While autovacuum dutifully launched anti-multixact-wraparound vacuums
when the multixact "age" was reached, the vacuum code was not aware that
it needed to make them be full table vacuums. As the resulting
partial-table vacuums aren't capable of actually increasing relminmxid,
autovacuum continued to launch anti-wraparound vacuums that didn't have
the intended effect, until age of relfrozenxid caused the vacuum to
finally be a full table one via vacuum_freeze_table_age.
To fix, introduce logic for multixacts similar to that for plain
TransactionIds, using the same GUCs.
Backpatch to 9.3, where permanent MultiXactIds were introduced.
Andres Freund, some cleanup by Álvaro
Parts of the code used autovacuum_freeze_max_age to determine whether
anti-multixact-wraparound vacuums are necessary, while others used a
hardcoded 200000000 value. This leads to problems when
autovacuum_freeze_max_age is set to a non-default value. Use the latter
everywhere.
Backpatch to 9.3, where vacuuming of multixacts was introduced.
Andres Freund
The various places that transferred fast-path locks to the main lock table
neglected to release the PGPROC's backendLock if SetupLockInTable failed
due to being out of shared memory. In most cases this is no big deal since
ensuing error cleanup would release all held LWLocks anyway. But there are
some hot-standby functions that don't consider failure of
FastPathTransferRelationLocks to be a hard error, and in those cases this
oversight could lead to system lockup. For consistency, make all of these
places look the same as FastPathTransferRelationLocks.
Noted while looking for the cause of Dan Wood's bugs --- this wasn't it,
but it's a bug anyway.
Prevent handle_sig_alarm from losing control partway through due to a query
cancel (either an asynchronous SIGINT, or a cancel triggered by one of the
timeout handler functions). That would at least result in failure to
schedule any required future interrupt, and might result in actual
corruption of timeout.c's data structures, if the interrupt happened while
we were updating those.
We could still lose control if an asynchronous SIGINT arrives just as the
function is entered. This wouldn't break any data structures, but it would
have the same effect as if the SIGALRM interrupt had been silently lost:
we'd not fire any currently-due handlers, nor schedule any new interrupt.
To forestall that scenario, forcibly reschedule any pending timer interrupt
during AbortTransaction and AbortSubTransaction. We can avoid any extra
kernel call in most cases by not doing that until we've allowed
LockErrorCleanup to kill the DEADLOCK_TIMEOUT and LOCK_TIMEOUT events.
Another hazard is that some platforms (at least Linux and *BSD) block a
signal before calling its handler and then unblock it on return. When we
longjmp out of the handler, the unblock doesn't happen, and the signal is
left blocked indefinitely. Again, we can fix that by forcibly unblocking
signals during AbortTransaction and AbortSubTransaction.
These latter two problems do not manifest when the longjmp reaches
postgres.c, because the error recovery code there kills all pending timeout
events anyway, and it uses sigsetjmp(..., 1) so that the appropriate signal
mask is restored. So errors thrown outside any transaction should be OK
already, and cleaning up in AbortTransaction and AbortSubTransaction should
be enough to fix these issues. (We're assuming that any code that catches
a query cancel error and doesn't re-throw it will do at least a
subtransaction abort to clean up; but that was pretty much required already
by other subsystems.)
Lastly, ProcSleep should not clear the LOCK_TIMEOUT indicator flag when
disabling that event: if a lock timeout interrupt happened after the lock
was granted, the ensuing query cancel is still going to happen at the next
CHECK_FOR_INTERRUPTS, and we want to report it as a lock timeout not a user
cancel.
Per reports from Dan Wood.
Back-patch to 9.3 where the new timeout handling infrastructure was
introduced. We may at some point decide to back-patch the signal
unblocking changes further, but I'll desist from that until we hear
actual field complaints about it.
Although user-defined relations can't be directly created in
pg_catalog, it's possible for them to end up there, because you can
create them in some other schema and then use ALTER TABLE .. SET SCHEMA
to move them there. Previously, such relations couldn't afterwards
be manipulated, because IsSystemRelation()/IsSystemClass() rejected
all attempts to modify objects in the pg_catalog schema, regardless
of their origin. With this patch, they now reject only those
objects in pg_catalog which were created at initdb-time, allowing
most operations on user-created tables in pg_catalog to proceed
normally.
This patch also adds new functions IsCatalogRelation() and
IsCatalogClass(), which is similar to IsSystemRelation() and
IsSystemClass() but with a slightly narrower definition: only TOAST
tables of system catalogs are included, rather than *all* TOAST tables.
This is currently used only for making decisions about when
invalidation messages need to be sent, but upcoming logical decoding
patches will find other uses for this information.
Andres Freund, with some modifications by me.
In the GIN incomplete-splits patch, I used BlockIdDatas to store the block
number of left and right children, when inserting a downlink after a split
to an internal page posting list page. But gin_desc thought they were stored
as BlockNumbers.
We have for a long time checked the head pointer of each of the backend's
proclock lists and skipped acquiring the corresponding locktable partition
lock if the head pointer was NULL. This was safe enough in the days when
proclock lists were changed only by the owning backend, but it is pretty
questionable now that the fast-path patch added cases where backends add
entries to other backends' proclock lists. However, we don't really wish
to revert to locking each partition lock every time, because in simple
transactions that would add a lot of useless lock/unlock cycles on
already-heavily-contended LWLocks. Fortunately, the only way that another
backend could be modifying our proclock list at this point would be if it
was promoting a formerly fast-path lock of ours; and any such lock must be
one that we'd decided not to delete in the previous loop over the locallock
table. So it's okay if we miss seeing it in this loop; we'd just decide
not to delete it again. However, once we've detected a non-empty list,
we'd better re-fetch the list head pointer after acquiring the partition
lock. This guards against possibly fetching a corrupt-but-non-null pointer
if pointer fetch/store isn't atomic. It's not clear if any practical
architectures are like that, but we've never assumed that before and don't
wish to start here. In any case, the situation certainly deserves a code
comment.
While at it, refactor the partition traversal loop to use a for() construct
instead of a while() loop with goto's.
Back-patch, just in case the risk is real and not hypothetical.
Instead of simply checking the KEYS_UPDATED bit, we need to check
whether each lock held on the future version of the tuple conflicts with
the lock we're trying to acquire.
Per bug report #8434 by Tomonari Katsumata
Not doing so causes us to traverse an update chain that has been broken
by concurrent page pruning. All other code that traverses update chains
uses this check as one of the cases in which to stop iterating, so
replicate it here too. Failure to do so leads to erroneous CLOG,
subtrans or multixact lookups.
Per discussion following the bug report by J Smith in
CADFUPgc5bmtv-yg9znxV-vcfkb+JPRqs7m2OesQXaM_4Z1JpdQ@mail.gmail.com
as diagnosed by Andres Freund.
If a transaction updates/deletes a tuple just before aborting, and a
concurrent transaction tries to prune the page concurrently, the pruner
may see HeapTupleSatisfiesVacuum return HEAPTUPLE_DELETE_IN_PROGRESS,
but a later call to HeapTupleGetUpdateXid() return InvalidXid. This
would cause an assertion failure in development builds, but would be
otherwise Mostly Harmless.
Fix by checking whether the updater Xid is valid before trying to apply
it as page prune point.
Reported by Andres in 20131124000203.GA4403@alap2.anarazel.de
The reason for the fetch failure is that the tuple was removed because
it was dead; so the failure is innocuous and can be ignored. Moreover,
there's no need for further work and we can return success to the caller
immediately. EvalPlanQualFetch is doing something very similar to this
already.
Report and test case from Andres Freund in
20131124000203.GA4403@alap2.anarazel.de
When acquiring a lock in fast-path mode, we must reset the locallock
object's lock and proclock fields to NULL. They are not necessarily that
way to start with, because the locallock could be left over from a failed
lock acquisition attempt earlier in the transaction. Failure to do this
led to all sorts of interesting misbehaviors when LockRelease tried to
clean up no-longer-related lock and proclock objects in shared memory.
Per report from Dan Wood.
In passing, modify LockRelease to elog not just Assert if it doesn't find
lock and proclock objects for a formerly fast-path lock, matching the code
in FastPathGetRelationLockEntry and LockRefindAndRelease. This isn't a
bug but it will help in diagnosing any future bugs in this area.
Also, modify FastPathTransferRelationLocks and FastPathGetRelationLockEntry
to break out of their loops over the fastpath array once they've found the
sole matching entry. This was inconsistently done in some search loops
and not others.
Improve assorted related comments, too.
Back-patch to 9.2 where the fast-path mechanism was introduced.
Correct an obsolete statement that no backend touches another backend's
PROCLOCK lists. This was probably wrong even when written (the deadlock
checker looks at everybody's lists), and it's certainly quite wrong now
that fast-path locking can require creation of lock and proclock objects
on behalf of another backend. Also improve some statements in the hot
standby explanation, and do one or two other trivial bits of wordsmithing/
reformatting.
Replace it with an approach similar to what GiST uses: when a page is split,
the left sibling is marked with a flag indicating that the parent hasn't been
updated yet. When the parent is updated, the flag is cleared. If an insertion
steps on a page with the flag set, it will finish split before proceeding
with the insertion.
The post-recovery cleanup mechanism was never totally reliable, as insertion
to the parent could fail e.g because of running out of memory or disk space,
leaving the tree in an inconsistent state.
This also divides the responsibility of WAL-logging more clearly between
the generic ginbtree.c code, and the parts specific to entry and posting
trees. There is now a common WAL record format for insertions and deletions,
which is written by ginbtree.c, followed by tree-specific payload, which is
returned by the placetopage- and split- callbacks.
Separate the insertion payload from the more static portions of GinBtree.
GinBtree now only contains information related to searching the tree, and
the information of what to insert is passed separately.
Add root block number to GinBtree, instead of passing it around all the
functions as argument.
Split off ginFinishSplit() from ginInsertValue(). ginFinishSplit is
responsible for finding the parent and inserting the downlink to it.
Vacuum recognizes that it can update relfrozenxid by checking whether it has
processed all pages of a relation. Unfortunately it performed that check
after truncating the dead pages at the end of the relation, and used the new
number of pages to decide whether all pages have been scanned. If the new
number of pages happened to be smaller or equal to the number of pages
scanned, it incorrectly decided that all pages were scanned.
This can lead to relfrozenxid being updated, even though some pages were
skipped that still contain old XIDs. That can lead to data loss due to xid
wraparounds with some rows suddenly missing. This likely has escaped notice
so far because it takes a large number (~2^31) of xids being used to see the
effect, while a full-table vacuum before that would fix the issue.
The incorrect logic was introduced by commit
b4b6923e03. Backpatch this fix down to 8.4,
like that commit.
Andres Freund, with some modifications by me.
Reviewed-by: Ali Dar <ali.munir.dar@gmail.com>
Reviewed-by: Amit Khandekar <amit.khandekar@enterprisedb.com>
Reviewed-by: Rodolfo Campero <rodolfo.campero@anachronics.com>
Change SET LOCAL/CONSTRAINTS/TRANSACTION behavior outside of a
transaction block from error (post-9.3) to warning. (Was nothing in <=
9.3.) Also change ABORT outside of a transaction block from notice to
warning.
Previously, messages were emitted at the LOG level every time a
backend preloaded a library. That was acceptable (though unnecessary)
for shared_preload_libraries; but it was excessive for
local_preload_libraries and session_preload_libraries. Reduce to
DEBUG1.
Also, there was logic in the EXEC_BACKEND case to avoid repeated
messages for shared_preload_libraries by demoting them to
DEBUG2. DEBUG1 seems more appropriate there, as well, so eliminate
that special case.
Peter Geoghegan.
These functions must be careful that they return the intended value of
errno to their callers. There were several scenarios where this might
not happen:
1. The recent SSL renegotiation patch added a hunk of code that would
execute after setting errno. In the first place, it's doubtful that we
should consider renegotiation to be successfully completed after a failure,
and in the second, there's no real guarantee that the called OpenSSL
routines wouldn't clobber errno. Fix by not executing that hunk except
during success exit.
2. errno was left in an unknown state in case of an unrecognized return
code from SSL_get_error(). While this is a "can't happen" case, it seems
like a good idea to be sure we know what would happen, so reset errno to
ECONNRESET in such cases. (The corresponding code in libpq's fe-secure.c
already did this.)
3. There was an (undocumented) assumption that client_read_ended() wouldn't
change errno. While true in the current state of the code, this seems less
than future-proof. Add explicit saving/restoring of errno to make sure
that changes in the called functions won't break things.
I see no need to back-patch, since #1 is new code and the other two issues
are mostly hypothetical.
Per discussion with Amit Kapila.
The previous coding labeled expressions such as pg_index.indkey[1:3] as
being of int2vector type; which is not right because the subscript bounds
of such a result don't, in general, satisfy the restrictions of int2vector.
To fix, implicitly promote the result of slicing int2vector to int2[],
or oidvector to oid[]. This is similar to what we've done with domains
over arrays, which is a good analogy because these types are very much
like restricted domains of the corresponding regular-array types.
A side-effect is that we now also forbid array-element updates on such
columns, eg while "update pg_index set indkey[4] = 42" would have worked
before if you were superuser (and corrupted your catalogs irretrievably,
no doubt) it's now disallowed. This seems like a good thing since, again,
some choices of subscripting would've led to results not satisfying the
restrictions of int2vector. The case of an array-slice update was
rejected before, though with a different error message than you get now.
We could make these cases work in future if we added a cast from int2[]
to int2vector (with a cast function checking the subscript restrictions)
but it seems unlikely that there's any value in that.
Per report from Ronan Dunklau. Back-patch to all supported branches
because of the crash risks involved.
Two call sites were apparently thinking that the last argument of
SPI_execute_plan() is the number of query parameters, but it is actually
the row limit. Change the calls to 0, since we don't care about the
limit there. The previous code didn't break anything, but it was still
wrong.
A pointer to a C string was treated as a pointer to a "name" datum and
passed to SPI_execute_plan(). This pointer would then end up being
passed through datumCopy(), which would try to copy the entire 64 bytes
of name data, thus running past the end of the C string. Fix by
converting the string to a proper name structure.
Found by LLVM AddressSanitizer.
pullup_replace_vars()'s decisions about whether a pulled-up replacement
expression needs to be wrapped in a PlaceHolderVar depend on the assumption
that what looks like a Var behaves like a Var. However, if the Var is a
join alias reference, later flattening of join aliases might replace the
Var with something that's not a Var at all, and should have been wrapped.
To fix, do a forcible pass of flatten_join_alias_vars() on the subquery
targetlist before we start to copy items out of it. We'll re-run that
processing on the pulled-up expressions later, but that's harmless.
Per report from Ken Tanzer; the added regression test case is based on his
example. This bug has been there since the PlaceHolderVar mechanism was
invented, but has escaped detection because the circumstances that trigger
it are fairly narrow. You need a flattenable query underneath an outer
join, which contains another flattenable query inside a join of its own,
with a dangerous expression (a constant or something else non-strict)
in that one's targetlist.
Having seen this, I'm wondering if it wouldn't be prudent to do all
alias-variable flattening earlier, perhaps even in the rewriter.
But that would probably not be a back-patchable change.
These bugs can cause data loss on standbys started with hot_standby=on at
the moment they start to accept read only queries, by marking committed
transactions as uncommited. The likelihood of such corruptions is small
unless the primary has a high transaction rate.
5a031a5556 fixed bugs in HS's startup logic
by maintaining less state until at least STANDBY_SNAPSHOT_PENDING state
was reached, missing the fact that both clog and subtrans are written to
before that. This only failed to fail in common cases because the usage
of ExtendCLOG in procarray.c was superflous since clog extensions are
actually WAL logged.
f44eedc3f0f347a856eea8590730769125964597/I then tried to fix the missing
extensions of pg_subtrans due to the former commit's changes - which are
not WAL logged - by performing the extensions when switching to a state
> STANDBY_INITIALIZED and not performing xid assignments before that -
again missing the fact that ExtendCLOG is unneccessary - but screwed up
twice: Once because latestObservedXid wasn't updated anymore in that
state due to the earlier commit and once by having an off-by-one error in
the loop performing extensions. This means that whenever a
CLOG_XACTS_PER_PAGE (32768 with default settings) boundary was crossed
between the start of the checkpoint recovery started from and the first
xl_running_xact record old transactions commit bits in pg_clog could be
overwritten if they started and committed in that window.
Fix this mess by not performing ExtendCLOG() in HS at all anymore since
it's unneeded and evidently dangerous and by performing subtrans
extensions even before reaching STANDBY_SNAPSHOT_PENDING.
Analysis and patch by Andres Freund. Reported by Christophe Pettus.
Backpatch down to 9.0, like the previous commit that caused this.
RecoveryIsInProgress() can be called very frequently. During normal
operation, it just checks a backend-local variable and returns quickly,
but during hot standby, it checks a spinlock-protected shared variable.
Those spinlock acquisitions can become a point of contention on a busy
hot standby system.
Replace the spinlock acquisition with a memory barrier.
Per discussion with Andres Freund, Ants Aasma and Merlin Moncure.
This patch adds the ability to write TABLE( function1(), function2(), ...)
as a single FROM-clause entry. The result is the concatenation of the
first row from each function, followed by the second row from each
function, etc; with NULLs inserted if any function produces fewer rows than
others. This is believed to be a much more useful behavior than what
Postgres currently does with multiple SRFs in a SELECT list.
This syntax also provides a reasonable way to combine use of column
definition lists with WITH ORDINALITY: put the column definition list
inside TABLE(), where it's clear that it doesn't control the ordinality
column as well.
Also implement SQL-compliant multiple-argument UNNEST(), by turning
UNNEST(a,b,c) into TABLE(unnest(a), unnest(b), unnest(c)).
The SQL standard specifies TABLE() with only a single function, not
multiple functions, and it seems to require an implicit UNNEST() which is
not what this patch does. There may be something wrong with that reading
of the spec, though, because if it's right then the spec's TABLE() is just
a pointless alternative spelling of UNNEST(). After further review of
that, we might choose to adopt a different syntax for what this patch does,
but in any case this functionality seems clearly worthwhile.
Andrew Gierth, reviewed by Zoltán Böszörményi and Heikki Linnakangas, and
significantly revised by me
Split off the portion of ginInsertValue that inserts the tuple to current
level into a separate function, ginPlaceToPage. ginInsertValue's charter
is now to recurse up the tree to insert the downlink, when a page split is
required.
This is in preparation for a patch to change the way incomplete splits are
handled, which will need to do these operations separately. And IMHO makes
the code more readable anyway.
This creates a new gin-btree callback function for creating a downlink for
a page. Previously, ginxlog.c duplicated the logic used during normal
operation.
Previously, if VACUUM skipped vacuuming a page because it's pinned, it
didn't count that page as scanned. However, that meant that relfrozenxid
was not bumped up either, which prevented anti-wraparound vacuum from
doing its job.
Report by Миша Тюрин, analysis and patch by Sergey Burladyn and Jeff Janes.
Backpatch to 9.2, where the skip-locked-pages behavior was introduced.
This patch improves performance of most built-in aggregates that formerly
used a NUMERIC or NUMERIC array as their transition type; this includes
not only aggregates on numeric inputs, but some aggregates on integer
inputs where overflow of an int8 value is a possibility. The code now
uses a special-purpose data structure to avoid array construction and
deconstruction overhead, as well as packing and unpacking overhead for
numeric values.
These aggregates' transition type is now declared as INTERNAL, since
it doesn't correspond to any SQL data type. To keep the planner from
thinking that that means a lot of storage will be used, we make use
of the just-added pg_aggregate.aggtransspace feature. The space estimate
is set to 128 bytes, which is at least in the right ballpark.
Hadi Moshayedi, reviewed by Pavel Stehule and Tomas Vondra
Formerly the planner had a hard-wired rule of thumb for guessing the amount
of space consumed by an aggregate function's transition state data. This
estimate is critical to deciding whether it's OK to use hash aggregation,
and in many situations the built-in estimate isn't very good. This patch
adds a column to pg_aggregate wherein a per-aggregate estimate can be
provided, overriding the planner's default, and infrastructure for setting
the column via CREATE AGGREGATE.
It may be that additional smarts will be required in future, perhaps even
a per-aggregate estimation function. But this is already a step forward.
This is extracted from a larger patch to improve the performance of numeric
and int8 aggregates. I (tgl) thought it was worth reviewing and committing
this infrastructure separately. In this commit, all built-in aggregates
are given aggtransspace = 0, so no behavior should change.
Hadi Moshayedi, reviewed by Pavel Stehule and Tomas Vondra
A couple of places that should have been iterating over WORDS_PER_CHUNK
words were iterating over WORDS_PER_PAGE words instead. This thinko
accidentally failed to fail, because (at least on common architectures
with default BLCKSZ) WORDS_PER_CHUNK is a bit less than WORDS_PER_PAGE,
and the extra words being looked at were always zero so nothing happened.
Still, it's a bug waiting to happen if anybody ever fools with the
parameters affecting TIDBitmap sizes, and it's a small waste of cycles
too. So back-patch to all active branches.
Etsuro Fujita
Bug #8591 from Claudio Freire demonstrates that get_eclass_for_sort_expr
must be able to compute valid em_nullable_relids for any new equivalence
class members it creates. I'd worried about this in the commit message
for db9f0e1d9a, but claimed that it wasn't a
problem because multi-member ECs should already exist when it runs. That
is transparently wrong, though, because this function is also called by
initialize_mergeclause_eclasses, which runs during deconstruct_jointree.
The example given in the bug report (which the new regression test item
is based upon) fails because the COALESCE() expression is first seen by
initialize_mergeclause_eclasses rather than process_equivalence.
Fixing this requires passing the appropriate nullable_relids set to
get_eclass_for_sort_expr, and it requires new code to compute that set
for top-level expressions such as ORDER BY, GROUP BY, etc. We store
the top-level nullable_relids in a new field in PlannerInfo to avoid
computing it many times. In the back branches, I've added the new
field at the end of the struct to minimize ABI breakage for planner
plugins. There doesn't seem to be a good alternative to changing
get_eclass_for_sort_expr's API signature, though. There probably aren't
any third-party extensions calling that function directly; moreover,
if there are, they probably need to think about what to pass for
nullable_relids anyway.
Back-patch to 9.2, like the previous patch in this area.
The old code entered a new hash table entry first, then scanned
pg_class to determine what value to fill in, and then populated the
entry. This fails to work properly if a cache invalidation happens
as a result of opening pg_class. Repair.
Along the way, get rid of the idea of blowing away the entire hash
table as a method of processing invalidations. Instead, just delete
all the entries one by one. This is probably not quite as cheap but
it's simpler, and shouldn't happen often.
Andres Freund
The root page is filled with as many items as fit, and the rest are inserted
using normal insertions. However, I fumbled the variable names, and the code
actually memcpy'd all the items on the page, overflowing the buffer. While
at it, rename the variable to make the distinction more clear.
Reported by Teodor Sigaev. This bug was introduced by my recent
refactorings, so no backpatching required.
Simple oversight in commit 1cb108efb0 ---
recursively examining a subquery output column is only sane if the
original Var refers to a single output column. Found by Kevin Grittner.
The pretty-printing logic in ruleutils.c operates by inserting a newline
and some indentation whitespace into strings that are already valid SQL.
This naturally results in leaving some trailing whitespace before the
newline in many cases; which can be annoying when processing the output
with other tools, as complained of by Joe Abbate. We can fix that in
a pretty localized fashion by deleting any trailing whitespace before
we append a pretty-printing newline. In addition, we have to modify the
code inserted by commit 2f582f76b1 so that
we also delete trailing whitespace when transposing items from temporary
buffers into the main result string, when a temporary item starts with a
newline.
This results in rather voluminous changes to the regression test results,
but it's easily verified that they are only removal of trailing whitespace.
Back-patch to 9.3, because the aforementioned commit resulted in many
more cases of trailing whitespace than had occurred in earlier branches.
Although the SQL spec forbids duplicate table aliases, historically
we've allowed queries like
SELECT ... FROM tab1 x CROSS JOIN (tab2 x CROSS JOIN tab3 y) z
on the grounds that the aliased join (z) hides the aliases within it,
therefore there is no conflict between the two RTEs named "x". The
LATERAL patch broke this, on the misguided basis that "x" could be
ambiguous if tab3 were a LATERAL subquery. To avoid breaking existing
queries, it's better to allow this situation and complain only if
tab3 actually does contain an ambiguous reference. We need only remove
the check that was throwing an error, because the column lookup code
is already prepared to handle ambiguous references. Per bug #8444.
Alvaro Herrera
Tom Lane
Robert Haas
Heikki Linnakangas
Peter Eisentraut
Jeff Davis
Bruce Momjian