]> granicus.if.org Git - postgresql/log
postgresql
15 years agoPartial fix for contrib vpath breakage.
Tom Lane [Thu, 2 Oct 2008 12:25:25 +0000 (12:25 +0000)]
Partial fix for contrib vpath breakage.

15 years agoMake the blkno arguments bigints instead of int4s. A signed int4 is not
Heikki Linnakangas [Thu, 2 Oct 2008 12:20:50 +0000 (12:20 +0000)]
Make the blkno arguments bigints instead of int4s. A signed int4 is not
large enough for block numbers higher than 2^31. The old pre-FSM-rewrite
pg_freespacemap implementation got this right. While we're at it, remove
some unnecessary #includes.

15 years agoFix function name in pg_freespacemap docs. It's "pg_freespace", not
Heikki Linnakangas [Thu, 2 Oct 2008 10:26:51 +0000 (10:26 +0000)]
Fix function name in pg_freespacemap docs. It's "pg_freespace", not
"pg_freespacemap". Reported by hubert depesz lubaczewski.

15 years agoUpdate pg_regress calls in PL checks to handle vpath builds.
Peter Eisentraut [Thu, 2 Oct 2008 08:11:11 +0000 (08:11 +0000)]
Update pg_regress calls in PL checks to handle vpath builds.

15 years agoAllow pg_regress to be run outside the build tree. Look for input files
Peter Eisentraut [Wed, 1 Oct 2008 22:38:57 +0000 (22:38 +0000)]
Allow pg_regress to be run outside the build tree.  Look for input files
in both input and output dir, to handle vpath builds more simply.

15 years agoImprove tuplestore.c to support multiple concurrent read positions.
Tom Lane [Wed, 1 Oct 2008 19:51:50 +0000 (19:51 +0000)]
Improve tuplestore.c to support multiple concurrent read positions.
This facility replaces the former mark/restore support but is otherwise
upward-compatible with previous uses.  It's expected to be needed for
single evaluation of CTEs and also for window functions, so I'm committing
it separately instead of waiting for either one of those patches to be
finished.  Per discussion with Greg Stark and Hitoshi Harada.

Note: I removed nodeFunctionscan's mark/restore support, instead of bothering
to update it for this change, because it was dead code anyway.

15 years agoLink libpq with libgssapi if configure finds it, as
Magnus Hagander [Wed, 1 Oct 2008 15:35:32 +0000 (15:35 +0000)]
Link libpq with libgssapi if configure finds it, as
required by at least NetBSD.

Markus Schaaf

15 years agoSuppress an uninitialized-variable warning (not all versions of gcc
Tom Lane [Wed, 1 Oct 2008 14:59:23 +0000 (14:59 +0000)]
Suppress an uninitialized-variable warning (not all versions of gcc
complain here, but some do)

15 years agoFix WAL redo of FSM truncation. We can't call smgrtruncate() during WAL
Heikki Linnakangas [Wed, 1 Oct 2008 08:12:14 +0000 (08:12 +0000)]
Fix WAL redo of FSM truncation. We can't call smgrtruncate() during WAL
replay, because it tries to XLogInsert().

15 years agoFix compiler warning (unportable sprintf usage)
Tom Lane [Tue, 30 Sep 2008 14:15:58 +0000 (14:15 +0000)]
Fix compiler warning (unportable sprintf usage)

15 years agoFix misspelled comment
Tom Lane [Tue, 30 Sep 2008 12:51:07 +0000 (12:51 +0000)]
Fix misspelled comment

15 years agoRewrite pg_freespacemap to match the new FSM implementation. I missed these
Heikki Linnakangas [Tue, 30 Sep 2008 11:17:07 +0000 (11:17 +0000)]
Rewrite pg_freespacemap to match the new FSM implementation. I missed these
changes from the main FSM commit for some reason.

15 years agoForgot to bump catalog version in the commit of FSM rewrite.
Heikki Linnakangas [Tue, 30 Sep 2008 11:11:28 +0000 (11:11 +0000)]
Forgot to bump catalog version in the commit of FSM rewrite.

15 years agoRewrite the FSM. Instead of relying on a fixed-size shared memory segment, the
Heikki Linnakangas [Tue, 30 Sep 2008 10:52:14 +0000 (10:52 +0000)]
Rewrite the FSM. Instead of relying on a fixed-size shared memory segment, the
free space information is stored in a dedicated FSM relation fork, with each
relation (except for hash indexes; they don't use FSM).

This eliminates the max_fsm_relations and max_fsm_pages GUC options; remove any
trace of them from the backend, initdb, and documentation.

Rewrite contrib/pg_freespacemap to match the new FSM implementation. Also
introduce a new variant of the get_raw_page(regclass, int4, int4) function in
contrib/pageinspect that let's you to return pages from any relation fork, and
a new fsm_page_contents() function to inspect the new FSM pages.

15 years agoDept of second thoughts: let's make sure that get_index_stats_hook is only
Tom Lane [Sun, 28 Sep 2008 20:42:12 +0000 (20:42 +0000)]
Dept of second thoughts: let's make sure that get_index_stats_hook is only
applied to expression indexes, not to plain relations.  The original coding
in btcostestimate conflated the two cases, but it's not hard to use
get_relation_stats_hook instead when we're looking to the underlying relation.

15 years agoAdd hooks to let plugins override the planner's lookups in pg_statistic.
Tom Lane [Sun, 28 Sep 2008 19:51:40 +0000 (19:51 +0000)]
Add hooks to let plugins override the planner's lookups in pg_statistic.
Simon Riggs, with some editorialization by me.

15 years agoCompare escaped chars case insensitively for ILIKE - per gripe from TGL.
Andrew Dunstan [Sat, 27 Sep 2008 16:53:54 +0000 (16:53 +0000)]
Compare escaped chars case insensitively for ILIKE - per gripe from TGL.

15 years agoFix pointer-advancement bugs in MS and US cases of new to_timestamp() code.
Tom Lane [Fri, 26 Sep 2008 15:35:28 +0000 (15:35 +0000)]
Fix pointer-advancement bugs in MS and US cases of new to_timestamp() code.
Alex Hunsaker

15 years agoMake LIKE throw an error if the escape character is at the end of the pattern
Tom Lane [Fri, 26 Sep 2008 02:16:40 +0000 (02:16 +0000)]
Make LIKE throw an error if the escape character is at the end of the pattern
(ie, has nothing to quote), rather than silently ignoring the character as has
been our historical behavior.  This is required by SQL spec and should help
reduce the sort of user confusion seen in bug #4436.  Per discussion.

This is not so much a bug fix as a definitional change, and it could break
existing applications; so not back-patched.  It might deserve being mentioned
as an incompatibility in the 8.4 release notes.

15 years agoEstablish the rule that array types should have the same typdelim as their
Tom Lane [Thu, 25 Sep 2008 03:28:56 +0000 (03:28 +0000)]
Establish the rule that array types should have the same typdelim as their
element types.  Since the backend doesn't actually pay attention to the array
type's delimiter, this has no functional effect, but it seems better for the
catalog entries to be consistent.  Per gripe from Greg Mullane and subsequent
discussion.

15 years agoAdd documentation about when trigger values NEW/OLD return NULL.
Bruce Momjian [Wed, 24 Sep 2008 19:51:22 +0000 (19:51 +0000)]
Add documentation about when trigger values NEW/OLD return NULL.

Jeff Davis

15 years agoFix integral timestamps so the output is consistent in all cases to
Bruce Momjian [Wed, 24 Sep 2008 19:46:44 +0000 (19:46 +0000)]
Fix integral timestamps so the output is consistent in all cases to
round:

select interval '0:0:0.7', interval '@ 0.70 secs', interval '0.7
seconds';

Ron Mayer

15 years agoFix pg_dump bug in the database-level collation patch. "datcollate" and
Heikki Linnakangas [Wed, 24 Sep 2008 19:33:15 +0000 (19:33 +0000)]
Fix pg_dump bug in the database-level collation patch. "datcollate" and
"datctype" columns were misspelled. Per report from Chris Browne.

15 years agoFix more problems with rewriter failing to set Query.hasSubLinks when inserting
Tom Lane [Wed, 24 Sep 2008 16:52:46 +0000 (16:52 +0000)]
Fix more problems with rewriter failing to set Query.hasSubLinks when inserting
a SubLink expression into a rule query.  We missed cases where the original
query contained a sub-SELECT in a function in FROM, a multi-row VALUES list,
or a RETURNING list.  Per bug #4434 from Dean Rasheed and subsequent
investigation.

Back-patch to 8.1; older releases don't have the issue because they didn't
try to be smart about setting hasSubLinks only when needed.

15 years agoFix markup tag error, envvar -> envar.
Bruce Momjian [Wed, 24 Sep 2008 16:30:26 +0000 (16:30 +0000)]
Fix markup tag error, envvar -> envar.

15 years agoFix unportable syntax used in recent patch. Per results from
Tom Lane [Wed, 24 Sep 2008 14:40:00 +0000 (14:40 +0000)]
Fix unportable syntax used in recent patch.  Per results from
buildfarm member 'bear'.

15 years agoSilence compiler warning caused by recent collation patch.
Magnus Hagander [Wed, 24 Sep 2008 09:00:44 +0000 (09:00 +0000)]
Silence compiler warning caused by recent collation patch.

15 years agoMake sure pg_control is opened in binary mode, to deal
Magnus Hagander [Wed, 24 Sep 2008 08:59:42 +0000 (08:59 +0000)]
Make sure pg_control is opened in binary mode, to deal
with situtations when the file contains an EOF maker
(0x1A) on Windows.

ITAGAKI Takahiro

15 years agoOnly show source file and line numbers to superusers, for consistent
Magnus Hagander [Tue, 23 Sep 2008 21:12:03 +0000 (21:12 +0000)]
Only show source file and line numbers to superusers, for consistent
security level with other parts of the system.

Per gripe from Tom

15 years agoAdd comment about the use of EXEC_BACKEND.
Bruce Momjian [Tue, 23 Sep 2008 20:35:38 +0000 (20:35 +0000)]
Add comment about the use of EXEC_BACKEND.

15 years agoMention battery-backed cache under hardware selection options.
Bruce Momjian [Tue, 23 Sep 2008 20:20:07 +0000 (20:20 +0000)]
Mention battery-backed cache under hardware selection options.

15 years agoTighten the check in initdb and CREATE DATABASE that the chosen encoding
Heikki Linnakangas [Tue, 23 Sep 2008 10:58:03 +0000 (10:58 +0000)]
Tighten the check in initdb and CREATE DATABASE that the chosen encoding
matches the encoding of the locale. LC_COLLATE is now checked in addition
to LC_CTYPE.

15 years agoMake LC_COLLATE and LC_CTYPE database-level settings. Collation and
Heikki Linnakangas [Tue, 23 Sep 2008 09:20:39 +0000 (09:20 +0000)]
Make LC_COLLATE and LC_CTYPE database-level settings. Collation and
ctype are now more like encoding, stored in new datcollate and datctype
columns in pg_database.

This is a stripped-down version of Radek Strnad's patch, with further
changes by me.

15 years agoGet rid of pgpass_from_client tracking inside libpq --- given the conclusion
Tom Lane [Mon, 22 Sep 2008 14:21:44 +0000 (14:21 +0000)]
Get rid of pgpass_from_client tracking inside libpq --- given the conclusion
that presence of the password in the conninfo string must be checked *before*
risking a connection attempt, there is no point in checking it afterwards.
This makes the specification of PQconnectionUsedPassword() a bit simpler
and perhaps more generally useful, too.

15 years agoFix dblink_connect() so that it verifies that a password is supplied in the
Tom Lane [Mon, 22 Sep 2008 13:55:14 +0000 (13:55 +0000)]
Fix dblink_connect() so that it verifies that a password is supplied in the
conninfo string *before* trying to connect to the remote server, not after.
As pointed out by Marko Kreen, in certain not-very-plausible situations
this could result in sending a password from the postgres user's .pgpass file,
or other places that non-superusers shouldn't have access to, to an
untrustworthy remote server.  The cleanest fix seems to be to expose libpq's
conninfo-string-parsing code so that dblink can check for a password option
without duplicating the parsing logic.

Joe Conway, with a little cleanup by Tom Lane

15 years agoSimplify the definitions of a couple of system views by using SELECT *
Tom Lane [Sun, 21 Sep 2008 19:38:56 +0000 (19:38 +0000)]
Simplify the definitions of a couple of system views by using SELECT *
instead of listing all the columns returned by the underlying function.

initdb not forced since this patch doesn't actually change anything about
the stored form of the views.  It just means there's one less place to change
if someone wants to add columns to them.

15 years agoAdd a PQfireResultCreateEvents function to allow applications to mimic the
Tom Lane [Fri, 19 Sep 2008 20:06:13 +0000 (20:06 +0000)]
Add a PQfireResultCreateEvents function to allow applications to mimic the
sequence of operations that libpq goes through while creating a PGresult.
Also, remove ill-considered "const" decoration on parameters passed to
event procedures.

15 years agoCreate a selectivity estimation function for the text search @@ operator.
Tom Lane [Fri, 19 Sep 2008 19:03:41 +0000 (19:03 +0000)]
Create a selectivity estimation function for the text search @@ operator.

Jan Urbanski

15 years agoImprove the recently-added libpq events code to provide more consistent
Tom Lane [Fri, 19 Sep 2008 16:40:40 +0000 (16:40 +0000)]
Improve the recently-added libpq events code to provide more consistent
guarantees about whether event procedures will receive DESTROY events.
They no longer need to defend themselves against getting a DESTROY
without a successful prior CREATE.

Andrew Chernow

15 years agoMark SessionReplicationRole as PGDLLIMPORT so it
Magnus Hagander [Fri, 19 Sep 2008 14:43:46 +0000 (14:43 +0000)]
Mark SessionReplicationRole as PGDLLIMPORT so it
can be used from Slony functions.

Per report from Hiroshi Saito.

15 years agoAdd additional rules so that make init-po for ecpg gets up to date flex and
Peter Eisentraut [Fri, 19 Sep 2008 13:08:08 +0000 (13:08 +0000)]
Add additional rules so that make init-po for ecpg gets up to date flex and
bison output.  Without these, make can sometimes be tempted to invoke its
built-in rules using lex and yacc, which can fail if those commands are not
available.

This was a main cause for the NLS web site breakage.

15 years agoOptimize CleanupTempFiles by having a boolean flag that keeps track of whether
Alvaro Herrera [Fri, 19 Sep 2008 04:57:10 +0000 (04:57 +0000)]
Optimize CleanupTempFiles by having a boolean flag that keeps track of whether
there are FD_XACT_TEMPORARY files to clean up at transaction end.

Per performance profiling results on AWeber's huge systems.

Patch by me after an idea suggested by Simon Riggs.

15 years agoPreliminary release notes for upcoming back branch releases.
Tom Lane [Wed, 17 Sep 2008 20:57:35 +0000 (20:57 +0000)]
Preliminary release notes for upcoming back branch releases.

15 years agoUpdate time zone data files to tzdata release 2008f (DST law changes in
Tom Lane [Wed, 17 Sep 2008 14:18:39 +0000 (14:18 +0000)]
Update time zone data files to tzdata release 2008f (DST law changes in
Argentina, Bahamas, Brazil, Mauritius, Morocco, Pakistan, Palestine, Paraguay).

15 years agoAllow ShowBufferUsage() to report the number of reads/writes that have
Tom Lane [Wed, 17 Sep 2008 13:15:55 +0000 (13:15 +0000)]
Allow ShowBufferUsage() to report the number of reads/writes that have
occurred to temporary files.  This replaces the unused
NDirectFileRead/NDirectFileWrite counters.

Itagaki Takahiro

15 years agoAdd an "events" system to libpq, whereby applications can get callbacks that
Tom Lane [Wed, 17 Sep 2008 04:31:08 +0000 (04:31 +0000)]
Add an "events" system to libpq, whereby applications can get callbacks that
enable them to manage private data associated with PGconns and PGresults.

Andrew Chernow and Merlin Moncure

15 years agoClean up a couple of weird corner cases in interval parsing: make -yyyy-mm be
Tom Lane [Tue, 16 Sep 2008 22:31:21 +0000 (22:31 +0000)]
Clean up a couple of weird corner cases in interval parsing: make -yyyy-mm be
interpreted as expected (the sign should affect months too), and get rid of
hard-wired assumption that unmarked signed values must be hours (if integers)
or seconds (if floats).  The former was just a bug in my previous patch,
while the latter may have made sense at one time but seems illogical now
that we support determination of the units from typmod information.
Ron Mayer and myself.

15 years agoWiden the nLocks counts in local lock tables from int to int64. This
Tom Lane [Tue, 16 Sep 2008 01:56:26 +0000 (01:56 +0000)]
Widen the nLocks counts in local lock tables from int to int64.  This
forestalls potential overflow when the same table (or other object, but
usually tables) is accessed by very many successive queries within a single
transaction.  Per report from Michael Milligan.

Back-patch to 8.0, which is as far back as the patch conveniently applies.
There have been no reports of overflow in pre-8.3 releases, but clearly the
risk existed all along.  (Michael's report suggests that 8.3 may consume lock
counts faster than prior releases, but with no test case to look at it's hard
to be sure about that.  Widening the counts seems a good future-proofing
measure in any event.)

15 years agoFix multiple memory leaks in xml_out(). Per report from Matt Magoffin.
Tom Lane [Tue, 16 Sep 2008 00:49:41 +0000 (00:49 +0000)]
Fix multiple memory leaks in xml_out().  Per report from Matt Magoffin.

15 years agoFix caching of foreign-key-checking queries so that when a replan is needed,
Tom Lane [Mon, 15 Sep 2008 23:37:40 +0000 (23:37 +0000)]
Fix caching of foreign-key-checking queries so that when a replan is needed,
we regenerate the SQL query text not merely the plan derived from it.  This
is needed to handle contingencies such as renaming of a table or column
used in an FK.  Pre-8.3, such cases worked despite the lack of replanning
(because the cached plan needn't actually change), so this is a regression.
Per bug #4417 from Benjamin Bihler.

15 years agoFix error messages from recent pg_hba parsing patch to use errcontext()
Magnus Hagander [Mon, 15 Sep 2008 20:55:04 +0000 (20:55 +0000)]
Fix error messages from recent pg_hba parsing patch to use errcontext()
to indicate where the error occurred.

15 years agoChange hash indexes to store only the hash code rather than the whole indexed
Tom Lane [Mon, 15 Sep 2008 18:43:41 +0000 (18:43 +0000)]
Change hash indexes to store only the hash code rather than the whole indexed
value.  This means that hash index lookups are always lossy and have to be
rechecked when the heap is visited; however, the gain in index compactness
outweighs this when the indexed values are wide.  Also, we only need to
perform datatype comparisons when the hash codes match exactly, rather than
for every entry in the hash bucket; so it could also win for datatypes that
have expensive comparison functions.  A small additional win is gained by
keeping hash index pages sorted by hash code and using binary search to reduce
the number of index tuples we have to look at.

Xiao Meng

This commit also incorporates Zdenek Kotala's patch to isolate hash metapages
and hash bitmaps a bit better from the page header datastructures.

15 years agoAdd missing quotes.
Magnus Hagander [Mon, 15 Sep 2008 12:41:54 +0000 (12:41 +0000)]
Add missing quotes.

15 years agoParse pg_hba.conf in postmaster, instead of once in each backend for
Magnus Hagander [Mon, 15 Sep 2008 12:32:57 +0000 (12:32 +0000)]
Parse pg_hba.conf in postmaster, instead of once in each backend for
each connection. This makes it possible to catch errors in the pg_hba
file when it's being reloaded, instead of silently reloading a broken
file and failing only when a user tries to connect.

This patch also makes the "sameuser" argument to ident authentication
optional.

15 years agoAvoid compiler warning about variable used before assigned.
Peter Eisentraut [Mon, 15 Sep 2008 12:18:00 +0000 (12:18 +0000)]
Avoid compiler warning about variable used before assigned.

15 years agoPolish and Turkish FAQ also needed converting to UTF8
Magnus Hagander [Mon, 15 Sep 2008 08:50:09 +0000 (08:50 +0000)]
Polish and Turkish FAQ also needed converting to UTF8

15 years agoConvert Russian FAQ to UTF8, to make it render properly on the website.
Magnus Hagander [Mon, 15 Sep 2008 08:40:46 +0000 (08:40 +0000)]
Convert Russian FAQ to UTF8, to make it render properly on the website.

15 years agoUpdate citext's documentation to match the recently-applied patch,
Tom Lane [Fri, 12 Sep 2008 18:29:49 +0000 (18:29 +0000)]
Update citext's documentation to match the recently-applied patch,
per David Wheeler.

15 years agoSkip opfamily check in eclass_matches_any_index() when the index isn't a
Tom Lane [Fri, 12 Sep 2008 14:56:13 +0000 (14:56 +0000)]
Skip opfamily check in eclass_matches_any_index() when the index isn't a
btree.  We can't easily tell whether clauses generated from the equivalence
class could be used with such an index, so just assume that they might be.
This bit of over-optimization prevented use of non-btree indexes for nestloop
inner indexscans, in any case where the join uses an equality operator that
is also a btree operator --- which in particular is typically true for hash
indexes.  Noted while trying to test the current hash index patch.

15 years agoUpdate man page build instructions.
Peter Eisentraut [Fri, 12 Sep 2008 08:43:53 +0000 (08:43 +0000)]
Update man page build instructions.

15 years agoAdd a duration option to pgbench, so that test length can be specified in seconds
Tom Lane [Thu, 11 Sep 2008 23:52:48 +0000 (23:52 +0000)]
Add a duration option to pgbench, so that test length can be specified in seconds
instead of by number of transactions to run.  Takahiro Itagaki

15 years agoTighten up to_date/to_timestamp so that they are more likely to reject
Tom Lane [Thu, 11 Sep 2008 17:32:34 +0000 (17:32 +0000)]
Tighten up to_date/to_timestamp so that they are more likely to reject
erroneous input, rather than silently producing bizarre results as formerly
happened.

Brendan Jurd

15 years agoAdjust the parser to accept the typename syntax INTERVAL ... SECOND(n)
Tom Lane [Thu, 11 Sep 2008 15:27:30 +0000 (15:27 +0000)]
Adjust the parser to accept the typename syntax INTERVAL ... SECOND(n)
and the literal syntax INTERVAL 'string' ... SECOND(n), as required by the
SQL standard.  Our old syntax put (n) directly after INTERVAL, which was
a mistake, but will still be accepted for backward compatibility as well
as symmetry with the TIMESTAMP cases.

Change intervaltypmodout to show it in the spec's way, too.  (This could
potentially affect clients, if there are any that analyze the typmod of an
INTERVAL in any detail.)

Also fix interval input to handle 'min:sec.frac' properly; I had overlooked
this case in my previous patch.

Document the use of the interval fields qualifier, which up to now we had
never mentioned in the docs.  (I think the omission was intentional because
it didn't work per spec; but it does now, or at least close enough to be
credible.)

15 years agoInitialize the minimum frozen Xid in vac_update_datfrozenxid using
Alvaro Herrera [Thu, 11 Sep 2008 14:01:10 +0000 (14:01 +0000)]
Initialize the minimum frozen Xid in vac_update_datfrozenxid using
GetOldestXmin() instead of RecentGlobalXmin; this is safer because we do not
depend on the latter being correctly set elsewhere, and while it is more
expensive, this code path is not performance-critical.  This is a real
risk for autovacuum, because it can execute whole cycles without doing
a single vacuum, which would mean that RecentGlobalXmin would stay at its
initialization value, FirstNormalTransactionId, causing a bogus value to be
inserted in pg_database.  This bug could explain some recent reports of
failure to truncate pg_clog.

At the same time, change the initialization of RecentGlobalXmin to
InvalidTransactionId, and ensure that it's set to something else whenever
it's going to be used.  Using it as FirstNormalTransactionId in HOT page
pruning could incur in data loss.  InitPostgres takes care of setting it
to a valid value, but the extra checks are there to prevent "special"
backends from behaving in unusual ways.

Per Tom Lane's detailed problem dissection in 29544.1221061979@sss.pgh.pa.us

15 years agoTweak newly added set_config_sourcefile() so that the target record
Tom Lane [Wed, 10 Sep 2008 19:16:22 +0000 (19:16 +0000)]
Tweak newly added set_config_sourcefile() so that the target record
isn't left corrupt if guc_strdup should fail.

15 years agoMake our parsing of INTERVAL literals spec-compliant (or at least a heck of
Tom Lane [Wed, 10 Sep 2008 18:29:41 +0000 (18:29 +0000)]
Make our parsing of INTERVAL literals spec-compliant (or at least a heck of
a lot closer than it was before).  To do this, tweak coerce_type() to pass
through the typmod information when invoking interval_in() on an UNKNOWN
constant; then fix DecodeInterval to pay attention to the typmod when deciding
how to interpret a units-less integer value.  I changed one or two other
details as well.  I believe the code now reacts as expected by spec for all
the literal syntaxes that are specifically enumerated in the spec.  There
are corner cases involving strings that don't exactly match the set of fields
called out by the typmod, for which we might want to tweak the behavior some
more; but I think this is an area of user friendliness rather than spec
compliance.  There remain some non-compliant details about the SQL syntax
(as opposed to what's inside the literal string); but at least we'll throw
error rather than silently doing the wrong thing in those cases.

15 years agoAdd "source file" and "source line" information to each GUC variable.
Alvaro Herrera [Wed, 10 Sep 2008 18:09:20 +0000 (18:09 +0000)]
Add "source file" and "source line" information to each GUC variable.

initdb forced due to changes in the pg_settings view.

Magnus Hagander and Alvaro Herrera.

15 years agoAvoid using sprintf() for a simple octal conversion in PQescapeByteaInternal.
Tom Lane [Wed, 10 Sep 2008 17:01:07 +0000 (17:01 +0000)]
Avoid using sprintf() for a simple octal conversion in PQescapeByteaInternal.
Improves performance, per suggestion from Rudolf Leitgeb (bug #4414).
The backend did this right already, but not libpq.

15 years agoFix a couple of places where the plpgsql grammar would produce an unhelpful
Tom Lane [Wed, 10 Sep 2008 01:09:45 +0000 (01:09 +0000)]
Fix a couple of places where the plpgsql grammar would produce an unhelpful
'syntax error' message, rather than something that might draw one's
attention to a missing or wrong-type variable declaration.  Per recent
gripe.

15 years agoImprove the plan cache invalidation mechanism to make it invalidate plans
Tom Lane [Tue, 9 Sep 2008 18:58:09 +0000 (18:58 +0000)]
Improve the plan cache invalidation mechanism to make it invalidate plans
when user-defined functions used in a plan are modified.  Also invalidate
plans when schemas, operators, or operator classes are modified; but for these
cases we just invalidate everything rather than tracking exact dependencies,
since these types of objects seldom change in a production database.

Tom Lane; loosely based on a patch by Martin Pihlak.

15 years agoImprove plpgsql's ability to report tuple incompatibility problems.
Alvaro Herrera [Tue, 9 Sep 2008 15:14:08 +0000 (15:14 +0000)]
Improve plpgsql's ability to report tuple incompatibility problems.

Volkan YAZICI

15 years agoFix a couple of problems pointed out by Fujii Masao in the 2008-Apr-05 patch
Tom Lane [Mon, 8 Sep 2008 16:42:15 +0000 (16:42 +0000)]
Fix a couple of problems pointed out by Fujii Masao in the 2008-Apr-05 patch
for pg_stop_backup.  First, it is possible that the history file name is not
alphabetically later than the last WAL file name, so we should explicitly
check that both have been archived.  Second, the previous coding would wait
forever if a checkpoint had managed to remove the WAL file before we look for
it.

Simon Riggs, plus some code cleanup by me.

15 years agoMake pg_dump --data-only try to order the table dumps so that foreign keys'
Tom Lane [Mon, 8 Sep 2008 15:26:23 +0000 (15:26 +0000)]
Make pg_dump --data-only try to order the table dumps so that foreign keys'
referenced tables are dumped before the referencing tables.  This avoids
failures when the data is loaded with the FK constraints already active.
If no such ordering is possible because of circular or self-referential
constraints, print a NOTICE to warn the user about it.

15 years agoCreate a separate grantable privilege for TRUNCATE, rather than having it be
Tom Lane [Mon, 8 Sep 2008 00:47:41 +0000 (00:47 +0000)]
Create a separate grantable privilege for TRUNCATE, rather than having it be
always owner-only.  The TRUNCATE privilege works identically to the DELETE
privilege so far as interactions with the rest of the system go.

Robert Haas

15 years agoSupport set-returning functions in the target lists of Agg and Group plan
Tom Lane [Mon, 8 Sep 2008 00:22:56 +0000 (00:22 +0000)]
Support set-returning functions in the target lists of Agg and Group plan
nodes.  This is a pretty ugly feature but since we don't yet have a
plausible substitute, we'd better support it everywhere.
Per gripe from Jeff Davis.

15 years agoClarify description of pg_restore's handling of large objects.
Tom Lane [Sun, 7 Sep 2008 19:12:57 +0000 (19:12 +0000)]
Clarify description of pg_restore's handling of large objects.

15 years agoReimplement text_position and related functions to use Boyer-Moore-Horspool
Tom Lane [Sun, 7 Sep 2008 04:20:00 +0000 (04:20 +0000)]
Reimplement text_position and related functions to use Boyer-Moore-Horspool
searching instead of naive matching.  In the worst case this has the same
O(M*N) complexity as the naive method, but the worst case is hard to hit,
and the average case is very fast, especially with longer patterns.

David Rowley

15 years agoAdd a few more details in the source-code-formatting documentation.
Tom Lane [Sun, 7 Sep 2008 02:01:04 +0000 (02:01 +0000)]
Add a few more details in the source-code-formatting documentation.
This isn't exhaustive but it covers some of the more common layout
mistakes I've seen in submitted patches.

15 years agoClarify documention workding for xip_list().
Bruce Momjian [Sun, 7 Sep 2008 01:29:36 +0000 (01:29 +0000)]
Clarify documention workding for xip_list().

Simon Riggs

15 years agoAdjust psql's new \ef command to present an empty CREATE FUNCTION template
Tom Lane [Sat, 6 Sep 2008 20:18:08 +0000 (20:18 +0000)]
Adjust psql's new \ef command to present an empty CREATE FUNCTION template
for editing if no function name is specified.  This seems a much cleaner way
to offer that functionality than the original patch had.  In passing,
de-clutter the error displays that are given for a bogus function-name
argument, and standardize on "$function$" as the default delimiter for the
function body.  (The original coding would use the shortest possible
dollar-quote delimiter, which seems to create unnecessarily high risk of
later conflicts with the user-modified function body.)

15 years agoImplement a psql command "\ef" to edit the definition of a function.
Tom Lane [Sat, 6 Sep 2008 00:01:25 +0000 (00:01 +0000)]
Implement a psql command "\ef" to edit the definition of a function.
In support of that, create a backend function pg_get_functiondef().
The psql command is functional but maybe a bit rough around the edges...

Abhijit Menon-Sen

15 years agoAdd comment about why pg_dump doesn't dump the public schema comment.
Bruce Momjian [Fri, 5 Sep 2008 23:53:42 +0000 (23:53 +0000)]
Add comment about why pg_dump doesn't dump the public schema comment.

15 years agoFix an oversight in the 8.2 patch that improved mergejoin performance by
Tom Lane [Fri, 5 Sep 2008 21:07:29 +0000 (21:07 +0000)]
Fix an oversight in the 8.2 patch that improved mergejoin performance by
inserting a materialize node above an inner-side sort node, when the sort is
expected to spill to disk.  (The materialize protects the sort from having
to support mark/restore, allowing it to do its final merge pass on-the-fly.)
We neglected to teach cost_mergejoin about that hack, so it was failing to
include the materialize's costs in the estimated cost of the mergejoin.
The materialize's costs are generally going to be pretty negligible in
comparison to the sort's, so this is only a small error and probably not
worth back-patching; but it's still wrong.

In the similar case where a materialize is inserted to protect an inner-side
node that can't do mark/restore at all, it's still true that the materialize
should not spill to disk, and so we should cost it cheaply rather than
expensively.

Noted while thinking about a question from Tom Raney.

15 years agoCheck for gcov and lcov only when coverage testing is enabled.
Peter Eisentraut [Fri, 5 Sep 2008 18:54:58 +0000 (18:54 +0000)]
Check for gcov and lcov only when coverage testing is enabled.

15 years agoAdd some additional casts and regression tests for the citext data type.
Tom Lane [Fri, 5 Sep 2008 18:25:17 +0000 (18:25 +0000)]
Add some additional casts and regression tests for the citext data type.
David Wheeler

15 years agoAdd Win32 MSVC code to support this recent patch:
Bruce Momjian [Fri, 5 Sep 2008 16:54:39 +0000 (16:54 +0000)]
Add Win32 MSVC code to support this recent patch:

Add missing descriptions for aggregates, functions and conversions.

Bernd Helmle

15 years agoCode coverage testing with gcov. Documentation is in the regression test
Peter Eisentraut [Fri, 5 Sep 2008 12:11:18 +0000 (12:11 +0000)]
Code coverage testing with gcov.  Documentation is in the regression test
chapter.

Author: Michelle Caisse <Michelle.Caisse@Sun.COM>

15 years agoAdd DSSSL stylesheet location for Mac OS X/Fink installation.
Peter Eisentraut [Fri, 5 Sep 2008 09:37:37 +0000 (09:37 +0000)]
Add DSSSL stylesheet location for Mac OS X/Fink installation.

15 years agoFix strategy propagation to scanEntry for partial match by moving propagation
Teodor Sigaev [Thu, 4 Sep 2008 11:47:05 +0000 (11:47 +0000)]
Fix strategy propagation to scanEntry for partial match by moving propagation
to initializaion of scanEntry.

15 years agoIf a loadable module has wrong values in its magic block, spell out
Tom Lane [Wed, 3 Sep 2008 22:34:50 +0000 (22:34 +0000)]
If a loadable module has wrong values in its magic block, spell out
exactly what they are in the complaint message.  Marko Kreen,
some editorialization by me.

15 years agoPrevent memory leaks in our various bison parsers when an error occurs
Tom Lane [Tue, 2 Sep 2008 20:37:55 +0000 (20:37 +0000)]
Prevent memory leaks in our various bison parsers when an error occurs
during parsing.  Formerly the parser's stack was allocated with malloc
and so wouldn't be reclaimed; this patch makes it use palloc instead,
so that flushing the current context will reclaim the memory.  Per
Marko Kreen.

15 years agoFix plpgsql's exec_move_row() to supply valid type OIDs to exec_assign_value()
Tom Lane [Mon, 1 Sep 2008 22:30:33 +0000 (22:30 +0000)]
Fix plpgsql's exec_move_row() to supply valid type OIDs to exec_assign_value()
whenever possible, as per bug report from Oleg Serov.  While at it, reorder
the operations in the RECORD case to avoid possible palloc failure while the
variable update is only partly complete.

Back-patch as far as 8.1.  Although the code of the particular function is
similar in 8.0, 8.0's support for composite fields in rows is sufficiently
broken elsewhere that it doesn't seem worth fixing this.

15 years agoSigh, I missed checking the ecpg tests ...
Tom Lane [Mon, 1 Sep 2008 21:42:18 +0000 (21:42 +0000)]
Sigh, I missed checking the ecpg tests ...

15 years agoAdd a variant expected-output file for the sequence regression test, to cover
Tom Lane [Mon, 1 Sep 2008 21:24:52 +0000 (21:24 +0000)]
Add a variant expected-output file for the sequence regression test, to cover
output that is seen when a checkpoint occurs at just the right time during
the test.  Per my report of 2008-08-31.

This could be back-patched but I'm not sure it's worth the trouble.

15 years agoAdd a bunch of new error location reports to parse-analysis error messages.
Tom Lane [Mon, 1 Sep 2008 20:42:46 +0000 (20:42 +0000)]
Add a bunch of new error location reports to parse-analysis error messages.
There are still some weak spots around JOIN USING and relation alias lists,
but most errors reported within backend/parser/ now have locations.

15 years agoHeapTupleHeaderAdjustCmax made the incorrect assumption that the raw
Heikki Linnakangas [Mon, 1 Sep 2008 18:52:45 +0000 (18:52 +0000)]
HeapTupleHeaderAdjustCmax made the incorrect assumption that the raw
command id is the cmin, when it can in fact be a combo cid. That made rows
incorrectly invisible to a transaction where a tuple was deleted by multiple
aborted subtransactions.

Report and patch Karl Schnaitter. Back-patch to 8.3, where combo cids was
introduced.

15 years agoSynchronize the shared object build rules in Makefile.port with Makefile.shlib
Peter Eisentraut [Mon, 1 Sep 2008 08:50:10 +0000 (08:50 +0000)]
Synchronize the shared object build rules in Makefile.port with Makefile.shlib
somewhat by adding CFLAGS where the compiler is used and Makefile.shlib
already used CFLAGS.

15 years agoFix (hopefully) some oversights in recent Bison cleanup patch.
Tom Lane [Sat, 30 Aug 2008 02:32:24 +0000 (02:32 +0000)]
Fix (hopefully) some oversights in recent Bison cleanup patch.
Per buildfarm results.

15 years agoFix the raw-parsetree representation of star (as in SELECT * FROM or
Tom Lane [Sat, 30 Aug 2008 01:39:14 +0000 (01:39 +0000)]
Fix the raw-parsetree representation of star (as in SELECT * FROM or
SELECT foo.*) so that it cannot be confused with a quoted identifier "*".
Instead create a separate node type A_Star to represent this notation.
Per pgsql-hackers discussion of 2007-Sep-27.

15 years agoIn GCC-based builds, use a better newNode() macro that relies on GCC-specific
Tom Lane [Fri, 29 Aug 2008 22:49:07 +0000 (22:49 +0000)]
In GCC-based builds, use a better newNode() macro that relies on GCC-specific
syntax to avoid a useless store into a global variable.  Per experimentation,
this works better than my original thought of trying to push the code into
an out-of-line subroutine.