]> granicus.if.org Git - postgresql/log
postgresql
19 years agoRemove registration message in all the supported back branches; REL7_2_STABLE github/REL7_2_STABLE
Tom Lane [Tue, 16 Aug 2005 03:33:39 +0000 (03:33 +0000)]
Remove registration message in all the supported back branches;
we had decided to drop it for 7.4, and no one misses it.

19 years agoMemSet() must not cast its pointer argument to int32* until after it has
Tom Lane [Mon, 18 Jul 2005 15:55:01 +0000 (15:55 +0000)]
MemSet() must not cast its pointer argument to int32* until after it has
checked that the pointer is actually word-aligned.  Casting a non-aligned
pointer to int32* is technically illegal per the C spec, and some recent
versions of gcc actually generate bad code for the memset() when given
such a pointer.  Per report from Andrew Morrow.

19 years agoBack-patch into 7.2 branch the 7.3 changes that made contrib/seg
Tom Lane [Sat, 16 Jul 2005 19:48:16 +0000 (19:48 +0000)]
Back-patch into 7.2 branch the 7.3 changes that made contrib/seg
error messages reasonably independent of the bison version used to
build segparse.c.   Needed to get this branch passing on buildfarm.

19 years agoBack-patch into 7.2 branch the 7.3 changes that made contrib/cube
Tom Lane [Sat, 16 Jul 2005 19:27:20 +0000 (19:27 +0000)]
Back-patch into 7.2 branch the 7.3 changes that made contrib/cube
error messages reasonably independent of the bison version used to
build cubeparse.c.  Needed to get this branch passing on buildfarm.

19 years agoMark the geometry regression test as an ignorable failure in the pre-7.4
Tom Lane [Sat, 16 Jul 2005 19:02:01 +0000 (19:02 +0000)]
Mark the geometry regression test as an ignorable failure in the pre-7.4
branches.  Per discussion, this seems saner than trying to maintain
last-significant-digit regression test outputs for all supported platforms.

19 years agoFix bogus "extern int errno;" in back branches, per Andrew Dunstan.
Tom Lane [Sat, 16 Jul 2005 15:24:08 +0000 (15:24 +0000)]
Fix bogus "extern int errno;" in back branches, per Andrew Dunstan.

19 years agoMake libpq_gettext save and restore errno in a Windows-compatible way.
Tom Lane [Fri, 8 Jul 2005 15:25:46 +0000 (15:25 +0000)]
Make libpq_gettext save and restore errno in a Windows-compatible way.
Also, back-patch fix into back branches.

19 years agoplpgsql's exec_assign_value() freed the old value of a variable before
Tom Lane [Mon, 20 Jun 2005 20:45:12 +0000 (20:45 +0000)]
plpgsql's exec_assign_value() freed the old value of a variable before
copying/converting the new value, which meant that it failed badly on
"var := var" if var is of pass-by-reference type.  Fix this and a similar
hazard in exec_move_row(); not sure that the latter can manifest before
8.0, but patch it all the way back anyway.  Per report from Dave Chapeskie.

19 years agoThe random selection in function linear() could deliver a value equal to max
Tom Lane [Tue, 14 Jun 2005 14:21:43 +0000 (14:21 +0000)]
The random selection in function linear() could deliver a value equal to max
if geqo_rand() returns exactly 1.0, resulting in failure due to indexing
off the end of the pool array.  Also, since this is using inexact float math,
it seems wise to guard against roundoff error producing values slightly
outside the expected range.  Per report from bug@zedware.org.

19 years agoAdd test to WAL replay to verify that xl_prev points back to the previous
Tom Lane [Tue, 31 May 2005 19:11:28 +0000 (19:11 +0000)]
Add test to WAL replay to verify that xl_prev points back to the previous
WAL record; this is necessary to be sure we recognize stale WAL records
when a WAL page was only partially written during a system crash.

19 years agoGuard against duplicate IDs in input file in SortTocFromFile().
Tom Lane [Tue, 17 May 2005 17:31:24 +0000 (17:31 +0000)]
Guard against duplicate IDs in input file in SortTocFromFile().
Per report from Brian Hackett.

19 years agoUpdate release notes for upcoming re-releases. REL7_2_8
Tom Lane [Mon, 9 May 2005 00:10:53 +0000 (00:10 +0000)]
Update release notes for upcoming re-releases.

19 years agoRepair very-low-probability race condition between relation extension
Tom Lane [Sat, 7 May 2005 21:34:20 +0000 (21:34 +0000)]
Repair very-low-probability race condition between relation extension
and VACUUM: in the interval between adding a new page to the relation
and formatting it, it was possible for VACUUM to come along and decide
it should format the page too.  Though not harmful in itself, this would
cause data loss if a third transaction were able to insert tuples into
the vacuumed page before the original extender got control back.

19 years agoAdjust time qual checking code so that we always check TransactionIdIsInProgress
Tom Lane [Sat, 7 May 2005 21:23:49 +0000 (21:23 +0000)]
Adjust time qual checking code so that we always check TransactionIdIsInProgress
before we check commit/abort status.  Formerly this was done in some paths
but not all, with the result that a transaction might be considered
committed for some purposes before it became committed for others.
Per example found by Jan Wieck.

19 years agoStamp release 7.2.8.
Tom Lane [Thu, 5 May 2005 20:10:08 +0000 (20:10 +0000)]
Stamp release 7.2.8.

19 years agoMake standalone backends ignore pg_database.datallowconn, so that there
Tom Lane [Thu, 5 May 2005 19:54:13 +0000 (19:54 +0000)]
Make standalone backends ignore pg_database.datallowconn, so that there
is a way to recover from disabling connections to all databases at once.

19 years agoGCC 4.0 includes a new warning option, -Wformat-literal, that emits
Neil Conway [Sat, 30 Apr 2005 09:08:14 +0000 (09:08 +0000)]
GCC 4.0 includes a new warning option, -Wformat-literal, that emits
a warning when a variable is used as a format string for printf()
and similar functions (if the variable is derived from untrusted
data, it could include unexpected formatting sequences). This
emits too many warnings to be enabled by default, but it does
flag a few dubious constructs in the Postgres tree. This patch
fixes up the obvious variants: functions that are passed a variable
format string but no additional arguments.

This patch fixes a bug in pg_dump (triggers with formatting sequences
in their names are not dumped correctly) and some related pg_dump
code that looks dubious; cleanups for more harmless instances have
been applied to more recent branches. This patch also fixes an
additional format string bug that is present in 7.2 but not in later
releases: pg_dump would also fail to correctly dump indexes with
formatting sequences in their names.

19 years agoThis patch fixes a bug in the error message emitted by pg_restore on an
Neil Conway [Sat, 30 Apr 2005 08:01:58 +0000 (08:01 +0000)]
This patch fixes a bug in the error message emitted by pg_restore on an
incorrect -F argument: write_msg() expects its first parameter to be a
"module name", not the format string.

19 years agoRepair two TIME WITH TIME ZONE bugs found by Dennis Vshivkov. Comparison
Tom Lane [Sat, 23 Apr 2005 22:54:23 +0000 (22:54 +0000)]
Repair two TIME WITH TIME ZONE bugs found by Dennis Vshivkov.  Comparison
of timetz values misbehaved in --enable-integer-datetime cases, and
EXTRACT(EPOCH) subtracted the zone instead of adding it in all cases.
Backpatch to all supported releases (except --enable-integer-datetime code
does not exist in 7.2).

19 years agoDon't try to run clauseless index scans on index types that don't support
Tom Lane [Wed, 20 Apr 2005 21:48:39 +0000 (21:48 +0000)]
Don't try to run clauseless index scans on index types that don't support
it.  Per report from Marinos Yannikos.

20 years agoPrevent to_char(interval) from dumping core on month-related formats
Tom Lane [Sat, 26 Mar 2005 00:42:56 +0000 (00:42 +0000)]
Prevent to_char(interval) from dumping core on month-related formats
when a zero-month interval is given.  Per discussion with Karel.

20 years agoSome builds (depends on crypto engine support?) of OpenSSL
Neil Conway [Sun, 13 Mar 2005 23:46:27 +0000 (23:46 +0000)]
Some builds (depends on crypto engine support?) of OpenSSL
0.9.7x have EVP_DigestFinal function which which clears all of
EVP_MD_CTX.  This makes pgcrypto crash in functions which
re-use one digest context several times: hmac() and crypt()
with md5 algorithm.

Following patch fixes it by carring the digest info around
EVP_DigestFinal and re-initializing cipher.

Marko Kreen.

20 years agoUpdate contrib/pgcrypto in 7.3 and 7.2 branches to avoid problems with
Neil Conway [Sun, 13 Mar 2005 23:42:07 +0000 (23:42 +0000)]
Update contrib/pgcrypto in 7.3 and 7.2 branches to avoid problems with
recent versions of OpenSSL. From Marko Kreen.

20 years agoBack-patch Neil's four additional buffer overrun checks.
Tom Lane [Tue, 8 Feb 2005 18:22:54 +0000 (18:22 +0000)]
Back-patch Neil's four additional buffer overrun checks.

20 years agoRecommend security@postgresql.org as the contact point for security-related bugs. REL7_2_7
Tom Lane [Sun, 30 Jan 2005 21:32:38 +0000 (21:32 +0000)]
Recommend security@postgresql.org as the contact point for security-related bugs.

20 years agoStamp release 7.2.7.
Tom Lane [Sun, 30 Jan 2005 20:32:42 +0000 (20:32 +0000)]
Stamp release 7.2.7.

20 years agoMake sure contrib C functions are marked strict where needed.
Tom Lane [Sat, 29 Jan 2005 22:36:03 +0000 (22:36 +0000)]
Make sure contrib C functions are marked strict where needed.
Kris Jurka

20 years agoBackpatch fix for buffer overrun in parsing refcursor parameters to
Neil Conway [Thu, 27 Jan 2005 01:52:34 +0000 (01:52 +0000)]
Backpatch fix for buffer overrun in parsing refcursor parameters to
REL7_2_STABLE.

20 years agoMark the text_soundex() function as "strict", to avoid crashing on NULL
Neil Conway [Wed, 26 Jan 2005 08:25:46 +0000 (08:25 +0000)]
Mark the text_soundex() function as "strict", to avoid crashing on NULL
input. Also, may as well mark it "cacheable" as well. From Kris Jurka.

20 years agoDisallow LOAD to non-superusers. Per report from John Heasman.
Tom Lane [Mon, 24 Jan 2005 17:46:58 +0000 (17:46 +0000)]
Disallow LOAD to non-superusers.  Per report from John Heasman.

20 years agoThe result of a FULL or RIGHT join can't be assumed to be sorted by the
Tom Lane [Sun, 23 Jan 2005 02:26:05 +0000 (02:26 +0000)]
The result of a FULL or RIGHT join can't be assumed to be sorted by the
left input's sorting, because null rows may be inserted at various points.
Per report from Ferenc Lutischá¸n.

20 years agointerval_out failed to mention 'ago' for negative intervals in SQL and
Tom Lane [Tue, 11 Jan 2005 18:34:41 +0000 (18:34 +0000)]
interval_out failed to mention 'ago' for negative intervals in SQL and
GERMAN datestyles.  Ancient bug reported by Terry Lee Tucker.

20 years agoBack-patch removal of extraneous semicolon, so that this will build
Tom Lane [Sun, 24 Oct 2004 23:02:41 +0000 (23:02 +0000)]
Back-patch removal of extraneous semicolon, so that this will build
with recent bisons (if anyone still cares).

20 years agoStamp release 7.2.6. REL7_2_6
Tom Lane [Fri, 22 Oct 2004 00:27:34 +0000 (00:27 +0000)]
Stamp release 7.2.6.

20 years agoPrevent pg_ctl from being run as root. Since it uses configuration files
Tom Lane [Fri, 22 Oct 2004 00:24:39 +0000 (00:24 +0000)]
Prevent pg_ctl from being run as root.  Since it uses configuration files
owned by postgres, doing "pg_ctl start" as root could allow a privilege
escalation attack, as pointed out by iDEFENSE.  Of course the postmaster would
fail, but we ought to fail a little sooner to protect sysadmins unfamiliar
with Postgres.  The chosen fix is to disable root use of pg_ctl in all cases,
just to be confident there are no other holes.

20 years agoBack-patch make_oidjoins_check security improvement.
Tom Lane [Thu, 21 Oct 2004 17:12:35 +0000 (17:12 +0000)]
Back-patch make_oidjoins_check security improvement.

20 years agoRepair possible failure to update hint bits back to disk, per
Tom Lane [Wed, 13 Oct 2004 22:22:41 +0000 (22:22 +0000)]
Repair possible failure to update hint bits back to disk, per
http://archives.postgresql.org/pgsql-hackers/2004-10/msg00464.php.
I plan a more permanent fix in HEAD, but for the back branches it seems
best to just touch the places that actually have a problem.

20 years agoFix breakage in hashjoin from recent backpatch of left-join bug fix.
Tom Lane [Wed, 13 Oct 2004 21:56:14 +0000 (21:56 +0000)]
Fix breakage in hashjoin from recent backpatch of left-join bug fix.
(That's what I get for not testing the back branches *before* committing.)

20 years agoHashed LEFT JOIN would miss outer tuples with no inner match if the join
Tom Lane [Fri, 17 Sep 2004 18:29:40 +0000 (18:29 +0000)]
Hashed LEFT JOIN would miss outer tuples with no inner match if the join
was large enough to be batched and the tuples fell into a batch where
there were no inner tuples at all.  Thanks to Xiaoyu Wang for finding a
test case that exposed this long-standing bug.

20 years agoUpdate 7.2 regression tests to match what you get when using a modern REL7_2_5
Tom Lane [Thu, 19 Aug 2004 20:03:49 +0000 (20:03 +0000)]
Update 7.2 regression tests to match what you get when using a modern
version of Bison.

20 years agoStamp 7.2.5.
Tom Lane [Sun, 15 Aug 2004 01:23:30 +0000 (01:23 +0000)]
Stamp 7.2.5.

20 years agoFix failure to guarantee that a checkpoint will write out pg_clog updates
Tom Lane [Wed, 11 Aug 2004 04:09:14 +0000 (04:09 +0000)]
Fix failure to guarantee that a checkpoint will write out pg_clog updates
for transaction commits that occurred just before the checkpoint.  This is
an EXTREMELY serious bug --- kudos to Satoshi Okada for creating a
reproducible test case to prove its existence.

21 years agoRepair multiple memory leaks in getTables(), including one that could
Tom Lane [Sat, 20 Mar 2004 18:12:32 +0000 (18:12 +0000)]
Repair multiple memory leaks in getTables(), including one that could
easily exhaust memory on databases with more than a few hundred triggers.
I don't expect any more releases of these old versions, but let's put the
fix in CVS just so it's archived.

21 years agoFix longstanding error in _bt_search(): should moveright at top of loop not
Tom Lane [Tue, 29 Jul 2003 22:18:53 +0000 (22:18 +0000)]
Fix longstanding error in _bt_search(): should moveright at top of loop not
bottom.  Otherwise we fail to moveright when the root page was split while
we were "in flight" to it.  This is not a significant problem when the root
is above the leaf level, but if the root was also a leaf (ie, a single-page
index just got split) we may return the wrong leaf page to the caller,
resulting in failure to find a key that is in fact present.  Bug has existed
at least since 7.1, probably forever.

21 years agoFix a *second* buffer overrun bug in to_ascii(). Grumble.
Tom Lane [Mon, 14 Jul 2003 16:41:56 +0000 (16:41 +0000)]
Fix a *second* buffer overrun bug in to_ascii().  Grumble.

21 years agoSecond try at avoiding conflicts with system isblank().
Tom Lane [Sun, 13 Apr 2003 04:07:43 +0000 (04:07 +0000)]
Second try at avoiding conflicts with system isblank().

21 years agoGuard against macro versions of isblank().
Tom Lane [Sat, 12 Apr 2003 22:28:45 +0000 (22:28 +0000)]
Guard against macro versions of isblank().

21 years agoAdd #include <errno.h> per recent reports that it's now necessary on
Tom Lane [Fri, 11 Apr 2003 20:51:27 +0000 (20:51 +0000)]
Add #include <errno.h> per recent reports that it's now necessary on
RHL 9.  (Not clear why it didn't break long before, actually...)

22 years agoFix buffer overrun in to_ascii(), per report from Guido Notari.
Tom Lane [Wed, 2 Apr 2003 21:08:14 +0000 (21:08 +0000)]
Fix buffer overrun in to_ascii(), per report from Guido Notari.

22 years agoTestConfiguration returns int, not bool. This mistake is relatively
Tom Lane [Mon, 31 Mar 2003 20:33:09 +0000 (20:33 +0000)]
TestConfiguration returns int, not bool.  This mistake is relatively
harmless on signed-char machines but would lead to core dump in the
deadlock detection code if char is unsigned.  Amazingly, this bug has
been here since 7.1 and yet wasn't reported till now.  Thanks to Robert
Bruccoleri for providing the opportunity to track it down.

22 years agoAsync_NotifyHandler must save and restore ImmediateInterruptOK. Fixes
Tom Lane [Thu, 13 Mar 2003 23:44:07 +0000 (23:44 +0000)]
Async_NotifyHandler must save and restore ImmediateInterruptOK.  Fixes
known problem with failure to respond to 'pg_ctl stop -m fast', and
probable problems if SIGINT or SIGTERM arrives while processing a
SIGUSR2 interrupt that arrived while waiting for a new client query.

22 years agoBack patch bug fix to quote function.
D'Arcy J.M. Cain [Thu, 27 Feb 2003 10:43:50 +0000 (10:43 +0000)]
Back patch bug fix to quote function.

22 years agoBrand 7.2.4. REL7_2_4
Tom Lane [Wed, 29 Jan 2003 22:13:19 +0000 (22:13 +0000)]
Brand 7.2.4.

22 years agoUpdate release history for 7.2.4.
Tom Lane [Wed, 29 Jan 2003 22:12:42 +0000 (22:12 +0000)]
Update release history for 7.2.4.

22 years agoBack-patch fix to avoid integer overflow in ExecHashJoinGetBatch(),
Tom Lane [Wed, 29 Jan 2003 19:37:23 +0000 (19:37 +0000)]
Back-patch fix to avoid integer overflow in ExecHashJoinGetBatch(),
which leads to core dump in large-enough hash joins.

22 years agoBack-patch fixes to detoast pg_group.grolist.
Tom Lane [Sun, 26 Jan 2003 23:16:23 +0000 (23:16 +0000)]
Back-patch fixes to detoast pg_group.grolist.

22 years agoBack-patch fixes to ensure t_ctid always has correct value (prevents
Tom Lane [Sun, 26 Jan 2003 23:09:37 +0000 (23:09 +0000)]
Back-patch fixes to ensure t_ctid always has correct value (prevents
some instances of 'No one parent tuple' VACUUM error, and perhaps
worse things).

22 years agoBack-patch fix for alphabetization mistakes in datetime token tables.
Tom Lane [Sun, 26 Jan 2003 22:33:16 +0000 (22:33 +0000)]
Back-patch fix for alphabetization mistakes in datetime token tables.

22 years agoBack-patch fix to ensure pg_clog updates are not only written but sync'ed
Tom Lane [Tue, 21 Jan 2003 19:51:42 +0000 (19:51 +0000)]
Back-patch fix to ensure pg_clog updates are not only written but sync'ed
before we consider the checkpoint to be done.

22 years agoBack-patch fixes for integer overflows in circle_poly(), path_encode(),
Tom Lane [Tue, 21 Jan 2003 19:41:26 +0000 (19:41 +0000)]
Back-patch fixes for integer overflows in circle_poly(), path_encode(),
and path_add() --- from Neil Conway.  Also, repair recently-detected
errors in lseg_eq(), lseg_ne(), lseg_center().

22 years agoBack-patch fix for VACUUM being confused by SELECT FOR UPDATE of tuple
Tom Lane [Tue, 21 Jan 2003 19:38:21 +0000 (19:38 +0000)]
Back-patch fix for VACUUM being confused by SELECT FOR UPDATE of tuple
that was previously outdated by a transaction that later aborted.
Also, prevent VACUUM from being called inside function.

22 years agoBrand 7.2.3. REL7_2_3
Bruce Momjian [Tue, 1 Oct 2002 03:34:29 +0000 (03:34 +0000)]
Brand 7.2.3.

22 years agoUpdate HISTORY/release.sgml for 7.2.3.
Bruce Momjian [Tue, 1 Oct 2002 03:24:17 +0000 (03:24 +0000)]
Update HISTORY/release.sgml for 7.2.3.

22 years agoBack-patch fixes to work around broken mktime() in recent glibc releases.
Tom Lane [Mon, 30 Sep 2002 20:57:11 +0000 (20:57 +0000)]
Back-patch fixes to work around broken mktime() in recent glibc releases.

22 years agoBack-patch fix for bad SIGUSR2 interrupt handling during backend shutdown.
Tom Lane [Mon, 30 Sep 2002 20:47:22 +0000 (20:47 +0000)]
Back-patch fix for bad SIGUSR2 interrupt handling during backend shutdown.

22 years agoBack-patch fix for correct TAS operation on multi-CPU PPC machines.
Tom Lane [Mon, 30 Sep 2002 20:24:53 +0000 (20:24 +0000)]
Back-patch fix for correct TAS operation on multi-CPU PPC machines.

22 years agoBack-patch fix for 'can't wait without a PROC structure' failures:
Tom Lane [Mon, 30 Sep 2002 20:18:59 +0000 (20:18 +0000)]
Back-patch fix for 'can't wait without a PROC structure' failures:
remove separate ShutdownBufferPoolAccess exit callback, and do the
work in ProcKill instead, before we delete MyProc.

22 years agoBack-patch fix to ensure a checkpoint occurs before truncating CLOG,
Tom Lane [Mon, 30 Sep 2002 19:55:08 +0000 (19:55 +0000)]
Back-patch fix to ensure a checkpoint occurs before truncating CLOG,
even if no recent WAL activity has occurred.

22 years agoBack-patch fix to not change pg_database.datvacuumxid or truncate CLOG
Tom Lane [Mon, 30 Sep 2002 19:45:57 +0000 (19:45 +0000)]
Back-patch fix to not change pg_database.datvacuumxid or truncate CLOG
when an unprivileged user runs VACUUM.

22 years agoBack-patch fix for failure to dump views containing FULL JOIN USING.
Tom Lane [Fri, 20 Sep 2002 21:37:07 +0000 (21:37 +0000)]
Back-patch fix for failure to dump views containing FULL JOIN USING.
The bug is not present in CVS tip due to cleanup of JOIN handling,
but 7.2.* is broken.

22 years agoUpdate HISTORY with Neil's changes. REL7_2_2
Bruce Momjian [Fri, 23 Aug 2002 02:33:06 +0000 (02:33 +0000)]
Update HISTORY with Neil's changes.

22 years agoStamp the 7.2.2 release.
Bruce Momjian [Thu, 22 Aug 2002 23:26:26 +0000 (23:26 +0000)]
Stamp the 7.2.2 release.

22 years agoFix regression-test breakage from (unnecessary) rewrite of show_datestyle. Grumble.
Tom Lane [Thu, 22 Aug 2002 22:17:22 +0000 (22:17 +0000)]
Fix regression-test breakage from (unnecessary) rewrite of show_datestyle.  Grumble.

22 years agoBack-patch fix to make partial indexes usable on relations other than
Tom Lane [Thu, 22 Aug 2002 16:20:38 +0000 (16:20 +0000)]
Back-patch fix to make partial indexes usable on relations other than
the first one listed in a query.  Per request from Oleg.

22 years agoApply Neil Conway's security patches to 7.2.X.
Bruce Momjian [Thu, 22 Aug 2002 05:27:41 +0000 (05:27 +0000)]
Apply Neil Conway's security patches to 7.2.X.

22 years agoFix bug in pg_convert() per report from MaC.Yui.
Tatsuo Ishii [Mon, 19 Aug 2002 04:58:26 +0000 (04:58 +0000)]
Fix bug in pg_convert() per report from MaC.Yui.
It pfree() wrong pointer.

22 years agoBack-patch 7.3 fix to fully parenthesize UNION/EXCEPT/INTERSECT queries
Tom Lane [Sat, 15 Jun 2002 18:38:10 +0000 (18:38 +0000)]
Back-patch 7.3 fix to fully parenthesize UNION/EXCEPT/INTERSECT queries
in ruleutils output.  The previous partial parenthesization was a hack
to get around grammar restrictions that have since been fixed; and as
Sam O'Connor pointed out, there were cases where it didn't work.

22 years agoPlease apply attached patch to contrib/intarray (7.2, 7.3).
Bruce Momjian [Fri, 7 Jun 2002 21:53:09 +0000 (21:53 +0000)]
Please apply attached patch to contrib/intarray (7.2, 7.3).

     Fixed bug with '=' operator for gist__int_ops and
     define '=' operator for gist__intbig_ops opclass.
     Now '=' operator is consistent with standard 'array' type.

     Thanks Achilleus Mantzios for bug report and suggestion.

Oleg Bartunov

22 years agoRepair incorrect dumping of user-defined aggregate with null initcond.
Tom Lane [Tue, 28 May 2002 15:40:36 +0000 (15:40 +0000)]
Repair incorrect dumping of user-defined aggregate with null initcond.
(Already fixed in current, but need a patch for 7.2.2.)

22 years agoFix a bug with building rtree_gist indexes.
Tom Lane [Tue, 28 May 2002 15:25:03 +0000 (15:25 +0000)]
Fix a bug with building rtree_gist indexes.
Patch from Teodor Sigaev.

22 years agoRepair error with not adjusting active scans properly after gistSplit.
Tom Lane [Tue, 28 May 2002 15:22:44 +0000 (15:22 +0000)]
Repair error with not adjusting active scans properly after gistSplit.
Patch from Teodor Sigaev.

22 years agoMake RelationForgetRelation error out if the relcache entry has nonzero
Tom Lane [Wed, 22 May 2002 17:29:45 +0000 (17:29 +0000)]
Make RelationForgetRelation error out if the relcache entry has nonzero
reference count.  This avoids leaving dangling pointers around, as in
recent bug report against sequences (bug# 671).

22 years agoRepair OPEN cursor(args), which I broke on 11/29/01 with a change to
Tom Lane [Tue, 21 May 2002 18:50:18 +0000 (18:50 +0000)]
Repair OPEN cursor(args), which I broke on 11/29/01 with a change to
be smarter about parentheses in read_sql_construct().  Sigh.

22 years agoRemove unnecessary pfree's in geometric operators. At least one of these
Tom Lane [Tue, 14 May 2002 18:16:54 +0000 (18:16 +0000)]
Remove unnecessary pfree's in geometric operators.  At least one of these
is actively dangerous, per bug report from Ewald Geschwinde 14-May-02,
and several of the rest look suspicious to me.  Since there is no longer
any significant value in retail pfree's in these functions, just get
rid of all of them for safety's sake.

22 years agoFix bug in pg_dump and psql (to reproduce the bug, just try pg_dump
Tatsuo Ishii [Tue, 14 May 2002 02:08:22 +0000 (02:08 +0000)]
Fix bug in pg_dump and psql (to reproduce the bug, just try pg_dump
--nonexistingoption).

per report from sugita@sra.co.jp on Thu, 09 May 2002 11:57:51 +0900
(JST) at pgsql-patches list.

Illegal long options to pg_dump makes core on some systems, since it
lacks the last null sentinel of struct option array.

Attached is a patch made by Mr. Ishida Akio <iakio@pjam.jpweb.net>.

22 years agoplpgsql_dstring_append was broken for long strings.
Tom Lane [Sun, 5 May 2002 17:38:28 +0000 (17:38 +0000)]
plpgsql_dstring_append was broken for long strings.

22 years agoChange heap_get_latest_tid() so that a transaction can
Hiroshi Inoue [Wed, 1 May 2002 01:27:31 +0000 (01:27 +0000)]
Change heap_get_latest_tid() so that a transaction can
see changes made by the transaction itself.

22 years agoA backport patch.
Tatsuo Ishii [Mon, 8 Apr 2002 06:21:31 +0000 (06:21 +0000)]
A backport patch.
Fix PQescapeBytea/PQunescapeBytea so that they handle bytes > 0x7f.
This is necessary for mulibyte character sequences.
See "[HACKERS] PQescapeBytea is not multibyte aware" thread posted around
2002/04/05 for more details.

23 years agoFix CLOG truncation code to not do the Wrong Thing when there are already
Tom Lane [Tue, 2 Apr 2002 05:12:00 +0000 (05:12 +0000)]
Fix CLOG truncation code to not do the Wrong Thing when there are already
wrapped-around databases.  The unvacuumed databases might be fine, or
they might not, but things will definitely not be fine if we remove the
wrong CLOG segments.  Per trouble report from Gary Wolfe, 1-Apr-2002.

23 years agoUpdate release list. REL7_2_1
Bruce Momjian [Tue, 26 Mar 2002 05:34:37 +0000 (05:34 +0000)]
Update release list.

23 years agoUpdate sgml version properly.
Bruce Momjian [Tue, 26 Mar 2002 05:33:47 +0000 (05:33 +0000)]
Update sgml version properly.

23 years agoNew wording:
Bruce Momjian [Mon, 25 Mar 2002 21:24:46 +0000 (21:24 +0000)]
New wording:

 Allow EXECUTE of "CREATE TABLE AS ... SELECT" in PL/PgSQL (Tom)

23 years agoAdd to HISTORY:
Bruce Momjian [Mon, 25 Mar 2002 20:58:40 +0000 (20:58 +0000)]
Add to HISTORY:

 Allow CREATE TABLE AS ... SELECT in PL/PgSQL (Tom)

23 years agoRe-allow CREATE AS (but not SELECT INTO) in EXECUTE.
Tom Lane [Mon, 25 Mar 2002 07:41:21 +0000 (07:41 +0000)]
Re-allow CREATE AS (but not SELECT INTO) in EXECUTE.

23 years agoRemove long-dead 'fix for SELECT NULL' to stop current coredump.
Tom Lane [Thu, 21 Mar 2002 06:21:14 +0000 (06:21 +0000)]
Remove long-dead 'fix for SELECT NULL' to stop current coredump.

23 years ago> > Users of contrib/tsearch needs after upgrading of module (compiling, installing)
Bruce Momjian [Tue, 19 Mar 2002 14:16:32 +0000 (14:16 +0000)]
> > Users of contrib/tsearch needs after upgrading of module (compiling, installing)
> > to perform sql command:
> >  update pg_amop set amopreqcheck = true where amopclaid =
> >       (select oid from pg_opclass where opcname = 'gist_txtidx_ops');
>
> Oleg, sorry, I don't understand where this should appear.  In the README
> file, and if so, where?  Is this something only for people upgrading
> from 7.2?

Sorry Bruce, I was unclear. I have attached patch to Readme.tsearch
Also, It'd be worth to mention in Changes to point users of tsearch
about importang upgrade notices.

Oleg Bartunov

23 years agoImprove wording of bug fix:
Bruce Momjian [Tue, 19 Mar 2002 00:12:41 +0000 (00:12 +0000)]
Improve wording of bug fix:

Ensure that sequence counters do not go backwards after a crash

23 years agoUpdate for 7.2.1.
Bruce Momjian [Mon, 18 Mar 2002 23:04:11 +0000 (23:04 +0000)]
Update for 7.2.1.

23 years agoBackpatch array I/O code and documentation fixes, also array slice
Tom Lane [Sun, 17 Mar 2002 20:05:59 +0000 (20:05 +0000)]
Backpatch array I/O code and documentation fixes, also array slice
lower subscript bounds change.

23 years agoRepair bad calculation in timetz_izone() which gave erroneous results.
Thomas G. Lockhart [Fri, 15 Mar 2002 23:37:48 +0000 (23:37 +0000)]
Repair bad calculation in timetz_izone() which gave erroneous results.
Will be patched separately but equivalently in HEAD for 7.3.