knows which subexpression of a plan node it's
working on at any instant, it's possible to adjust the estimated
number of executions of the subplan based on that. For example,
we should usually estimate more executions of a qual expression
than a targetlist expression. The implementation used here is
pretty simplistic, because we don't want to expend a lot of cycles
on the issue; but it's better than ignoring the point entirely,
as the executor had to.
That last point might possibly result in shifting the choice
between hashed and non-hashed EXISTS subplans in a few cases,
but in general this patch isn't meant to change planner choices.
Since we're doing the resolution so late, it's really impossible
to change any plan choices outside the AlternativeSubPlan itself.
Patch by me; thanks to David Rowley for review.
Discussion: https://postgr.es/m/1992952.1592785225@sss.pgh.pa.us
This function leaked some memory while loading qual clauses for
an RLS policy. While ordinarily negligible, that could build up
in some repeated-reload cases, as reported by Konstantin Knizhnik.
We can improve matters by borrowing the coding long used in
RelationBuildRuleLock: build stringToNode's result directly in
the target context, and remember to explicitly pfree the
input string.
This patch by no means completely guarantees zero leaks within
this function, since we have no real guarantee that the catalog-
reading subroutines it calls don't leak anything. However,
practical tests suggest that this is enough to resolve the issue.
In any case, any remaining leaks are similar to those risked by
RelationBuildRuleLock and other relcache-loading subroutines.
If we need to fix them, we should adopt a more global approach
such as that used by the RECOVER_RELATION_BUILD_MEMORY hack.
While here, let's remove the need for an expensive PG_TRY block by
using MemoryContextSetParent to reparent an initially-short-lived
context for the RLS data.
Back-patch to all supported branches.
Discussion: https://postgr.es/m/21356c12-8917-8249-b35f-1c447231922b@postgrespro.ru
Commit 464824323e changed the logical replication protocol to allow the
streaming of in-progress transactions and used the new version of protocol
irrespective of the server version. Use the appropriate version of the
protocol based on the server version.
Reported-by: Ashutosh Sharma
Author: Dilip Kumar
Reviewed-by: Ashutosh Sharma and Amit Kapila
Discussion: https://postgr.es/m/CAE9k0P=9OpXcNrcU5Gsvd5MZ8GFpiN833vNHzX6Uc=8+h1ft1Q@mail.gmail.com
Previously, we called fsync() after writing out individual pg_xact,
pg_multixact and pg_commit_ts pages due to cache pressure, leading to
regular I/O stalls in user backends and recovery. Collapse requests for
the same file into a single system call as part of the next checkpoint,
as we already did for relation files, using the infrastructure developed
by commit 3eb77eba. This can cause a significant improvement to
recovery performance, especially when it's otherwise CPU-bound.
Hoist ProcessSyncRequests() up into CheckPointGuts() to make it clearer
that it applies to all the SLRU mini-buffer-pools as well as the main
buffer pool. Rearrange things so that data collected in CheckpointStats
includes SLRU activity.
Also remove the Shutdown{CLOG,CommitTS,SUBTRANS,MultiXact}() functions,
because they were redundant after the shutdown checkpoint that
immediately precedes them. (I'm not sure if they were ever needed, but
they aren't now.)
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us> (parts)
Tested-by: Jakub Wartak <Jakub.Wartak@tomtom.com>
Discussion: https://postgr.es/m/CA+hUKGLJ=84YT+NvhkEEDAuUtVHMfQ9i-N7k_o50JmQ6Rpj_OQ@mail.gmail.com
The previous coding was confused about whether head_timestamp was
intended to represent the timestamp for the newest bucket in the
mapping or the oldest timestamp for the oldest bucket in the mapping.
Decide that it's intended to be the oldest one, and repair
accordingly.
To do that, we need to do two things. First, when advancing to a
new bucket, don't categorically set head_timestamp to the new
timestamp. Do this only if we're blowing out the map completely
because a lot of time has passed since we last maintained it. If
we're replacing entries one by one, advance head_timestamp by
1 minute for each; if we're filling in unused entries, don't
advance head_timestamp at all.
Second, fix the computation of how many buckets we need to advance.
The previous formula would be correct if head_timestamp were the
timestamp for the new bucket, but we're now making all the code
agree that it's the timestamp for the oldest bucket, so adjust the
formula accordingly.
This is certainly a bug fix, but I don't feel good about
back-patching it without the introspection tools added by commit
aecf5ee2bb, and perhaps also some
actual tests. Since back-patching the introspection tools might
not attract sufficient support and since there are no automated
tests of these fixes yet, I'm just committing this to master for
now.
Patch by me, reviewed by Thomas Munro, Dilip Kumar, Hamid Akhtar.
Discussion: http://postgr.es/m/CA+TgmoY=aqf0zjTD+3dUWYkgMiNDegDLFjo+6ze=Wtpik+3XqA@mail.gmail.com
Existing code used various inconsistent ways to printf struct stat's
st_size member. The type of that is off_t, which is in most cases a
signed 64-bit integer, so use the long long int format for it.
This makes it possible for code outside snapmgr.c to examine the
contents of this data structure. This commit does not add any code
which actually does so; a subsequent commit will make that change.
Patch by me, reviewed by Thomas Munro, Dilip Kumar, Hamid Akhtar.
Discussion: http://postgr.es/m/CA+TgmoY=aqf0zjTD+3dUWYkgMiNDegDLFjo+6ze=Wtpik+3XqA@mail.gmail.com
Commit fbeb9da22, which added the tsearch_readline APIs, left
t_readline() in place as a compatibility measure. But that function
has been unused and deprecated for twelve years now, so that seems
like enough time to remove it. Doing so, and merging t_readline's
code into tsearch_readline, aids in making several useful
improvements:
* The hard-wired 4K limit on line length in tsearch data files is
removed, by using a StringInfo buffer instead of a fixed-size buffer.
* We can buy back the per-line palloc/pfree added by 3ea7e9550
in the common case where encoding conversion is not required.
* We no longer need a separate pg_verify_mbstr call, as that
functionality was folded into encoding conversion some time ago.
(We could have done some of this stuff while keeping t_readline as a
separate API, but there seems little point, since there's no reason
for anyone to still be using t_readline directly.)
Discussion: https://postgr.es/m/48A4FA71-524E-41B9-953A-FD04EF36E2E7@yesql.se
Harmonize behavior by moving reponsibility for fsyncing directories down
into slru.c. In 10 and later, only the multixact directories were
missed (see commit 1b02be21), and in older branches all SLRUs were
missed.
Back-patch to all supported releases.
Reviewed-by: Andres Freund <andres@anarazel.de>
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Discussion: https://postgr.es/m/CA%2BhUKGLtsTUOScnNoSMZ-2ZLv%2BwGh01J6kAo_DM8mTRq1sKdSQ%40mail.gmail.com
We failed to pass down the query string to check_new_partition_bound,
so that its attempts to provide error cursor positions were for naught;
one must have the query string to get parser_errposition to do anything.
Adjust its API to require a ParseState to be passed down.
Also, improve the logic inside check_new_partition_bound so that the
cursor points at the partition bound for the specific column causing
the issue, when one can be identified.
That part is also for naught if we can't determine the query position of
the column with the problem. Improve transformPartitionBoundValue so
that it makes sure that const-simplified partition expressions will be
properly labeled with positions. In passing, skip calling evaluate_expr
if the value is already a Const, which is surely the most common case.
Alexandra Wang, Ashwin Agrawal, Amit Langote; reviewed by Ashutosh Bapat
Discussion: https://postgr.es/m/CACiyaSopZoqssfMzgHk6fAkp01cL6vnqBdmTw2C5_KJaFR_aMg@mail.gmail.com
Discussion: https://postgr.es/m/CAJV4CdrZ5mKuaEsRSbLf2URQ3h6iMtKD=hik8MaF5WwdmC9uZw@mail.gmail.com
tsearch_readline() saves the string pointer it returns to the caller
for possible use in the associated error context callback. However,
the caller will usually pfree that string sometime before it next
calls tsearch_readline(), so that there is a window where an ereport
will try to print an already-freed string.
The built-in users of tsearch_readline() happen to all do that pfree
at the bottoms of their loops, so that the window is effectively
empty for them. However, this is not documented as a requirement,
and contrib/dict_xsyn doesn't do it like that, so it seems likely
that third-party dictionaries might have live bugs here.
The practical consequences of this seem pretty limited in any case,
since production builds wouldn't clobber the freed string immediately,
besides which you'd not expect syntax errors in dictionary files
being used in production. Still, it's clearly a bug waiting to bite
somebody.
Fix by pstrdup'ing the string to be saved for the error callback,
and then pfree'ing it next time through. It's been like this for
a long time, so back-patch to all supported branches.
Discussion: https://postgr.es/m/48A4FA71-524E-41B9-953A-FD04EF36E2E7@yesql.se
The existing message about "a column definition list is only allowed for
functions returning "record"" could be given in some cases where it was
fairly confusing; in particular, a function with multiple OUT parameters
*does* return record according to pg_proc. Break it down into a couple
more cases to deliver a more on-point complaint. Per complaint from
Bruce Momjian.
Discussion: https://postgr.es/m/798909.1600562993@sss.pgh.pa.us
This completes the project of making all our derived files be
pgindent-clean (or else explicitly excluded from indentation),
so that no surprises result when running pgindent in a built-out
development tree.
Discussion: https://postgr.es/m/79ed5348-be7a-b647-dd40-742207186a22@2ndquadrant.com
Since we're bypassing the buffer manager, we need to call
PageSetChecksumInplace() directly. As reported by Justin Pryzby.
In the passing, add RelationOpenSmgr() calls before all smgrwrite() and
smgrextend() calls. Tom added one before the first smgrextend() call in
commit c2bb287025, which seems to be enough, but let's play it safe and
do it before each one. That's how it's done in the similar code in
nbtsort.c, too.
Discussion: https://www.postgresql.org/message-id/20200920224446.GF30557@telsasoft.com
It's no longer necessary to assign explicit precedences to GENERATED,
NULL_P, PRESERVE, or STRIP_P.
Actually, we don't need to assign precedence to IDENT either; that was
really just there to govern the behavior of target_el's "a_expr IDENT"
production, which no longer ends with that terminal. However, it seems
like a good idea to continue to do so, because it provides a reference
point for a precedence level that we can assign to other unreserved
keywords that lack a natural precedence level.
Research by Peter Eisentraut and John Naylor; comment rewrite by me.
Discussion: https://postgr.es/m/38ca86db-42ab-9b48-2902-337a0d6b8311@2ndquadrant.com
Commit be0a6666 left behind a comment about the order of some tests that
didn't make sense without the expensive division, and in fact we might
as well change the order to one that fails more cheaply most of the time
as a micro-optimization. Also, remove the "+ 1" applied to max_bucket,
to drop an instruction and match the original behavior. Per review
from Tom Lane.
Discussion: https://postgr.es/m/VI1PR0701MB696044FC35013A96FECC7AC8F62D0%40VI1PR0701MB6960.eurprd07.prod.outlook.com
Since ancient times we have had support for a fill factor (maximum load
factor) to be set for a dynahash hash table, but:
1. It was an integer, whereas for in-memory hash tables interesting
load factor targets are probably somewhere near the 0.75-1.0 range.
2. It was implemented in a way that performed an expensive division
operation that regularly showed up in profiles.
3. We are not aware of anyone ever having used a non-default value.
Therefore, remove support, effectively fixing it at 1.
Author: Jakub Wartak <Jakub.Wartak@tomtom.com>
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Reviewed-by: Tomas Vondra <tomas.vondra@2ndquadrant.com>
Reviewed-by: Thomas Munro <thomas.munro@gmail.com>
Reviewed-by: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/VI1PR0701MB696044FC35013A96FECC7AC8F62D0%40VI1PR0701MB6960.eurprd07.prod.outlook.com
Up to now, if you tried to omit "AS" before a column label in a SELECT
list, it would only work if the column label was an IDENT, that is not
any known keyword. This is rather unfriendly considering that we have
so many keywords and are constantly growing more. In the wake of commit
1ed6b8956 it's possible to improve matters quite a bit.
We'd originally tried to make this work by having some of the existing
keyword categories be allowed without AS, but that didn't work too well,
because each category contains a few special cases that don't work
without AS. Instead, invent an entirely orthogonal keyword property
"can be bare column label", and mark all keywords that way for which
we don't get shift/reduce errors by doing so.
It turns out that of our 450 current keywords, all but 39 can be made
bare column labels, improving the situation by over 90%. This number
might move around a little depending on future grammar work, but it's
a pretty nice improvement.
Mark Dilger, based on work by myself and Robert Haas;
review by John Naylor
Discussion: https://postgr.es/m/38ca86db-42ab-9b48-2902-337a0d6b8311@2ndquadrant.com
After commits 85f6b49c2c and 3ba59ccc89, we can allow parallel inserts
which was earlier not possible as parallel group members won't conflict
for relation extension and page lock. In those commits, we forgot to
update comments at few places.
Author: Amit Kapila
Reviewed-by: Robert Haas and Dilip Kumar
Backpatch-through: 13
Discussion: https://postgr.es/m/CAFiTN-tMrQh5FFMPx5aWJ+1gi1H6JxktEhq5mDwCHgnEO5oBkA@mail.gmail.com
This feature has been a thorn in our sides for a long time, causing
many grammatical ambiguity problems. It doesn't seem worth the
pain to continue to support it, so remove it.
There are some follow-on improvements we can make in the grammar,
but this commit only removes the bare minimum number of productions,
plus assorted backend support code.
Note that pg_dump and psql continue to have full support, since
they may be used against older servers. However, pg_dump warns
about postfix operators. There is also a check in pg_upgrade.
Documentation-wise, I (tgl) largely removed the "left unary"
terminology in favor of saying "prefix operator", which is
a more standard and IMO less confusing term.
I included a catversion bump, although no initial catalog data
changes here, to mark the boundary at which oprkind = 'r'
stopped being valid in pg_operator.
Mark Dilger, based on work by myself and Robert Haas;
review by John Naylor
Discussion: https://postgr.es/m/38ca86db-42ab-9b48-2902-337a0d6b8311@2ndquadrant.com
For parallel btree scan to work for array of scan keys, it should reach
BTPARALLEL_DONE state once for every distinct combination of array keys.
This is required to ensure that the parallel workers don't try to seize
blocks at the same time for different scan keys. We missed to update this
state when we discovered that the scan keys can't be satisfied.
Author: James Hunter
Reviewed-by: Amit Kapila
Tested-by: Justin Pryzby
Backpatch-through: 10, where it was introduced
Discussion: https://postgr.es/m/4248CABC-25E3-4809-B4D0-128E1BAABC3C@amazon.com
In the particular case of GRANTED BY, this is specified in the SQL
standard. Since in PostgreSQL, CURRENT_ROLE is equivalent to
CURRENT_USER, and CURRENT_USER is already supported here, adding
CURRENT_ROLE is trivial. The other cases are PostgreSQL extensions,
but for the same reason it also makes sense there.
Reviewed-by: Vik Fearing <vik@postgresfriends.org>
Reviewed-by: Asif Rehman <asifr.rehman@gmail.com>
Reviewed-by: Alvaro Herrera <alvherre@2ndquadrant.com>
Discussion: https://www.postgresql.org/message-id/flat/f2feac44-b4c5-f38f-3699-2851d6a76dc9%402ndquadrant.com
This adds a new optional support function to the GiST access method:
sortsupport. If it is defined, the GiST index is built by sorting all data
to the order defined by the sortsupport's comparator function, and packing
the tuples in that order to GiST pages. This is similar to how B-tree
index build works, and is much faster than inserting the tuples one by
one. The resulting index is smaller too, because the pages are packed more
tightly, upto 'fillfactor'. The normal build method works by splitting
pages, which tends to lead to more wasted space.
The quality of the resulting index depends on how good the opclass-defined
sort order is. A good order preserves locality of the input data.
As the first user of this facility, add 'sortsupport' function to the
point_ops opclass. It sorts the points in Z-order (aka Morton Code), by
interleaving the bits of the X and Y coordinates.
Author: Andrey Borodin
Reviewed-by: Pavel Borisov, Thomas Munro
Discussion: https://www.postgresql.org/message-id/1A36620E-CAD8-4267-9067-FB31385E7C0D%40yandex-team.ru
Because the code path taken for SQL commands executed in a walsender
will update the process title, we pretty much have to update the
title for replication commands as well. Otherwise, the title shows
"idle" for the rest of a logical walsender's lifetime once it's
executed any SQL command.
Playing with this, I confirm that a walsender now typically spends
most of its life reporting
walsender postgres [local] START_REPLICATION
Considering this in isolation, it might be better to have it say
walsender postgres [local] sending replication data
However, consistency with the other cases seems to be a stronger
argument.
In passing, remove duplicative pgstat_report_activity call.
Discussion: https://postgr.es/m/880181.1600026471@sss.pgh.pa.us
Since commit fd5942c18f (2012, 9.3-era), walsender has been sending
completion tags for certain replication commands twice -- and they're
not even consistent. Apparently neither libpq nor JDBC have a problem
with it, but it's not kosher. Fix by remove the EndCommand() call in
the common code path for them all, and inserting specific calls to
EndReplicationCommand() specifically in those places where it's needed.
EndReplicationCommand() is a new simple function to send the completion
tag for replication commands. Do this instead of sending a generic
SELECT completion tag for them all, which was also pretty bogus (if
innocuous). While at it, change StartReplication() to use
EndReplicationCommand() instead of pg_puttextmessage().
In commit 2f9661311b, I failed to realize that replication commands
are not close-enough kin of regular SQL commands, so the
DROP_REPLICATION_SLOT tag I added is undeserved and a type pun. Take it
out.
Backpatch to 13, where the latter commit appeared. The duplicate tag
has been sent since 9.3, but since nothing is broken, it doesn't seem
worth fixing.
Per complaints from Tom Lane.
Discussion: https://postgr.es/m/1347966.1600195735@sss.pgh.pa.us
We decided that the policy established in commit 7634bd4f6 for
the bgwriter, checkpointer, walwriter, and walreceiver processes,
namely that they should accept SIGQUIT at all times, really ought
to apply uniformly to all postmaster children. Therefore, get
rid of the duplicative and inconsistent per-process code for
establishing that signal handler and removing SIGQUIT from BlockSig.
Instead, make InitPostmasterChild do it.
The handler set up by InitPostmasterChild is SignalHandlerForCrashExit,
which just summarily does _exit(2). In interactive backends, we
almost immediately replace that with quickdie, since we would prefer
to try to tell the client that we're dying. However, this patch is
changing the behavior of autovacuum (both launcher and workers), as
well as walsenders. Those processes formerly also used quickdie,
but AFAICS that was just mindless copy-and-paste: they don't have
any interactive client that's likely to benefit from being told this.
The stats collector continues to be an outlier, in that it thinks
SIGQUIT means normal exit. That should probably be changed for
consistency, but there's another patch set where that's being
dealt with, so I didn't do so here.
Discussion: https://postgr.es/m/644875.1599933441@sss.pgh.pa.us
Since there is only one place that actually needs the partition check
expression, namely ExecPartitionCheck, it's better to fetch it from
the relcache there. In this way we will never fetch it at all if
the query never has use for it, and we still fetch it just once when
we do need it.
The reason for taking an interest in this is that if the relcache
doesn't already have the check expression cached, fetching it
requires obtaining AccessShareLock on the partition root. That
means that operations that look like they should only touch the
partition itself will also take a lock on the root. In particular
we observed that TRUNCATE on a partition may take a lock on the
partition's root, contributing to a deadlock situation in parallel
pg_restore.
As written, this patch does have a small cost, which is that we
are microscopically reducing efficiency for the case where a partition
has an empty check expression. ExecPartitionCheck will be called,
and will go through the motions of setting up and checking an empty
qual, where before it would not have been called at all. We could
avoid that by adding a separate boolean flag to track whether there
is a partition expression to test. However, this case only arises
for a default partition with no siblings, which surely is not an
interesting case in practice. Hence adding complexity for it
does not seem like a good trade-off.
Amit Langote, per a suggestion by me
Discussion: https://postgr.es/m/VI1PR03MB31670CA1BD9625C3A8C5DD05EB230@VI1PR03MB3167.eurprd03.prod.outlook.com
If a partitioned table's column is already marked NOT NULL, there is
no need to examine its partitions, because we can rely on previous
DDL to have enforced that the child columns are NOT NULL as well.
(Unfortunately, the same cannot be said for traditional inheritance,
so for now we have to restrict the optimization to partitioned tables.)
Hence, we may skip recursing to child tables in this situation.
The reason this case is worth worrying about is that when pg_dump dumps
a partitioned table having a primary key, it will include the requisite
NOT NULL markings in the CREATE TABLE commands, and then add the
primary key as a separate step. The primary key addition generates a
SET NOT NULL as a subcommand, just to be sure. So the situation where
a SET NOT NULL is redundant does arise in the real world.
Skipping the recursion does more than just save a few cycles: it means
that a command such as "ALTER TABLE ONLY partition_parent ADD PRIMARY
KEY" will take locks only on the partition parent table, not on the
partitions. It turns out that parallel pg_restore is effectively
assuming that that's true, and has little choice but to do so because
the dependencies listed for such a TOC entry don't include the
partitions. pg_restore could thus issue this ALTER while data restores
on the partitions are still in progress. Taking unnecessary locks on
the partitions not only hurts concurrency, but can lead to actual
deadlock failures, as reported by Domagoj Smoljanovic.
(A contributing factor in the deadlock is that TRUNCATE on a child
partition wants a non-exclusive lock on the parent. This seems
likewise unnecessary, but the fix for it is more invasive so we
won't consider back-patching it. Fortunately, getting rid of one
of these two poor behaviors is enough to remove the deadlock.)
Although support for partitioned primary keys came in with v11,
this patch is dependent on the SET NOT NULL refactoring done by
commit f4a3fdfbd, so we can only patch back to v12.
Patch by me; thanks to Alvaro Herrera and Amit Langote for review.
Discussion: https://postgr.es/m/VI1PR03MB31670CA1BD9625C3A8C5DD05EB230@VI1PR03MB3167.eurprd03.prod.outlook.com
The code recorded cache invalidation events by zeroing the "localreloid"
field of affected cache entries. However, it's possible for an inval
event to occur even while we have the entry open and locked. So an
ill-timed inval could result in "cache lookup failed for relation 0"
errors, if the worker's code tried to use the cleared field. We can
fix that by creating a separate bool field to record whether the entry
needs to be revalidated. (In the back branches, cram the bool into
what had been padding space, to avoid an ABI break in the somewhat
unlikely event that any extension is looking at this struct.)
Also, rearrange the logic in logicalrep_rel_open so that it
does the right thing in cases where table_open would fail.
We should retry the lookup by name in that case, but we didn't.
The real-world impact of this is probably small. In the first place,
the error conditions are very low probability, and in the second place,
the worker would just exit and get restarted. We only noticed because
in a CLOBBER_CACHE_ALWAYS build, the failure can occur repeatedly,
preventing the worker from making progress. Nonetheless, it's clearly
a bug, and it impedes a useful type of testing; so back-patch to v10
where this code was introduced.
Discussion: https://postgr.es/m/1032727.1600096803@sss.pgh.pa.us
Previously, it was based on nBlocksAllocated to account for tapes with
open write buffers that may not have made it to the BufFile yet.
That was unnecessary, because callers do not need to get the number of
blocks while a tape has an open write buffer; and it also conflicted
with the preallocation logic added for HashAgg.
Reviewed-by: Peter Geoghegan
Discussion: https://postgr.es/m/ce5af05900fdbd0e9185747825a7423c48501964.camel@j-davis.com
Backpatch-through: 13
In commit 464824323e, for each RelationSyncEntry we maintained the list
of xids (streamed_txns) for which we have already sent the schema. This
helps us to track when to send the schema to the downstream node for
replication of streaming transactions. Before this list got initialized,
we were processing invalidation messages which access this list and led
to an assertion failure.
In passing, clean up the nearby code:
* Initialize the list of xids with NIL instead of NULL which is our usual
coding practice.
* Remove the MemoryContext switch for creating a RelationSyncEntry in dynahash.
Diagnosed-by: Amit Kapila and Tom Lane
Author: Amit Kapila
Reviewed-by: Tom Lane and Dilip Kumar
Discussion: https://postgr.es/m/904373.1600033123@sss.pgh.pa.us
This function could often be seen in profiles of vacuum and could often
be a significant bottleneck during recovery. The problem was that a qsort
was performed in order to sort an array of item pointers in reverse offset
order so that we could use that to safely move tuples up to the end of the
page without overwriting the memory of yet-to-be-moved tuples. i.e. we
used to compact the page starting at the back of the page and move towards
the front. The qsort that this required could be expensive for pages with
a large number of tuples.
In this commit, we take another approach to tuple compactification.
Now, instead of sorting the remaining item pointers array we first check
if the array is presorted and only memmove() the tuples that need to be
moved. This presorted check can be done very cheaply in the calling
functions when the array is being populated. This presorted case is very
fast.
When the item pointer array is not presorted we must copy tuples that need
to be moved into a temp buffer before copying them back into the page
again. This differs from what we used to do here as we're now copying the
tuples back into the page in reverse line pointer order. Previously we
left the existing order alone. Reordering the tuples results in an
increased likelihood of hitting the pre-sorted case the next time around.
Any newly added tuple which consumes a new line pointer will also maintain
the correct sort order of tuples in the page which will also result in the
presorted case being hit the next time. Only consuming an unused line
pointer can cause the order of tuples to go out again, but that will be
corrected next time the function is called for the page.
Benchmarks have shown that the non-presorted case is at least equally as
fast as the original qsort method even when the page just has a few
tuples. As the number of tuples becomes larger the new method maintains
its performance whereas the original qsort method became much slower when
the number of tuples on the page became large.
Author: David Rowley
Reviewed-by: Thomas Munro
Tested-by: Jakub Wartak
Discussion: https://postgr.es/m/CA+hUKGKMQFVpjr106gRhwk6R-nXv0qOcTreZuQzxgpHESAL6dw@mail.gmail.com
ALTER TABLE commands in an extension script are added to an event
trigger command list; but starting with commit b5810de3f4 they do so in
a memory context that's too short-lived, so when execution ends and time
comes to use the entries, they've already been freed.
(This would also be a problem with ALTER TABLE commands in a
multi-command query string, but these serendipitously end in
PortalContext -- which probably explains why it took so long for this to
be reported.)
Fix by using the memory context specifically set for that, instead.
Backpatch to 13, where the aforementioned commit appeared.
Reported-by: Philippe Beaudoin
Author: Jehan-Guillaume de Rorthais <jgdr@dalibo.com>
Discussion: https://postgr.es/m/20200902193715.6e0269d4@firost
Introduced in 0aa8f7640.
MSVC warned about performing 32-bit bit shifting when it appeared like we
might like a 64-bit result. We did, but it just so happened that none of
the calls to this function could have caused the 32-bit shift to overflow.
Here we just cast the constant to int64 to make the compiler happy.
Discussion: https://postgr.es/m/CAApHDvofA_vsrpC13mq_hZyuye5B-ssKEaer04OouXYCO5-uXQ@mail.gmail.com
A walsender process that has executed a SQL command left the text of
that command in pg_stat_activity.query indefinitely, which is quite
confusing if it's in RUNNING state but not doing that query. An easy
and useful fix is to treat replication commands as if they were SQL
queries, and show them in pg_stat_activity according to the same rules
as for regular queries. While we're at it, it seems also sensible to
set debug_query_string, allowing error logging and debugging to see
the replication command.
While here, clean up assorted silliness in exec_replication_command:
* The SQLCmd path failed to restore CurrentMemoryContext to the caller's
value, and failed to delete the temp context created in this routine.
It's only through great good fortune that these oversights did not
result in long-term memory leaks or other problems. It seems cleaner
to code SQLCmd as a separate early-exit path, so do it like that.
* Remove useless duplicate call of SnapBuildClearExportedSnapshot().
* replication_scanner_finish() was never called.
None of those things are significant enough to merit a backpatch,
so this is for HEAD only.
Discussion: https://postgr.es/m/880181.1600026471@sss.pgh.pa.us
3c84046 is the original commit that introduced index_set_state_flags(),
where the presence of SnapshotNow made necessary the use of an in-place
update. SnapshotNow has been removed in 813fb03, so there is no actual
reasons to not make this operation transactional.
Note that while making the operation more robust, using a transactional
operation in this routine was not strictly necessary as there was no use
case for it yet. However, some future features are going to need a
transactional behavior, like support for CREATE/DROP INDEX CONCURRENTLY
with partitioned tables, where indexes in a partition tree need to have
all their pg_index.indis* flags updated in the same transaction to make
the operation stable to the end-user by keeping partition trees
consistent, even with a failure mid-flight.
REINDEX CONCURRENTLY uses already transactional updates when swapping
the old and new indexes, making this change more consistent with the
index-swapping logic.
Author: Michael Paquier
Reviewed-by: Anastasia Lubennikova
Discussion: https://postgr.es/m/20200903080440.GA8559@paquier.xyz
transformCreateStmt() adjusts the transformed statement's RangeVar
to specify the target schema explicitly, for the express reason
of making sure that auxiliary statements derived by parse
transformation operate on the right table. But the refactoring
I did in commit 502898192 got this wrong and passed the untransformed
RangeVar to expandTableLikeClause(). This could lead to assertion
failures or weird misbehavior if the wrong table was accessed.
Per report from Alexander Lakhin. Like the previous patch, back-patch
to all supported branches.
Discussion: https://postgr.es/m/05051f9d-b32b-cb35-6735-0e9f2ab86b5f@gmail.com
We use the timestamp of the global statfile if we are not able to
determine it for a particular database in case the entry for that database
doesn't exist. However, we were using it even when the statfile is
corrupt.
As there is no user reported issue and it is not clear if there is any
impact of this on actual application so decided not to backpatch.
Reported-by: Amit Kapila
Author: Amit Kapila
Reviewed-by: Sawada Masahiko, Magnus Hagander and Alvaro Herrera
Discussion: https://postgr.es/m/CAA4eK1J3oTJKyVq6v7K4d3jD+vtnruG9fHRib6UuWWsrwAR6Aw@mail.gmail.com
The bgwriter, checkpointer, walwriter, and walreceiver processes
claimed to allow SIGQUIT "at all times". In reality SIGQUIT
would get re-blocked during error recovery, because we didn't
update the actual signal mask immediately, so sigsetjmp() would
save and reinstate a mask that includes SIGQUIT.
This appears to be simply a coding oversight. There's never a
good reason to hold off SIGQUIT in these processes, because it's
going to just call _exit(2) which should be safe enough, especially
since the postmaster is going to tear down shared memory afterwards.
Hence, stick in PG_SETMASK() calls to install the modified BlockSig
mask immediately.
Also try to improve the comments around sigsetjmp blocks. Most of
them were just referencing postgres.c, which is misleading because
actually postgres.c manages the signals differently.
No back-patch, since there's no evidence that this is causing any
problems in the field.
Discussion: https://postgr.es/m/CALDaNm1d1hHPZUg3xU4XjtWBOLCrA+-2cJcLpw-cePZ=GgDVfA@mail.gmail.com
Currently, no useful trace is left in the logs when the postmaster
is forced to use SIGKILL to shut down children that failed to respond
to SIGQUIT. Some questions were raised about how often that scenario
happens in the buildfarm, so let's add a LOG-level message showing
that it happened.
Discussion: https://postgr.es/m/1850884.1599601164@sss.pgh.pa.us
Although 58c6feccf fixed the case for SIGQUIT, we were still calling
proc_exit() from signal handlers for SIGTERM and timeout failures in
ProcessStartupPacket. Fortunately, at the point where that code runs,
we haven't yet connected to shared memory in any meaningful way, so
there is nothing we need to undo in shared memory. This means it
should be safe to use _exit(1) here, ie, not run any atexit handlers
but also inform the postmaster that it's not a crash exit.
To make sure nobody breaks the "nothing to undo" expectation, add
a cross-check that no on-shmem-exit or before-shmem-exit handlers
have been registered yet when we finish using these signal handlers.
This change is simple enough that maybe it could be back-patched,
but I won't risk that right now.
Discussion: https://postgr.es/m/1850884.1599601164@sss.pgh.pa.us
Bring the signal handling for startup-packet collection into line
with the policy established in commits bedadc732 and 8e19a8264,
namely don't risk running atexit callbacks when handling SIGQUIT.
Ideally, we'd not do so for SIGTERM or timeout interrupts either,
but that change seems a bit too risky for the back branches.
For now, just improve the comments in this area to describe the risk.
Also relocate where BackendInitialize re-disables these interrupts,
to minimize the code span where they're active. This doesn't buy
a whole lot of safety, but it can't hurt.
In passing, rename startup_die() to remove confusion about whether
it is for the startup process.
Like the previous commits, back-patch to all supported branches.
Discussion: https://postgr.es/m/1850884.1599601164@sss.pgh.pa.us
Move applicable code out of RelationBuildDesc(), which nailed relations
bypass. Non-assert builds experienced no known problems. Back-patch to
v13, where commit c6b92041d3 introduced
rd_firstRelfilenodeSubid.
Kyotaro Horiguchi. Reported by Justin Pryzby.
Discussion: https://postgr.es/m/20200907023737.GA7158@telsasoft.com
Commit 8e19a8264 changed the SIGQUIT handlers of almost all server
processes not to run atexit callbacks. The archiver process was
skipped, perhaps because it's not connected to shared memory; but
it's just as true here that running atexit callbacks in a signal
handler is unsafe. So let's make it work like the rest.
In HEAD and v13, we can use the common SignalHandlerForCrashExit
handler. Before that, just tweak pgarch_exit to use _exit(2)
explicitly.
Like the previous commit, back-patch to all supported branches.
Kyotaro Horiguchi, back-patching by me
Discussion: https://postgr.es/m/1850884.1599601164@sss.pgh.pa.us
Partitioning tuple route code assumes that the partition chosen while
descending the partition hierarchy is always the correct one. This is
true except when the partition is the default partition and another
partition has been added concurrently: the partition constraint changes
and we don't recheck it. This can lead to tuples mistakenly being added
to the default partition that should have been rejected.
Fix by rechecking the default partition constraint while descending the
hierarchy.
An isolation test based on the reproduction steps described by Hao Wu
(with tweaks for extra coverage) is included.
Backpatch to 12, where this bug came in with 898e5e3290.
Reported by: Hao Wu <hawu@vmware.com>
Author: Amit Langote <amitlangote09@gmail.com>
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/CA+HiwqFqBmcSSap4sFnCBUEL_VfOMmEKaQ3gwUhyfa4c7J_-nA@mail.gmail.com
Discussion: https://postgr.es/m/DM5PR0501MB3910E97A9EDFB4C775CF3D75A42F0@DM5PR0501MB3910.namprd05.prod.outlook.com
Historically, cancel_before_shmem_exit() just silently did nothing
if the specified callback wasn't the top-of-stack. The folly of
ignoring this case was exposed by the bugs fixed in 303640199 and
bab150045, so let's make it throw elog(ERROR) instead.
There is a decent argument to be made that PG_ENSURE_ERROR_CLEANUP
should use some separate infrastructure, so it wouldn't break if
something inside the guarded code decides to register a new
before_shmem_exit callback. However, a survey of the surviving
uses of before_shmem_exit() and PG_ENSURE_ERROR_CLEANUP doesn't
show any plausible conflicts of that sort today, so for now we'll
forgo the extra complexity. (It will almost certainly become
necessary if anyone ever wants to wrap PG_ENSURE_ERROR_CLEANUP
around arbitrary user-defined actions, though.)
No backpatch, since this is developer support not a production issue.
Bharath Rupireddy, per advice from Andres Freund, Robert Haas, and myself
Discussion: https://postgr.es/m/CALj2ACWk7j4F2v2fxxYfrroOF=AdFNPr1WsV+AGtHAFQOqm_pw@mail.gmail.com
The problem is caused by me (Andres) having ProcSleep() look at the
wrong PGPROC entry in 5788e258bb.
Unfortunately it seems hard to write a reliable test for autovacuum
cancellations. Perhaps somebody will come up with a good approach, but
it seems worth fixing the issue even without a test.
Reported-By: Jeff Janes <jeff.janes@gmail.com>
Author: Jeff Janes <jeff.janes@gmail.com>
Discussion: https://postgr.es/m/CAMkU=1wH2aUy+wDRDz+5RZALdcUnEofV1t9PzXS_gBJO9vZZ0Q@mail.gmail.com
This essentially reverts a micro-optimization I made years ago,
as part of the much larger commit d72f6c750. It's doubtful
that there was any hard evidence for it being helpful even then,
and the case is even more dubious now that modern compilers
are so much smarter about inlining memset().
The proximate reason for undoing it is to get rid of the type punning
inherent in MemSet, for fear that that may cause problems now that
we're applying additional optimization switches to numeric.c.
At the very least this'll silence some warnings from a few old
buildfarm animals.
(It's probably past time for another look at whether MemSet is still
worth anything at all, but I do not propose to tackle that question
right now.)
Discussion: https://postgr.es/m/CAJ3gD9evtA_vBo+WMYMyT-u=keHX7-r8p2w7OSRfXf42LTwCZQ@mail.gmail.com
Until now, REINDEX was not able to work with partitioned tables and
indexes, forcing users to reindex partitions one by one. This extends
REINDEX INDEX and REINDEX TABLE so as they can accept a partitioned
index and table in input, respectively, to reindex all the partitions
assigned to them with physical storage (foreign tables, partitioned
tables and indexes are then discarded).
This shares some logic with schema and database REINDEX as each
partition gets processed in its own transaction after building a list of
relations to work on. This choice has the advantage to minimize the
number of invalid indexes to one partition with REINDEX CONCURRENTLY in
the event a cancellation or failure in-flight, as the only indexes
handled at once in a single REINDEX CONCURRENTLY loop are the ones from
the partition being working on.
Isolation tests are added to emulate some cases I bumped into while
developing this feature, particularly with the concurrent drop of a
leaf partition reindexed. However, this is rather limited as LOCK would
cause REINDEX to block in the first transaction building the list of
partitions.
Per its multi-transaction nature, this new flavor cannot run in a
transaction block, similarly to REINDEX SCHEMA, SYSTEM and DATABASE.
Author: Justin Pryzby, Michael Paquier
Reviewed-by: Anastasia Lubennikova
Discussion: https://postgr.es/m/db12e897-73ff-467e-94cb-4af03705435f.adger.lj@alibaba-inc.com
Tomas Vondra observed that the IO behavior for HashAgg tends to be
worse than for Sort. Penalize HashAgg IO costs accordingly.
Also, account for the CPU effort of spilling the tuples and reading
them back.
Discussion: https://postgr.es/m/20200906212112.nzoy5ytrzjjodpfh@development
Reviewed-by: Tomas Vondra
Backpatch-through: 13
Experimentation shows that clang will auto-vectorize the critical
multiplication loop if the termination condition is written "i2 < limit"
rather than "i2 <= limit". This seems unbelievably stupid, but I've
reproduced it on both clang 9.0.1 (RHEL8) and 11.0.3 (macOS Catalina).
gcc doesn't care, so tweak the code to do it that way.
Discussion: https://postgr.es/m/CAJ3gD9evtA_vBo+WMYMyT-u=keHX7-r8p2w7OSRfXf42LTwCZQ@mail.gmail.com
Some kernels can tell us the type of a "dirent", so we can avoid a call
to stat() or lstat() in many cases. Define a new function
get_dirent_type() to contain that logic, for use by the backend and
frontend versions of walkdir(), and perhaps other callers in future.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Reviewed-by: Juan José Santamaría Flecha <juanjo.santamaria@gmail.com>
Discussion: https://postgr.es/m/CA%2BhUKG%2BFzxupGGN4GpUdbzZN%2Btn6FQPHo8w0Q%2BAPH5Wz8RG%2Bww%40mail.gmail.com
Compile numeric.c with -ftree-vectorize where available, and adjust
the innermost loop of mul_var() so that it is amenable to being
auto-vectorized. (Mainly, that involves making it process the arrays
left-to-right not right-to-left.)
Applying -ftree-vectorize actually makes numeric.o smaller, at least
with my compiler (gcc 8.3.1 on x86_64), and it's a little faster too.
Independently of that, fixing the inner loop to be vectorizable also
makes things a bit faster. But doing both is a huge win for
multiplications with lots of digits. For me, the numeric regression
test is the same speed to within measurement noise, but numeric_big
is a full 45% faster.
We also looked into applying -funroll-loops, but that makes numeric.o
bloat quite a bit, and the additional speed improvement is very
marginal.
Amit Khandekar, reviewed and edited a little by me
Discussion: https://postgr.es/m/CAJ3gD9evtA_vBo+WMYMyT-u=keHX7-r8p2w7OSRfXf42LTwCZQ@mail.gmail.com
Replace CFLAGS_VECTOR with CFLAGS_UNROLL_LOOPS and CFLAGS_VECTORIZE,
allowing us to distinguish whether we want to apply -funroll-loops,
-ftree-vectorize, or both to a particular source file. Up to now
the only consumer of the symbol has been checksum.c which wants
both, so that there was no need to distinguish; but that's about
to change.
Amit Khandekar, reviewed and edited a little by me
Discussion: https://postgr.es/m/CAJ3gD9evtA_vBo+WMYMyT-u=keHX7-r8p2w7OSRfXf42LTwCZQ@mail.gmail.com
Letting the caller provide a StringInfo to read into is helpful when
the caller needs to merge lines or otherwise modify the data after
it's been read. Notably, now the code added by commit 8f8154a50
can use pg_get_line_append() instead of having its own copy of that
logic. A follow-on commit will also make use of this.
Also, since StringInfo buffers are a minimum of 1KB long, blindly
using pg_get_line() in a loop can eat a lot more memory than one would
expect. I discovered for instance that commit e0f05cd5b caused initdb
to consume circa 10MB to read postgres.bki, even though that's under
1MB worth of data. A less memory-hungry alternative is to re-use the
same StringInfo for all lines and pg_strdup the results.
Discussion: https://postgr.es/m/1315832.1599345736@sss.pgh.pa.us
We reported the wrong types when complaining that an aggregate's
moving-aggregate implementation is inconsistent with its regular
implementation.
This was wrong since the feature was introduced, so back-patch
to all supported branches.
Jeff Janes
Discussion: https://postgr.es/m/CAMkU=1x808LH=LPhZp9mNSP0Xd1xDqEd+XeGcvEe48dfE6xV=A@mail.gmail.com
At some point back in the last century, somebody felt that reading
all of pg_type twice was cheaper, or at least easier, than using
repalloc() to resize the Typ[] array dynamically. That seems like an
entirely wacko proposition, so rewrite the code to do it the other
way. (To add insult to injury, there were two not-quite-identical
copies of said code.)
initdb.c's readfile() function had the same disease of preferring
to do double the I/O to avoid resizing its output array. Here,
we can make things easier by using the just-invented pg_get_line()
function to handle reading individual lines without a predetermined
notion of how long they are.
On my machine, it's difficult to detect any net change in the
overall runtime of initdb from these changes; but they should
help on slower buildfarm machines (especially since a buildfarm
cycle involves a lot of initdb's these days).
My attention was drawn to these places by scan-build complaints,
but on inspection they needed a lot more work than just suppressing
dead stores :-(
This commit improves the dependency registrations by taking advantage of
the preliminary work done in 63110c62, to group together the insertion
of dependencies of the same type to pg_depend. With the current layer
of routines available, and as only dependencies of the same type can be
grouped, there are code paths still doing more than one multi-insert
when it is necessary to register dependencies of multiple types
(constraint and index creation are two cases doing that).
While on it, this refactors some of the code to use ObjectAddressSet()
when manipulating object addresses.
Author: Daniel Gustafsson, Michael Paquier
Reviewed-by: Andres Freund, Álvaro Herrera
Discussion: https://postgr.es/m/20200807061619.GA23955@paquier.xyz
This is a follow-up of the work done in e3931d01. This case is a bit
different than pg_attribute and pg_shdepend: the maximum number of items
to insert is known in advance, but there is no need to handle pinned
dependencies. Hence, the base allocation for slots is done based on the
number of items and the maximum allowed with a cap at 64kB. Slots are
initialized once used to minimize the overhead of the operation.
The insertions can be done for dependencies of the same type. More
could be done by grouping the insertion of multiple dependency types in
a single batch. This is left as future work.
Some of the multi-insert logic is also simplified for pg_shdepend, as
per the feedback discussed for this specific patch. This also moves to
indexing.h the variable capping the maximum amount of data that can be
used at once for a multi-insert, instead of having separate definitions
for pg_attribute, pg_depend and pg_shdepend.
Author: Daniel Gustafsson, Michael Paquier
Reviewed-by: Andres Freund, Álvaro Herrera
Discussion: https://postgr.es/m/20200807061619.GA23955@paquier.xyz
Commit 3f60f690f only partially fixed the broken-status-tracking
issue in LogicalRepApplyLoop: we need ping_sent to have the same
lifetime as last_recv_timestamp. The effects are much less serious
than what that commit fixed, though. AFAICS this would just lead to
extra ping requests being sent, once per second until the sender
responds. Still, it's a bug, so backpatch to v10 as before.
Discussion: https://postgr.es/m/959627.1599248476@sss.pgh.pa.us
When reading a WAL record fails to find continuation record(s) of the
proper length, report what it expects, for clarity.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://postgr.es/m/20200903212152.GA15319@alvherre.pgsql
This node already handles multiple options using a bitmask, so having a
separate boolean flag is not necessary. This simplifies the code a bit
with less arguments to give to the reindex routines, by replacing the
boolean with an equivalent bitmask value.
Reviewed-by: Julien Rouhaud
Discussion: https://postgr.es/m/20200902110326.GA14963@paquier.xyz
This patch started out with the goal of harmonizing various arbitrary
limits on password length, but after awhile a better idea emerged:
let's just get rid of those fixed limits.
recv_password_packet() has an arbitrary limit on the packet size,
which we don't really need, so just drop it. (Note that this doesn't
really affect anything for MD5 or SCRAM password verification, since
those will hash the user's password to something shorter anyway.
It does matter for auth methods that require a cleartext password.)
Likewise remove the arbitrary error condition in pg_saslprep().
The remaining limits are mostly in client-side code that prompts
for passwords. To improve those, refactor simple_prompt() so that
it allocates its own result buffer that can be made as big as
necessary. Actually, it proves best to make a separate routine
pg_get_line() that has essentially the semantics of fgets(), except
that it allocates a suitable result buffer and hence will never
return a truncated line. (pg_get_line has a lot of potential
applications to replace randomly-sized fgets buffers elsewhere,
but I'll leave that for another patch.)
I built pg_get_line() atop stringinfo.c, which requires moving
that code to src/common/; but that seems fine since it was a poor
fit for src/port/ anyway.
This patch is mostly mine, but it owes a good deal to Nathan Bossart
who pressed for a solution to the password length problem and
created a predecessor patch. Also thanks to Peter Eisentraut and
Stephen Frost for ideas and discussion.
Discussion: https://postgr.es/m/09512C4F-8CB9-4021-B455-EF4C4F0D55A0@amazon.com
Because sigsetjmp() will restore the initial state with signals blocked,
the code path in bgworker.c for reporting an error and exiting would
execute that way. Usually this is fairly harmless; but if a parallel
worker had an error message exceeding the shared-memory communication
buffer size (16K) it would lock up, because it would wait for a
resume-sending signal from its parallel leader which it would never
detect.
To fix, just unblock signals at the appropriate point.
This can be shown to fail back to 9.6. The lack of parallel query
infrastructure makes it difficult to provide a simple test case for
9.5; but I'm pretty sure the issue exists in some form there as well,
so apply the code change there too.
Vignesh C, reviewed by Bharath Rupireddy, Robert Haas, and myself
Discussion: https://postgr.es/m/CALDaNm1d1hHPZUg3xU4XjtWBOLCrA+-2cJcLpw-cePZ=GgDVfA@mail.gmail.com
A backslash at the end of a line now causes the next line to be appended
to the current one (effectively, the backslash and newline are discarded).
This allows long HBA entries to be created without legibility problems.
While we're here, get rid of the former hard-wired length limit on
pg_hba.conf lines, by using an expansible StringInfo buffer instead
of a fixed-size local variable.
Since the same code is used to read the ident map file, these changes
apply there as well.
Fabien Coelho, reviewed by Justin Pryzby and David Zhang
Discussion: https://postgr.es/m/alpine.DEB.2.21.2003251906140.15243@pseudo
To add support for streaming of in-progress transactions into the
built-in logical replication, we need to do three things:
* Extend the logical replication protocol, so identify in-progress
transactions, and allow adding additional bits of information (e.g.
XID of subtransactions).
* Modify the output plugin (pgoutput) to implement the new stream
API callbacks, by leveraging the extended replication protocol.
* Modify the replication apply worker, to properly handle streamed
in-progress transaction by spilling the data to disk and then
replaying them on commit.
We however must explicitly disable streaming replication during
replication slot creation, even if the plugin supports it. We
don't need to replicate the changes accumulated during this phase,
and moreover we don't have a replication connection open so we
don't have where to send the data anyway.
Author: Tomas Vondra, Dilip Kumar and Amit Kapila
Reviewed-by: Amit Kapila, Kuntal Ghosh and Ajin Cherian
Tested-by: Neha Sharma, Mahendra Singh Thalor and Ajin Cherian
Discussion: https://postgr.es/m/688b0b7f-2f6c-d827-c27b-216a8e3ea700@2ndquadrant.com
This splits a string at occurrences of a delimiter. It is exactly like
string_to_array() except for producing a set of values instead of an
array of values. Thus, the relationship of these two functions is
the same as between regexp_split_to_table() and regexp_split_to_array().
Although the same results could be had from unnest(string_to_array()),
this is somewhat faster than that, and anyway it seems reasonable to
have it for symmetry with the regexp functions.
Pavel Stehule, reviewed by Peter Smith
Discussion: https://postgr.es/m/CAFj8pRD8HOpjq2TqeTBhSo_QkzjLOhXzGCpKJ4nCs7Y9SQkuPw@mail.gmail.com
In SyncRepWaitForLSN() routine called in transaction commit time,
SyncRepLock is necessary to atomically both check the shared
sync_standbys_defined flag and operate the sync replication wait-queue.
On the other hand, when the flag is false, the lock is not necessary
because the wait-queue is not touched. But due to the changes by
commit 48c9f49265, previously the lock was taken whatever the flag was.
This could cause unnecessary performance overhead in every transaction
commit time. Therefore this commit avoids that unnecessary aquisition
of SyncRepLock.
Author: Fujii Masao
Reviewed-by: Asim Praveen, Masahiko Sawada,
Discussion: https://postgr.es/m/20200406050332.nsscfqjzk2d57zyx@alap3.anarazel.de
When multiple relations are reindexed, a scan of pg_class is done first
to build the list of relations to work on. However the REINDEX logic
has never checked if a relation listed still exists when beginning the
work on it, causing for example sudden cache lookup failures.
This commit adds safeguards against dropped relations for REINDEX,
similarly to VACUUM or CLUSTER where we try to open the relation,
ignoring it if it is missing. A new option is added to the REINDEX
routines to control if a missed relation is OK to ignore or not.
An isolation test, based on REINDEX SCHEMA, is added for the concurrent
and non-concurrent cases.
Author: Michael Paquier
Reviewed-by: Anastasia Lubennikova
Discussion: https://postgr.es/m/20200813043805.GE11663@paquier.xyz
Since other sessions aren't allowed to look into a temporary table
of our own session, we do not need to worry about the global xmin
horizon when setting the vacuum XID cutoff. Indeed, if we're not
inside a transaction block, we may set oldestXmin to be the next
XID, because there cannot be any in-doubt tuples in a temp table,
nor any tuples that are dead but still visible to some snapshot of
our transaction. (VACUUM, of course, is never inside a transaction
block; but we need to test that because CLUSTER shares the same code.)
This approach allows us to always clean out a temp table completely
during VACUUM, independently of concurrent activity. Aside from
being useful in its own right, that simplifies building reproducible
test cases.
Discussion: https://postgr.es/m/3490536.1598629609@sss.pgh.pa.us
We were already raising an error for DROP INDEX CONCURRENTLY on a
partitioned table, albeit a different and confusing one:
ERROR: DROP INDEX CONCURRENTLY must be first action in transaction
Change that to throw a more comprehensible error:
ERROR: cannot drop partitioned index \"%s\" concurrently
Michael Paquier authored the test case for indexes on temporary
partitioned tables.
Backpatch to 11, where indexes on partitioned tables were added.
Reported-by: Jan Mussler <jan.mussler@zalando.de>
Reviewed-by: Michael Paquier <michael@paquier.xyz>
Discussion: https://postgr.es/m/16594-d2956ca909585067@postgresql.org
Commit 808e13b282 introduced a few APIs to extend the existing Buffile
interface. In SharedFileSetDeleteOnProcExit, it tries to delete the list
element while traversing the list with 'foreach' construct which makes the
behavior of list traversal unpredictable.
Author: Amit Kapila
Reviewed-by: Dilip Kumar
Tested-by: Dilip Kumar and Neha Sharma
Discussion: https://postgr.es/m/CAA4eK1JhLatVcQ2OvwA_3s0ih6Hx9+kZbq107cXVsSWWukH7vA@mail.gmail.com
Historically, we've considered the state with relpages and reltuples
both zero as indicating that we do not know the table's tuple density.
This is problematic because it's impossible to distinguish "never yet
vacuumed" from "vacuumed and seen to be empty". In particular, a user
cannot use VACUUM or ANALYZE to override the planner's normal heuristic
that an empty table should not be believed to be empty because it is
probably about to get populated. That heuristic is a good safety
measure, so I don't care to abandon it, but there should be a way to
override it if the table is indeed intended to stay empty.
Hence, represent the initial state of ignorance by setting reltuples
to -1 (relpages is still set to zero), and apply the minimum-ten-pages
heuristic only when reltuples is still -1. If the table is empty,
VACUUM or ANALYZE (but not CREATE INDEX) will override that to
reltuples = relpages = 0, and then we'll plan on that basis.
This requires a bunch of fiddly little changes, but we can get rid of
some ugly kluges that were formerly needed to maintain the old definition.
One notable point is that FDWs' GetForeignRelSize methods will see
baserel->tuples = -1 when no ANALYZE has been done on the foreign table.
That seems like a net improvement, since those methods were formerly
also in the dark about what baserel->tuples = 0 really meant. Still,
it is an API change.
I bumped catversion because code predating this change would get confused
by seeing reltuples = -1.
Discussion: https://postgr.es/m/F02298E0-6EF4-49A1-BCB6-C484794D9ACC@thebuild.com
A failure when dropping concurrently an index used in a replica identity
could leave in pg_index an index marked as !indisvalid and
indisreplident. Reindexing this index would switch back indisvalid to
true, and if the replica identity of the parent relation was switched to
use a different index, it would be possible to finish with more than one
index marked as indisreplident. If that were to happen, this could mess
up with the relation cache as an incorrect index could be used for the
replica identity.
Indexes marked as invalid are discarded as candidates for the replica
identity, as of RelationGetIndexList(), so similarly to what is done
with indisclustered, resetting indisreplident when the index is marked
as invalid keeps things consistent. REINDEX CONCURRENTLY's swapping
already resets the flag for the old index, while the new index inherits
the value of the old index to-be-dropped, so only DROP INDEX was an
issue.
Even if this is a bug, the sequence able to reproduce a problem requires
a failure while running DROP INDEX CONCURRENTLY, something unlikely
going to happen in the field, so no backpatch is done.
Author: Michael Paquier
Reviewed-by: Dmitry Dolgov
Discussion: https://postgr.es/m/20200827025721.GN2017@paquier.xyz
collectMatchBitmap() needs to re-find the index tuple it was previously
looking at, after transiently dropping lock on the index page it's on.
The tuple should still exist and be at its prior position or somewhere
to the right of that, since ginvacuum never removes tuples but
concurrent insertions could add one. However, there was a thinko in
that logic, to the effect of expecting any inserted tuples to have the
same index "attnum" as what we'd been scanning. Since there's no
physical separation of tuples with different attnums, it's not terribly
hard to devise scenarios where this fails, leading to transient "lost
saved point in index" errors. (While I've duplicated this with manual
testing, it seems impossible to make a reproducible test case with our
available testing technology.)
Fix by just continuing the scan when the attnum doesn't match.
While here, improve the error message used if we do fail, so that it
matches the wording used in btree for a similar case.
collectMatchBitmap()'s posting-tree code path was previously not
exercised at all by our regression tests. While I can't make
a regression test that exhibits the bug, I can at least improve
the code coverage here, so do that. The test case I made for this
is an extension of one added by 4b754d6c1, so it only works in
HEAD and v13; didn't seem worth trying hard to back-patch it.
Per bug #16595 from Jesse Kinkead. This has been broken since
multicolumn capability was added to GIN (commit 27cb66fdf),
so back-patch to all supported branches.
Discussion: https://postgr.es/m/16595-633118be8eef9ce2@postgresql.org
The additional information added will be an offset number for heap
operations. This information will help us in finding the exact tuple due
to which the error has occurred.
Author: Mahendra Singh Thalor and Amit Kapila
Reviewed-by: Sawada Masahiko, Justin Pryzby and Amit Kapila
Discussion: https://postgr.es/m/CAKYtNApK488TDF4bMbw+1QH8HJf9cxdNDXquhU50TK5iv_FtCQ@mail.gmail.com
Allow BufFile to support temporary files that can be used by the single
backend when the corresponding files need to be survived across the
transaction and need to be opened and closed multiple times. Such files
need to be created as a member of a SharedFileSet.
Additionally, this commit implements the interface for BufFileTruncate to
allow files to be truncated up to a particular offset and extends the
BufFileSeek API to support the SEEK_END case. This also adds an option to
provide a mode while opening the shared BufFiles instead of always opening
in read-only mode.
These enhancements in BufFile interface are required for the upcoming
patch to allow the replication apply worker, to handle streamed
in-progress transactions.
Author: Dilip Kumar, Amit Kapila
Reviewed-by: Amit Kapila
Tested-by: Neha Sharma
Discussion: https://postgr.es/m/688b0b7f-2f6c-d827-c27b-216a8e3ea700@2ndquadrant.com
Previously the codes for pg_backend_memory_contexts were in
src/backend/utils/mmgr/mcxt.c. This commit moves them to
src/backend/utils/adt/mcxtfuncs.c so that mcxt.c basically includes
only the low-level interface for memory contexts.
Author: Atsushi Torikoshi
Reviewed-by: Michael Paquier, Fujii Masao
Discussion: https://postgr.es/m/20200819135545.GC19121@paquier.xyz
pg_backend_memory_contexts view contains some internal information of
memory contexts. Since exposing them to any users by default may cause
security issue, this commit allows only superusers to read this view,
by default, like we do for pg_shmem_allocations view.
Bump catalog version.
Author: Atsushi Torikoshi
Reviewed-by: Michael Paquier, Fujii Masao
Discussion: https://postgr.es/m/1414992.1597849297@sss.pgh.pa.us
It's a bit inefficient to test if a Bitmapset is empty by counting all the
members and seeing if that number is zero. It's much better just to use
bms_is_empty(). Likewise for checking if there are at least two members,
just use bms_membership(), which does not need to do anything more after
finding two members.
Discussion: https://postgr.es/m/CAApHDvpvwm_QjbDOb5xga%2BKmX9XkN9xQavNGm3SvDbVnCYOerQ%40mail.gmail.com
Reviewed-by: Tomas Vondra
We were displaying the wrong phase information for 'info' message in the
index clean up phase because we were switching to the previous phase a bit
early. We were also not displaying context information for heap phase
unless the block number is valid which is fine for error cases but for
messages at 'info' or lower error level it appears to be inconsistent with
index phase information.
Reported-by: Sawada Masahiko
Author: Sawada Masahiko
Reviewed-by: Amit Kapila
Backpatch-through: 13, where it was introduced
Discussion: https://postgr.es/m/CA+fd4k4HcbhPnCs7paRTw1K-AHin8y4xKomB9Ru0ATw0UeTy2w@mail.gmail.com
The trouble with doing this is that an apparently-constant subquery
output column isn't really constant if it is a grouping column that
appears in only some of the grouping sets. A qual using such a
column would be subject to incorrect const-folding after push-down,
as seen in bug #16585 from Paul Sivash.
To fix, just disable qual pushdown altogether if the sub-query has
nonempty groupingSets. While we could imagine far less restrictive
solutions, there is not much point in working harder right now,
because subquery_planner() won't move HAVING clauses to WHERE within
such a subquery. If the qual stays in HAVING it's not going to be
a lot more useful than if we'd kept it at the outer level.
Having said that, this restriction could be removed if we used a
parsetree representation that distinguished such outputs from actual
constants, which is something I hope to do in future. Hence, make
the patch a minimal addition rather than integrating it more tightly
(e.g. by renumbering the existing items in subquery_is_pushdown_safe's
comment).
Back-patch to 9.5 where grouping sets were introduced.
Discussion: https://postgr.es/m/16585-9d8c340d23ade8c1@postgresql.org
Commit 1281a5c90 rearranged the logic in this area rather drastically,
and it broke the case of adding a foreign key constraint in the same
ALTER that adds the pkey or unique constraint it depends on. While
self-referential fkeys are surely a pretty niche case, this used to
work so we shouldn't break it.
To fix, reorganize the scheduling rules in ATParseTransformCmd so
that a transformed AT_AddConstraint subcommand will be delayed into
a later pass in all cases, not only when it's been spit out as a
side-effect of parsing some other command type.
Also tweak the logic so that we won't run ATParseTransformCmd twice
while doing this. It seems to work even without that, but it's
surely wasting cycles to do so.
Per bug #16589 from Jeremy Evans. Back-patch to v13 where the new
code was introduced.
Discussion: https://postgr.es/m/16589-31c8d981ca503896@postgresql.org
If a CREATE TABLE command uses both LIKE and traditional inheritance,
Vars in CHECK constraints and expression indexes that are absorbed
from a LIKE parent table tended to get mis-numbered, resulting in
wrong answers and/or bizarre error messages (though probably not any
actual crashes, thanks to validation occurring in the executor).
In v12 and up, the same could happen to Vars in GENERATED expressions,
even in cases with no LIKE clause but multiple traditional-inheritance
parents.
The cause of the problem for LIKE is that parse_utilcmd.c supposed
it could renumber such Vars correctly during transformCreateStmt(),
which it cannot since we have not yet accounted for columns added via
inheritance. Fix that by postponing processing of LIKE INCLUDING
CONSTRAINTS, DEFAULTS, GENERATED, INDEXES till after we've performed
DefineRelation().
The error with GENERATED and multiple inheritance is a simple oversight
in MergeAttributes(); it knows it has to renumber Vars in inherited
CHECK constraints, but forgot to apply the same processing to inherited
GENERATED expressions (a/k/a defaults).
Per bug #16272 from Tom Gottfried. The non-GENERATED variants of the
issue are ancient, presumably dating right back to the addition of
CREATE TABLE LIKE; hence back-patch to all supported branches.
Discussion: https://postgr.es/m/16272-6e32da020e9a9381@postgresql.org
Commit ce77abe63c allowed EXPLAIN (BUFFERS) to report the information
on buffer usage during planning phase. However three issues were
reported regarding this feature.
(1) Previously, EXPLAIN option BUFFERS required ANALYZE. So the query
had to be actually executed by specifying ANALYZE even when we
want to see only the planner's buffer usage. This was inconvenient
especially when the query was write one like DELETE.
(2) EXPLAIN included the planner's buffer usage in summary
information. So SUMMARY option had to be enabled to report that.
Also this format was confusing.
(3) The output structure for planning information was not consistent
between TEXT format and the others. For example, "Planning" tag
was output in JSON format, but not in TEXT format.
For (1), this commit allows us to perform EXPLAIN (BUFFERS) without
ANALYZE to report the planner's buffer usage.
For (2), this commit changed EXPLAIN output so that the planner's
buffer usage is reported before summary information.
For (3), this commit made the output structure for planning
information more consistent between the formats.
Back-patch to v13 where the planner's buffer usage was allowed to
be reported in EXPLAIN.
Reported-by: Pierre Giraud, David Rowley
Author: Fujii Masao
Reviewed-by: David Rowley, Julien Rouhaud, Pierre Giraud
Discussion: https://postgr.es/m/07b226e6-fa49-687f-b110-b7c37572f69e@dalibo.com
This corrects an oversight by me in 2072932407, which made
ProcArrayClearTransaction() increment xactCompletionCount. That requires an
exclusive lock, obviously.
There's other approaches that avoid the exclusive acquisition, but given that a
2PC commit is fairly heavyweight, it doesn't seem worth doing so. I've not been
able to measure a performance difference, unsurprisingly. I did add a
comment documenting that we could do so, should it ever become a bottleneck.
Reported-By: Tom Lane <tgl@sss.pgh.pa.us>
Author: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/1355915.1597794204@sss.pgh.pa.us
Commit a477bfc1d fixed eval_const_expressions() to ensure that it
didn't generate unnecessary RelabelType nodes, but I failed to notice
that some other places in the planner had the same issue. Really
noplace in the planner should be using plain makeRelabelType(), for
fear of generating expressions that should be equal() to semantically
equivalent trees, but aren't.
An example is that because canonicalize_ec_expression() failed
to be careful about this, we could end up with an equivalence class
containing both a plain Const, and a Const-with-RelabelType
representing exactly the same value. So far as I can tell this led to
no visible misbehavior, but we did waste a bunch of cycles generating
and evaluating "Const = Const-with-RelabelType" to prove such entries
are redundant.
Hence, move the support function added by a477bfc1d to where it can
be more generally useful, and use it in the places where planner code
previously used makeRelabelType.
Back-patch to v12, like the previous patch. While I have no concrete
evidence of any real misbehavior here, it's certainly possible that
I overlooked a case where equivalent expressions that aren't equal()
could cause a user-visible problem. In any case carrying extra
RelabelType nodes through planning to execution isn't very desirable.
Discussion: https://postgr.es/m/1311836.1597781384@sss.pgh.pa.us
This view displays the usages of all the memory contexts of the server
process attached to the current session. This information is useful to
investigate the cause of backend-local memory bloat.
This information can be also collected by calling
MemoryContextStats(TopMemoryContext) via a debugger. But this technique
cannot be uesd in some environments because no debugger is available there.
And it outputs lots of text messages and it's not easy to analyze them.
So, pg_backend_memory_contexts view allows us to access to backend-local
memory contexts information more easily.
Bump catalog version.
Author: Atsushi Torikoshi, Fujii Masao
Reviewed-by: Tatsuhito Kasahara, Andres Freund, Daniel Gustafsson, Robert Haas, Michael Paquier
Discussion: https://postgr.es/m/72a656e0f71d0860161e0b3f67e4d771@oss.nttdata.com
When preparing a transaction xactCompletionCount needs to be
incremented, even though the transaction has not committed
yet. Otherwise the snapshot used within the transaction otherwise can
get reused outside of the prepared transaction. As GetSnapshotData()
does not include the current xid when building a snapshot, reuse would
not be correct.
Somewhat surprisingly the regression tests only rarely show incorrect
results without the fix. The reason for that is that often the
snapshot's xmax will be >= the backend xid, yielding a snapshot that
is correct, despite the bug.
I'm working on a reliable test for the bug, but it seems worth seeing
whether this fixes all the BF failures while I do.
Author: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/E1k7tGP-0005V0-5k@gemulon.postgresql.org
Previous commits made it faster/more scalable to compute snapshots. But not
building a snapshot is still faster. Now that GetSnapshotData() does not
maintain RecentGlobal* anymore, that is actually not too hard:
This commit introduces xactCompletionCount, which tracks the number of
top-level transactions with xids (i.e. which may have modified the database)
that completed in some form since the start of the server.
We can avoid rebuilding the snapshot's contents whenever the current
xactCompletionCount is the same as it was when the snapshot was
originally built. Currently this check happens while holding
ProcArrayLock. While it's likely possible to perform the check without
acquiring ProcArrayLock, it seems better to do that separately /
later, some careful analysis is required. Even with the lock this is a
significant win on its own.
On a smaller two socket machine this gains another ~1.03x, on a larger
machine the effect is roughly double (earlier patch version tested
though). If we were able to safely avoid the lock there'd be another
significant gain on top of that.
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Robert Haas <robertmhaas@gmail.com>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
If a commit or abort record includes "dropped relfilenodes", then replaying
the record will remove data files. That is surely a "special rel update",
but the records were not marked as such. Fix that, teach pg_rewind to
expect and ignore them, and add a test case to cover it.
It's always been like this, but no backporting for fear of breaking
existing applications. If an application parsed the WAL but was not
handling commit/abort records, it would stop working. That might be a good
thing if it really needed to handle the dropped rels, but it will be caught
when the application is updated to work with PostgreSQL v14 anyway.
Discussion: https://www.postgresql.org/message-id/07b33e2c-46a6-86a1-5f9e-a7da73fddb95%40iki.fi
Reviewed-by: Amit Kapila, Michael Paquier
xact.h included utils/datetime.h, which cannot be used in the frontend
(it includes fmgr.h, which needs Datum). But xact.h only needs the
definition of TimestampTz from it, which is available directly in
datatypes/timestamp.h. Change xact.h to include that instead of
utils/datetime.h, so that it can be used in client programs.
This bug, recently introduced in 941697c3c1, at least lead to vacuum
failing because it found tuples inserted by a running transaction, but
below the freeze limit. The freeze limit in turn is directly affected
by the aforementioned bug.
Thanks to Tom Lane figuring how to make the bug reproducible.
We should add a few more assertions to make sure this type of bug
isn't as hard to notice, but it's not yet clear how to best do so.
Co-Diagnosed-By: Tom Lane <tgl@sss.pgh.pa.us>
Author: Andres Freund <andres@anarazel.de>
Discussion: https://postgr.es/m/1013484.1597609043@sss.pgh.pa.us
Reuse cautionary language from src/test/ssl/README in
src/test/kerberos/README. SLRUs have had access to six-character
segments names since commit 73c986adde,
and recovery stopped calling HeapTupleHeaderAdvanceLatestRemovedXid() in
commit 558a9165e0. The other corrections
are more self-evident.
The SimpleLruTruncate() header comment states the new coding rule. To
achieve this, add locktype "frozenid" and two LWLocks. This closes a
rare opportunity for data loss, which manifested as "apparent
wraparound" or "could not access status of transaction" errors. Data
loss is more likely in pg_multixact, due to released branches' thin
margin between multiStopLimit and multiWrapLimit. If a user's physical
replication primary logged ": apparent wraparound" messages, the user
should rebuild standbys of that primary regardless of symptoms. At less
risk is a cluster having emitted "not accepting commands" errors or
"must be vacuumed" warnings at some point. One can test a cluster for
this data loss by running VACUUM FREEZE in every database. Back-patch
to 9.5 (all supported versions).
Discussion: https://postgr.es/m/20190218073103.GA1434723@rfd.leadboat.com
nodeSubplan.c expects that the testexpr for a hashable ANY SubPlan
has the form of one or more OpExprs whose LHS is an expression of the
outer query's, while the RHS is an expression over Params representing
output columns of the subquery. However, the planner only went as far
as verifying that the clauses were all binary OpExprs. This works
99.99% of the time, because the clauses have the right shape when
emitted by the parser --- but it's possible for function inlining to
break that, as reported by PegoraroF10. To fix, teach the planner
to check that the LHS and RHS contain the right things, or more
accurately don't contain the wrong things. Given that this has been
broken for years without anyone noticing, it seems sufficient to just
give up hashing when it happens, rather than go to the trouble of
commuting the clauses back again (which wouldn't necessarily work
anyway).
While poking at that, I also noticed that nodeSubplan.c had a baked-in
assumption that the number of hash clauses is identical to the number
of subquery output columns. Again, that's fine as far as parser output
goes, but it's not hard to break it via function inlining. There seems
little reason for that assumption though --- AFAICS, the only thing
it's buying us is not having to store the number of hash clauses
explicitly. Adding code to the planner to reject such cases would take
more code than getting nodeSubplan.c to cope, so I fixed it that way.
This has been broken for as long as we've had hashable SubPlans,
so back-patch to all supported branches.
Discussion: https://postgr.es/m/1549209182255-0.post@n3.nabble.com
Similar to the previous changes this increases the chance that data
frequently needed by GetSnapshotData() stays in l2 cache. In many
workloads subtransactions are very rare, and this makes the check for
that considerably cheaper.
As this removes the last member of PGXACT, there is no need to keep it
around anymore.
On a larger 2 socket machine this and the two preceding commits result
in a ~1.07x performance increase in read-only pgbench. For read-heavy
mixed r/w workloads without row level contention, I see about 1.1x.
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Robert Haas <robertmhaas@gmail.com>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
Similar to the previous commit this increases the chance that data
frequently needed by GetSnapshotData() stays in l2 cache. As we now
take care to not unnecessarily write to ProcGlobal->vacuumFlags, there
should be very few modifications to the ProcGlobal->vacuumFlags array.
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Robert Haas <robertmhaas@gmail.com>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
The new array contains the xids for all connected backends / in-use
PGPROC entries in a dense manner (in contrast to the PGPROC/PGXACT
arrays which can have unused entries interspersed).
This improves performance because GetSnapshotData() always needs to
scan the xids of all live procarray entries and now there's no need to
go through the procArray->pgprocnos indirection anymore.
As the set of running top-level xids changes rarely, compared to the
number of snapshots taken, this substantially increases the likelihood
of most data required for a snapshot being in l2 cache. In
read-mostly workloads scanning the xids[] array will sufficient to
build a snapshot, as most backends will not have an xid assigned.
To keep the xid array dense ProcArrayRemove() needs to move entries
behind the to-be-removed proc's one further up in the array. Obviously
moving array entries cannot happen while a backend sets it
xid. I.e. locking needs to prevent that array entries are moved while
a backend modifies its xid.
To avoid locking ProcArrayLock in GetNewTransactionId() - a fairly hot
spot already - ProcArrayAdd() / ProcArrayRemove() now needs to hold
XidGenLock in addition to ProcArrayLock. Adding / Removing a procarray
entry is not a very frequent operation, even taking 2PC into account.
Due to the above, the dense array entries can only be read or modified
while holding ProcArrayLock and/or XidGenLock. This prevents a
concurrent ProcArrayRemove() from shifting the dense array while it is
accessed concurrently.
While the new dense array is very good when needing to look at all
xids it is less suitable when accessing a single backend's xid. In
particular it would be problematic to have to acquire a lock to access
a backend's own xid. Therefore a backend's xid is not just stored in
the dense array, but also in PGPROC. This also allows a backend to
only access the shared xid value when the backend had acquired an
xid.
The infrastructure added in this commit will be used for the remaining
PGXACT fields in subsequent commits. They are kept separate to make
review easier.
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Robert Haas <robertmhaas@gmail.com>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
Up to now, upon receipt of a SIGTERM ("smart shutdown" command), the
postmaster has immediately killed all "optional" background processes,
and subsequently refused to launch new ones while it's waiting for
foreground client processes to exit. No doubt this seemed like an OK
policy at some point; but it's a pretty bad one now, because it makes
for a seriously degraded environment for the remaining clients:
* Parallel queries are killed, and new ones fail to launch. (And our
parallel-query infrastructure utterly fails to deal with the case
in a reasonable way --- it just hangs waiting for workers that are
not going to arrive. There is more work needed in that area IMO.)
* Autovacuum ceases to function. We can tolerate that for awhile,
but if bulk-update queries continue to run in the surviving client
sessions, there's eventually going to be a mess. In the worst case
the system could reach a forced shutdown to prevent XID wraparound.
* The bgwriter and walwriter are also stopped immediately, likely
resulting in performance degradation.
Hence, let's rearrange things so that the only immediate change in
behavior is refusing to let in new normal connections. Once the last
normal connection is gone, shut everything down as though we'd received
a "fast" shutdown. To implement this, remove the PM_WAIT_BACKUP and
PM_WAIT_READONLY states, instead staying in PM_RUN or PM_HOT_STANDBY
while normal connections remain. A subsidiary state variable tracks
whether or not we're letting in new connections in those states.
This also allows having just one copy of the logic for killing child
processes in smart and fast shutdown modes. I moved that logic into
PostmasterStateMachine() by inventing a new state PM_STOP_BACKENDS.
Back-patch to 9.6 where parallel query was added. In principle
this'd be a good idea in 9.5 as well, but the risk/reward ratio
is not as good there, since lack of autovacuum is not a problem
during typical uses of smart shutdown.
Per report from Bharath Rupireddy.
Patch by me, reviewed by Thomas Munro
Discussion: https://postgr.es/m/CALj2ACXAZ5vKxT9P7P89D87i3MDO9bfS+_bjMHgnWJs8uwUOOw@mail.gmail.com
Now that xmin isn't needed for GetSnapshotData() anymore, it leads to
unnecessary cacheline ping-pong to have it in PGXACT, as it is updated
considerably more frequently than the other PGXACT members.
After the changes in dc7420c2c9, this is a very straight-forward change.
For highly concurrent, snapshot acquisition heavy, workloads this change alone
can significantly increase scalability. E.g. plain pgbench on a smaller 2
socket machine gains 1.07x for read-only pgbench, 1.22x for read-only pgbench
when submitting queries in batches of 100, and 2.85x for batches of 100
'SELECT';. The latter numbers are obviously not to be expected in the
real-world, but micro-benchmark the snapshot computation
scalability (previously spending ~80% of the time in GetSnapshotData()).
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Robert Haas <robertmhaas@gmail.com>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
When a table is scanned by heapam_index_build_range_scan (née
IndexBuildHeapScan) and the table lock being held allows concurrent data
changes, it is possible for new HOT chains to sprout in a page that were
unknown when the scan of a page happened. This leads to an error such
as
ERROR: failed to find parent tuple for heap-only tuple at (X,Y) in table "tbl"
because the root tuple was not present when we first obtained the list
of the page's root tuples. This can be fixed by re-obtaining the list
of root tuples, if we see that a heap-only tuple appears to point to a
non-existing root.
This was reported by Anastasia as occurring for BRIN summarization
(which exists since 9.5), but I think it could theoretically also happen
with CREATE INDEX CONCURRENTLY (much older) or REINDEX CONCURRENTLY
(very recent). It seems a happy coincidence that BRIN forces us to
backpatch this all the way to 9.5.
Reported-by: Anastasia Lubennikova <a.lubennikova@postgrespro.ru>
Diagnosed-by: Anastasia Lubennikova <a.lubennikova@postgrespro.ru>
Co-authored-by: Anastasia Lubennikova <a.lubennikova@postgrespro.ru>
Co-authored-by: Álvaro Herrera <alvherre@alvh.no-ip.org>
Discussion: https://postgr.es/m/602d8487-f0b2-5486-0088-0f372b2549fa@postgrespro.ru
Backpatch: 9.5 - master
To make GetSnapshotData() more scalable, it cannot not look at at each proc's
xmin: While snapshot contents do not need to change whenever a read-only
transaction commits or a snapshot is released, a proc's xmin is modified in
those cases. The frequency of xmin modifications leads to, particularly on
higher core count systems, many cache misses inside GetSnapshotData(), despite
the data underlying a snapshot not changing. That is the most
significant source of GetSnapshotData() scaling poorly on larger systems.
Without accessing xmins, GetSnapshotData() cannot calculate accurate horizons /
thresholds as it has so far. But we don't really have to: The horizons don't
actually change that much between GetSnapshotData() calls. Nor are the horizons
actually used every time a snapshot is built.
The trick this commit introduces is to delay computation of accurate horizons
until there use and using horizon boundaries to determine whether accurate
horizons need to be computed.
The use of RecentGlobal[Data]Xmin to decide whether a row version could be
removed has been replaces with new GlobalVisTest* functions. These use two
thresholds to determine whether a row can be pruned:
1) definitely_needed, indicating that rows deleted by XIDs >= definitely_needed
are definitely still visible.
2) maybe_needed, indicating that rows deleted by XIDs < maybe_needed can
definitely be removed
GetSnapshotData() updates definitely_needed to be the xmin of the computed
snapshot.
When testing whether a row can be removed (with GlobalVisTestIsRemovableXid())
and the tested XID falls in between the two (i.e. XID >= maybe_needed && XID <
definitely_needed) the boundaries can be recomputed to be more accurate. As it
is not cheap to compute accurate boundaries, we limit the number of times that
happens in short succession. As the boundaries used by
GlobalVisTestIsRemovableXid() are never reset (with maybe_needed updated by
GetSnapshotData()), it is likely that further test can benefit from an earlier
computation of accurate horizons.
To avoid regressing performance when old_snapshot_threshold is set (as that
requires an accurate horizon to be computed), heap_page_prune_opt() doesn't
unconditionally call TransactionIdLimitedForOldSnapshots() anymore. Both the
computation of the limited horizon, and the triggering of errors (with
SetOldSnapshotThresholdTimestamp()) is now only done when necessary to remove
tuples.
This commit just removes the accesses to PGXACT->xmin from
GetSnapshotData(), but other members of PGXACT residing in the same
cache line are accessed. Therefore this in itself does not result in a
significant improvement. Subsequent commits will take advantage of the
fact that GetSnapshotData() now does not need to access xmins anymore.
Note: This contains a workaround in heap_page_prune_opt() to keep the
snapshot_too_old tests working. While that workaround is ugly, the tests
currently are not meaningful, and it seems best to address them separately.
Author: Andres Freund <andres@anarazel.de>
Reviewed-By: Robert Haas <robertmhaas@gmail.com>
Reviewed-By: Thomas Munro <thomas.munro@gmail.com>
Reviewed-By: David Rowley <dgrowleyml@gmail.com>
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
If a page range is desummarized at just the right time concurrently with
an index walk, BRIN would raise an error indicating index corruption.
This is scary and unhelpful; silently returning that the page range is
not summarized is sufficient reaction.
This bug was introduced by commit 975ad4e602 as additional protection
against a bug whose actual fix was elsewhere. Backpatch equally.
Reported-By: Anastasia Lubennikova <a.lubennikova@postgrespro.ru>
Diagnosed-By: Alexander Lakhin <exclusion@gmail.com>
Discussion: https://postgr.es/m/2588667e-d07d-7e10-74e2-7e1e46194491@postgrespro.ru
Backpatch: 9.5 - master
This had gotten a little disjointed over time, and some of the grammar
was sloppy. Rewrite for more clarity.
In passing, re-pgindent some recently added comments.
No code changes.
The reason for doing so is that a subsequent commit will need that to
avoid wraparound issues. As the subsequent change is large this was
split out for easier review.
The reason this is not a perfect straight-forward change is that we do
not want track 64bit xids in the procarray or the WAL. Therefore we
need to advance lastestCompletedXid in relation to 32 bit xids. The
code for that is now centralized in MaintainLatestCompletedXid*.
Author: Andres Freund
Reviewed-By: Thomas Munro, Robert Haas, David Rowley
Discussion: https://postgr.es/m/20200301083601.ews6hz5dduc3w2se@alap3.anarazel.de
Including Full in variable names duplicates the type information and
leads to overly long names. As FullTransactionId cannot accidentally
be casted to TransactionId that does not seem necessary.
Author: Andres Freund
Discussion: https://postgr.es/m/20200724011143.jccsyvsvymuiqfxu@alap3.anarazel.de
They are equivalent, except that StrNCpy() zero-fills the entire
destination buffer instead of providing just one trailing zero. For
all but a tiny number of callers, that's just overhead rather than
being desirable.
Remove StrNCpy() as it is now unused.
In some cases, namestrcpy() is the more appropriate function to use.
While we're here, simplify the API of namestrcpy(): Remove the return
value, don't check for NULL input. Nothing was using that anyway.
Also, remove a few unused name-related functions.
Reviewed-by: Tom Lane <tgl@sss.pgh.pa.us>
Discussion: https://www.postgresql.org/message-id/flat/44f5e198-36f6-6cdb-7fa9-60e34784daae%402ndquadrant.com
This is like CVE-2018-1058 commit
582edc369c. Today, a malicious user of a
publisher or subscriber database can invoke arbitrary SQL functions
under an identity running replication, often a superuser. This fix may
cause "does not exist" or "no schema has been selected to create in"
errors in a replication process. After upgrading, consider watching
server logs for these errors. Objects accruing schema qualification in
the wake of the earlier commit are unlikely to need further correction.
Back-patch to v10, which introduced logical replication.
Security: CVE-2020-14349
Hostile objects located within the installation-time search_path could
capture references in an extension's installation or upgrade script.
If the extension is being installed with superuser privileges, this
opens the door to privilege escalation. While such hazards have existed
all along, their urgency increases with the v13 "trusted extensions"
feature, because that lets a non-superuser control the installation path
for a superuser-privileged script. Therefore, make a number of changes
to make such situations more secure:
* Tweak the construction of the installation-time search_path to ensure
that references to objects in pg_catalog can't be subverted; and
explicitly add pg_temp to the end of the path to prevent attacks using
temporary objects.
* Disable check_function_bodies within installation/upgrade scripts,
so that any security gaps in SQL-language or PL-language function bodies
cannot create a risk of unwanted installation-time code execution.
* Adjust lookup of type input/receive functions and join estimator
functions to complain if there are multiple candidate functions. This
prevents capture of references to functions whose signature is not the
first one checked; and it's arguably more user-friendly anyway.
* Modify various contrib upgrade scripts to ensure that catalog
modification queries are executed with secure search paths. (These
are in-place modifications with no extension version changes, since
it is the update process itself that is at issue, not the end result.)
Extensions that depend on other extensions cannot be made fully secure
by these methods alone; therefore, revert the "trusted" marking that
commit eb67623c9 applied to earthdistance and hstore_plperl, pending
some better solution to that set of issues.
Also add documentation around these issues, to help extension authors
write secure installation scripts.
Patch by me, following an observation by Andres Freund; thanks
to Noah Misch for review.
Security: CVE-2020-14350
Testing that an unsigned variable is >= 0 is pretty pointless,
as noted by Coverity and numerous buildfarm members.
In passing, add comment about new uses of "volatile" --- Coverity
doesn't much like that either, but it seems probably necessary.
Ashutosh Bapat noticed that when logical walsender needs to wait for
WAL, and it realizes that it must send a keepalive message to
walreceiver to update the sent-LSN, which *does not* request a reply
from walreceiver, it wrongly sets the flag that it's going to wait for
that reply. That means that any future would-be sender of feedback
messages ends up not sending a feedback message, because they all
believe that a reply is expected.
With built-in logical replication there's not much harm in this, because
WalReceiverMain will send a ping-back every wal_receiver_timeout/2
anyway; but with other logical replication systems (e.g. pglogical) it
can cause significant pain.
This problem was introduced in commit 41d5f8ad73, where the
request-reply flag was changed from true to false to WalSndKeepalive,
without at the same time removing the line that sets
waiting_for_ping_response.
Just removing that line would be a sufficient fix, but it seems better
to shift the responsibility of setting the flag to WalSndKeepalive
itself instead of requiring caller to do it; this is clearly less
error-prone.
Author: Álvaro Herrera <alvherre@alvh.no-ip.org>
Reported-by: Ashutosh Bapat <ashutosh.bapat@2ndquadrant.com>
Backpatch: 9.5 and up
Discussion: https://postgr.es/m/20200806225558.GA22401@alvherre.pgsql
Instead of serializing the transaction to disk after reaching the
logical_decoding_work_mem limit in memory, we consume the changes we have
in memory and invoke stream API methods added by commit 45fdc9738b.
However, sometimes if we have incomplete toast or speculative insert we
spill to the disk because we can't generate the complete tuple and stream.
And, as soon as we get the complete tuple we stream the transaction
including the serialized changes.
We can do this incremental processing thanks to having assignments
(associating subxact with toplevel xacts) in WAL right away, and
thanks to logging the invalidation messages at each command end. These
features are added by commits 0bead9af48 and c55040ccd0 respectively.
Now that we can stream in-progress transactions, the concurrent aborts
may cause failures when the output plugin consults catalogs (both system
and user-defined).
We handle such failures by returning ERRCODE_TRANSACTION_ROLLBACK
sqlerrcode from system table scan APIs to the backend or WALSender
decoding a specific uncommitted transaction. The decoding logic on the
receipt of such a sqlerrcode aborts the decoding of the current
transaction and continue with the decoding of other transactions.
We have ReorderBufferTXN pointer in each ReorderBufferChange by which we
know which xact it belongs to. The output plugin can use this to decide
which changes to discard in case of stream_abort_cb (e.g. when a subxact
gets discarded).
We also provide a new option via SQL APIs to fetch the changes being
streamed.
Author: Dilip Kumar, Tomas Vondra, Amit Kapila, Nikhil Sontakke
Reviewed-by: Amit Kapila, Kuntal Ghosh, Ajin Cherian
Tested-by: Neha Sharma, Mahendra Singh Thalor and Ajin Cherian
Discussion: https://postgr.es/m/688b0b7f-2f6c-d827-c27b-216a8e3ea700@2ndquadrant.com
Make the nbtree page split REDO routine consistent with original
execution in its approach to acquiring and releasing buffer locks (at
least for pages on the tree level of the page being split). This brings
btree_xlog_split() in line with btree_xlog_unlink_page(), which was
taught to couple buffer locks by commit 9a9db08a.
Note that the precise order in which we both acquire and release sibling
buffer locks in btree_xlog_split() now matches original execution
exactly (the precise order in which the locks are released probably
doesn't matter much, but we might as well be consistent about it).
The rule for nbtree REDO routines from here on is that same-level locks
should be acquired in an order that's consistent with original
execution. It's not practical to have a similar rule for cross-level
page locks, since for the most part original execution holds those locks
for a period that spans multiple atomic actions/WAL records. It's also
not necessary, because clearly the cross-level lock coupling is only
truly needed during original execution because of the presence of
concurrent inserters.
This is not a bug fix (unlike the similar aforementioned commit, commit
9a9db08a). The immediate reason to tighten things up in this area is to
enable an upcoming enhancement to contrib/amcheck that allows it to
verify that sibling links are in agreement with only an AccessShareLock
(this check produced false positives when run on a replica server on
account of the inconsistency fixed by this commit). But that's not the
only reason to be stricter here.
It is generally useful to make locking on replicas be as close to what
happens during original execution as practically possible. It makes it
less likely that hard to catch bugs will slip in in the future. The
previous state of affairs seems to be a holdover from before the
introduction of Hot Standby, when buffer lock acquisitions during
recovery were totally unnecessary. See also: commit 3bbf668d, which
tightened things up in this area a few years after the introduction of
Hot Standby.
Discussion: https://postgr.es/m/CAH2-Wz=465cJj11YXD9RKH8z=nhQa2dofOZ_23h67EXUGOJ00Q@mail.gmail.com
Make the nbtree page split REDO routine variable names consistent with
_bt_split() (which handles the original execution of page splits).
These names make the code easier to follow by making the distinction
between the original page and the left half of the split clear. (The
left half of the split page is a temp page that REDO creates to replace
the origpage contents.)
Also reduce the elevel used when adding a new high key to the temp page
from PANIC to ERROR to be consistent. We already only raise an ERROR
when data item PageAddItem() temp page calls fail.
Commit 13838740f fixed some issues with step generation in partition
pruning, but there was yet another one: get_steps_using_prefix() assumes
that clauses in the passed-in prefix list are sorted in ascending order
of their partition key numbers, but the caller failed to ensure this for
range partitioning, which led to an assertion failure in debug builds.
Adjust the caller function to arrange the clauses in the prefix list in
the required order for range partitioning.
Back-patch to v11, like the previous commit.
Patch by me, reviewed by Amit Langote.
Discussion: https://postgr.es/m/CAPmGK16jkXiFG0YqMbU66wte-oJTfW6D1HaNvQf%3D%2B5o9%3Dm55wQ%40mail.gmail.com
9bdb300de modified the EXPLAIN output for Hash Aggregate to show details
from parallel workers. However, it neglected to consider that a given
parallel worker may not have assisted with the given Hash Aggregate. This
can occur when workers fail to start or during Parallel Append with
enable_partitionwise_join enabled when only a single worker is working on
a non-parallel aware sub-plan. It could also happen if a worker simply
wasn't fast enough to get any work done before other processes went and
finished all the work.
The bogus output came from the fact that ExplainOpenWorker() skipped
showing any details for non-initialized workers but show_hashagg_info()
did show details from the worker. This meant that the worker properties
that were shown were not properly attributed to the worker that they
belong to.
In passing, we also now don't show Hash Aggregate properties for the
leader process when it did not contribute any work to the Hash Aggregate.
This can occur either during Parallel Append when only a parallel worker
worked on a given sub plan or with parallel_leader_participation set to
off. This aims to make the behavior of Hash Aggregate's EXPLAIN output
more similar to Sort's.
Reported-by: Justin Pryzby
Discussion: https://postgr.es/m/20200805012105.GZ28072%40telsasoft.com
Backpatch-through: 13, where the original breakage was introduced
This seems more correct, because other before_shmem_exit calls may
expect the infrastructure that is needed to run queries and access the
database to be working, and also because this cleanup has nothing to
do with shared memory.
There are no known user-visible consequences to this, though, apart
from what was previous fixed by commit
303640199d and back-patched as commit
bcbc27251d and commit
f7013683d9, so for now, no back-patch.
Bharath Rupireddy
Discussion: http://postgr.es/m/CALj2ACWk7j4F2v2fxxYfrroOF=AdFNPr1WsV+AGtHAFQOqm_pw@mail.gmail.com
Since we no longer require AccessExclusiveLock to add a partition,
the executor may see that a partitioned table has more partitions
than the planner saw. ExecCreatePartitionPruneState's code for
matching up the partition lists in such cases was faulty, and would
misbehave if the planner had successfully pruned any partitions from
the query. (Thus, trouble would occur only if a partition addition
happens concurrently with a query that uses both static and dynamic
partition pruning.) This led to an Assert failure in debug builds,
and probably to crashes or query misbehavior in production builds.
To repair the bug, just explicitly skip zeroes in the plan's
relid_map[] list. I also made some cosmetic changes to make the code
more readable (IMO anyway). Also, convert the cross-checking Assert
to a regular test-and-elog, since it's now apparent that this logic
is more fragile than one would like.
Currently, there's no way to repeatably exercise this code, except
with manual use of a debugger to stop the backend between planning
and execution. Hence, no test case in this patch. We oughta do
something about that testability gap, but that's for another day.
Amit Langote and Tom Lane, per report from Justin Pryzby. Oversight
in commit 898e5e329; backpatch to v12 where that appeared.
Discussion: https://postgr.es/m/20200802181131.GA27754@telsasoft.com
Currently, page unlink leaves remaining items "as is", but replay of
corresponding WAL-record re-initializes page leaving it with no items.
For the sake of consistency, this commit makes primary delete all the items
during page unlink as well.
Thanks to this change, we now don't mask contents of deleted btree page for
WAL consistency checking.
Discussion: https://postgr.es/m/CAPpHfdt_OTyQpXaPJcWzV2N-LNeNJseNB-K_A66qG%3DL518VTFw%40mail.gmail.com
Author: Alexander Korotkov
Reviewed-by: Peter Geoghegan
It was possible for the logic used by backward scans (which must reason
about concurrent page splits/deletions in its own peculiar way) to
become confused when running on a replica. Concurrent replay of a WAL
record that describes the second phase of page deletion could cause
_bt_walk_left() to get confused. btree_xlog_unlink_page() simply failed
to adhere to the same locking protocol that we use on the primary, which
is obviously wrong once you consider these two disparate functions
together. This bug is present in all stable branches.
More concretely, the problem was that nothing stopped _bt_walk_left()
from observing inconsistencies between the deletion's target page and
its original sibling pages when running on a replica. This is true even
though the second phase of page deletion is supposed to work as a single
atomic action. Queries running on replicas raised "could not find left
sibling of block %u in index %s" can't-happen errors when they went back
to their scan's "original" page and observed that the page has not been
marked deleted (even though it really was concurrently deleted).
There is no evidence that this actually happened in the real world. The
issue came to light during unrelated feature development work. Note
that _bt_walk_left() is the only code that cares about the difference
between a half-dead page and a fully deleted page that isn't also
exclusively used by nbtree VACUUM (unless you include contrib/amcheck
code). It seems very likely that backward scans are the only thing that
could become confused by the inconsistency. Even amcheck's complex
bt_right_page_check_scankey() dance was unaffected.
To fix, teach btree_xlog_unlink_page() to lock the left sibling, target,
and right sibling pages in that order before releasing any locks (just
like _bt_unlink_halfdead_page()). This is the simplest possible
approach. There doesn't seem to be any opportunity to be more clever
about lock acquisition in the REDO routine, and it hardly seems worth
the trouble in any case.
This fix might enable contrib/amcheck verification of leaf page sibling
links with only an AccessShareLock on the relation. An amcheck patch
from Andrey Borodin was rejected back in January because it clashed with
btree_xlog_unlink_page()'s lax approach to locking pages. It now seems
likely that the real problem was with btree_xlog_unlink_page(), not the
patch.
This is a low severity, low likelihood bug, so no backpatch.
Author: Michail Nikolaev
Diagnosed-By: Michail Nikolaev
Discussion: https://postgr.es/m/CANtu0ohkR-evAWbpzJu54V8eCOtqjJyYp3PQ_SGoBTRGXWhWRw@mail.gmail.com
Add a documenting assertion that's similar to the nearby assertion added
by commit cd8c73a3. This conveys that the entire call to _bt_pagedel()
does no work if it isn't possible to get a descent stack for the initial
scanblkno page.
This is useful for monitoring purposes with log parsing. Similarly to
pg_stat_activity, the leader's PID is shown only for active parallel
workers, minimizing the log footprint for the leaders as the equivalent
shared memory field is set as long as a backend is alive.
Author: Justin Pryzby
Reviewed-by: Álvaro Herrera, Michael Paquier, Julien Rouhaud, Tom Lane
Discussion: https://postgr.es/m/20200315111831.GA21492@telsasoft.com
We have edge-case bugs when assigning values in the last few dozen pages
before the wrap limit. We may introduce similar bugs in the future. At
default BLCKSZ, this makes such bugs unreachable outside of single-user
mode. Also, when VACUUM began to consume mxacts, multiStopLimit did not
change to compensate.
pg_upgrade may fail on a cluster that was already printing "must be
vacuumed" warnings. Follow the warning's instructions to clear the
warning, then run pg_upgrade again. One can still, peacefully consume
98% of XIDs or mxacts, so DBAs need not change routine VACUUM settings.
Discussion: https://postgr.es/m/20200621083513.GA3074645@rfd.leadboat.com
This allows AM-specific knowledge to be applied during creation of
pg_amop and pg_amproc entries. Specifically, the AM knows better than
core code which entries to consider as required or optional. Giving
the latter entries the appropriate sort of dependency allows them to
be dropped without taking out the whole opclass or opfamily; which
is something we'd like to have to correct obsolescent entries in
extensions.
This callback also opens the door to performing AM-specific validity
checks during opclass creation, rather than hoping than an opclass
developer will remember to test with "amvalidate". For the most part
I've not actually added any such checks yet; that can happen in a
follow-on patch. (Note that we shouldn't remove any tests from
"amvalidate", as those are still needed to cross-check manually
constructed entries in the initdb data. So adding tests to
"amadjustmembers" will be somewhat duplicative, but it seems like
a good idea anyway.)
Patch by me, reviewed by Alexander Korotkov, Hamid Akhtar, and
Anastasia Lubennikova.
Discussion: https://postgr.es/m/4578.1565195302@sss.pgh.pa.us
When doing multiple insertions in pg_shdepend for the copy of
dependencies from a template database in CREATE DATABASE, the same
number of slots would have been created and used all the time. As the
number of items to insert is not known in advance, this makes most of
the slots created for nothing. This improves the slot handling so as
slot creation only happens when needed, minimizing the overhead of the
operation.
Author: Michael Paquier
Reviewed-by: Daniel Gustafsson
Discussion: https://postgr.es/m/20200731024148.GB3317@paquier.xyz
When reading the code it's not obvious when one should prefer dynahash
over simplehash and vice-versa, so, for programmer-friendliness, add
comments to inform that decision.
Show sample simplehash method signatures.
Author: James Coleman <jtc331@gmail.com>
Discussion: https://postgr.es/m/CAAaqYe_dOF39gAJ8rL-a3YO3Qo96MHMRQ2whFjK5ZcU6YvMQSA%40mail.gmail.com
If a base type supports typmods, its array type does too, with the
same interpretation. Hence changes in pg_type.typmodin/typmodout
must be propagated to the array type.
While here, improve AlterTypeRecurse to not recurse to domains if
there is nothing we'd need to change.
Oversight in fe30e7ebf. Back-patch to v13 where that came in.
The new hlCover() algorithm that I introduced in commit c9b0c678d
turns out to potentially take O(N^2) or worse time on long documents,
if there are many occurrences of individual query words but few or no
substrings that actually satisfy the query. (One way to hit this
behavior is with a "common_word & rare_word" type of query.) This
seems unavoidable given the original goal of checking every substring
of the document, so we have to back off that idea. Fortunately, it
seems unlikely that anyone would really want headlines spanning all of
a long document, so we can avoid the worse-than-linear behavior by
imposing a maximum length of substring that we'll consider.
For now, just hard-wire that maximum length as a multiple of max_words
times max_fragments. Perhaps at some point somebody will argue for
exposing it as a ts_headline parameter, but I'm hesitant to make such
a feature addition in a back-patched bug fix.
I also noted that the hlFirstIndex() function I'd added in that
commit was unnecessarily stupid: it really only needs to check whether
a HeadlineWordEntry's item pointer is null or not. This wouldn't make
all that much difference in typical cases with queries having just
a few terms, but a cycle shaved is a cycle earned.
In addition, add a CHECK_FOR_INTERRUPTS call in TS_execute_recurse.
This ensures that hlCover's loop is cancellable if it manages to take
a long time, and it may protect some other TS_execute callers as well.
Back-patch to 9.6 as the previous commit was. I also chose to add the
CHECK_FOR_INTERRUPTS call to 9.5. The old hlCover() algorithm seems
to avoid the O(N^2) behavior, at least on the test case I tried, but
nonetheless it's not very quick on a long document.
Per report from Stephen Frost.
Discussion: https://postgr.es/m/20200724160535.GW12375@tamriel.snowman.net
Create an optional region in the main shared memory segment that can be
used to acquire and release "fast" DSM segments, and can benefit from
huge pages allocated at cluster startup time, if configured. Fall back
to the existing mechanisms when that space is full. The size is
controlled by a new GUC min_dynamic_shared_memory, defaulting to 0.
Main region DSM segments initially contain whatever garbage the memory
held last time they were used, rather than zeroes. That change revealed
that DSA areas failed to initialize themselves correctly in memory that
wasn't zeroed first, so fix that problem.
Discussion: https://postgr.es/m/CA%2BhUKGLAE2QBv-WgGp%2BD9P_J-%3Dyne3zof9nfMaqq1h3EGHFXYQ%40mail.gmail.com
Avoid repeatedly calling lseek(SEEK_END) during recovery by caching
the size of each fork. For now, we can't use the same technique in
other processes, because we lack a shared invalidation mechanism.
Do this by generalizing the pre-existing caching used by FSM and VM
to support all forks.
Discussion: https://postgr.es/m/CAEepm%3D3SSw-Ty1DFcK%3D1rU-K6GSzYzfdD4d%2BZwapdN7dTa6%3DnQ%40mail.gmail.com
For pg_attribute, this allows to insert at once a full set of attributes
for a relation (roughly 15% of WAL reduction in extreme cases). For
pg_shdepend, this reduces the work done when creating new shared
dependencies from a database template. The number of slots used for the
insertion is capped at 64kB of data inserted for both, depending on the
number of items to insert and the length of the rows involved.
More can be done for other catalogs, like pg_depend. This part requires
a different approach as the number of slots to use depends also on the
number of entries discarded as pinned dependencies. This is also
related to the rework or dependency handling for ALTER TABLE and CREATE
TABLE, mainly.
Author: Daniel Gustafsson
Reviewed-by: Andres Freund, Michael Paquier
Discussion: https://postgr.es/m/20190213182737.mxn6hkdxwrzgxk35@alap3.anarazel.de
Create LatchWaitSet at backend startup time, and use it to implement
WaitLatch(). This avoids repeated epoll/kqueue setup and teardown
system calls.
Reorder SubPostmasterMain() slightly so that we restore the postmaster
pipe and Windows signal emulation before we reach InitPostmasterChild(),
to make this work in EXEC_BACKEND builds.
Reviewed-by: Kyotaro Horiguchi <horikyota.ntt@gmail.com>
Discussion: https://postgr.es/m/CA%2BhUKGJAC4Oqao%3DqforhNey20J8CiG2R%3DoBPqvfR0vOJrFysGw%40mail.gmail.com
Add a GUC that acts as a multiplier on work_mem. It gets applied when
sizing executor node hash tables that were previously size constrained
using work_mem alone.
The new GUC can be used to preferentially give hash-based nodes more
memory than the generic work_mem limit. It is intended to enable admin
tuning of the executor's memory usage. Overall system throughput and
system responsiveness can be improved by giving hash-based executor
nodes more memory (especially over sort-based alternatives, which are
often much less sensitive to being memory constrained).
The default value for hash_mem_multiplier is 1.0, which is also the
minimum valid value. This means that hash-based nodes continue to apply
work_mem in the traditional way by default.
hash_mem_multiplier is generally useful. However, it is being added now
due to concerns about hash aggregate performance stability for users
that upgrade to Postgres 13 (which added disk-based hash aggregation in
commit 1f39bce0). While the old hash aggregate behavior risked
out-of-memory errors, it is nevertheless likely that many users actually
benefited. Hash agg's previous indifference to work_mem during query
execution was not just faster; it also accidentally made aggregation
resilient to grouping estimate problems (at least in cases where this
didn't create destabilizing memory pressure).
hash_mem_multiplier can provide a certain kind of continuity with the
behavior of Postgres 12 hash aggregates in cases where the planner
incorrectly estimates that all groups (plus related allocations) will
fit in work_mem/hash_mem. This seems necessary because hash-based
aggregation is usually much slower when only a small fraction of all
groups can fit. Even when it isn't possible to totally avoid hash
aggregates that spill, giving hash aggregation more memory will reliably
improve performance (the same cannot be said for external sort
operations, which appear to be almost unaffected by memory availability
provided it's at least possible to get a single merge pass).
The PostgreSQL 13 release notes should advise users that increasing
hash_mem_multiplier can help with performance regressions associated
with hash aggregation. That can be taken care of by a later commit.
Author: Peter Geoghegan
Reviewed-By: Álvaro Herrera, Jeff Davis
Discussion: https://postgr.es/m/20200625203629.7m6yvut7eqblgmfo@alap3.anarazel.de
Discussion: https://postgr.es/m/CAH2-WzmD%2Bi1pG6rc1%2BCjc4V6EaFJ_qSuKCCHVnH%3DoruqD-zqow%40mail.gmail.com
Backpatch: 13-, where disk-based hash aggregation was introduced.
When fast promotion was supported in 9.3, non-fast promotion became
undocumented feature and it's basically not available for ordinary users.
However we decided not to remove non-fast promotion at that moment,
to leave it for a release or two for debugging purpose or as an emergency
method because fast promotion might have some issues, and then to
remove it later. Now, several versions were released since that decision
and there is no longer reason to keep supporting non-fast promotion.
Therefore this commit removes non-fast promotion.
Author: Fujii Masao
Reviewed-by: Hamid Akhtar, Kyotaro Horiguchi
Discussion: https://postgr.es/m/76066434-648f-f567-437b-54853b43398f@oss.nttdata.com
Use HyperLogLog to estimate the group cardinality in a spilled
partition. This estimate is used to choose the number of partitions if
we recurse.
The previous behavior was to use the number of tuples in a spilled
partition as the estimate for the number of groups, which lead to
overpartitioning. That could cause the number of batches to be much
higher than expected (with each batch being very small), which made it
harder to interpret EXPLAIN ANALYZE results.
Reviewed-by: Peter Geoghegan
Discussion: https://postgr.es/m/a856635f9284bc36f7a77d02f47bbb6aaf7b59b3.camel@j-davis.com
Backpatch-through: 13
There were various unnecessary differences between Hash Agg's EXPLAIN
ANALYZE output and Hash Join's. Here we modify the Hash Agg output so
that it's better aligned to Hash Join's.
The following changes have been made:
1. Start batches counter at 1 instead of 0.
2. Always display the "Batches" property, even when we didn't spill to
disk.
3. Use the text "Batches" instead of "HashAgg Batches" for text format.
4. Use the text "Memory Usage" instead of "Peak Memory Usage" for text
format.
5. Include "Batches" before "Memory Usage" in both text and non-text
formats.
In passing also modify the "Planned Partitions" property so that we show
it regardless of if the value is 0 or not for non-text EXPLAIN formats.
This was pointed out by Justin Pryzby and probably should have been part
of 40efbf870.
Reviewed-by: Justin Pryzby, Jeff Davis
Discussion: https://postgr.es/m/CAApHDvrshRnA6C0VFnu7Fb9TVvgGo80PUMm5+2DiaS1gEkPvtw@mail.gmail.com
Backpatch-through: 13, where HashAgg batching was introduced
Alvaro Herrera
Tom Lane
Alexander Korotkov
Fujii Masao
Michael Paquier
Amit Kapila
Thomas Munro
Robert Haas
Peter Eisentraut
Heikki Linnakangas
Jeff Davis
David Rowley
Etsuro Fujita
Noah Misch
Andres Freund
Magnus Hagander
Bruce Momjian
Peter Geoghegan