Heikki Linnakangas [Wed, 18 Feb 2009 14:19:08 +0000 (16:19 +0200)] Replace two remaining calls of IsRecoveryProcessingMode() with
RecoveryInProgress().
Heikki Linnakangas [Wed, 18 Feb 2009 14:06:34 +0000 (16:06 +0200)] Comment fixes. Don't try to skip AbsorbFsyncRequests in bgwriter when in
recovery. We do it elsewhere anyway, and it's a very inexpensive call.
Heikki Linnakangas [Wed, 18 Feb 2009 13:21:04 +0000 (15:21 +0200)] Move DEBUG2 message inside if section, per Simon's suggestion.
Heikki Linnakangas [Wed, 18 Feb 2009 09:02:10 +0000 (11:02 +0200)] Rename IsRecoveryProcessingMode -> RecoveryInProgress, to avoid confusion
with the actual "processing modes", like BootstrapProcessingMode.
Heikki Linnakangas [Wed, 18 Feb 2009 08:00:49 +0000 (10:00 +0200)] Merge commit 'origin/master' into recoveryinfra
Heikki Linnakangas [Wed, 18 Feb 2009 07:59:44 +0000 (09:59 +0200)] Remove BgWriterRecoveryProcessingMode, in favor of a local variable
just where we decide whether to perform a checkpoint or restartpoint.
Bruce Momjian [Tue, 17 Feb 2009 22:32:54 +0000 (22:32 +0000)] Add missing newline.
Heikki Linnakangas [Tue, 17 Feb 2009 18:53:37 +0000 (20:53 +0200)] Merge commit 'origin/master' into recoveryinfra
Bruce Momjian [Tue, 17 Feb 2009 15:41:50 +0000 (15:41 +0000)] Add pg_dump --binary-upgrade flag to be used by binary upgrade
utilities.
The new code allows transfer of dropped column information to the
upgraded server.
Peter Eisentraut [Tue, 17 Feb 2009 13:01:13 +0000 (13:01 +0000)] Add _() calls for the argument of plpgsql_yyerror() so it actually gets
translated somehow.
Peter Eisentraut [Tue, 17 Feb 2009 12:59:35 +0000 (12:59 +0000)] Add plpgsql_yyerror to gettext triggers
Peter Eisentraut [Tue, 17 Feb 2009 12:51:59 +0000 (12:51 +0000)] Mark 3rd argument of validate_tupdesc_compat() for translation, instead of
marking up each instance separately.
Peter Eisentraut [Tue, 17 Feb 2009 11:34:34 +0000 (11:34 +0000)] Redefine _() to dgettext() instead of gettext() so that it uses the plpgsql
text domain, instead of the postgres one (or whatever the default may be).
Peter Eisentraut [Tue, 17 Feb 2009 09:24:57 +0000 (09:24 +0000)] Remove gettext trigger write_stderr(), which isn't used by PLs.
Bruce Momjian [Mon, 16 Feb 2009 23:06:55 +0000 (23:06 +0000)] Wrap some long queries.
Bruce Momjian [Mon, 16 Feb 2009 22:50:41 +0000 (22:50 +0000)] Uppercase keywords in pg_dump.c
Tom Lane [Sun, 15 Feb 2009 20:16:21 +0000 (20:16 +0000)] Teach the planner to treat a partial unique index as proving a variable is
unique for a particular query, if the index predicate is satisfied. This
requires a bit of reordering of operations so that we check the predicates
before doing any selectivity estimates, but shouldn't really cause any
noticeable slowdown. Per a comment from Michal Politowski.
Tom Lane [Sun, 15 Feb 2009 18:28:48 +0000 (18:28 +0000)] Minor wordsmithing.
Magnus Hagander [Sun, 15 Feb 2009 13:58:18 +0000 (13:58 +0000)] Loop calling CallNamedPipe() several times in case it fails,
since it can be transient failures, causing kill() to not
properly send signals.
Original from Steve Marshall, modified by me.
Tom Lane [Sat, 14 Feb 2009 20:48:36 +0000 (20:48 +0000)] A couple of marginal performance hacks for the information_schema views:
replace the old recursive-SQL-function implementation of _pg_keysequal()
with use of the built-in array containment operators, and change
table_constraints' UNION to UNION ALL. Per discussion with Octavio Alvarez.
initdb not forced since this doesn't affect results, but you'd need to
initdb or reload the information_schema to see the new definitions.
Tom Lane [Fri, 13 Feb 2009 17:12:04 +0000 (17:12 +0000)] Fix UNLISTEN to fall out quickly if the current backend has never executed
any LISTEN command. This is more important than it used to be because
DISCARD ALL invokes UNLISTEN. Connection-pooled applications making heavy
use of DISCARD ALL were seeing significant contention for pg_listener,
as reported by Matteo Beccati. It seems unlikely that clients using LISTEN
would use pooled connections, so this simple tweak seems sufficient,
especially since the pg_listener implementation is slated to go away soon
anyway.
Back- to 8.3, where DISCARD ALL was introduced.
Andrew Dunstan [Thu, 12 Feb 2009 15:12:47 +0000 (15:12 +0000)] do not use unsetenv.c on mingw
Peter Eisentraut [Thu, 12 Feb 2009 15:11:44 +0000 (15:11 +0000)] The Czech (cs_CZ) and Slovak (sk_SK) locales sort numbers after letters,
instead of vice versa. Update the regression test expectations to support
that. In the plpgsql test, adjust the test data so that this isn't an
issue. In the char and varchar tests, add new expected files.
Peter Eisentraut [Thu, 12 Feb 2009 13:26:03 +0000 (13:26 +0000)] Unset language-related locale settings in any case, otherwise psql will
speak in tongues and mess up the regression test diffs.
Bruce Momjian [Thu, 12 Feb 2009 13:25:33 +0000 (13:25 +0000)] Remove tabs from SGML.
Magnus Hagander [Thu, 12 Feb 2009 12:53:34 +0000 (12:53 +0000)] Don't call SetEnvironmentVariable() when removing an environment variable,
as this seems to crash on at least some versions of MingW. Our current usage
of this function does not require it, so it should be ok to ignore.
Tom Lane [Wed, 11 Feb 2009 21:11:16 +0000 (21:11 +0000)] Change ALTER TABLE SET WITHOUT OIDS to rewrite the whole table to physically
get rid of the OID column. This eliminates the problem discovered by Heikki
back in November that 8.4's suppression of "unnecessary" junk filtering in
INSERT/SELECT could lead to an Assert failure, or storing of oids into a table
that shouldn't have them if Asserts are off. While that particular problem
could have been solved in other ways, it seems likely to be just a forerunner
of things to come if we continue to allow tables to contain rows that disagree
with the pg_class.relhasoids setting. It's better to make this operation slow
than to sacrifice performance or risk bugs in more common code paths.
Also, add ALTER TABLE SET WITH OIDS to rewrite the table to add oids.
This was a bit more controversial, but in view of the very small amount of
extra code needed given the current ALTER TABLE infrastructure, it seems best
to eliminate the asymmetry in features.
Tom Lane [Wed, 11 Feb 2009 20:02:40 +0000 (20:02 +0000)] Tweak configure to attempt to add -qnoansialias to CFLAGS whenever running
on AIX with a non-gcc compiler. The previous coding would do this only if
CC was exactly "xlc"; which is a bad idea, as demonstrated by trouble report
from Mihai Criveti.
Alvaro Herrera [Wed, 11 Feb 2009 19:12:04 +0000 (19:12 +0000)] Improve psql \d+ to show TOAST reloptions too. Per complaint from
ITAGAKI Takahiro.
Peter Eisentraut [Wed, 11 Feb 2009 14:03:42 +0000 (14:03 +0000)] Only unset the locale environment when --no-locale is used (the way it was
presumably designed, but didn't act). This allows running the temp install
tests in a non-C locale, thus exercising users' real environments better.
Document how to change locales for test runs.
Bruce Momjian [Wed, 11 Feb 2009 04:08:47 +0000 (04:08 +0000)] Clarify PQinitSSL() documentation to mention what the argument controls.
Peter Eisentraut [Tue, 10 Feb 2009 19:29:39 +0000 (19:29 +0000)] Support for KOI8U encoding
Peter Eisentraut [Tue, 10 Feb 2009 16:44:44 +0000 (16:44 +0000)] Remove the encoding *numbers* from the comments. They are useless, and
make maintenance harder.
Peter Eisentraut [Tue, 10 Feb 2009 16:36:55 +0000 (16:36 +0000)] Add possibility to generate only some files, by passing command-line
arguments.
Bruce Momjian [Tue, 10 Feb 2009 00:55:21 +0000 (00:55 +0000)] Update wording of how to prevent pg_dump from affecting statistics
collection.
Tom Lane [Mon, 9 Feb 2009 21:18:28 +0000 (21:18 +0000)] Adopt Bob Jenkins' improved hash function for hash_any(). This changes the
contents of hash indexes (again), so bump catversion.
Kenneth Marshall
Alvaro Herrera [Mon, 9 Feb 2009 20:57:59 +0000 (20:57 +0000)] Update autovacuum to use reloptions instead of a system catalog, for
per-table overrides of parameters.
This removes a whole class of problems related to misusing the catalog,
and perhaps more importantly, gives us pg_dump support for the parameters.
Based on a by Euler Taveira de Oliveira, heavily reworked by me.
Heikki Linnakangas [Mon, 9 Feb 2009 15:00:31 +0000 (17:00 +0200)] Comment and error/debug message enhancements
Bruce Momjian [Mon, 9 Feb 2009 14:09:56 +0000 (14:09 +0000)] Fix spelling error in docs.
Erik Rijkers
Heikki Linnakangas [Mon, 9 Feb 2009 12:54:41 +0000 (14:54 +0200)] Add IsRecoveryProcessingMode() quick exits to XLogNeedsFlush,
XLogAsyncCommitFlush and XLogBackgroundFlush. Fix restore command
invocation so that fast shutdown requests are not lost. Update
minRecoveryPoint in CreateRestartPoint when we can't create a
restart point.
Heikki Linnakangas [Mon, 9 Feb 2009 09:52:19 +0000 (11:52 +0200)] If a restartpoint is in progress at end of recovery, hurry it.
Heikki Linnakangas [Mon, 9 Feb 2009 08:13:19 +0000 (10:13 +0200)] If bgwriter skips a restartpoint because we haven't seen any new checkpoint
records since last restartpoint, try again in 15 s, instead of
checkpoint_timeout. Plus some comment changes
Tom Lane [Sun, 8 Feb 2009 18:02:27 +0000 (18:02 +0000)] Ensure that INSERT ... SELECT into a table with OIDs never copies row OIDs
from the source table. This could never happen anyway before 8.4 because
the executor invariably applied a "junk filter" to rows due to be inserted;
but now that we skip doing that when it's not necessary, the case can occur.
Problem noted 2008-11-27 by KaiGai Kohei, though I misunderstood what he
was on about at the time (the opacity of the he proposed didn't help).
Bruce Momjian [Sat, 7 Feb 2009 20:11:16 +0000 (20:11 +0000)] Document that LIMIT NULL is the same as no LIMIT clause.
Bruce Momjian [Sat, 7 Feb 2009 20:05:44 +0000 (20:05 +0000)] Remove documentation mention that hash indexes perform no better than
btree; keep mention about missing WAL logging.
Kenneth Marshall
Bruce Momjian [Sat, 7 Feb 2009 19:27:25 +0000 (19:27 +0000)] Uppercase CHECK mention in relchecks documentation mention.
Bruce Momjian [Sat, 7 Feb 2009 17:17:34 +0000 (17:17 +0000)] ecpg requires libpq; add Makefile rules to require libpq to be built
first.
Alvaro Herrera
Bruce Momjian [Sat, 7 Feb 2009 15:25:51 +0000 (15:25 +0000)] Properly wrap new pg_dump doc text.
Bruce Momjian [Sat, 7 Feb 2009 15:00:27 +0000 (15:00 +0000)] Update pg_constraint.conkey documentation description.
Bruce Momjian [Sat, 7 Feb 2009 14:31:30 +0000 (14:31 +0000)] Document disabling the statistics collector pg_dump activity, and give a
bit more visibility to the PGOPTIONS environment variable supported by
libpq.
Bryce Nesbitt
Bruce Momjian [Sat, 7 Feb 2009 14:16:46 +0000 (14:16 +0000)] Fix to_timestamp() to not require upper/lower case matching for meridian
designations (AM/PM). Also separate out matching of a meridian with
periods (e.g. A.M.) and with those without.
Do the same for AD/BC.
Brendan Jurd
Heikki Linnakangas [Sat, 7 Feb 2009 10:49:36 +0000 (10:49 +0000)] Fix obsolete comment. Zdenek Kotala
Peter Eisentraut [Sat, 7 Feb 2009 01:02:55 +0000 (01:02 +0000)] Set column privileges to supported
Tom Lane [Fri, 6 Feb 2009 23:43:24 +0000 (23:43 +0000)] Fix cost_mergejoin's failure to adjust for rescanning of non-unique merge join
keys when considering a semi or anti join. This requires estimating the
selectivity of the merge qual as though it were a regular inner join condition.
To allow caching both that and the real outer-join-aware selectivity, split
RestrictInfo.this_selec into two fields.
This fixes one of the problems reported by Kevin Grittner.
Tom Lane [Fri, 6 Feb 2009 21:22:49 +0000 (21:22 +0000)] Ooops, forgot to update COPY reference page to mention column privileges.
Tom Lane [Fri, 6 Feb 2009 21:15:12 +0000 (21:15 +0000)] Clean up some loose ends from the column privileges : add
has_column_privilege and has_any_column_privilege SQL functions; fix the
information_schema views that are supposed to pay attention to column
privileges; adjust pg_stats to show stats for any column you have select
privilege on; and fix COPY to allow copying a subset of columns if the user
has suitable per-column privileges for all the columns.
To improve efficiency of some of the information_schema views, extend the
has_xxx_privilege functions to allow inquiring about the OR of a set of
privileges in just one call. This is just exposing capability that already
existed in the underlying aclcheck routines.
In passing, make the information_schema views report the owner's own
privileges as being grantable, since Postgres assumes this even when the grant
option bit is not set in the ACL. This is a longstanding oversight.
Also, make the new has_xxx_privilege functions for foreign data objects follow
the same coding conventions used by the older ones.
Stephen Frost and Tom Lane
Bruce Momjian [Fri, 6 Feb 2009 19:24:12 +0000 (19:24 +0000)] Fix typo in docs.
Bruce Momjian [Fri, 6 Feb 2009 18:18:54 +0000 (18:18 +0000)] Document ways to avoid libpq WSACleanup() overhead on Windows.
Andrew Chernow
Heikki Linnakangas [Thu, 5 Feb 2009 17:02:06 +0000 (19:02 +0200)] comment fixex
Heikki Linnakangas [Thu, 5 Feb 2009 16:23:36 +0000 (18:23 +0200)] Merge commit 'origin/master' into recoveryinfra
Heikki Linnakangas [Thu, 5 Feb 2009 16:21:47 +0000 (18:21 +0200)] Fix crash handling, so that we don't try to reinitialize after startup
process crash.
Bruce Momjian [Thu, 5 Feb 2009 15:25:49 +0000 (15:25 +0000)] Add PL/PgSQL FOUND and GET DIAGNOSTICS support for RETURN QUERY
statement
Pavel Stehule
Heikki Linnakangas [Thu, 5 Feb 2009 12:28:31 +0000 (14:28 +0200)] Revert some more changes back the way it's in CVS HEAD, now that we
do startup checkpoints again.
Heikki Linnakangas [Thu, 5 Feb 2009 11:38:45 +0000 (13:38 +0200)] Change error message if recovery ends before reaching consistency.
Heikki Linnakangas [Thu, 5 Feb 2009 11:21:20 +0000 (13:21 +0200)] Bring back startup checkpoints. Plus some other small changes
Michael Meskes [Thu, 5 Feb 2009 08:28:47 +0000 (08:28 +0000)] Mixing long and int entries of a union only works on some architectures.
Tom Lane [Thu, 5 Feb 2009 01:24:55 +0000 (01:24 +0000)] Fix an old corner-case error in match_unsorted_outer(): don't consider
the cheapest-total inner path as a new candidate while truncating the
sort key list, if it already matched the full sort key list. This is
too much of a corner case to be worth back-ing, since it's unusual
for the cheapest total path to be sorted, and anyway no real harm is
done (except in JOIN_SEMI/ANTI cases where cost_mergejoin is a bit
broken at the moment). But it wasn't behaving as intended, so fix it.
Noted while examining a test case from Kevin Grittner. This error doesn't
explain his issue, but it does explain why "set enable_seqscan = off"
seemed to reproduce it for me.
Alvaro Herrera [Wed, 4 Feb 2009 21:30:41 +0000 (21:30 +0000)] Fix some SGML-compiler warnings on -wxml mode.
Heikki Linnakangas [Wed, 4 Feb 2009 18:15:44 +0000 (20:15 +0200)] Don't rename recovery.conf out of the way until the first checkpoint, like
it was done in Simon's original . And fix some other comments pointed out
by Simon.
Bruce Momjian [Wed, 4 Feb 2009 16:05:50 +0000 (16:05 +0000)] Update read committed documentation to better explain undesirable
behavior of concurrent commands in cases where rows are being added and
removed from matching query criteria.
Minor word-smithing.
Heikki Linnakangas [Wed, 4 Feb 2009 12:34:01 +0000 (14:34 +0200)] Merge commit 'origin/master' into recoveryinfra
Heikki Linnakangas [Wed, 4 Feb 2009 12:33:28 +0000 (14:33 +0200)] Fix FatalError handling. Bug pointed out by Fujii-san.
Michael Meskes [Wed, 4 Feb 2009 08:51:10 +0000 (08:51 +0000)] Added result checks for calls to gmtime().
Heikki Linnakangas [Tue, 3 Feb 2009 12:20:59 +0000 (14:20 +0200)] Remove minSafeStartPoint, advance minRecoveryPoint instead. Advance it in
XLogFlush, instead of when a WAL file is restored.
Michael Meskes [Tue, 3 Feb 2009 08:55:45 +0000 (08:55 +0000)] Argh, wrong line copied again.
Michael Meskes [Tue, 3 Feb 2009 08:48:29 +0000 (08:48 +0000)] Fixed copy&paste mistake that made library use uninitialized variable.
Andrew Dunstan [Tue, 3 Feb 2009 01:24:57 +0000 (01:24 +0000)] allow alternative names for tclsh used on Windows
Tom Lane [Tue, 3 Feb 2009 00:59:26 +0000 (00:59 +0000)] Add a failure check for syslogger's use of _beginthreadex(), and remove
unnecessary thread address output parameter, to make this code look more
like that in pg_restore.
Tom Lane [Mon, 2 Feb 2009 20:42:57 +0000 (20:42 +0000)] Document the longstanding behavior of LIMIT NULL and OFFSET NULL,
per gripe from David Wheeler that this was mentioned nowhere.
In passing, editorialize a bit on the description of the SQL:2008
equivalent syntax.
Tom Lane [Mon, 2 Feb 2009 20:25:38 +0000 (20:25 +0000)] Fix plpgsql to not treat INSERT INTO as an INTO-variables clause anywhere
in the string, not just at the start. Per bug #4629 from Martin Blazek.
Back- to 8.2; prior versions don't have the problem, at least not in
the reported case, because they don't try to recognize INTO in non-SELECT
statements. (IOW, this is really fallout from the RETURNING .)
Andrew Dunstan [Mon, 2 Feb 2009 20:07:37 +0000 (20:07 +0000)] Provide for parallel restoration from a custom format archive. Each data and
post-data step is run in a separate worker child (a thread on Windows, a child
process elsewhere) up to the concurrent number specified by the new pg_restore
command-line --multi-thread | -m switch.
Andrew Dunstan, with some editing by Tom Lane.
Alvaro Herrera [Mon, 2 Feb 2009 19:31:40 +0000 (19:31 +0000)] Allow reloption names to have qualifiers, initially supporting a TOAST
qualifier, and add support for this in pg_dump.
This allows TOAST tables to have user-defined fillfactor, and will also
enable us to move the autovacuum parameters to reloptions without taking
away the possibility of setting values for TOAST tables.
Michael Meskes [Mon, 2 Feb 2009 16:54:28 +0000 (16:54 +0000)] When changing a regression test file one should also change the expected output file.
Michael Meskes [Mon, 2 Feb 2009 16:49:50 +0000 (16:49 +0000)] Set memory segment to '0', just to be sure.
Michael Meskes [Mon, 2 Feb 2009 16:14:06 +0000 (16:14 +0000)] Fixed auto allocation for binary data types.
Michael Meskes [Mon, 2 Feb 2009 15:35:28 +0000 (15:35 +0000)] Fixed bug in %s handling.
Peter Eisentraut [Mon, 2 Feb 2009 09:49:29 +0000 (09:49 +0000)] Put pg_user_mapping section into the right (alphabetical) order.
Tom Lane [Fri, 30 Jan 2009 21:21:18 +0000 (21:21 +0000)] Fix unnest() to handle a toasted input array safely. Per report from
Alvaro.
Heikki Linnakangas [Fri, 30 Jan 2009 17:24:47 +0000 (17:24 +0000)] Check that connection limit is within valid range. IOW, not < -1.
It's missing in older versions too, but it doesn't seem worth
back-porting. All negative are just harmlessly treated as "no limit", and
tightening the check might even brake an application that relies on it.
Heikki Linnakangas [Fri, 30 Jan 2009 14:43:47 +0000 (16:43 +0200)] Comment enhancements
Heikki Linnakangas [Fri, 30 Jan 2009 14:24:55 +0000 (16:24 +0200)] Change the signaling between startup process and post Startup process
now tells postmaster about the start, consistent point, and completion of
WAL redo, using signals. It now returns with exit code 0 when WAL replay is
aborted with SIGTERM. Background writer is launched as soon as we start WAL
redo, so that it can perform restartpoints before reaching consistent state.
Postmaster handles unexpected death of the startup process gracefully. In
case of a fast shutdown, bgwriter performs a "shutdown restartpoint", similar
to a shutdown checkpoint.
Peter Eisentraut [Fri, 30 Jan 2009 12:53:43 +0000 (12:53 +0000)] Make messages consistent with existing ones
Peter Eisentraut [Fri, 30 Jan 2009 12:53:12 +0000 (12:53 +0000)] Whitespace fix in message
Tom Lane [Fri, 30 Jan 2009 00:37:29 +0000 (00:37 +0000)] Update back-branch release notes.
Tom Lane [Thu, 29 Jan 2009 19:59:59 +0000 (19:59 +0000)] Update time zone data files to tzdata release 2009a: introduces Asia/Kathmandu
as the preferred spelling of that zone name, corrects historical DST
information for Switzerland and Cuba.
Heikki Linnakangas [Thu, 29 Jan 2009 19:46:43 +0000 (21:46 +0200)] Move the check for whether we've reached a safe starting point *before*
checking if we've reached the recovery target. Otherwise, if the first safe
starting point is at the exact same record as the recovery target, we'd
fail to start up.
Tom Lane [Thu, 29 Jan 2009 19:23:42 +0000 (19:23 +0000)] Replace argument-checking Asserts with regular test-and-elog checks in all
encoding conversion functions. These are not can't-happen cases because
it's possible to create a conversion with the wrong conversion function
for the specified encoding pair. That would lead to an Assert crash in
an Assert-enabled build, or incorrect conversion otherwise, neither of
which is desirable. This would be a DOS issue if production databases
were customarily built with asserts enabled, but fortunately that's not so.
Per an observation by Heikki.
Back- to all supported branches.
Heikki Linnakangas [Thu, 29 Jan 2009 18:16:00 +0000 (20:16 +0200)] Wait for bgwriter to die after startup process ends in PM_RECOVERY. More
needs to be done, but it's a start.
Teodor Sigaev [Thu, 29 Jan 2009 16:22:10 +0000 (16:22 +0000)] Fix incorrect dereferencing of char* to array's index.
Per Tommy Gildseth <
[email protected]> report
Heikki Linnakangas [Thu, 29 Jan 2009 12:09:05 +0000 (14:09 +0200)] Make IsRecoveryProcessingMode() init ThisTimeLineID, to make sure
it's always set before it's used.