That was otherwise harmless, but tripped the new assertion in
PageGetSpecialPointer().
Reported by Amit Langote. Backpatch to 9.5, where the assertion was added.
Some of the older OS X critters in the buildfarm are failing regression,
with symptoms showing that a request for 100% sampling in BERNOULLI or
SYSTEM methods actually gets only around 50% of the table. gdb revealed
that the computation of the "cutoff" number was producing 0x7FFFFFFF
rather than the expected 0x100000000. Inspecting the assembly code,
it looks like gcc is trying to use lrint() instead of rint() and then
fumbling the conversion from long double to uint64. This seems like a
clear compiler bug, but assigning the intermediate result into a plain
double variable works around it, so let's just do that. (Another idea
would be to give up one bit of hash width so that we don't need to use
a uint64 cutoff, but let's see if this is enough.)
The original implementation of TABLESAMPLE modeled the tablesample method
API on index access methods, which wasn't a good choice because, without
specialized DDL commands, there's no way to build an extension that can
implement a TSM. (Raw inserts into system catalogs are not an acceptable
thing to do, because we can't undo them during DROP EXTENSION, nor will
pg_upgrade behave sanely.) Instead adopt an API more like procedural
language handlers or foreign data wrappers, wherein the only SQL-level
support object needed is a single handler function identified by having
a special return type. This lets us get rid of the supporting catalog
altogether, so that no custom DDL support is needed for the feature.
Adjust the API so that it can support non-constant tablesample arguments
(the original coding assumed we could evaluate the argument expressions at
ExecInitSampleScan time, which is undesirable even if it weren't outright
unsafe), and discourage sampling methods from looking at invisible tuples.
Make sure that the BERNOULLI and SYSTEM methods are genuinely repeatable
within and across queries, as required by the SQL standard, and deal more
honestly with methods that can't support that requirement.
Make a full code-review pass over the tablesample additions, and fix
assorted bugs, omissions, infelicities, and cosmetic issues (such as
failure to put the added code stanzas in a consistent ordering).
Improve EXPLAIN's output of tablesample plans, too.
Back-patch to 9.5 so that we don't have to support the original API
in production.
If there were no subtransactions (or multixacts) active, we would calculate
the oldestxid == next xid. That's correct, but if next XID happens to be
on the next pg_subtrans (pg_multixact) page, the page does not exist yet,
and SimpleLruTruncate will produce an "apparent wraparound" warning. The
warning is harmless in this case, but looks very alarming to users.
Backpatch to all supported versions. Patch and analysis by Thomas Munro.
Remove HeapScanDescData.rs_initblock, which wasn't being used for anything
in the final version of the patch.
Fix IndexBuildHeapScan so that it supports syncscan again; the patch
broke synchronous scanning for index builds by forcing rs_startblk
to zero even when the caller did not care about that and had asked
for syncscan.
Add some commentary and usage defenses to heap_setscanlimits().
Fix heapam so that asking for rs_numblocks == 0 does what you would
reasonably expect. As coded it amounted to requesting a whole-table
scan, because those "--x <= 0" tests on an unsigned variable would
behave surprisingly.
There was already a sanity-check in the other direction: if a page was
marked with WILL_INIT, it had to be initialized by the redo routine. It's
not strictly necessary for correctness that a page is marked with WILL_INIT
if it's going to be initialized at redo, but it's a missed optimization if
nothing else.
Fix a few instances of this issue in SP-GiST, where a block in WAL record
was not marked with WILL_INIT, but was in fact always initialized at redo.
We were creating a full-page image of the page unnecessarily in those
cases.
Backpatch to 9.5, where the new WILL_INIT flag was added.
This removes some info about support procedures being used, which was
obsoleted by commit db5f98ab4f, as well as add some more documentation
on how to create new opclasses using the Minmax infrastructure.
(Hopefully we can get something similar for Inclusion as well.)
In passing, fix some obsolete mentions of "mmtuples" in source code
comments.
Backpatch to 9.5, where BRIN was introduced.
These variants used the old-style 'n'/' ' NULL indicators. The new-style
functions have been available since version 8.1. That should be long enough
that if there is still any old external code using these functions, they
can just switch to the new functions without worrying about backwards
compatibility
Peter Geoghegan
Patch by David Rowley. Backpatch to 9.5, as some of the calls were new in
9.5, and keeping the code in sync with master makes future backpatching
easier.
XLogFileCopy() was changed heavily in commit de76884. However it was
partially reverted in commit 7abc685 and most of those changes to
XLogFileCopy() were no longer needed. Then commit 7cbee7c removed
those unnecessary code, but XLogFileCopy() looked different in master
and 9.4 though the contents are almost the same.
This patch makes XLogFileCopy() look the same in master and back-branches,
which makes back-patching easier, per discussion on pgsql-hackers.
Back-patch to 9.5.
Discussion: 55760844.7090703@iki.fi
Michael Paquier
After calling XLogInitBufferForRedo(), the page might be all-zeros if it was
not in page cache already. btree_xlog_unlink_page initialized the page
correctly, but it called PageGetSpecialPointer before initializing it, which
would lead to a corrupt page at WAL replay, if the unlinked page is not in
page cache.
Backpatch to 9.4, the bug came with the rewrite of B-tree page deletion.
I broke this with my WAL format refactoring patch. Before that, the metapage
was read from disk, and modified in-place regardless of the LSN. That was
always a bit silly, as there's no need to read the old page version from
disk disk when we're overwriting it anyway. So that was changed in 9.5, but
I failed to add a GinInitPage call to initialize the page-headers correctly.
Usually you wouldn't notice, because the metapage is already in the page
cache and is not zeroed.
One way to reproduce this is to perform a VACUUM on an already vacuumed
table (so that the vacuum has no real work to do), immediately after a
checkpoint, and then perform an immediate shutdown. After recovery, the
page headers of the metapage will be incorrectly all-zeroes.
Reported by Jeff Janes
When archive recovery and restartpoints were initially introduced,
checkpoint_segments was ignored on the grounds that the files restored from
archive don't consume any space in the recovery server. That was changed in
later releases, but even then it was arguably a feature rather than a bug,
as performing restartpoints as often as checkpoints during normal operation
might be excessive, but you might nevertheless not want to waste a lot of
space for pre-allocated WAL by setting checkpoint_segments to a high value.
But now that we have separate min_wal_size and max_wal_size settings, you
can bound WAL usage with max_wal_size, and still avoid consuming excessive
space usage by setting min_wal_size to a lower value, so that argument is
moot.
There are still some issues with actually limiting the space usage to
max_wal_size: restartpoints in recovery can only start after seeing the
checkpoint record, while a checkpoint starts flushing buffers as soon as
the redo-pointer is set. Restartpoint is paced to happen at the same
leisurily speed, determined by checkpoint_completion_target, as checkpoints,
but because they are started later, max_wal_size can be exceeded by upto
one checkpoint cycle's worth of WAL, depending on
checkpoint_completion_target. But that seems better than not trying at all,
and max_wal_size is a soft limit anyway.
The documentation already claimed that max_wal_size is obeyed in recovery,
so this just fixes the behaviour to match the docs. However, add some
weasel-words there to mention that max_wal_size may well be exceeded by
some amount in recovery.
Seems like cheap insurance for WAL bugs. A spurious call to
XLogBeginInsert() in itself would be fairly harmless, but if there is any
data registered and the insertion is not completed/cancelled properly, there
is a risk that the data ends up in a wrong WAL record.
Per Jeff Janes's suggestion.
If data checksums or wal_log_hints is on, and a GIN page is split, the code
to find a new, empty, block was called after having already called
XLogBeginInsert(). That causes an assertion failure or PANIC, if finding the
new block involves updating a FSM page that had not been modified since last
checkpoint, because that update is WAL-logged, which calls XLogBeginInsert
again. Nested XLogBeginInsert calls are not supported.
To fix, rearrange GIN code so that XLogBeginInsert is called later, after
finding the victim buffers.
Reported by Jeff Janes.
VACUUM FREEZE generated false cancelations of standby queries on an
otherwise idle master. Caused by an off-by-one error on cutoff_xid
which goes back to original commit.
Backpatch to all versions 9.0+
Analysis and report by Marco Nenciarini
Bug fix by Simon Riggs
There was a confusion about which block number to use when storing an
item's pointer in the revmap -- the revmap page's blkno was being used,
not the data page's blkno.
Spotted-by: Jeff Janes
Don't apply rmtree(), which will gleefully remove an entire subtree,
and don't even apply unlink() unless it's symlink or a directory,
the only things that we expect to find.
Amit Kapila, with minor tweaks by me, per extensive discussions
involving Andrew Dunstan, Fujii Masao, and Heikki Linnakangas,
at least some of whom also reviewed the code.
1. Replay of the WAL record for setting a bit in the visibility map
contained an assertion that a full-page image of that record type can only
occur with checksums enabled. But it can also happen with wal_log_hints, so
remove the assertion. Unlike checksums, wal_log_hints can be changed on the
fly, so it would be complicated to figure out if it was enabled at the time
that the WAL record was generated.
2. wal_log_hints has the same effect on the locking needed to read the LSN
of a page as data checksums. BufferGetLSNAtomic() didn't get the memo.
Backpatch to 9.4, where wal_log_hints was added.
Previously autovacuum was not necessarily triggered if space in the
members slru got tight. The first problem was that the signalling was
tied to values in the offsets slru, but members can advance much
faster. Thats especially a problem if old sessions had been around that
previously prevented the multixact horizon to increase. Secondly the
skipping logic doesn't work if the database was restarted after
autovacuum was triggered - that knowledge is not preserved across
restart. This is especially a problem because it's a common
panic-reaction to restart the database if it gets slow to
anti-wraparound vacuums.
Fix the first problem by separating the logic for members from
offsets. Trigger autovacuum whenever a multixact crosses a segment
boundary, as the current member offset increases in irregular values, so
we can't use a simple modulo logic as for offsets. Add a stopgap for
the second problem, by signalling autovacuum whenver ERRORing out
because of boundaries.
Discussion: 20150608163707.GD20772@alap3.anarazel.de
Backpatch into 9.3, where it became more likely that multixacts wrap
around.
9a20a9b2 added a new elog(), enabled when WAL_DEBUG is defined. The
other WAL_DEBUG dependant messages check for the wal_debug GUC, but this
one did not. While at it replace 'upto' with 'up to'.
Discussion: 20150610110253.GF3832@alap3.anarazel.de
Backpatch to 9.4, the first release containing 9a20a9b2.
Since find_multixact_start() relies on SimpleLruDoesPhysicalPageExist(),
and that function looks only at the on-disk state, it's possible for it
to fail to find a page that exists in the in-memory SLRU that has not
been written yet. If that happens, SetOffsetVacuumLimit() will
erroneously decide to force emergency autovacuuming immediately.
We should probably fix find_multixact_start() to consider the data
cached in memory as well as on the on-disk state, but that's no excuse
for SetOffsetVacuumLimit() to be stupid about the case where it can
no longer read the value after having previously succeeded in doing so.
Report by Andres Freund.
* Remove unused argument "dstfname" and related code from XLogFileCopy().
* Previously XLogFileCopy() returned a pstrdup'd string so that
InstallXLogFileSegment() used it later. Since the pstrdup'd string was never
free'd, there could be a risk of memory leak. It was almost harmless because
the startup process exited just after calling XLogFileCopy(), it existed.
This commit changes XLogFileCopy() so that it directly calls
InstallXLogFileSegment() and doesn't call pstrdup() at all. Which fixes that
memory leak problem.
* Extend InstallXLogFileSegment() so that the caller can specify the log level.
Which allows us to emit an error when InstallXLogFileSegment() fails a disk
file access like link() and rename(). Previously it was always logged with
LOG level and additionally needed to be logged with ERROR when we wanted
to treat it as an error.
Michael Paquier
HotStandbyActiveInReplay, introduced in 061b079f, only allowed WAL
replay to happen in the startup process, missing the single user case.
This buglet is fairly harmless as it only causes problems when single
user mode in an assertion enabled build is used to replay a btree vacuum
record.
Backpatch to 9.2. 061b079f was backpatched further, but the assertion
was not.
Recent commits, mainly b69bf30b9b and
53bb309d2d, introduced mechanisms to
protect against wraparound of the MultiXact member space: the number
of multixacts that can exist at one time is limited to 2^32, but the
total number of members in those multixacts is also limited to 2^32,
and older code did not take care to enforce the second limit,
potentially allowing old data to be overwritten while it was still
needed.
Unfortunately, these new mechanisms failed to account for the fact
that the code paths in which they run might be executed during
recovery or while the cluster was in an inconsistent state. Also,
they failed to account for the fact that users who used pg_upgrade
to upgrade a PostgreSQL version between 9.3.0 and 9.3.4 might have
might oldestMultiXid = 1 in the control file despite the true value
being larger.
To fix these problems, first, avoid unnecessarily examining the
mmembers of MultiXacts when the cluster is not known to be consistent.
TruncateMultiXact has done this for a long time, and this patch does
not fix that. But the new calls used to prevent member wraparound
are not needed until we reach normal running, so avoid calling them
earlier. (SetMultiXactIdLimit is actually called before InRecovery
is set, so we can't rely on that; we invent our own multixact-specific
flag instead.)
Second, make failure to look up the members of a MultiXact a non-fatal
error. Instead, if we're unable to determine the member offset at
which wraparound would occur, postpone arming the member wraparound
defenses until we are able to do so. If we're unable to determine the
member offset that should force autovacuum, force it continuously
until we are able to do so. If we're unable to deterine the member
offset at which we should truncate the members SLRU, log a message and
skip truncation.
An important consequence of these changes is that anyone who does have
a bogus oldestMultiXid = 1 value in pg_control will experience
immediate emergency autovacuuming when upgrading to a release that
contains this fix. The release notes should highlight this fact. If
a user has no pg_multixact/offsets/0000 file, but has oldestMultiXid = 1
in the control file, they may wish to vacuum any tables with
relminmxid = 1 prior to upgrading in order to avoid an immediate
emergency autovacuum after the upgrade. This must be done with a
PostgreSQL version 9.3.5 or newer and with vacuum_multixact_freeze_min_age
and vacuum_multixact_freeze_table_age set to 0.
This patch also adds an additional log message at each database server
startup, indicating either that protections against member wraparound
have been engaged, or that they have not. In the latter case, once
autovacuum has advanced oldestMultiXid to a sane value, the message
indicating that the guards have been engaged will appear at the next
checkpoint. A few additional messages have also been added at the DEBUG1
level so that the correct operation of this code can be properly audited.
Along the way, this patch fixes another, related bug in TruncateMultiXact
that has existed since PostgreSQL 9.3.0: when no MultiXacts exist at
all, the truncation code looks up NextMultiXactId, which doesn't exist
yet. This can lead to TruncateMultiXact removing every file in
pg_multixact/offsets instead of keeping one around, as it should.
This in turn will cause the database server to refuse to start
afterwards.
Patch by me. Review by Álvaro Herrera, Andres Freund, Noah Misch, and
Thomas Munro.
Commit 2ce439f337 introduced a rather serious
regression, namely that if its scan of the data directory came across any
un-fsync-able files, it would fail and thereby prevent database startup.
Worse yet, symlinks to such files also caused the problem, which meant that
crash restart was guaranteed to fail on certain common installations such
as older Debian.
After discussion, we agreed that (1) failure to start is worse than any
consequence of not fsync'ing is likely to be, therefore treat all errors
in this code as nonfatal; (2) we should not chase symlinks other than
those that are expected to exist, namely pg_xlog/ and tablespace links
under pg_tblspc/. The latter restriction avoids possibly fsync'ing a
much larger part of the filesystem than intended, if the user has left
random symlinks hanging about in the data directory.
This commit takes care of that and also does some code beautification,
mainly moving the relevant code into fd.c, which seems a much better place
for it than xlog.c, and making sure that the conditional compilation for
the pre_sync_fname pass has something to do with whether pg_flush_data
works.
I also relocated the call site in xlog.c down a few lines; it seems a
bit silly to be doing this before ValidateXLOGDirectoryStructure().
The similar logic in initdb.c ought to be made to match this, but that
change is noncritical and will be dealt with separately.
Back-patch to all active branches, like the prior commit.
Abhijit Menon-Sen and Tom Lane
brin_form_tuple calculated an exact tuple size, then palloc'd and
filled just that much. Later, brin_doinsert or brin_doupdate would
MAXALIGN the tuple size and tell PageAddItem that that was the size
of the tuple to insert. If the original tuple size wasn't a multiple
of MAXALIGN, the net result would be that PageAddItem would memcpy
a few more bytes than the palloc request had been for.
AFAICS, this is totally harmless in the real world: the error is a
read overrun not a write overrun, and palloc would certainly have
rounded the request up to a MAXALIGN multiple internally, so there's
no chance of the memcpy fetching off the end of memory. Valgrind,
however, is picky to the byte level not the MAXALIGN level.
Fix it by pushing the MAXALIGN step back to brin_form_tuple. (The other
possible source of tuples in this code, brin_form_placeholder_tuple,
was already producing a MAXALIGN'd result.)
In passing, be a bit more paranoid about internal allocations in
brin_form_tuple.
Multixact truncation is now handled differently, and this file hadn't
gotten the memo.
Per note from Amit Langote. I didn't use his patch, though.
Also update the description of infomask bits, which weren't completely up
to date either. This commit also propagates b01a4f6838 back to 9.3 and
9.4, which apparently I failed to do back then.
Fix some places where pgindent did silly stuff, often because project
style wasn't followed to begin with. (I've not touched the atomics
headers, though.)
Typo in commit 7cbee7c0a. No practical effect since the buffer should
never actually be overrun, but various compilers and static analyzers will
whine about it.
Petr Jelinek
Fix confusion in documentation, substantial memory leakage if float8 or
float4 are pass-by-reference, and assorted comments that were obsoleted
by commit 98edd617f3.
With commit de768844, a copy of the partial segment was archived with the
.partial suffix, but the original file was still left in pg_xlog, so it
didn't actually solve the problems with archiving the partial segment that
it was supposed to solve. With this patch, the partial segment is renamed
rather than copied, so we only archive it with the .partial suffix.
Also be more robust in detecting if the last segment is already being
archived. Previously I used XLogArchiveIsBusy() for that, but that's not
quite right. With archive_mode='always', there might be a .ready file for
it, and we don't want to rename it to .partial in that case.
The old segment is needed until we're fully committed to the new timeline,
i.e. until we've written the end-of-recovery WAL record and updated the
min recovery point and timeline in the control file. So move the renaming
later in the startup sequence, after all that's been done.
Previously even if recovery_target_action was set to pause and
the recovery target was reached, the recovery could never be paused.
Because the setting of pause was *always* overridden with that of
shutdown unexpectedly. This override is valid and intentional
if hot_standby is not enabled because there is no way to resume
the paused recovery in this case and the setting of pause is
completely useless. But not if hot_standby is enabled.
This patch changes the code so that the setting of pause is overridden
with that of shutdown only when hot_standby is not enabled.
Bug reported by Andres Freund
Use "a" and "an" correctly, mostly in comments. Two error messages were
also fixed (they were just elogs, so no translation work required). Two
function comments in pg_proc.h were also fixed. Etsuro Fujita reported one
of these, but I found a lot more with grep.
Also fix a few other typos spotted while grepping for the a/an typos.
For example, "consists out of ..." -> "consists of ...". Plus a "though"/
"through" mixup reported by Euler Taveira.
Many of these typos were in old code, which would be nice to backpatch to
make future backpatching easier. But much of the code was new, and I didn't
feel like crafting separate patches for each branch. So no backpatching.
This lets BRIN be used with R-Tree-like indexing strategies.
Also provided are operator classes for range types, box and inet/cidr.
The infrastructure provided here should be sufficient to create operator
classes for similar datatypes; for instance, opclasses for PostGIS
geometries should be doable, though we didn't try to implement one.
(A box/point opclass was also submitted, but we ripped it out before
commit because the handling of floating point comparisons in existing
code is inconsistent and would generate corrupt indexes.)
Author: Emre Hasegeli. Cosmetic changes by me
Review: Andreas Karlsson
For upcoming BRIN opclasses, it's convenient to have strategy numbers
defined in a single place. Since there's nothing appropriate, create
it. The StrategyNumber typedef now lives there, as well as existing
strategy numbers for B-trees (from skey.h) and R-tree-and-friends (from
gist.h). skey.h is forced to include stratnum.h because of the
StrategyNumber typedef, but gist.h is not; extensions that currently
rely on gist.h for rtree strategy numbers might need to add a new
A few .c files can stop including skey.h and/or gist.h, which is a nice
side benefit.
Per discussion:
https://www.postgresql.org/message-id/20150514232132.GZ2523@alvh.no-ip.org
Authored by Emre Hasegeli and Álvaro.
(It's not clear to me why bootscanner.l has any #include lines at all.)
Add a TABLESAMPLE clause to SELECT statements that allows
user to specify random BERNOULLI sampling or block level
SYSTEM sampling. Implementation allows for extensible
sampling functions to be written, using a standard API.
Basic version follows SQLStandard exactly. Usable
concrete use cases for the sampling API follow in later
commits.
Petr Jelinek
Reviewed by Michael Paquier and Simon Riggs
We can only support a lossy distance function when the distance function's
datatype is comparable with the original ordering operator's datatype.
The distance function always returns a float8, so we are limited to float8,
and float4 (by a hard-coded cast of the float8 to float4).
In light of this limitation, it seems like a good idea to have a separate
'recheck' flag for the ORDER BY expressions, so that if you have a non-lossy
distance function, it still works with lossy quals. There are cases like
that with the build-in or contrib opclasses, but it's plausible.
There was a hidden assumption that the ORDER BY values returned by GiST
match the original ordering operator's return type, but there are plenty
of examples where that's not true, e.g. in btree_gist and pg_trgm. As long
as the distance function is not lossy, we can tolerate that and just not
return the distance to the executor (or rather, always return NULL). The
executor doesn't need the distances if there are no lossy results.
There was another little bug: the recheck variable was not initialized
before calling the distance function. That revealed the bigger issue,
as the executor tried to reorder tuples that didn't need reordering, and
that failed because of the datatype mismatch.
The distance function can now set *recheck = false, like index quals. The
executor will then re-check the ORDER BY expressions, and use a queue to
reorder the results on the fly.
This makes it possible to do kNN-searches on polygons and circles, which
don't store the exact value in the index, but just a bounding box.
Alexander Korotkov and me
This patch introduces the ability for complex datatypes to have an
in-memory representation that is different from their on-disk format.
On-disk formats are typically optimized for minimal size, and in any case
they can't contain pointers, so they are often not well-suited for
computation. Now a datatype can invent an "expanded" in-memory format
that is better suited for its operations, and then pass that around among
the C functions that operate on the datatype. There are also provisions
(rudimentary as yet) to allow an expanded object to be modified in-place
under suitable conditions, so that operations like assignment to an element
of an array need not involve copying the entire array.
The initial application for this feature is arrays, but it is not hard
to foresee using it for other container types like JSON, XML and hstore.
I have hopes that it will be useful to PostGIS as well.
In this initial implementation, a few heuristics have been hard-wired
into plpgsql to improve performance for arrays that are stored in
plpgsql variables. We would like to generalize those hacks so that
other datatypes can obtain similar improvements, but figuring out some
appropriate APIs is left as a task for future work. (The heuristics
themselves are probably not optimal yet, either, as they sometimes
force expansion of arrays that would be better left alone.)
Preliminary performance testing shows impressive speed gains for plpgsql
functions that do element-by-element access or update of large arrays.
There are other cases that get a little slower, as a result of added array
format conversions; but we can hope to improve anything that's annoyingly
bad. In any case most applications should see a net win.
Tom Lane, reviewed by Andres Freund
If a postgres_fdw foreign table is a non-locked source relation in an
UPDATE, DELETE, or SELECT FOR UPDATE/SHARE, and the query selects its
ctid column, the wrong value would be returned if an EvalPlanQual
recheck occurred. This happened because the foreign table's result row
was copied via the ROW_MARK_COPY code path, and EvalPlanQualFetchRowMarks
just unconditionally set the reconstructed tuple's t_self to "invalid".
To fix that, we can have EvalPlanQualFetchRowMarks copy the composite
datum's t_ctid field, and be sure to initialize that along with t_self
when postgres_fdw constructs a tuple to return.
If we just did that much then EvalPlanQualFetchRowMarks would start
returning "(0,0)" as ctid for all other ROW_MARK_COPY cases, which perhaps
does not matter much, but then again maybe it might. The cause of that is
that heap_form_tuple, which is the ultimate source of all composite datums,
simply leaves t_ctid as zeroes in newly constructed tuples. That seems
like a bad idea on general principles: a field that's really not been
initialized shouldn't appear to have a valid value. So let's eat the
trivial additional overhead of doing "ItemPointerSetInvalid(&(td->t_ctid))"
in heap_form_tuple.
This closes out our handling of Etsuro Fujita's report that tableoid and
ctid weren't correctly set in postgres_fdw EvalPlanQual cases. Along the
way we did a great deal of work to improve FDWs' ability to control row
locking behavior; which was not wasted effort by any means, but it didn't
end up being a fix for this problem because that feature would be too
expensive for postgres_fdw to use all the time.
Although the fix for the tableoid misbehavior was back-patched, I'm
hesitant to do so here; it seems far less likely that people would care
about remote ctid than tableoid, and even such a minor behavioral change
as this in heap_form_tuple is perhaps best not back-patched. So commit
to HEAD only, at least for the moment.
Etsuro Fujita, with some adjustments by me
Windows can't reliably restore symbolic links from a tar format, so
instead during backup start we create a tablespace_map file, which is
used by the restoring postgres to create the correct links in pg_tblspc.
The backup protocol also now has an option to request this file to be
included in the backup stream, and this is used by pg_basebackup when
operating in tar mode.
This is done on all platforms, not just Windows.
This means that pg_basebackup will not not work in tar mode against 9.4
and older servers, as this protocol option isn't implemented there.
Amit Kapila, reviewed by Dilip Kumar, with a little editing from me.
Analysis by Noah Misch shows that the 25% threshold set by commit
53bb309d2d is lower than any other,
similar autovac threshold. While we don't know exactly what value
will be optimal for all users, it is better to err a little on the
high side than on the low side. A higher value increases the risk
that users might exhaust the available space and start seeing errors
before autovacuum can clean things up sufficiently, but a user who
hits that problem can compensate for it by reducing
autovacuum_multixact_freeze_max_age to a value dependent on their
average multixact size. On the flip side, if the emergency cap
imposed by that patch kicks in too early, the user will experience
excessive wraparound scanning and will be unable to mitigate that
problem by configuration. The new value will hopefully reduce the
risk of such bad experiences while still providing enough headroom
to avoid multixact member exhaustion for most users.
Along the way, adjust the documentation to reflect the effects of
commit 04e6d3b877, which taught
autovacuum to run for multixact wraparound even when autovacuum
is configured off.
Commit b69bf30b9b advanced the stop point
at vacuum time, but this has subsequently been shown to be unsafe as a
result of analysis by myself and Thomas Munro and testing by Thomas
Munro. The crux of the problem is that the SLRU deletion logic may
get confused about what to remove if, at exactly the right time during
the checkpoint process, the head of the SLRU crosses what used to be
the tail.
This patch, by me, fixes the problem by advancing the stop point only
following a checkpoint. This has the additional advantage of making
the removal logic work during recovery more like the way it works during
normal running, which is probably good.
At least one of the calls to DetermineSafeOldestOffset which this patch
removes was already dead, because MultiXactAdvanceOldest is called only
during recovery and DetermineSafeOldestOffset was set up to do nothing
during recovery. That, however, is inconsistent with the principle that
recovery and normal running should work similarly, and was confusing to
boot.
Along the way, fix some comments that previous patches in this area
neglected to update. It's not clear to me whether there's any
concrete basis for the decision to use only half of the multixact ID
space, but it's neither necessary nor sufficient to prevent multixact
member wraparound, so the comments should not say otherwise.
Commit b69bf30b9b failed to take into
account the possibility that there might be no multixacts in existence
at all.
Report by Thomas Munro; patch by me.
Previously, we would archive the possible-incomplete WAL segment with its
normal filename, but that causes trouble if the server owning that timeline
is still running, and tries to archive the same segment later. It's not nice
for the standby to trip up the master's archival like that. And it's pretty
confusing, anyway, to have an incomplete segment in the archive that's
indistinguishable from a normal, complete segment.
To avoid such confusion, add a .partial suffix to the file. Or to be more
precise, make a copy of the old segment under the .partial suffix, and
archive that instead of the original file. pg_receivexlog also uses the
.partial suffix for the same purpose, to tell apart incompletely streamed
files from complete ones.
There is no automatic mechanism to use the .partial files at recovery, so
they will go unused, unless the administrator manually copies to them to
the pg_xlog directory (and removes the .partial suffix). Recovery won't
normally need the WAL - when recovering to the new timeline, it will find
the same WAL on the first segment on the new timeline instead - but it
nevertheless feels better to archive the file with the .partial suffix, for
debugging purposes if nothing else.
The logic introduced in commit b69bf30b9b
and repaired in commits 669c7d20e6 and
7be47c56af helps to ensure that we don't
overwrite old multixact member information while it is still needed,
but a user who creates many large multixacts can still exhaust the
member space (and thus start getting errors) while autovacuum stands
idly by.
To fix this, progressively ramp down the effective value (but not the
actual contents) of autovacuum_multixact_freeze_max_age as member space
utilization increases. This makes autovacuum more aggressive and also
reduces the threshold for a manual VACUUM to perform a full-table scan.
This patch leaves unsolved the problem of ensuring that emergency
autovacuums are triggered even when autovacuum=off. We'll need to fix
that via a separate patch.
Thomas Munro and Robert Haas
The newly added ON CONFLICT clause allows to specify an alternative to
raising a unique or exclusion constraint violation error when inserting.
ON CONFLICT refers to constraints that can either be specified using a
inference clause (by specifying the columns of a unique constraint) or
by naming a unique or exclusion constraint. DO NOTHING avoids the
constraint violation, without touching the pre-existing row. DO UPDATE
SET ... [WHERE ...] updates the pre-existing tuple, and has access to
both the tuple proposed for insertion and the existing tuple; the
optional WHERE clause can be used to prevent an update from being
executed. The UPDATE SET and WHERE clauses have access to the tuple
proposed for insertion using the "magic" EXCLUDED alias, and to the
pre-existing tuple using the table name or its alias.
This feature is often referred to as upsert.
This is implemented using a new infrastructure called "speculative
insertion". It is an optimistic variant of regular insertion that first
does a pre-check for existing tuples and then attempts an insert. If a
violating tuple was inserted concurrently, the speculatively inserted
tuple is deleted and a new attempt is made. If the pre-check finds a
matching tuple the alternative DO NOTHING or DO UPDATE action is taken.
If the insertion succeeds without detecting a conflict, the tuple is
deemed inserted.
To handle the possible ambiguity between the excluded alias and a table
named excluded, and for convenience with long relation names, INSERT
INTO now can alias its target table.
Bumps catversion as stored rules change.
Author: Peter Geoghegan, with significant contributions from Heikki
Linnakangas and Andres Freund. Testing infrastructure by Jeff Janes.
Reviewed-By: Heikki Linnakangas, Andres Freund, Robert Haas, Simon Riggs,
Dean Rasheed, Stephen Frost and many others.
The minmax opclass was using the wrong support functions when
cross-datatypes queries were run. Instead of trying to fix the
pg_amproc definitions (which apparently is not possible), use the
already correct pg_amop entries instead. This requires jumping through
more hoops (read: extra syscache lookups) to obtain the underlying
functions to execute, but it is necessary for correctness.
Author: Emre Hasegeli, tweaked by Álvaro
Review: Andreas Karlsson
Also change BrinOpcInfo to record each stored type's typecache entry
instead of just the OID. Turns out that the full type cache is
necessary in brin_deform_tuple: the original code used the indexed
type's byval and typlen properties to extract the stored tuple, which is
correct in Minmax; but in other implementations that want to store
something different, that's wrong. The realization that this is a bug
comes from Emre also, but I did not use his patch.
I also adopted Emre's regression test code (with smallish changes),
which is more complete.
The old formula didn't have enough parentheses, so it would do the wrong
thing, and it used / rather than % to find a remainder. The effect of
these oversights is that the stop point chosen by the logic introduced in
commit b69bf30b9b might be rather
meaningless.
Thomas Munro, reviewed by Kevin Grittner, with a whitespace tweak by me.
Otherwise, if there's another crash, some writes from after the first
crash might make it to disk while writes from before the crash fail
to make it to disk. This could lead to data corruption.
Back-patch to all supported versions.
Abhijit Menon-Sen, reviewed by Andres Freund and slightly revised
by me.
ParseCommitRecord() accessed xl_xact_origin directly. But the chunks in
the commit record's data only have 4 byte alignment, whereas
xl_xact_origin's members require 8 byte alignment on some
platforms. Update comments to make not of that and copy the record to
stack local storage before reading.
With help from Stefan Kaltenbrunner in pinning down the buildfarm and
verifying the fix.
This does four basic things. First, it provides convenience routines
to coordinate the startup and shutdown of parallel workers. Second,
it synchronizes various pieces of state (e.g. GUCs, combo CID
mappings, transaction snapshot) from the parallel group leader to the
worker processes. Third, it prohibits various operations that would
result in unsafe changes to that state while parallelism is active.
Finally, it propagates events that would result in an ErrorResponse,
NoticeResponse, or NotifyResponse message being sent to the client
from the parallel workers back to the master, from which they can then
be sent on to the client.
Robert Haas, Amit Kapila, Noah Misch, Rushabh Lathia, Jeevan Chalke.
Suggestions and review from Andres Freund, Heikki Linnakangas, Noah
Misch, Simon Riggs, Euler Taveira, and Jim Nasby.
We need to create the pg_multixact/offsets file deleted by pg_upgrade
much earlier than we originally were: it was in TrimMultiXact(), which
runs after we exit recovery, but it actually needs to run earlier than
the first call to SetMultiXactIdLimit (before recovery), because that
routine already wants to read the first offset segment.
Per pg_upgrade trouble report from Jeff Janes.
While at it, silence a compiler warning about a pointless assert that an
unsigned variable was being tested non-negative. This was a signed
constant in Thomas Munro's patch which I changed to unsigned before
commit. Pointed out by Andres Freund.
When implementing a replication solution ontop of logical decoding, two
related problems exist:
* How to safely keep track of replication progress
* How to change replication behavior, based on the origin of a row;
e.g. to avoid loops in bi-directional replication setups
The solution to these problems, as implemented here, consist out of
three parts:
1) 'replication origins', which identify nodes in a replication setup.
2) 'replication progress tracking', which remembers, for each
replication origin, how far replay has progressed in a efficient and
crash safe manner.
3) The ability to filter out changes performed on the behest of a
replication origin during logical decoding; this allows complex
replication topologies. E.g. by filtering all replayed changes out.
Most of this could also be implemented in "userspace", e.g. by inserting
additional rows contain origin information, but that ends up being much
less efficient and more complicated. We don't want to require various
replication solutions to reimplement logic for this independently. The
infrastructure is intended to be generic enough to be reusable.
This infrastructure also replaces the 'nodeid' infrastructure of commit
timestamps. It is intended to provide all the former capabilities,
except that there's only 2^16 different origins; but now they integrate
with logical decoding. Additionally more functionality is accessible via
SQL. Since the commit timestamp infrastructure has also been introduced
in 9.5 (commit 73c986add) changing the API is not a problem.
For now the number of origins for which the replication progress can be
tracked simultaneously is determined by the max_replication_slots
GUC. That GUC is not a perfect match to configure this, but there
doesn't seem to be sufficient reason to introduce a separate new one.
Bumps both catversion and wal page magic.
Author: Andres Freund, with contributions from Petr Jelinek and Craig Ringer
Reviewed-By: Heikki Linnakangas, Petr Jelinek, Robert Haas, Steve Singer
Discussion: 20150216002155.GI15326@awork2.anarazel.de,
20140923182422.GA15776@alap3.anarazel.de,
20131114172632.GE7522@alap2.anarazel.de
Multixact member files are subject to early wraparound overflow and
removal: if the average multixact size is above a certain threshold (see
note below) the protections against offset overflow are not enough:
during multixact truncation at checkpoint time, some
pg_multixact/members files would be removed because the server considers
them to be old and not needed anymore. This leads to loss of files that
are critical to interpret existing tuples's Xmax values.
To protect against this, since we don't have enough info in pg_control
and we can't modify it in old branches, we maintain shared memory state
about the oldest value that we need to keep; we use this during new
multixact creation to abort if an old still-needed file would get
overwritten. This value is kept up to date by checkpoints, which makes
it not completely accurate but should be good enough. We start emitting
warnings sometime earlier, so that the eventual multixact-shutdown
doesn't take DBAs completely by surprise (more precisely: once 20
members SLRU segments are remaining before shutdown.)
On troublesome average multixact size: The threshold size depends on the
multixact freeze parameters. The oldest age is related to the greater of
multixact_freeze_table_age and multixact_freeze_min_age: anything
older than that should be removed promptly by autovacuum. If autovacuum
is keeping up with multixact freezing, the troublesome multixact average
size is
(2^32-1) / Max(freeze table age, freeze min age)
or around 28 members per multixact. Having an average multixact size
larger than that will eventually cause new multixact data to overwrite
the data area for older multixacts. (If autovacuum is not able to keep
up, or there are errors in vacuuming, the actual maximum is
multixact_freeeze_max_age instead, at which point multixact generation
is stopped completely. The default value for this limit is 400 million,
which means that the multixact size that would cause trouble is about 10
members).
Initial bug report by Timothy Garnett, bug #12990
Backpatch to 9.3, where the problem was introduced.
Authors: Álvaro Herrera, Thomas Munro
Reviews: Thomas Munro, Amit Kapila, Robert Haas, Kevin Grittner
When the startup process recovers transactions by scanning pg_twophase
directory, it should clear MyLockedGxact after it's done processing each
transaction. Like we do during normal operation, at PREPARE TRANSACTION.
Otherwise, if the startup process exits due to an error, it will try to
clear the locking_backend field of the last recovered transaction. That's
usually harmless, but if the error happens in MarkAsPreparing, while
holding TwoPhaseStateLock, the shmem-exit hook will try to acquire
TwoPhaseStateLock again, and deadlock with itself.
This fixes bug #13128 reported by Grant McAlister. The bug was introduced
by commit bb38fb0d, so backpatch to all supported versions like that
commit.
After the WAL format changes, the calculation of the size of a checkpoint
record became incorrect. Instead of trying to fix the math, check that the
previous record, i.e. the xl_prev value that we'd write for the next
record, matches the last checkpoint's redo pointer. That way it's not
dependent on the size of the checkpoint record at all.
The old logic was actually slightly wrong all along: if the previous
checkpoint record crossed a page boundary, the page headers threw off the
record size calculation, and the checkpoint was not skipped. The new
checkpoint would not cross a page boundary, so this only resulted in at
most one extra checkpoint after the system became idle. The new logic fixes
that. (It's not worth fixing in backbranches).
However, it makes some sense to try to keep the latest checkpoint contained
fully in a page, or at least in a single WAL segment, just on general
robustness grounds. If something goes awfully wrong, it's more likely that
you can recover the latest WAL segment, than the last two WAL segments. So
I added an extra check that the checkpoint is not skipped if the previous
checkpoint crossed a WAL segment.
Reported by Jeff Janes.
SLRU_SEGMENTS_PER_PAGE -> SLRU_PAGES_PER_SEGMENT
I introduced this ancient typo in subtrans.c and later propagated it to
multixact.c. I fixed the latter in f741300c, but only back to 9.3;
backpatch to all supported branches for consistency.
Now that we use CRC-32C in WAL and the control file, the "traditional" and
"legacy" CRC-32 variants are not used in any frontend programs anymore.
Move the code for those back from src/common to src/backend/utils/hash.
Also move the slicing-by-8 implementation (back) to src/port. This is in
preparation for next patch that will add another implementation that uses
Intel SSE 4.2 instructions to calculate CRC-32C, where available.
After a timeline switch, we would leave behind recycled WAL segments that
are in the future, but on the old timeline. After promotion, and after they
become old enough to be recycled again, we would notice that they don't have
a .ready or .done file, create a .ready file for them, and archive them.
That's bogus, because the files contain garbage, recycled from an older
timeline (or prealloced as zeros). We shouldn't archive such files.
This could happen when we're following a timeline switch during replay, or
when we switch to new timeline at end-of-recovery.
To fix, whenever we switch to a new timeline, scan the data directory for
WAL segments on the old timeline, but with a higher segment number, and
remove them. Those don't belong to our timeline history, and are most
likely bogus recycled or preallocated files. They could also be valid files
that we streamed from the primary ahead of time, but in any case, they're
not needed to recover to the new timeline.
Locking and updating the same tuple repeatedly led to some strange
multixacts being created which had several subtransactions of the same
parent transaction holding locks of the same strength. However,
once a subxact of the current transaction holds a lock of a given
strength, it's not necessary to acquire the same lock again. This made
some coding patterns much slower than required.
The fix is twofold. First we change HeapTupleSatisfiesUpdate to return
HeapTupleBeingUpdated for the case where the current transaction is
already a single-xid locker for the given tuple; it used to return
HeapTupleMayBeUpdated for that case. The new logic is simpler, and the
change to pgrowlocks is a testament to that: previously we needed to
check for the single-xid locker separately in a very ugly way. That
test is simpler now.
As fallout from the HTSU change, some of its callers need to be amended
so that tuple-locked-by-own-transaction is taken into account in the
BeingUpdated case rather than the MayBeUpdated case. For many of them
there is no difference; but heap_delete() and heap_update now check
explicitely and do not grab tuple lock in that case.
The HTSU change also means that routine MultiXactHasRunningRemoteMembers
introduced in commit 11ac4c73cb is no longer necessary and can be
removed; the case that used to require it is now handled naturally as
result of the changes to heap_delete and heap_update.
The second part of the fix to the performance issue is to adjust
heap_lock_tuple to avoid the slowness:
1. Previously we checked for the case that our own transaction already
held a strong enough lock and returned MayBeUpdated, but only in the
multixact case. Now we do it for the plain Xid case as well, which
saves having to LockTuple.
2. If the current transaction is the only locker of the tuple (but with
a lock not as strong as what we need; otherwise it would have been
caught in the check mentioned above), we can skip sleeping on the
multixact, and instead go straight to create an updated multixact with
the additional lock strength.
3. Most importantly, make sure that both the single-xid-locker case and
the multixact-locker case optimization are applied always. We do this
by checking both in a single place, rather than them appearing in two
separate portions of the routine -- something that is made possible by
the HeapTupleSatisfiesUpdate API change. Previously we would only check
for the single-xid case when HTSU returned MayBeUpdated, and only
checked for the multixact case when HTSU returned BeingUpdated. This
was at odds with what HTSU actually returned in one case: if our own
transaction was locker in a multixact, it returned MayBeUpdated, so the
optimization never applied. This is what led to the large multixacts in
the first place.
Per bug report #8470 by Oskari Saarenmaa.
Commit ed9cc2b5df made it unnecessary to pass
start_nblkno to _hash_splitbucket(), and for that matter unnecessary to
have the internal nblkno variable either. My compiler didn't complain
about that, but some did. I also rearranged the use of oblkno a bit to
make that case more parallel.
Report and initial patch by Petr Jelinek, rearranged a bit by me.
Back-patch to all branches, like the previous patch.
This is useful to control autovacuum log volume, for situations where
monitoring only a set of tables is necessary.
Author: Michael Paquier
Reviewed by: A team led by Naoya Anzai (also including Akira Kurosawa,
Taiki Kondo, Huong Dangminh), Fujii Masao.
Similarly to previous fix 9b8d478, commit 2c03216 has switched
XLogReaderAllocate() to use a set of palloc calls instead of malloc,
causing any callers of this function to fail with an error instead of
receiving a NULL pointer in case of out-of-memory error. Fix this by
using palloc_extended with MCXT_ALLOC_NO_OOM that will safely return
NULL in case of an OOM.
Michael Paquier, slightly modified by me.
Commit 2c03216 changed allocate_recordbuf() so that it uses a palloc to
allocate the read buffer and fails immediately when an out-of-memory error
shows up, even though its callers still expect that NULL is returned in that
case. This bug is fixed making allocate_recordbuf() use a palloc_extended
with MCXT_ALLOC_NO_OOM flag and return NULL in OOM case.
Michael Paquier
In 83ff1618 we defined integer limits iff they're not provided by the
system. That turns out not to be the greatest idea because there's
different ways some datatypes can be represented. E.g. on OSX PG's 64bit
datatype will be a 'long int', but OSX unconditionally uses 'long
long'. That disparity then can lead to warnings, e.g. around printf
formats.
One way to fix that would be to back int64 using stdint.h's
int64_t. While a good idea it's not that easy to implement. We would
e.g. need to include stdint.h in our external headers, which we don't
today. Also computing the correct int64 printf formats in that case is
nontrivial.
Instead simply prefix the integer limits with PG_ and define them
unconditionally. I've adjusted all the references to them in code, but
not the ones in comments; the latter seems unnecessary to me.
Discussion: 20150331141423.GK4878@alap3.anarazel.de
_hash_splitbucket() obtained the base page of the new bucket by calling
_hash_getnewbuf(), but it held no exclusive lock that would prevent some
other process from calling _hash_getnewbuf() at the same time. This is
contrary to _hash_getnewbuf()'s API spec and could in fact cause failures.
In practice, we must only call that function while holding write lock on
the hash index's metapage.
An additional problem was that we'd already modified the metapage's bucket
mapping data, meaning that failure to extend the index would leave us with
a corrupt index.
Fix both issues by moving the _hash_getnewbuf() call to just before we
modify the metapage in _hash_expandtable().
Unfortunately there's still a large problem here, which is that we could
also incur ENOSPC while trying to get an overflow page for the new bucket.
That would leave the index corrupt in a more subtle way, namely that some
index tuples that should be in the new bucket might still be in the old
one. Fixing that seems substantially more difficult; even preallocating as
many pages as we could possibly need wouldn't entirely guarantee that the
bucket split would complete successfully. So for today let's just deal
with the base case.
Per report from Antonin Houska. Back-patch to all active branches.
Previously the funcCtx was a child of the tmpCtx, but that was broken
by commit eaa5808e8e, which made
MemoryContextReset() delete, not reset, child contexts. The behavior of
having a tmpCtx reset also clear the other context seems rather dubious
anyway, so let's just disentangle them. Per report from Erik Rijkers.
In passing, fix badly-inaccurate comments about these contexts.
We cannot use the index's tuple descriptor directly to describe the index
tuples returned in an index-only scan. That's because the index might use
a different datatype for the values stored on disk than the type originally
indexed. As long as they were both pass-by-ref, it worked, but will not work
for pass-by-value types of different sizes. I noticed this as a crash when I
started hacking a patch to add fetch methods to btree_gist.
This improves on commit bbfd7edae5 by
making two simple changes:
* pg_attribute_noreturn now takes parentheses, ie pg_attribute_noreturn().
Likewise pg_attribute_unused(), pg_attribute_packed(). This reduces
pgindent's tendency to misformat declarations involving them.
* attributes are now always attached to function declarations, not
definitions. Previously some places were taking creative shortcuts,
which were not merely candidates for bad misformatting by pgindent
but often were outright wrong anyway. (It does little good to put a
noreturn annotation where callers can't see it.) In any case, if
we would like to believe that these macros can be used with non-gcc
compilers, we should avoid gratuitous variance in usage patterns.
I also went through and manually improved the formatting of a lot of
declarations, and got rid of excessively repetitive (and now obsolete
anyway) comments informing the reader what pg_attribute_printf is for.
This adds a new GiST opclass method, 'fetch', which is used to reconstruct
the original Datum from the value stored in the index. Also, the 'canreturn'
index AM interface function gains a new 'attno' argument. That makes it
possible to use index-only scans on a multi-column index where some of the
opclasses support index-only scans but some do not.
This patch adds support in the box and point opclasses. Other opclasses
can added later as follow-on patches (btree_gist would be particularly
interesting).
Anastasia Lubennikova, with additional fixes and modifications by me.
Several submitted and even committed patches have run into the problem
that C89, our baseline, does not provide minimum/maximum values for
various integer datatypes. C99's stdint.h does, but we can't rely on
it.
Several parts of the code defined limits locally, so instead centralize
the definitions to c.h.
This patch also changes the more obvious usages of literal limit values;
there's more places that could be changed, but it's less clear whether
it's beneficial to change those.
Author: Andrew Gierth
Discussion: 87619tc5wc.fsf@news-spur.riddles.org.uk
Even though the main benefit of the Lehman and Yao algorithm for
btrees is that no locks need be held between page reads in an
index search, we were holding a buffer pin on each leaf page after
it was read until we were ready to read the next one. The reason
was so that we could treat this as a weak lock to create an
"interlock" with vacuum's deletion of heap line pointers, even
though our README file pointed out that this was not necessary for
a scan using an MVCC snapshot.
The main goal of this patch is to reduce the blocking of vacuum
processes by in-progress btree index scans (including a cursor
which is idle), but the code rearrangement also allows for one
less buffer content lock to be taken when a forward scan steps from
one page to the next, which results in a small but consistent
performance improvement in many workloads.
This patch leaves behavior unchanged for some cases, which can be
addressed separately so that each case can be evaluated on its own
merits. These unchanged cases are when a scan uses a non-MVCC
snapshot, an index-only scan, and a scan of a btree index for which
modifications are not WAL-logged. If later patches allow all of
these cases to drop the buffer pin after reading a leaf page, then
the btree vacuum process can be simplified; it will no longer need
the "super-exclusive" lock to delete tuples from a page.
Reviewed by Heikki Linnakangas and Kyotaro Horiguchi
Recovery delays are implemented by waiting on a latch, and latches take
milliseconds as a parameter. The required amount of waiting was computed
using microsecond resolution though and the wait loop's abort condition
was checking the delay in microseconds as well. This could lead to
short spurts of busy looping when the overall wait time was below a
millisecond, but above 0 microseconds.
Instead just formulate the wait loop's abort condition in millisecond
granularity as well. Given that that's recovery_min_apply_delay
resolution, it seems harmless to not wait for less than a millisecond.
Backpatch to 9.4 where recovery_min_apply_delay was introduced.
Discussion: 20150323141819.GH26995@alap3.anarazel.de
Since 465883b0a two versions of commit records have existed. A compact
version that was used when no cache invalidations, smgr unlinks and
similar were needed, and a full version that could deal with all
that. Additionally the full version was embedded into twophase commit
records.
That resulted in a measurable reduction in the size of the logged WAL in
some workloads. But more recently additions like logical decoding, which
e.g. needs information about the database something was executed on,
made it applicable in fewer situations. The static split generally made
it hard to expand the commit record, because concerns over the size made
it hard to add anything to the compact version.
Additionally it's not particularly pretty to have twophase.c insert
RM_XACT records.
Rejigger things so that the commit and abort records only have one form
each, including the twophase equivalents. The presence of the various
optional (in the sense of not being in every record) pieces is indicated
by a bits in the 'xinfo' flag. That flag previously was not included in
compact commit records. To prevent an increase in size due to its
presence, it's only included if necessary; signalled by a bit in the
xl_info bits available for xact.c, similar to heapam.c's
XLOG_HEAP_OPMASK/XLOG_HEAP_INIT_PAGE.
Twophase commit/aborts are now the same as their normal
counterparts. The original transaction's xid is included in an optional
data field.
This means that commit records generally are smaller, except in the case
of a transaction with subtransactions, but no other special cases; the
increase there is four bytes, which seems acceptable given that the more
common case of not having subtransactions shrank. The savings are
especially measurable for twophase commits, which previously always used
the full version; but will in practice only infrequently have required
that.
The motivation for this work are not the space savings and and
deduplication though; it's that it makes it easier to extend commit
records with additional information. That's just a few lines of code
now; without impacting the common case where that information is not
needed.
Discussion: 20150220152150.GD4149@awork2.anarazel.de,
235610.92468.qm%40web29004.mail.ird.yahoo.com
Reviewed-By: Heikki Linnakangas, Simon Riggs
The introduction of min_wal_size & max_wal_size in 88e9823026 makes it
feasible to increase the default upper bound in checkpoint
size. Previously raising the default would lead to a increased disk
footprint, even if more segments weren't beneficial. The low default of
checkpoint size is one of common performance problem users have thus
increasing the default makes sense. Setups where the increase in
maximum disk usage is a problem will very likely have to run with a
modified configuration anyway.
Discussion: 54F4EFB8.40202@agliodbs.com,
CA+TgmoZEAgX5oMGJOHVj8L7XOkAe05Gnf45rP40m-K3FhZRVKg@mail.gmail.com
Author: Josh Berkus, after a discussion involving lots of people.
The new recovery_target_action (introduced in aedccb1f6/b8e33a85d4)
replaces it's functionality. Having both seems likely to cause more
confusion than it saves worry due to the incompatibility.
Discussion: 5484FC53.2060903@2ndquadrant.com
Author: Petr Jelinek
Until now __attribute__() was defined to be empty for all compilers but
gcc. That's problematic because it prevents using it in other compilers;
which is necessary e.g. for atomics portability. It's also just
generally dubious to do so in a header as widely included as c.h.
Instead add pg_attribute_format_arg, pg_attribute_printf,
pg_attribute_noreturn macros which are implemented in the compilers that
understand them. Also add pg_attribute_noreturn and pg_attribute_packed,
but don't provide fallbacks, since they can affect functionality.
This means that external code that, possibly unwittingly, relied on
__attribute__ defined to be empty on !gcc compilers may now run into
warnings or errors on those compilers. But there shouldn't be many
occurances of that and it's hard to work around...
Discussion: 54B58BA3.8040302@ohmu.fi
Author: Oskari Saarenmaa, with some minor changes by me.
When newly-added GUC parameter, wal_compression, is on, the PostgreSQL server
compresses a full page image written to WAL when full_page_writes is on or
during a base backup. A compressed page image will be decompressed during WAL
replay. Turning this parameter on can reduce the WAL volume without increasing
the risk of unrecoverable data corruption, but at the cost of some extra CPU
spent on the compression during WAL logging and on the decompression during
WAL replay.
This commit changes the WAL format (so bumping WAL version number) so that
the one-byte flag indicating whether a full page image is compressed or not is
included in its header information. This means that the commit increases the
WAL volume one-byte per a full page image even if WAL compression is not used
at all. We can save that one-byte by borrowing one-bit from the existing field
like hole_offset in the header and using it as the flag, for example. But which
would reduce the code readability and the extensibility of the feature.
Per discussion, it's not worth paying those prices to save only one-byte, so we
decided to add the one-byte flag to the header.
This commit doesn't introduce any new compression algorithm like lz4.
Currently a full page image is compressed using the existing PGLZ algorithm.
Per discussion, we decided to use it at least in the first version of the
feature because there were no performance reports showing that its compression
ratio is unacceptably lower than that of other algorithm. Of course,
in the future, it's worth considering the support of other compression
algorithm for the better compression.
Rahila Syed and Michael Paquier, reviewed in various versions by myself,
Andres Freund, Robert Haas, Abhijit Menon-Sen and many others.
... which is the usual convention among AMs, so that pg_filedump and
similar utilities can tell apart pages of different AMs. It was also
the intent of the original code, but I failed to realize that alignment
considerations would move the whole thing to the previous-to-last word
in the page.
The new definition of the associated macro makes surrounding code a bit
leaner, too.
Per note from Heikki at
http://www.postgresql.org/message-id/546A16EF.9070005@vmware.com
We allow this module to be turned off on restarts, so a restart time
check is enough to activate or deactivate the module; however, if there
is a standby replaying WAL emitted from a master which is restarted, but
the standby isn't, the state in the standby becomes inconsistent and can
easily be crashed.
Fix by activating and deactivating the module during WAL replay on
parameter change as well as on system start.
Problem reported by Fujii Masao in
http://www.postgresql.org/message-id/CAHGQGwFhJ3CnHo1CELEfay18yg_RA-XZT-7D8NuWUoYSZ90r4Q@mail.gmail.com
Author: Petr Jelínek
This makes it easier to write frontend programs that needs to understand
the WAL record format of CREATE/DROP DATABASE. dbcommands.h cannot easily
be #included in a frontend program, because it pulls in other header files
that need backend stuff, but the new dbcommands_xlog.h header file has
fewer dependencies.
Up to now RecordTransactionCommit() waited for WAL to be flushed (if
synchronous_commit != off) and to be synchronously replicated (if
enabled), even if a transaction did not have a xid assigned. The primary
reason for that is that sequence's nextval() did not assign a xid, but
are worthwhile to wait for on commit.
This can be problematic because sometimes read only transactions do
write WAL, e.g. HOT page prune records. That then could lead to read only
transactions having to wait during commit. Not something people expect
in a read only transaction.
This lead to such strange symptoms as backends being seemingly stuck
during connection establishment when all synchronous replicas are
down. Especially annoying when said stuck connection is the standby
trying to reconnect to allow syncrep again...
This behavior also is involved in a rather complicated <= 9.4 bug where
the transaction started by catchup interrupt processing waited for
syncrep using latches, but didn't get the wakeup because it was already
running inside the same overloaded signal handler. Fix the issue here
doesn't properly solve that issue, merely papers over the problems. In
9.5 catchup interrupts aren't processed out of signal handlers anymore.
To fix all this, make nextval() acquire a top level xid, and only wait for
transaction commit if a transaction both acquired a xid and emitted WAL
records. If only a xid has been assigned we don't uselessly want to
wait just because of writes to temporary/unlogged tables; if only WAL
has been written we don't want to wait just because of HOT prunes.
The xid assignment in nextval() is unlikely to cause overhead in
real-world workloads. For one it only happens SEQ_LOG_VALS/32 values
anyway, for another only usage of nextval() without using the result in
an insert or similar is affected.
Discussion: 20150223165359.GF30784@awork2.anarazel.de,
369698E947874884A77849D8FE3680C2@maumau,
5CF4ABBA67674088B3941894E22A0D25@maumau
Per complaint from maumau and Thom Brown
Backpatch all the way back; 9.0 doesn't have syncrep, but it seems
better to be consistent behavior across all maintained branches.
Instead of having a single knob (checkpoint_segments) that both triggers
checkpoints, and determines how many checkpoints to recycle, they are now
separate concerns. There is still an internal variable called
CheckpointSegments, which triggers checkpoints. But it no longer determines
how many segments to recycle at a checkpoint. That is now auto-tuned by
keeping a moving average of the distance between checkpoints (in bytes),
and trying to keep that many segments in reserve. The advantage of this is
that you can set max_wal_size very high, but the system won't actually
consume that much space if there isn't any need for it. The min_wal_size
sets a floor for that; you can effectively disable the auto-tuning behavior
by setting min_wal_size equal to max_wal_size.
The max_wal_size setting is now the actual target size of WAL at which a
new checkpoint is triggered, instead of the distance between checkpoints.
Previously, you could calculate the actual WAL usage with the formula
"(2 + checkpoint_completion_target) * checkpoint_segments + 1". With this
patch, you set the desired WAL usage with max_wal_size, and the system
calculates the appropriate CheckpointSegments with the reverse of that
formula. That's a lot more intuitive for administrators to set.
Reviewed by Amit Kapila and Venkata Balaji N.
Previously when the standby server failed to retrieve WAL files from any sources
(i.e., streaming replication, local pg_xlog directory or WAL archive), it always
waited for five seconds (hard-coded) before the next attempt. For example,
this is problematic in warm-standby because restore_command can fail
every five seconds even while new WAL file is expected to be unavailable for
a long time and flood the log files with its error messages.
This commit adds new parameter, wal_retrieve_retry_interval, to control that
wait time.
Alexey Vasiliev and Michael Paquier, reviewed by Andres Freund and me.
This requires changing quite a few places that were depending on
sizeof(HeapTupleHeaderData), but it seems for the best.
Michael Paquier, some adjustments by me
The part of the comparison function that was supposed to keep heap tuples
ahead of index items was backwards. It would not lead to incorrect results,
but it is more efficient to return heap tuples first, before scanning more
index pages, when both have the same distance.
Alexander Korotkov
Fix some issues I noticed while fooling with an extension to allow an
additional kind of toast pointer. Much of this is just comment
improvement, but there are a couple of actual bugs, which might or might
not be reachable today depending on what can happen during logical
decoding. An example is that toast_flatten_tuple() failed to cover the
possibility of an indirection pointer in its input. Back-patch to 9.4
just in case that is reachable now.
In HEAD, also correct some really minor issues with recent compression
reorganization, such as dangerously underparenthesized macros.
The meta data of PGLZ symbolized by PGLZ_Header is removed, to make
the compression and decompression code independent on the backend-only
varlena facility. PGLZ_Header is being used to store some meta data
related to the data being compressed like the raw length of the uncompressed
record or some varlena-related data, making it unpluggable once PGLZ is
stored in src/common as it contains some backend-only code paths with
the management of varlena structures. The APIs of PGLZ are reworked
at the same time to do only compression and decompression of buffers
without the meta-data layer, simplifying its use for a more general usage.
On-disk format is preserved as well, so there is no incompatibility with
previous major versions of PostgreSQL for TOAST entries.
Exposing compression and decompression APIs of pglz makes possible its
use by extensions and contrib modules. Especially this commit is required
for upcoming WAL compression feature so that the WAL reader facility can
decompress the WAL data by using pglz_decompress.
Michael Paquier, reviewed by me.
It was getting tedious to track and release all the different things that
form a scan key. We were leaking at least the queryCategories array, and
possibly more, on a rescan. That was visible if a GIN index was used in a
nested loop join. This also protects from leaks in extractQuery method.
No backpatching, given the lack of complaints from the field. Maybe later,
after this has received more field testing.
If an insertion or update had to wait for another transaction to finish,
because there was another insertion with conflicting key in progress,
we would pass a just-free'd item pointer to XactLockTableWait().
All calls to XactLockTableWait() and MultiXactIdWait() had similar issues.
Some passed a pointer to a buffer in the buffer cache, after already
releasing the lock. The call in EvalPlanQualFetch had already released the
pin too. All but the call in execUtils.c would merely lead to reporting a
bogus ctid, however (or an assertion failure, if enabled).
All the callers that passed HeapTuple->t_data->t_ctid were slightly bogus
anyway: if the tuple was updated (again) in the same transaction, its ctid
field would point to the next tuple in the chain, not the tuple itself.
Backpatch to 9.4, where the 'ctid' argument to XactLockTableWait was added
(in commit f88d4cfc)
The requiredEntries / additionalEntries arrays were not freed in
freeScanKeys() like other per-key stuff.
It's not obvious, but startScanKey() was only ever called after the keys
have been initialized with ginNewScanKey(). That's why it doesn't need to
worry about freeing existing arrays. The ginIsNewKey() test in gingetbitmap
was never true, because ginrescan free's the existing keys, and it's not OK
to call gingetbitmap twice in a row without calling ginrescan in between.
To make that clear, remove the unnecessary ginIsNewKey(). And just to be
extra sure that nothing funny happens if there is an existing key after all,
call freeScanKeys() to free it if it exists. This makes the code more
straightforward.
(I'm seeing other similar leaks in testing a query that rescans an GIN index
scan, but that's a different issue. This just fixes the obvious leak with
those two arrays.)
Backpatch to 9.4, where GIN fast scan was added.
In 804b6b6db4 we modified
BuildIndexValueDescription to pay attention to which columns are visible
to the user, but unfortunatley that commit neglected to consider indexes
which are built on expressions.
Handle error-reporting of violations of constraint indexes based on
expressions by not returning any detail when the user does not have
table-level SELECT rights.
Backpatch to 9.0, as the prior commit was.
Pointed out by Tom.
When gin_fuzzy_search_limit was used, we could jump out of startScan()
without calling startScanKey(). That was harmless in 9.3 and below, because
startScanKey()() didn't do anything interesting, but in 9.4 it initializes
information needed for skipping entries (aka GIN fast scans), and you
readily get a segfault if it's not done. Nevertheless, it was clearly wrong
all along, so backpatch all the way to 9.1 where the early return was
introduced.
(AFAICS startScanKey() did nothing useful in 9.3 and below, because the
fields it initialized were already initialized in ginFillScanKey(), but I
don't dare to change that in a minor release. ginFillScanKey() is always
called in gingetbitmap() even though there's a check there to see if the
scan keys have already been initialized, because they never are; ginrescan()
free's them.)
In the passing, remove unnecessary if-check from the second inner loop in
startScan(). We already check in the first loop that the condition is true
for all entries.
Reported by Olaf Gawenda, bug #12694, Backpatch to 9.1 and above, although
AFAICS it causes a live bug only in 9.4.
While building error messages to return to the user,
BuildIndexValueDescription, ExecBuildSlotValueDescription and
ri_ReportViolation would happily include the entire key or entire row in
the result returned to the user, even if the user didn't have access to
view all of the columns being included.
Instead, include only those columns which the user is providing or which
the user has select rights on. If the user does not have any rights
to view the table or any of the columns involved then no detail is
provided and a NULL value is returned from BuildIndexValueDescription
and ExecBuildSlotValueDescription. Note that, for key cases, the user
must have access to all of the columns for the key to be shown; a
partial key will not be returned.
Further, in master only, do not return any data for cases where row
security is enabled on the relation and row security should be applied
for the user. This required a bit of refactoring and moving of things
around related to RLS- note the addition of utils/misc/rls.c.
Back-patch all the way, as column-level privileges are now in all
supported versions.
This has been assigned CVE-2014-8161, but since the issue and the patch
have already been publicized on pgsql-hackers, there's no point in trying
to hide this commit.
gist_poly_compress() and gist_circle_compress() checked for a NULL-pointer
key argument, but that was dead code; the gist code never passes a
NULL-pointer to the "compress" method.
This commit also removes a documentation note added in commit a0a3883,
about doing NULL-pointer checks in the "compress" method. It was added
based on the fact that some implementations were doing NULL-pointer
checks, but those checks were unnecessary in the first place.
The NULL-pointer check in gbt_var_same() function was also unnecessary.
The arguments to the "same" method come from the "compress", "union", or
"picksplit" methods, but none of them return a NULL pointer.
None of this is to be confused with SQL NULL values. Those are dealt with
by the gist machinery, and are never passed to the GiST opclass methods.
Michael Paquier
In the union support proc, we were not checking the hasnulls flag of
value A early enough, so it could be skipped if the "allnulls" flag in
value B is set. Also, a check on the allnulls flag of value "B" was
redundant, so remove it.
Also change inet_minmax_ops to not be the default opclass for type inet,
as a future inclusion operator class would be more useful and it's
pretty difficult to change default opclass for a datatype later on.
(There is no catversion bump for this catalog change; this shouldn't be
a problem.)
Extracted from a larger patch to add an "inclusion" operator class.
Author: Emre Hasegeli
This commit extends the SortSupport infrastructure to allow operator
classes the option to provide abbreviated representations of Datums;
in the case of text, we abbreviate by taking the first few characters
of the strxfrm() blob. If the abbreviated comparison is insufficent
to resolve the comparison, we fall back on the normal comparator.
This can be much faster than the old way of doing sorting if the
first few bytes of the string are usually sufficient to resolve the
comparison.
There is the potential for a performance regression if all of the
strings to be sorted are identical for the first 8+ characters and
differ only in later positions; therefore, the SortSupport machinery
now provides an infrastructure to abort the use of abbreviation if
it appears that abbreviation is producing comparatively few distinct
keys. HyperLogLog, a streaming cardinality estimator, is included in
this commit and used to make that determination for text.
Peter Geoghegan, reviewed by me.
The flag is supposed to be copied from the record. Same issue with
track_commit_timestamps, but that's master-only.
Report and fix by Petr Jalinek. Backpatch to 9.4, where wal_log_hints was
added.
Some spaces were missing, and putting the affected tuple offset first in
the lock cases instead of the locking data makes more sense.
No backpatch since this is cosmetic and surrounding code has changed.
Since commit ba94518a, we used XLogFileOpen to open the next segment for
writing, but if the end-of-recovery happens exactly at a segment boundary,
the new segment might not exist yet. (Before ba94518a, XLogFileOpen was
correct, because we would open the previous segment if the switch happened
at the boundary.)
Instead of trying to create it if necessary, it's simpler to not bother
opening the segment at all. XLogWrite() will open or create it soon anyway,
after writing the checkpoint or end-of-recovery record.
Reported by Andres Freund.
Commit 0e5680f473 contained a thinko
mixing LOCKMODE with LockTupleMode. This caused misbehavior in the case
where a tuple is marked with a multixact with at most a FOR SHARE lock,
and another transaction tries to acquire a FOR NO KEY EXCLUSIVE lock;
this case should block but doesn't.
Include a new isolation tester spec file to explicitely try all the
tuple lock combinations; without the fix it shows the problem:
starting permutation: s1_begin s1_lcksvpt s1_tuplock2 s2_tuplock3 s1_commit
step s1_begin: BEGIN;
step s1_lcksvpt: SELECT * FROM multixact_conflict FOR KEY SHARE; SAVEPOINT foo;
a
1
step s1_tuplock2: SELECT * FROM multixact_conflict FOR SHARE;
a
1
step s2_tuplock3: SELECT * FROM multixact_conflict FOR NO KEY UPDATE;
a
1
step s1_commit: COMMIT;
With the fixed code, step s2_tuplock3 blocks until session 1 commits,
which is the correct behavior.
All other cases behave correctly.
Backpatch to 9.3, like the commit that introduced the problem.
At one point in the development of this feature, it was claimed that
allowing negative values would be useful to compensate for timezone
differences between master and slave servers. That was based on a mistaken
assumption that commit timestamps are recorded in local time; but of course
they're in UTC. Nor is a negative apply delay likely to be a sane way of
coping with server clock skew. However, the committed patch still treated
negative delays as doing something, and the timezone misapprehension
survived in the user documentation as well.
If recovery_min_apply_delay were a proper GUC we'd just set the minimum
allowed value to be zero; but for the moment it seems better to treat
negative settings as if they were zero.
In passing do some extra wordsmithing on the parameter's documentation,
including correcting a second misstatement that the parameter affects
processing of Restore Point records.
Issue noted by Michael Paquier, who also provided the code patch; doc
changes by me. Back-patch to 9.4 where the feature was introduced.
We were trying to acquire the lock even when we were subsequently
not sleeping in some other transaction, which opens us up unnecessarily
to deadlocks. In particular, this is troublesome if an update tries to
lock an updated version of a tuple and finds itself doing EvalPlanQual
update chain walking; more than two sessions doing this concurrently
will find themselves sleeping on each other because the HW tuple lock
acquisition in heap_lock_tuple called from EvalPlanQualFetch races with
the same tuple lock being acquired in heap_update -- one of these
sessions sleeps on the other one to finish while holding the tuple lock,
and the other one sleeps on the tuple lock.
Per trouble report from Andrew Sackville-West in
http://www.postgresql.org/message-id/20140731233051.GN17765@andrew-ThinkPad-X230
His scenario can be simplified down to a relatively simple
isolationtester spec file which I don't include in this commit; the
reason is that the current isolationtester is not able to deal with more
than one blocked session concurrently and it blocks instead of raising
the expected deadlock. In the future, if we improve isolationtester, it
would be good to include the spec file in the isolation schedule. I
posted it in
http://www.postgresql.org/message-id/20141212205254.GC1768@alvh.no-ip.org
Hat tip to Mark Kirkwood, who helped diagnose the trouble.
This reverts commit 60838df922.
That change needs a bit more thought to be workable. In view of
the potentially machine-dependent stuff that went in today,
we need all of the buildfarm to be testing those other changes.
Besides being shorter and much easier to read it changes the logic in
LWLockRelease() to release all shared lockers when waking up any. This
can yield some significant performance improvements - and the fairness
isn't really much worse than before, as we always allowed new shared
lockers to jump the queue.
Exposing compression and decompression APIs of pglz makes possible its
use by extensions and contrib modules. pglz_decompress contained a call
to elog to emit an error message in case of corrupted data. This function
is changed to return a status code to let its callers return an error instead.
This commit is required for upcoming WAL compression feature so that
the WAL reader facility can decompress the WAL data by using pglz_decompress.
Michael Paquier
This reverts commit 1826987a46.
The overall design was deemed unacceptable, in discussion following the
previous commit message; we might find some parts of it still
salvageable, but I don't want to be on the hook for fixing it, so let's
wait until we have a new patch.
The previous representation using a boolean column for each attribute
would not scale as well as we want to add further attributes.
Extra auxilliary functions are added to go along with this change, to
make up for the lost convenience of access of the old representation.
Catalog version bumped due to change in catalogs and the new functions.
Author: Adam Brightwell, minor tweaks by Álvaro
Reviewed by: Stephen Frost, Andres Freund, Álvaro Herrera
This performs slightly better, uses less memory, and needs slightly less
code in GiST, than the Red-Black tree previously used.
Reviewed by Peter Geoghegan
XLogFileInit() returns a file descriptor, which needs to be closed. The leak
was short-lived, since the startup process exits shortly afterwards, but it
was clearly a bug, nevertheless.
Per Coverity report.
We used time(null) to set a TimestampTz field, which gave bogus results.
Noticed while looking at pg_xlogdump output.
Backpatch to 9.3 and above, where the fast promotion was introduced.
Previously, if you wanted anything besides C-string hash keys, you had to
specify a custom hashing function to hash_create(). Nearly all such
callers were specifying tag_hash or oid_hash; which is tedious, and rather
error-prone, since a caller could easily miss the opportunity to optimize
by using hash_uint32 when appropriate. Replace this with a design whereby
callers using simple binary-data keys just specify HASH_BLOBS and don't
need to mess with specific support functions. hash_create() itself will
take care of optimizing when the key size is four bytes.
This nets out saving a few hundred bytes of code space, and offers
a measurable performance improvement in tidbitmap.c (which was not
exploiting the opportunity to use hash_uint32 for its 4-byte keys).
There might be some wins elsewhere too, I didn't analyze closely.
In future we could look into offering a similar optimized hashing function
for 8-byte keys. Under this design that could be done in a centralized
and machine-independent fashion, whereas getting it right for keys of
platform-dependent sizes would've been notationally painful before.
For the moment, the old way still works fine, so as not to break source
code compatibility for loadable modules. Eventually we might want to
remove tag_hash and friends from the exported API altogether, since there's
no real need for them to be explicitly referenced from outside dynahash.c.
Teodor Sigaev and Tom Lane
Two changes:
1. When copying a WAL segment from old timeline to create the first segment
on the new timeline, only copy up to the point where the timeline switch
happens, and zero-fill the rest. This avoids corner cases where we might
think that the copied WAL from the previous timeline belong to the new
timeline.
2. If the timeline switch happens at a segment boundary, don't copy the
whole old segment to the new timeline. It's pointless, because it's 100%
identical to the old segment.
When starting up from a basebackup taken off a standby extra logic has
to be applied to compute the point where the data directory is
consistent. Normal base backups use a WAL record for that purpose, but
that isn't possible on a standby.
That logic had a error check ensuring that the cluster's control file
indicates being in recovery. Unfortunately that check was too strict,
disregarding the fact that the control file could also indicate that
the cluster was shut down while in recovery.
That's possible when the a cluster starting from a basebackup is shut
down before the backup label has been removed. When everything goes
well that's a short window, but when either restore_command or
primary_conninfo isn't configured correctly the window can get much
wider. That's because inbetween reading and unlinking the label we
restore the last checkpoint from WAL which can need additional WAL.
To fix simply also allow starting when the control file indicates
"shutdown in recovery". There's nicer fixes imaginable, but they'd be
more invasive.
Backpatch to 9.2 where support for taking basebackups from standbys
was added.
In passing, also make some debugging elog's in pgstat.c a bit more
consistently worded.
Back-patch as far as applicable (9.3 or 9.4; none of these mistakes are
really old).
Mark Dilger identified and patched the type violations; the message
rewordings are mine.
Rename parameter action_at_recovery_target to
recovery_target_action suggested by Christoph Berg.
Place into recovery.conf suggested by Fujii Masao,
replacing (deprecating) earlier parameters, per
Michael Paquier.
It was an oversight in the original commit.
Also note in the sample config file that changing wal_log_hints requires a
restart.
Michael Paquier. Backpatch to 9.4, where wal_log_hints was added.
Transactions can now set their commit timestamp directly as they commit,
or an external transaction commit timestamp can be fed from an outside
system using the new function TransactionTreeSetCommitTsData(). This
data is crash-safe, and truncated at Xid freeze point, same as pg_clog.
This module is disabled by default because it causes a performance hit,
but can be enabled in postgresql.conf requiring only a server restart.
A new test in src/test/modules is included.
Catalog version bumped due to the new subdirectory within PGDATA and a
couple of new SQL functions.
Authors: Álvaro Herrera and Petr Jelínek
Reviewed to varying degrees by Michael Paquier, Andres Freund, Robert
Haas, Amit Kapila, Fujii Masao, Jaime Casanova, Simon Riggs, Steven
Singer, Peter Eisentraut
Get rid of PG_FUNCTION_INFO_V1() macros, which are quite inappropriate
for built-in functions (possibly leftovers from testing as a loadable
module?). Also, fix gratuitous inconsistency between SQL-level and
C-level names of the minmax support functions.
InitXLogInsert() cannot be called in a critical section, because it
allocates memory. But CreateCheckPoint() did that, when called for the
end-of-recovery checkpoint by the startup process.
In the passing, fix the scratch space allocation in InitXLogInsert to go to
the right memory context. Also update the comment at InitXLOGAccess, which
hasn't been totally accurate since hot standby was introduced (in a hot
standby backend, InitXLOGAccess isn't called at backend startup).
Reported by Michael Paquier
This gives an overview of what Lehman & Yao's paper is all about, so that
you can understand the rest of the README without having to read the paper.
Per discussion with Peter Geoghegan and others.
Add a new XLOG_FPI_FOR_HINT record type, and use that for full-page images
generated for hint bit updates, when checksums are enabled. The new record
type is replayed exactly the same as XLOG_FPI, but allows them to be tallied
separately e.g. in pg_xlogdump.
There seems no prospect that any of this will ever be useful, and indeed
it's questionable whether some of it would work if it ever got called;
it's certainly not been exercised in a very long time, if ever. So let's
get rid of it, and make the comments about mark/restore in execAmi.c less
wishy-washy.
The mark/restore support for Result nodes is also currently dead code,
but that's due to planner limitations not because it's impossible that
it could be useful. So I left it in.
It's a false positive - the variable is only used when 'onleft' is true,
and it is initialized in that case. But the compiler doesn't necessarily
see that.
Each WAL record now carries information about the modified relation and
block(s) in a standardized format. That makes it easier to write tools that
need that information, like pg_rewind, prefetching the blocks to speed up
recovery, etc.
There's a whole new API for building WAL records, replacing the XLogRecData
chains used previously. The new API consists of XLogRegister* functions,
which are called for each buffer and chunk of data that is added to the
record. The new API also gives more control over when a full-page image is
written, by passing flags to the XLogRegisterBuffer function.
This also simplifies the XLogReadBufferForRedo() calls. The function can dig
the relation and block number from the WAL record, so they no longer need to
be passed as arguments.
For the convenience of redo routines, XLogReader now disects each WAL record
after reading it, copying the main data part and the per-block data into
MAXALIGNed buffers. The data chunks are not aligned within the WAL record,
but the redo routines can assume that the pointers returned by XLogRecGet*
functions are. Redo routines are now passed the XLogReaderState, which
contains the record in the already-disected format, instead of the plain
XLogRecord.
The new record format also makes the fixed size XLogRecord header smaller,
by removing the xl_len field. The length of the "main data" portion is now
stored at the end of the WAL record, and there's a separate header after
XLogRecord for it. The alignment padding at the end of XLogRecord is also
removed. This compansates for the fact that the new format would otherwise
be more bulky than the old format.
Reviewed by Andres Freund, Amit Kapila, Michael Paquier, Alvaro Herrera,
Fujii Masao.
For <, <=, > and >= strategies, mark the first scan key
as already matched if scanning in an appropriate direction.
If index tuple contains no nulls we can skip the first
re-check for each tuple.
Author: Rajeev Rastogi
Reviewer: Haribabu Kommi
Rework of the code and comments by Simon Riggs
There was some confusion on how to record the case that the operation
unlinks the last non-leaf page in the branch being deleted.
_bt_unlink_halfdead_page set the "topdead" field in the WAL record to
the leaf page, but the redo routine assumed that it would be an invalid
block number in that case. This commit fixes _bt_unlink_halfdead_page to
do what the redo routine expected.
This code is new in 9.4, so backpatch there.
Unlogged relations are reset at the end of crash recovery as they're
only synced to disk during a proper shutdown. Unfortunately that and
later steps can fail, e.g. due to running out of space. This reset
was, up to now performed after marking the database as having finished
crash recovery successfully. As out of space errors trigger a crash
restart that could lead to the situation that not all unlogged
relations are reset.
Once that happend usage of unlogged relations could yield errors like
"could not open file "...": No such file or directory". Luckily
clusters that show the problem can be fixed by performing a immediate
shutdown, and starting the database again.
To fix, just call ResetUnloggedRelations(UNLOGGED_RELATION_INIT)
earlier, before marking the database as having successfully recovered.
Discussion: 20140912112246.GA4984@alap3.anarazel.de
Backpatch to 9.1 where unlogged tables were introduced.
Abhijit Menon-Sen and Andres Freund
The initial patch for RLS mistakenly included headers associated with
the executor and planner bits in rewrite/rowsecurity.h. Per policy and
general good sense, executor headers should not be included in planner
headers or vice versa.
The include of execnodes.h was a mistaken holdover from previous
versions, while the include of relation.h was used for Relation's
definition, which should have been coming from utils/relcache.h. This
patch cleans these issues up, adds comments to the RowSecurityPolicy
struct and the RowSecurityConfigType enum, and changes Relation->rsdesc
to Relation->rd_rsdesc to follow Relation field naming convention.
Additionally, utils/rel.h was including rewrite/rowsecurity.h, which
wasn't a great idea since that was pulling in things not really needed
in utils/rel.h (which gets included in quite a few places). Instead,
use 'struct RowSecurityDesc' for the rd_rsdesc field and add comments
explaining why.
Lastly, add an include into access/nbtree/nbtsort.c for
utils/sortsupport.h, which was evidently missed due to the above mess.
Pointed out by Tom in 16970.1415838651@sss.pgh.pa.us; note that the
concerns regarding a similar situation in the custom-path commit still
need to be addressed.
This function has a loop which can lead to uninterruptible process
"stalls" (actually infinite loops) when some bugs are triggered. Avoid
that unpleasant situation by adding a check for interrupts in a place
that shouldn't degrade performance in the normal case.
Backpatch to 9.3. Older branches have an identical loop here, but the
aforementioned bugs are only a problem starting in 9.3 so there doesn't
seem to be any point in backpatching any further.
There was a window in RestoreBackupBlock where a page would be zeroed out,
but not yet locked. If a backend pinned and locked the page in that window,
it saw the zeroed page instead of the old page or new page contents, which
could lead to missing rows in a result set, or errors.
To fix, replace RBM_ZERO with RBM_ZERO_AND_LOCK, which atomically pins,
zeroes, and locks the page, if it's not in the buffer cache already.
In stable branches, the old RBM_ZERO constant is renamed to RBM_DO_NOT_USE,
to avoid breaking any 3rd party extensions that might use RBM_ZERO. More
importantly, this avoids renumbering the other enum values, which would
cause even bigger confusion in extensions that use ReadBufferExtended, but
haven't been recompiled.
Backpatch to all supported versions; this has been racy since hot standby
was introduced.
Previously the maximum size of GIN pending list was controlled only by
work_mem. But the reasonable value of work_mem and the reasonable size
of the list are basically not the same, so it was not appropriate to
control both of them by only one GUC, i.e., work_mem. This commit
separates new GUC, pending_list_cleanup_size, from work_mem to allow
users to control only the size of the list.
Also this commit adds pending_list_cleanup_size as new storage parameter
to allow users to specify the size of the list per index. This is useful,
for example, when users want to increase the size of the list only for
the GIN index which can be updated heavily, and decrease it otherwise.
Reviewed by Etsuro Fujita.
The code that generates the BRIN_XLOG_UPDATE removes the buffer
reference when the page that's target for the updated tuple is freshly
initialized. This is a pretty usual optimization, but was breaking the
case where the revmap buffer, which is referenced in the same WAL
record, is getting a backup block: the replay code was using backup
block index 1, which is not valid when the update target buffer gets
pruned; the revmap buffer gets assigned 0 instead. Make sure to use the
correct backup block index for revmap when replaying.
Bug reported by Fujii Masao.
Besides a couple of typo fixes, per David Rowley, Thom Brown, and Amit
Langote, and mentions of BRIN in the general CREATE INDEX page again per
David, this includes silencing MSVC compiler warnings (thanks Microsoft)
and an additional variable initialization per Coverity scanner.
Reported by David Rowley: variadic macros are a problem. Get rid of
them using a trick suggested by Tom Lane: add extra parentheses where
needed. In the future we might decide we don't need the calls at all
and remove them, but it seems appropriate to keep them while this code
is still new.
Also from David Rowley: brininsert() was trying to use a variable before
initializing it. Fix by moving the brin_form_tuple call (which
initializes the variable) to within the locked section.
Reported by Peter Eisentraut: can't use "new" as a struct member name,
because C++ compilers will choke on it, as reported by cpluspluscheck.
Now that the backup blocks are appended to the WAL record in xloginsert.c,
XLogInsert doesn't see them anymore and cannot remove them from the version
reconstructed for xlog_outdesc. This makes running with wal_debug=on more
expensive, as we now make (unnecessary) temporary copies of the backup
blocks, but it doesn't seem worth convoluting the code to keep that
optimization.
Reported by Alvaro Herrera.
BRIN is a new index access method intended to accelerate scans of very
large tables, without the maintenance overhead of btrees or other
traditional indexes. They work by maintaining "summary" data about
block ranges. Bitmap index scans work by reading each summary tuple and
comparing them with the query quals; all pages in the range are returned
in a lossy TID bitmap if the quals are consistent with the values in the
summary tuple, otherwise not. Normal index scans are not supported
because these indexes do not store TIDs.
As new tuples are added into the index, the summary information is
updated (if the block range in which the tuple is added is already
summarized) or not; in the latter case, a subsequent pass of VACUUM or
the brin_summarize_new_values() function will create the summary
information.
For data types with natural 1-D sort orders, the summary info consists
of the maximum and the minimum values of each indexed column within each
page range. This type of operator class we call "Minmax", and we
supply a bunch of them for most data types with B-tree opclasses.
Since the BRIN code is generalized, other approaches are possible for
things such as arrays, geometric types, ranges, etc; even for things
such as enum types we could do something different than minmax with
better results. In this commit I only include minmax.
Catalog version bumped due to new builtin catalog entries.
There's more that could be done here, but this is a good step forwards.
Loosely based on ideas from Simon Riggs; code mostly by Álvaro Herrera,
with contribution by Heikki Linnakangas.
Patch reviewed by: Amit Kapila, Heikki Linnakangas, Robert Haas.
Testing help from Jeff Janes, Erik Rijkers, Emanuel Calvo.
PS:
The research leading to these results has received funding from the
European Union's Seventh Framework Programme (FP7/2007-2013) under
grant agreement n° 318633.
The code that generated a record to clear the F_TUPLES_DELETED flag hasn't
existed since we got rid of old-style VACUUM FULL. I kept the code that sets
the flag, although it's not used for anything anymore, because it might
still be interesting information for debugging purposes that some tuples
have been deleted from a page.
Likewise, the code to turn the root page from non-leaf to leaf page was
removed when we got rid of old-style VACUUM FULL. Remove the code to replay
that action, too.
Previously .ready file was created for the timeline history file at the end
of an archive recovery even when WAL archiving was not enabled.
This creation is unnecessary and causes .ready file to remain infinitely.
This commit changes an archive recovery so that it creates .ready file for
the timeline history file only when WAL archiving is enabled.
Backpatch to all supported versions.
xlog.c is huge, this makes it a little bit smaller, which is nice. Functions
related to putting together the WAL record are in xloginsert.c, and the
lower level stuff for managing WAL buffers and such are in xlog.c.
Also move the definition of XLogRecord to a separate header file. This
causes churn in the #includes of all the files that write WAL records, and
redo routines, but it avoids pulling in xlog.h into most places.
Reviewed by Michael Paquier, Alvaro Herrera, Andres Freund and Amit Kapila.
The old algorithm was found to not be the usual CRC-32 algorithm, used by
Ethernet et al. We were using a non-reflected lookup table with code meant
for a reflected lookup table. That's a strange combination that AFAICS does
not correspond to any bit-wise CRC calculation, which makes it difficult to
reason about its properties. Although it has worked well in practice, seems
safer to use a well-known algorithm.
Since we're changing the algorithm anyway, we might as well choose a
different polynomial. The Castagnoli polynomial has better error-correcting
properties than the traditional CRC-32 polynomial, even if we had
implemented it correctly. Another reason for picking that is that some new
CPUs have hardware support for calculating CRC-32C, but not CRC-32, let
alone our strange variant of it. This patch doesn't add any support for such
hardware, but a future patch could now do that.
The old algorithm is kept around for tsquery and pg_trgm, which use the
values in indexes that need to remain compatible so that pg_upgrade works.
While we're at it, share the old lookup table for CRC-32 calculation
between hstore, ltree and core. They all use the same table, so might as
well.
Instead of initializing a new TransInvalidationInfo for every
transaction or subtransaction, we can just do it for those
transactions or subtransactions that actually need to queue
invalidation messages. That also avoids needing to free those
entries at the end of a transaction or subtransaction that does
not generate any invalidation messages, which is by far the
common case.
Patch by me. Review by Simon Riggs and Andres Freund.
Previously the archive recovery always created .ready file for
the last WAL file of the old timeline at the end of recovery even when
it's restored from the archive and has .done file. That is, there was
the case where the WAL file had both .ready and .done files.
This caused the already-archived WAL file to be archived again.
This commit prevents the archive recovery from creating .ready file
for the last WAL file if it has .done file, in order to prevent it from
being archived again.
This bug was added when cascading replication feature was introduced,
i.e., the commit 5286105800.
So, back-patch to 9.2, where cascading replication was added.
Reviewed by Michael Paquier
The duplication originated in cdd46c765, where restartpoints were
introduced.
In LogCheckpointStart's case the duplication actually lead to the
compiler's format string checking not to be effective because the
format string wasn't constant.
Arguably these messages shouldn't be elog(), but ereport() style
messages. That'd even allow to translate the messages... But as
there's more mistakes of that kind in surrounding code, it seems
better to change that separately.
Commit 7dbb606938 added a new CHECKPOINT_FLUSH_ALL flag. As that
commit needed to be backpatched I didn't change the numeric values of
the existing flags as that could lead to nastly problems if any
external code issued checkpoints. That's not a concern on master, so
renumber them there.
Also add a comment about CHECKPOINT_FLUSH_ALL above
CreateCheckPoint().
CREATE DATABASE and ALTER DATABASE .. SET TABLESPACE copy the source
database directory on the filesystem level. To ensure the on disk
state is consistent they block out users of the affected database and
force a checkpoint to flush out all data to disk. Unfortunately, up to
now, that checkpoint didn't flush out dirty buffers from unlogged
relations.
That bug means there could be leftover dirty buffers in either the
template database, or the database in its old location. Leading to
problems when accessing relations in an inconsistent state; and to
possible problems during shutdown in the SET TABLESPACE case because
buffers belonging files that don't exist anymore are flushed.
This was reported in bug #10675 by Maxim Boguk.
Fix by Pavan Deolasee, modified somewhat by me. Reviewed by MauMau and
Fujii Masao.
Backpatch to 9.1 where unlogged tables were introduced.
The new header contains many prototypes for functions in ruleutils.c
that are not exposed to the SQL level.
Reviewed by Andres Freund and Michael Paquier.
This clause changes the behavior of SELECT locking clauses in the
presence of locked rows: instead of causing a process to block waiting
for the locks held by other processes (or raise an error, with NOWAIT),
SKIP LOCKED makes the new reader skip over such rows. While this is not
appropriate behavior for general purposes, there are some cases in which
it is useful, such as queue-like tables.
Catalog version bumped because this patch changes the representation of
stored rules.
Reviewed by Craig Ringer (based on a previous attempt at an
implementation by Simon Riggs, who also provided input on the syntax
used in the current patch), David Rowley, and Álvaro Herrera.
Author: Thomas Munro
The page splitting code would go into infinite recursion if you try to
insert an index tuple that doesn't fit even on an empty page.
Per analysis and suggested fix by Andrew Gierth. Fixes bug #11555, reported
by Bryan Seitz (analysis happened over IRC). Backpatch to all supported
versions.
I left the GUC in place for the beta period, so that people could experiment
with different values. No-one's come up with any data that a different value
would be better under some circumstances, so rather than try to document to
users what the GUC, let's just hard-code the current value, 8.
As noted in http://bugs.debian.org/763098 there is a conflict between
postgres' definition of CACHE_LINE_SIZE and the definition by various
*bsd platforms. It's debatable who has the right to define such a
name, but postgres' use was only introduced in 375d8526f2 (9.4), so
it seems like a good idea to rename it.
Discussion: 20140930195756.GC27407@msg.df7cb.de
Per complaint of Christoph Berg in the above email, although he's not
the original bug reporter.
Backpatch to 9.4 where the define was introduced.
For the reason outlined in df4077cda2 also remove volatile qualifiers
from xlog.c. Some of these uses of volatile have been added after
noticing problems back when spinlocks didn't imply compiler
barriers. So they are a good test - in fact removing the volatiles
breaks when done without the barriers in spinlocks present.
Several uses of volatile remain where they are explicitly used to
access shared memory without locks. These locations are ok with
slightly out of date data, but removing the volatile might lead to the
variables never being reread from memory. These uses could also be
replaced by barriers, but that's a separate change of doubtful value.
There are four weaknesses in728f152e07f998d2cb4fe5f24ec8da2c3bda98f2:
* append_init() in heapdesc.c was ugly and required that rm_identify
return values are only valid till the next call. Instead just add a
couple more switch() cases for the INIT_PAGE cases. Now the returned
value will always be valid.
* a couple rm_identify() callbacks missed masking xl_info with
~XLR_INFO_MASK.
* pg_xlogdump didn't map a NULL rm_identify to UNKNOWN or a similar
string.
* append_init() was called when id=NULL - which should never actually
happen. But it's better to be careful.
This is primarily useful for the upcoming pg_xlogdump --stats feature,
but also allows to remove some duplicated code in the rmgr_desc
routines.
Due to the separation and harmonization, the output of dipsplayed
records changes somewhat. But since this isn't enduser oriented
content that's ok.
It's potentially desirable to further change pg_xlogdump's display of
records. It previously wasn't possible to show the record type
separately from the description forcing it to be in the last
column. But that's better done in a separate commit.
Author: Abhijit Menon-Sen, slightly editorialized by me
Reviewed-By: Álvaro Herrera, Andres Freund, and Heikki Linnakangas
Discussion: 20140604104716.GA3989@toroid.org
The code that tried to split a page at 75/25 ratio, when appending to the
end of an index, was buggy in two ways. First, there was a silly typo that
caused it to just fill the left page as full as possible. But the logic as
it was intended wasn't correct either, and would actually have given a ratio
closer to 60/40 than 75/25.
Gaetano Mendola spotted the typo. Backpatch to 9.4, where this code was added.
Every redo routine uses the same idiom to determine what to do to a page:
check if there's a backup block for it, and if not read, the buffer if the
block exists, and check its LSN. Refactor that into a common function,
XLogReadBufferForRedo, making all the redo routines shorter and more
readable.
This has no user-visible effect, and makes no changes to the WAL format.
Reviewed by Andres Freund, Alvaro Herrera, Michael Paquier.
The list of posting lists it's dealing with can contain placeholders for
deleted posting lists. The placeholders are kept around so that they can
be WAL-logged, but we must be careful to not try to access them.
This fixes bug #11280, reported by Mårten Svantesson. Backpatch to 9.4,
where the compressed data leaf page code was added.
This reverts commit e23014f3d4.
As the side effect of the reverted commit, when the unit is
specified, the reloption was stored in the catalog with the unit.
This broke pg_dump (specifically, it prevented pg_dump from
outputting restorable backup regarding the reloption) and
turned the buildfarm red. Revert the commit until the fixed
version is ready.
This introduces an infrastructure which allows us to specify the units
like ms (milliseconds) in integer relation option, like GUC parameter.
Currently only autovacuum_vacuum_cost_delay reloption can accept
the units.
Reviewed by Michael Paquier
There's no point in setting up a context error callback when doing
conditional lock acquisition, because we never actually wait and so the
user wouldn't be able to see the context message anywhere. In fact,
this is more in line with what ConditionalXactLockTableWait is doing.
Backpatch to 9.4, where this was added.
The symbol was added by 71901ab6d; the original code was introduced by
6868ed749. Development of both overlapped which is why we apparently
failed to notice.
This is a (very slight) behavior change, so I'm not backpatching this to
9.4 for now, even though the symbol does exist there.
strncmp() is a specialized API unsuited for routine copying into
fixed-size buffers. On a system where the length of a single filename
can exceed MAXPGPATH, the pg_archivecleanup change prevents a simple
crash in the subsequent strlen(). Few filesystems support names that
long, and calling pg_archivecleanup with untrusted input is still not a
credible use case. Therefore, no back-patch.
David Rowley
parseRelOptions() tended to leak memory in the caller's context. Most
of the time this doesn't really matter since the caller's context is
at most query-lifespan, and the function won't be invoked very many times.
However, when testing with CLOBBER_CACHE_RECURSIVELY, the same relcache
entry can get rebuilt a *lot* of times in one query, leading to significant
intraquery memory bloat if it has any reloptions. Noted while
investigating a related report from Tomas Vondra.
In passing, get rid of some Asserts that are redundant with the one
done by deconstruct_array().
As with other patches to avoid leaks in CLOBBER_CACHE testing, it doesn't
really seem worth back-patching this.
log_newpage is used by many indexams, in addition to heap, but for
historical reasons it's always been part of the heapam rmgr. Starting with
9.3, we have another WAL record type for logging an image of a page,
XLOG_FPI. Simplify things by moving log_newpage and log_newpage_buffer to
xlog.c, and switch to using the XLOG_FPI record type.
Bump the WAL version number because the code to replay the old HEAP_NEWPAGE
records is removed.
Commit 0ac5ad5134 removed an optimization in multixact.c that skipped
fetching members of MultiXactId that were older than our
OldestVisibleMXactId value. The reason this was removed is that it is
possible for multixacts that contain updates to be older than that
value. However, if the caller is certain that the multi does not
contain an update (because the infomask bits say so), it can pass this
info down to GetMultiXactIdMembers, enabling it to use the old
optimization.
Pointed out by Andres Freund in 20131121200517.GM7240@alap2.anarazel.de
Testing for abortedness of a multixact member that's being frozen is
unnecessary: we only need to know whether the transaction is still in
progress or committed to determine whether it must be kept or not. This
let us simplify the code a bit and avoid a useless TransactionIdDidAbort
test.
Suggested by Andres Freund awhile back.
There were several oversights in recovery code where COMMIT/ABORT PREPARED
records were ignored:
* pg_last_xact_replay_timestamp() (wasn't updated for 2PC commits)
* recovery_min_apply_delay (2PC commits were applied immediately)
* recovery_target_xid (recovery would not stop if the XID used 2PC)
The first of those was reported by Sergiy Zuban in bug #11032, analyzed by
Tom Lane and Andres Freund. The bug was always there, but was masked before
commit d19bd29f07, because COMMIT PREPARED
always created an extra regular transaction that was WAL-logged.
Backpatch to all supported versions (older versions didn't have all the
features and therefore didn't have all of the above bugs).
Nothing in the checkpointer calls InitXLOGAccess(), so WALInsertLocks
never got initialized there. Without EXEC_BACKEND, it works anyway
because the correct value is inherited from the postmaster, but
with EXEC_BACKEND we've got a problem. The problem appears to have
been introduced by commit 68a2e52bba.
To fix, move the relevant initialization steps from InitXLOGAccess()
to XLOGShmemInit(), making this more parallel to what we do
elsewhere.
Amit Kapila
Commit 1b86c81d2d fixed the decoding of toasted columns for the rows
contained in one xl_heap_multi_insert record. But that's not actually
enough, because heap_multi_insert() will actually first toast all
passed in rows and then emit several *_multi_insert records; one for
each page it fills with tuples.
Add a XLOG_HEAP_LAST_MULTI_INSERT flag which is set in
xl_heap_multi_insert->flag denoting that this multi_insert record is
the last emitted by one heap_multi_insert() call. Then use that flag
in decode.c to only set clear_toast_afterwards in the right situation.
Expand the number of rows inserted via COPY in the corresponding
regression test to make sure that more than one heap page is filled
with tuples by one heap_multi_insert() call.
Backpatch to 9.4 like the previous commit.
The old name wasn't very descriptive as of actual contents of the
directory, which are historical snapshots in the snapshots/
subdirectory and mappingdata for rewritten tuples in
mappings/. There's been a fair amount of discussion what would be a
good name. I'm settling for pg_logical because it's likely that
further data around logical decoding and replication will need saving
in the future.
Also add the missing entry for the directory into storage.sgml's list
of PGDATA contents.
Bumps catversion as the data directories won't be compatible.
The previous code, perhaps out of concern for avoid memory leaks, formed
the tuple in one memory context and then copied it to another memory
context. However, this doesn't appear to be necessary, since
index_form_tuple and the functions it calls take precautions against
leaking memory. In my testing, building the tuple directly inside the
sort context shaves several percent off the index build time.
Rearrange things so we do that.
Patch by me. Review by Amit Kapila, Tom Lane, Andres Freund.
The assertion failed if WAL_DEBUG or LWLOCK_STATS was enabled; fix that by
using separate memory contexts for the allocations made within those code
blocks.
This patch introduces a mechanism for marking any memory context as allowed
in a critical section. Previously ErrorContext was exempt as a special case.
Instead of a blanket exception of the checkpointer process, only exempt the
memory context used for the pending ops hash table.
Instead of truncating pg_multixact at vacuum time, do it only at
checkpoint time. The reason for doing it this way is twofold: first, we
want it to delete only segments that we're certain will not be required
if there's a crash immediately after the removal; and second, we want to
do it relatively often so that older files are not left behind if
there's an untimely crash.
Per my proposal in
http://www.postgresql.org/message-id/20140626044519.GJ7340@eldon.alvh.no-ip.org
we now execute the truncation in the checkpointer process rather than as
part of vacuum. Vacuum is in only charge of maintaining in shared
memory the value to which it's possible to truncate the files; that
value is stored as part of checkpoints also, and so upon recovery we can
reuse the same value to re-execute truncate and reset the
oldest-value-still-safe-to-use to one known to remain after truncation.
Per bug reported by Jeff Janes in the course of his tests involving
bug #8673.
While at it, update some comments that hadn't been updated since
multixacts were changed.
Backpatch to 9.3, where persistency of pg_multixact files was
introduced by commit 0ac5ad5134.
Don't assert MultiXactIdIsRunning if the multi came from a tuple that
had been share-locked and later copied over to the new cluster by
pg_upgrade. Doing that causes an error to be raised unnecessarily:
MultiXactIdIsRunning is not open to the possibility that its argument
came from a pg_upgraded tuple, and all its other callers are already
checking; but such multis cannot, obviously, have transactions still
running, so the assert is pointless.
Noticed while investigating the bogus pg_multixact/offsets/0000 file
left over by pg_upgrade, as reported by Andres Freund in
http://www.postgresql.org/message-id/20140530121631.GE25431@alap3.anarazel.de
Backpatch to 9.3, as the commit that introduced the buglet.
Normally, this won't matter too much; but if I/O is really slow, for
example because the system is overloaded, we might write many pages
before checking for interrupts. A single toast insertion might
write up to 1GB of data, and a multi-insert could write hundreds
of tuples (and their corresponding TOAST data).
The existance of the assert_enabled variable (backing the
debug_assertions GUC) reduced the amount of knowledge some static code
checkers (like coverity and various compilers) could infer from the
existance of the assertion. That could have been solved by optionally
removing the assertion_enabled variable from the Assert() et al macros
at compile time when some special macro is defined, but the resulting
complication doesn't seem to be worth the gain from having
debug_assertions. Recompiling is fast enough.
The debug_assertions GUC is still available, but readonly, as it's
useful when diagnosing problems. The commandline/client startup option
-A, which previously also allowed to enable/disable assertions, has
been removed as it doesn't serve a purpose anymore.
While at it, reduce code duplication in bufmgr.c and localbuf.c
assertions checking for spurious buffer pins. That code had to be
reindented anyway to cope with the assert_enabled removal.
Change the order of checks in similar functions to be the same; remove
a parameter that's not needed anymore; rename a memory context and
expand a couple of comments.
Per review comments from Amit Kapila
Previously, the code used a node label of zero both for strings that
contain no bytes beyond the inner tuple's prefix, and for cases where an
"allTheSame" inner tuple has to be split to allow a string with a different
next byte to be inserted into it. Failing to distinguish these cases meant
that if a string ending with the current prefix needed to be inserted into
an allTheSame tuple, we got into an infinite loop, because after splitting
the tuple we'd descend into the child allTheSame tuple and then find we
need to split again.
To fix, instead use -1 and -2 as the node labels for these two cases.
This requires widening the node label type from "char" to int2, but
fortunately SPGiST stores all pass-by-value node label types in their
Datum representation, which means that this change is transparently upward
compatible so far as the on-disk representation goes. We continue to
recognize zero as a dummy node label for reading purposes, but will not
attempt to push new index entries down into such a label, so that the loop
won't occur even when dealing with an existing index.
Per report from Teodor Sigaev. Back-patch to 9.2 where the faulty
code was introduced.
In a50d976254 I already changed this, but got it wrong for the case
where the number of members is larger than the number of entries that
fit in the last page of the last segment.
As reported by Serge Negodyuck in a followup to bug #8673.
It's critical that the backend's idea of LOBLKSIZE match the way data has
actually been divided up in pg_largeobject. While we don't provide any
direct way to adjust that value, doing so is a one-line source code change
and various people have expressed interest recently in changing it. So,
just as with TOAST_MAX_CHUNK_SIZE, it seems prudent to record the value in
pg_control and cross-check that the backend's compiled-in setting matches
the on-disk data.
Also tweak the code in inv_api.c so that fetches from pg_largeobject
explicitly verify that the length of the data field is not more than
LOBLKSIZE. Formerly we just had Asserts() for that, which is no protection
at all in production builds. In some of the call sites an overlength data
value would translate directly to a security-relevant stack clobber, so it
seems worth one extra runtime comparison to be sure.
In the back branches, we can't change the contents of pg_control; but we
can still make the extra checks in inv_api.c, which will offer some amount
of protection against running with the wrong value of LOBLKSIZE.
Previously there's been a mix between 'slotname' and 'slot_name'. It's
not nice to be unneccessarily inconsistent in a new feature. As a post
beta1 initdb now is required in the wake of eeca4cd35e, fix the
inconsistencies.
Most the changes won't affect usage of replication slots because the
majority of changes is around function parameter names. The prominent
exception to that is that the recovery.conf parameter
'primary_slotname' is now named 'primary_slot_name'.
The way the code was written, the padding was copied from uninitialized
memory areas.. Because the structs are local variables in the code where
the WAL records are constructed, making them larger and zeroing the padding
bytes would not make the code very pretty, so rather than fixing this
directly by zeroing out the padding bytes, it seems more clear to not try to
align the tuples in the WAL records. The redo functions are taught to copy
the tuple header to a local variable to avoid unaligned access.
Stable-branches have the same problem, but we can't change the WAL format
there, so fix in master only. Reading a few random extra bytes at the stack
is harmless in practice, so it's not worth crafting a different
back-patchable fix.
Per reports from Kevin Grittner and Andres Freund, using clang static
analyzer and Valgrind, respectively.
The new page deletion code didn't cope with the case the target page's
right sibling was marked half-dead. It failed a sanity check which checked
that the downlinks in the parent page match the lower level, because a
half-dead page has no downlink. To cope, check for that condition, and
just give up on the deletion if it happens. The vacuum will finish the
deletion of the half-dead page when it gets there, and on the next vacuum
after that the empty can be deleted.
Reported by Jeff Janes.
I got the backup block numbers off-by-one in the commit that changed the
way incomplete-splits are handled. I blame the comments, which said
"backup block 1" and "backup block 2", even though the backup blocks
are numbered starting from 0, in the macros and functions used in replay.
Fix the comments and the code.
Per Jeff Janes' bug report about corruption caused by torn page writes.
The incorrect code is new in git master, but backpatch the comment change
down to 9.0, where the numbering in the redo-side macros was changed.
transam/README explained how B-tree incomplete splits were tracked and
fixed after recovery, as an example of handling complex actions that need
multiple WAL records, but that's not how it works anymore. Explain the new
paradigm.
Commit dd428c79 added dbId and tsId to the xl_xact_commit struct but missed
that prepared transaction commits reuse that struct. Fix that.
Because those fields were left unitialized, replaying a commit prepared WAL
record in a hot standby node would fail to remove the relcache init file.
That can lead to "could not open file" errors on the standby. Relcache init
file only needs to be removed when a system table/index is rewritten in the
transaction using two phase commit, so that should be rare in practice. In
HEAD, the incorrect dbId/tsId values are also used for filtering in logical
replication code, causing the transaction to always be filtered out.
Analysis and fix by Andres Freund. Backpatch to 9.0 where hot standby was
introduced.
To lock a prepared transaction's shared memory entry, we used to mark it
with the XID of the backend. When the XID was no longer active according
to the proc array, the entry was implicitly considered as not locked
anymore. However, when preparing a transaction, the backend's proc array
entry was cleared before transfering the locks (and some other state) to
the prepared transaction's dummy PGPROC entry, so there was a window where
another backend could finish the transaction before it was in fact fully
prepared.
To fix, rewrite the locking mechanism of global transaction entries. Instead
of an XID, just have simple locked-or-not flag in each entry (we store the
locking backend's backend id rather than a simple boolean, but that's just
for debugging purposes). The backend is responsible for explicitly unlocking
the entry, and to make sure that that happens, install a callback to unlock
it on abort or process exit.
Backpatch to all supported versions.
rd_replidindex should be managed the same as rd_oidindex, and rd_keyattr
and rd_idattr should be managed like rd_indexattr. Omissions in this area
meant that the bitmapsets computed for rd_keyattr and rd_idattr would be
leaked during any relcache flush, resulting in a slow but permanent leak in
CacheMemoryContext. There was also a tiny probability of relcache entry
corruption if we ran out of memory at just the wrong point in
RelationGetIndexAttrBitmap. Otherwise, the fields were not zeroed where
expected, which would not bother the code any AFAICS but could greatly
confuse anyone examining the relcache entry while debugging.
Also, create an API function RelationGetReplicaIndex rather than letting
non-relcache code be intimate with the mechanisms underlying caching of
that value (we won't even mention the memory leak there).
Also, fix a relcache flush hazard identified by Andres Freund:
RelationGetIndexAttrBitmap must not assume that rd_replidindex stays valid
across index_open.
The aspects of this involving rd_keyattr date back to 9.3, so back-patch
those changes.
Per discussion, this seems like a more consistent choice of name.
Fabrízio de Royes Mello, after a suggestion by Peter Eisentraut;
some additional documentation wordsmithing by me
When returning rows from a bitmap, as done with partial match queries, we
would get stuck in an infinite loop if the bitmap contained a lossy page
reference.
This bug is new in master, it was introduced by the patch to allow skipping
items refuted by other entries in GIN scans.
Report and fix by Alexander Korotkov
To-be-deleted list pages contain no useful information, as they are being
deleted, but we must still protect the writes from being torn by a crash
after a partial write. To do that, re-initialize the pages on WAL replay.
Jeff Janes caught this with a test program to test partial writes.
Backpatch to all supported versions.
If they were not 'oldtup.t_data' would be dereferenced while set to NULL
in case of a full page image for block 0.
Do so primarily to silence coverity; but also to make sure this prerequisite
isn't changed without adapting the replay routine as that would appear to
work in many cases.
Andres Freund
If we have an array of records stored on disk, the individual record fields
cannot contain out-of-line TOAST pointers: the tuptoaster.c mechanisms are
only prepared to deal with TOAST pointers appearing in top-level fields of
a stored row. The same applies for ranges over composite types, nested
composites, etc. However, the existing code only took care of expanding
sub-field TOAST pointers for the case of nested composites, not for other
structured types containing composites. For example, given a command such
as
UPDATE tab SET arraycol = ARRAY[(ROW(x,42)::mycompositetype] ...
where x is a direct reference to a field of an on-disk tuple, if that field
is long enough to be toasted out-of-line then the TOAST pointer would be
inserted as-is into the array column. If the source record for x is later
deleted, the array field value would become a dangling pointer, leading
to errors along the line of "missing chunk number 0 for toast value ..."
when the value is referenced. A reproducible test case for this was
provided by Jan Pecek, but it seems likely that some of the "missing chunk
number" reports we've heard in the past were caused by similar issues.
Code-wise, the problem is that PG_DETOAST_DATUM() is not adequate to
produce a self-contained Datum value if the Datum is of composite type.
Seen in this light, the problem is not just confined to arrays and ranges,
but could also affect some other places where detoasting is done in that
way, for example form_index_tuple().
I tried teaching the array code to apply toast_flatten_tuple_attribute()
along with PG_DETOAST_DATUM() when the array element type is composite,
but this was messy and imposed extra cache lookup costs whether or not any
TOAST pointers were present, indeed sometimes when the array element type
isn't even composite (since sometimes it takes a typcache lookup to find
that out). The idea of extending that approach to all the places that
currently use PG_DETOAST_DATUM() wasn't attractive at all.
This patch instead solves the problem by decreeing that composite Datum
values must not contain any out-of-line TOAST pointers in the first place;
that is, we expand out-of-line fields at the point of constructing a
composite Datum, not at the point where we're about to insert it into a
larger tuple. This rule is applied only to true composite Datums, not
to tuples that are being passed around the system as tuples, so it's not
as invasive as it might sound at first. With this approach, the amount
of code that has to be touched for a full solution is greatly reduced,
and added cache lookup costs are avoided except when there actually is
a TOAST pointer that needs to be inlined.
The main drawback of this approach is that we might sometimes dereference
a TOAST pointer that will never actually be used by the query, imposing a
rather large cost that wasn't there before. On the other side of the coin,
if the field value is used multiple times then we'll come out ahead by
avoiding repeat detoastings. Experimentation suggests that common SQL
coding patterns are unaffected either way, though. Applications that are
very negatively affected could be advised to modify their code to not fetch
columns they won't be using.
In future, we might consider reverting this solution in favor of detoasting
only at the point where data is about to be stored to disk, using some
method that can drill down into multiple levels of nested structured types.
That will require defining new APIs for structured types, though, so it
doesn't seem feasible as a back-patchable fix.
Note that this patch changes HeapTupleGetDatum() from a macro to a function
call; this means that any third-party code using that macro will not get
protection against creating TOAST-pointer-containing Datums until it's
recompiled. The same applies to any uses of PG_RETURN_HEAPTUPLEHEADER().
It seems likely that this is not a big problem in practice: most of the
tuple-returning functions in core and contrib produce outputs that could
not possibly be toasted anyway, and the same probably holds for third-party
extensions.
This bug has existed since TOAST was invented, so back-patch to all
supported branches.
Commit a730183926 created rather a mess by
putting dependencies on backend-only include files into include/common.
We really shouldn't do that. To clean it up:
* Move TABLESPACE_VERSION_DIRECTORY back to its longtime home in
catalog/catalog.h. We won't consider this symbol part of the FE/BE API.
* Push enum ForkNumber from relfilenode.h into relpath.h. We'll consider
relpath.h as the source of truth for fork numbers, since relpath.c was
already partially serving that function, and anyway relfilenode.h was
kind of a random place for that enum.
* So, relfilenode.h now includes relpath.h rather than vice-versa. This
direction of dependency is fine. (That allows most, but not quite all,
of the existing explicit #includes of relpath.h to go away again.)
* Push forkname_to_number from catalog.c to relpath.c, just to centralize
fork number stuff a bit better.
* Push GetDatabasePath from catalog.c to relpath.c; it was rather odd
that the previous commit didn't keep this together with relpath().
* To avoid needing relfilenode.h in common/, redefine the underlying
function (now called GetRelationPath) as taking separate OID arguments,
and make the APIs using RelFileNode or RelFileNodeBackend into macro
wrappers. (The macros have a potential multiple-eval risk, but none of
the existing call sites have an issue with that; one of them had such a
risk already anyway.)
* Fix failure to follow the directions when "init" fork type was added;
specifically, the errhint in forkname_to_number wasn't updated, and neither
was the SGML documentation for pg_relation_size().
* Fix tablespace-path-too-long check in CreateTableSpace() to account for
fork-name component of maximum-length pathnames. This requires putting
FORKNAMECHARS into a header file, but it was rather useless (and
actually unreferenced) where it was.
The last couple of items are potentially back-patchable bug fixes,
if anyone is sufficiently excited about them; but personally I'm not.
Per a gripe from Christoph Berg about how include/common wasn't
self-contained.
In writeListPage, never take a full-page image of the page, because we
have all the information required to re-initialize in the WAL record
anyway. Before this fix, a full-page image was always generated, unless
full_page_writes=off, because when the page is initialized its LSN is
always 0. In stable-branches, keep the code to restore the backup blocks
if they exist, in case that the WAL is generated with an older minor
version, but in master Assert that there are no full-page images.
In the redo routine, add missing "off++". Otherwise the tuples are added
to the page in reverse order. That happens to be harmless because we
always scan and remove all the tuples together, but it was clearly wrong.
Also, it was masked by the first bug unless full_page_writes=off, because
the page was always restored from a full-page image.
Backpatch to all supported versions.
As noted some time ago, the original coding had a typo ("|" for "^")
that made the result less unique than intended. Even the intended
behavior is obsolete since it was based on wanting to produce a
usable value even if we didn't have int64 arithmetic --- a limitation
we stopped supporting years ago. Instead, let's redefine the system
identifier as tv_sec in the upper 32 bits (same as before), tv_usec
in the next 20 bits, and the low 12 bits of getpid() in the remaining
bits. This is still hardly guaranteed-universally-unique, but it's
noticeably better than before. Per my proposal at
<29019.1374535940@sss.pgh.pa.us>
If a tuple is locked, and this lock is later upgraded either to an
update or to a stronger lock, and in the meantime some other process
tries to lock, update or delete the same tuple, it (the tuple) could end
up being updated twice, or having conflicting locks held.
The reason for this is that the second updater checks for a change in
Xmax value, or in the HEAP_XMAX_IS_MULTI infomask bit, after noticing
the first lock; and if there's a change, it restarts and re-evaluates
its ability to update the tuple. But it neglected to check for changes
in lock strength or in lock-vs-update status when those two properties
stayed the same. This would lead it to take the wrong decision and
continue with its own update, when in reality it shouldn't do so but
instead restart from the top.
This could lead to either an assertion failure much later (when a
multixact containing multiple updates is detected), or duplicate copies
of tuples.
To fix, make sure to compare the other relevant infomask bits alongside
the Xmax value and HEAP_XMAX_IS_MULTI bit, and restart from the top if
necessary.
Also, in the belt-and-suspenders spirit, add a check to
MultiXactCreateFromMembers that a multixact being created does not have
two or more members that are claimed to be updates. This should protect
against other bugs that might cause similar bogus situations.
Backpatch to 9.3, where the possibility of multixacts containing updates
was introduced. (In prior versions it was possible to have the tuple
lock upgraded from shared to exclusive, and an update would not restart
from the top; yet we're protected against a bug there because there's
always a sleep to wait for the locking transaction to complete before
continuing to do anything. Really, the fact that tuple locks always
conflicted with concurrent updates is what protected against bugs here.)
Per report from Andrew Dunstan and Josh Berkus in thread at
http://www.postgresql.org/message-id/534C8B33.9050807@pgexperts.com
Bug analysis by Andres Freund.
Once we've completed a PREPARE, our session is not running a transaction,
so its entry in pg_stat_activity should show xact_start as null, rather
than leaving the value as the start time of the now-prepared transaction.
I think possibly this oversight was triggered by faulty extrapolation
from the adjacent comment that says PrepareTransaction should not call
AtEOXact_PgStat, so tweak the wording of that comment.
Noted by Andres Freund while considering bug #10123 from Maxim Boguk,
although this error doesn't seem to explain that report.
Back-patch to all active branches.
When marking a branch as half-dead, a pointer to the top of the branch is
stored in the leaf block's hi-key. During normal operation, the high key
was left in place, and the block number was just stored in the ctid field
of the high key tuple, but in WAL replay, the high key was recreated as a
truncated tuple with zero columns. For the sake of easier debugging, also
truncate the tuple in normal operation, so that the page is identical
after WAL replay. Also, rename the 'downlink' field in the WAL record to
'topparent', as that seems like a more descriptive name. And make sure
it's set to invalid when unlinking the leaf page.
Forgot to update LSN of left sibling's page, when creating a new root.
I fixed this for regular insertions and page splits earlier, but missed
new root creation.
The README incorrectly claimed that GIN posting tree pages contain an array
of uncompressed items in addition to compressed posting lists. Earlier
versions of the GIN posting list compression patch worked that way, but not
the one that was committed.
I mixed up BLCKSZ and XLOG_BLCKSZ when I changed the way the buffer is
allocated a couple of weeks ago. With the default settings, they are both
8k, but they can be changed at compile-time.
This allows squeezing out the unused space in full-page writes. And more
importantly, it can be a useful debugging aid.
In hindsight we should've done this back when GIN was added - we wouldn't
need the 'maxoff' field in the page opaque struct if we had used pd_lower
and pd_upper like on normal pages. But as long as there can be pages in the
index that have been binary-upgraded from pre-9.4 versions, we can't rely
on that, and have to continue using 'maxoff'.
Most of the code churn comes from renaming some macros, now that they're
used on internal pages, too.
This change is completely backwards-compatible, no effect on pg_upgrade.
Make sure we throw an error instead of silently doing the wrong thing when
fed a strategy number we don't recognize. Also, in the places that did
already throw an error, spell the error message in a way more consistent
with our message style guidelines.
Per report from Paul Jones. Although this is a bug, it won't occur unless
a superuser tries to do something he shouldn't, so it doesn't seem worth
back-patching.
The entry B-tree pages all follow the standard page layout. The 9.3 code has
this right. I inadvertently changed this at some point during the big
refactorings in git master.
There were a couple of bugs here. First, if the fuzzy limit was exceeded,
the loop in entryGetItem might drop out too soon if a whole block needs to
be skipped because it's < advancePast ("continue" in a while-loop checks the
loop condition too). Secondly, the loop checked when stepping to a new page
that there is at least one offset on the page < advancePast, but we cannot
rely on that on subsequent calls of entryGetItem, because advancePast might
change in between. That caused the skipping loop to read bogus items in the
TbmIterateResult's offset array.
First item and fix by Alexander Korotkov, second bug pointed out by Fabrízio
de Royes Mello, by a small variation of Alexander's test query.
Don't reset the rightlink of a page when replaying a page update record.
This was a leftover from pre-hot standby days, when it was not possible to
have scans concurrent with WAL replay. Resetting the right-link was not
necessary back then either, but it was done for the sake of tidiness. But
with hot standby, it's wrong, because a concurrent scan might still need it.
Backpatch all versions with hot standby, 9.0 and above.
Forgot to set the incomplete-split flag on the left page half, in redo of a
page split.
Spotted this by comparing the page contents on master and standby, after
inserting/applying each WAL record.
Also add a regression test for a GIN index with enough items with the same
key, so that a GIN posting tree gets created. Apparently none of the
existing GIN tests were large enough for that.
This code is new, no backpatching required.
GetVirtualXIDsDelayingChkpt calls palloc, which isn't safe in a critical
section. I thought I covered this case with the exemption for the
checkpointer, but CreateCheckPoint is also called from the startup process.
Memory allocation can fail if you run out of memory, and inside a critical
section that will lead to a PANIC. Use conservatively-sized arrays in stack
instead.
There was previously no explicit limit on the number of pages a GiST split
can produce, it was only limited by the number of LWLocks that can be held
simultaneously (100 at the moment). This patch adds an explicit limit of 75
pages. That should be plenty, a typical split shouldn't produce more than
2-3 page halves.
The bug has been there forever, but only backpatch down to 9.1. The code
was changed significantly in 9.1, and it doesn't seem worth the risk or
trouble to adapt this for 9.0 and 8.4.
Inserting a downlink to an internal page clears the incomplete-split flag
of the child's left sibling, so the left sibling's LSN also needs to be
updated and it needs to be marked dirty. The codepath for an insertion got
this right, but the case where the internal node is split because of
inserting the new downlink missed that.
Inserting a downlink to an internal page clears the incomplete-split flag
of the child's left sibling, so the left sibling's LSN also needs to be
updated.
Inserting (in retail) into the new 9.4 format GIN posting tree created much
larger WAL records than in 9.3. The previous strategy to WAL logging was
basically to log the whole page on each change, with the exception of
completely unmodified segments up to the first modified one. That was not
too bad when appending to the end of the page, as only the last segment had
to be WAL-logged, but per Fujii Masao's testing, even that produced 2x the
WAL volume that 9.3 did.
The new strategy is to keep track of changes to the posting lists in a more
fine-grained fashion, and also make the repacking" code smarter to avoid
decoding and re-encoding segments unnecessarily.
If you compile with WAL_DEBUG and enable it with wal_debug=on, we used to
only pass the first XLogRecData entry to the rm_desc routine. I think the
original assumprion was that the first XLogRecData entry contains all the
necessary information for the rm_desc routine, but that's a pretty shaky
assumption. At least standby_redo didn't get the memo.
To fix, piece together all the data in a temporary buffer, and pass that to
the rm_desc routine.
It's been like this forever, but the patch didn't apply cleanly to
back-branches. Probably wouldn't be hard to fix the conflicts, but it's
not worth the trouble.
Heikki Linnakangas
Tom Lane
Alvaro Herrera
Fujii Masao
Simon Riggs
Robert Haas
Andres Freund
Bruce Momjian
Magnus Hagander
Peter Eisentraut
Andrew Dunstan
Kevin Grittner
Stephen Frost
Noah Misch