OSDN Git Service
Bruce Momjian [Wed, 27 Oct 2004 02:01:14 +0000 (02:01 +0000)]
Update German FAQ.
Ian Barwick
Bruce Momjian [Tue, 26 Oct 2004 23:42:49 +0000 (23:42 +0000)]
Add reminder to update copyright in sgml file.
Tom Lane [Tue, 26 Oct 2004 22:16:12 +0000 (22:16 +0000)]
Document IS DISTINCT FROM in a more obvious place, and add some more
index entries for IS-foo constructs.
Bruce Momjian [Tue, 26 Oct 2004 19:41:22 +0000 (19:41 +0000)]
Add:
> * Allow TIMESTAMP WITH TIME ZONE to store the original timezone
> information, either by name or offset from UTC
Bruce Momjian [Tue, 26 Oct 2004 17:28:27 +0000 (17:28 +0000)]
Update Russian FAQ.
Viktor Vislobokov
Tom Lane [Tue, 26 Oct 2004 16:05:03 +0000 (16:05 +0000)]
Make heap_fetch API more consistent by having the buffer remain pinned
in all cases when keep_buf = true. This allows ANALYZE's inner loop to
use heap_release_fetch, which saves multiple buffer lookups for the same
page and avoids overestimation of cost by the vacuum cost mechanism.
Tom Lane [Mon, 25 Oct 2004 15:42:02 +0000 (15:42 +0000)]
In the new dispensation where REINDEX doesn't take exclusive lock on
the parent table, it's essential that all index accesses take some kind
of lock on the index. I had missed vacuumlazy.c :-( ...
Neil Conway [Mon, 25 Oct 2004 06:40:39 +0000 (06:40 +0000)]
Update copyright information in documentation. Also, add a note
to RELEASE_CHANGES so that this is not forgotten in the future.
Neil Conway [Mon, 25 Oct 2004 06:27:21 +0000 (06:27 +0000)]
Remove three unnecessary casts from a pointer type to char * when calling
pfree().
Neil Conway [Mon, 25 Oct 2004 03:23:02 +0000 (03:23 +0000)]
Fix tyop in comment.
Tom Lane [Mon, 25 Oct 2004 03:08:29 +0000 (03:08 +0000)]
Make error message more verbose, in hopes of avoiding misunderstandings
such as bug #1293.
Tom Lane [Mon, 25 Oct 2004 02:30:29 +0000 (02:30 +0000)]
Fix some more 'old-style parameter declaration' warnings.
Tom Lane [Mon, 25 Oct 2004 02:15:02 +0000 (02:15 +0000)]
Fix a bunch of 'old-style parameter declaration' warnings induced by
writing 'foo()' rather than 'foo(void)'.
Neil Conway [Mon, 25 Oct 2004 00:46:43 +0000 (00:46 +0000)]
Modify hash_create() to elog(ERROR) if an error occurs, rather than
returning a NULL pointer (some callers remembered to check the return
value, but some did not -- it is safer to just bail out).
Also, cleanup pgstat.c to use elog(ERROR) rather than elog(LOG) followed
by exit().
PostgreSQL Daemon [Mon, 25 Oct 2004 00:11:04 +0000 (00:11 +0000)]
make sure we tag configure.in as beta4 as well ...
Tom Lane [Sun, 24 Oct 2004 22:52:04 +0000 (22:52 +0000)]
Update for 8.0: pg_tablespace is now also a shared catalog.
Tom Lane [Sun, 24 Oct 2004 22:43:56 +0000 (22:43 +0000)]
Add note that REINDEX takes a weaker lock than it used to.
Tom Lane [Sun, 24 Oct 2004 22:11:37 +0000 (22:11 +0000)]
Add comparison file for exp-three-digits formatting.
Tom Lane [Sun, 24 Oct 2004 22:09:33 +0000 (22:09 +0000)]
Adjust message to not be misleading on Windows.
Andrew Dunstan
Tom Lane [Sun, 24 Oct 2004 22:08:19 +0000 (22:08 +0000)]
On Windows, cause get_progname to strip any .EXE suffix.
Andrew Dunstan
Tom Lane [Sun, 24 Oct 2004 21:54:57 +0000 (21:54 +0000)]
Add comparison file for exp-three-digits formatting.
Tom Lane [Sun, 24 Oct 2004 21:47:07 +0000 (21:47 +0000)]
Replace ad-hoc atof() code with call to float4in, per Andrew Dunstan.
Tom Lane [Sun, 24 Oct 2004 15:55:29 +0000 (15:55 +0000)]
Write config files as text not binary, per Magnus Hagander.
Tom Lane [Sun, 24 Oct 2004 15:09:57 +0000 (15:09 +0000)]
Sync timezone data with latest zic database (dated Oct 11 2004).
Tom Lane [Sun, 24 Oct 2004 15:01:54 +0000 (15:01 +0000)]
Update RELEASE_CHANGES to mention updating the timezone database as
a routine part of release prep.
Dennis Bjorklund [Sun, 24 Oct 2004 14:55:04 +0000 (14:55 +0000)]
Translation update
Tom Lane [Sun, 24 Oct 2004 00:54:12 +0000 (00:54 +0000)]
If we're going to test for switch validity by observing whether the
compiler emits any warnings, the test program had better be 100%
correct, not only 90% correct. The recent addition of -Wold-style-definition
broke thread-safety detection on every platform that has that switch,
because the test program used an old-style definition.
Dennis Bjorklund [Sat, 23 Oct 2004 14:48:03 +0000 (14:48 +0000)]
Translation update
Dennis Bjorklund [Sat, 23 Oct 2004 07:29:27 +0000 (07:29 +0000)]
Translation updates
Dennis Bjorklund [Sat, 23 Oct 2004 06:48:59 +0000 (06:48 +0000)]
Translation update
Dennis Bjorklund [Sat, 23 Oct 2004 06:40:24 +0000 (06:40 +0000)]
Translation updates
Dennis Bjorklund [Sat, 23 Oct 2004 06:32:53 +0000 (06:32 +0000)]
Translation updates
Tom Lane [Sat, 23 Oct 2004 00:05:27 +0000 (00:05 +0000)]
Avoid overflow in cost_sort when work_mem exceeds 1Gb.
Tom Lane [Fri, 22 Oct 2004 22:33:58 +0000 (22:33 +0000)]
Fix tsearch build problems.
Magnus Hagander
Tom Lane [Fri, 22 Oct 2004 22:30:57 +0000 (22:30 +0000)]
Don't use LC_MESSAGES value on WIN32, since it doesn't work.
Magnus Hagander
Tom Lane [Fri, 22 Oct 2004 19:48:19 +0000 (19:48 +0000)]
Add a GUC_SUPERUSER_ONLY flag to mark GUC variables that should not be
examinable by non-superusers, and use it to protect the recently-added
GUC variables for data directory and config files. For now I have only
flagged those variables that could be used to deduce something about
the server's filesystem layout, but possibly we should also mark vars
related to logging settings and other admin-only information?
Tom Lane [Fri, 22 Oct 2004 17:20:05 +0000 (17:20 +0000)]
In ALTER COLUMN TYPE, strip any implicit coercion operations appearing
at the top level of the column's old default expression before adding
an implicit coercion to the new column type. This seems to satisfy the
principle of least surprise, as per discussion of bug #1290.
Peter Eisentraut [Fri, 22 Oct 2004 16:05:02 +0000 (16:05 +0000)]
Translation updates
Peter Eisentraut [Fri, 22 Oct 2004 16:04:35 +0000 (16:04 +0000)]
Message styling
Dennis Bjorklund [Fri, 22 Oct 2004 14:14:24 +0000 (14:14 +0000)]
Translation update
Dennis Bjorklund [Fri, 22 Oct 2004 14:13:04 +0000 (14:13 +0000)]
New swedish translation of pg_config
Neil Conway [Fri, 22 Oct 2004 14:12:41 +0000 (14:12 +0000)]
Fix documentation typo.
Dennis Bjorklund [Fri, 22 Oct 2004 13:49:15 +0000 (13:49 +0000)]
Translation updates
Dennis Bjorklund [Fri, 22 Oct 2004 13:41:47 +0000 (13:41 +0000)]
Translation updates
Dennis Bjorklund [Fri, 22 Oct 2004 12:04:13 +0000 (12:04 +0000)]
Translation updates.
Neil Conway [Fri, 22 Oct 2004 07:21:06 +0000 (07:21 +0000)]
Minor code cleanup: hdefault() only ever returned "true", so it may as
well be declared to return "void" to save callers the trouble of
checking for errors.
Tom Lane [Fri, 22 Oct 2004 00:25:18 +0000 (00:25 +0000)]
Update release history for releases 7.4.6, 7.3.8, 7.2.6.
Tom Lane [Fri, 22 Oct 2004 00:24:18 +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.
Tom Lane [Thu, 21 Oct 2004 22:48:54 +0000 (22:48 +0000)]
Update obsolete comments about COPY vs INSERT options, per Uwe Schroeder.
Tom Lane [Thu, 21 Oct 2004 21:33:59 +0000 (21:33 +0000)]
Disallow referential integrity actions from being deferred; only the
NO ACTION check is deferrable. This seems to be a closer approximation
to what the SQL spec says than what we were doing before, and it prevents
some anomalous behaviors that are possible now that triggers can fire
during the execution of PL functions.
Stephan Szabo.
Tom Lane [Thu, 21 Oct 2004 20:23:19 +0000 (20:23 +0000)]
Set the close-on-exec flag for libpq's socket to the backend, to avoid
any possible problems from child programs executed by the client app.
Per suggestion from Elliot Lee of Red Hat.
Tom Lane [Thu, 21 Oct 2004 19:49:27 +0000 (19:49 +0000)]
Avoid macro-redefinition warnings on Windows, per Andrew Dunstan.
Tom Lane [Thu, 21 Oct 2004 19:28:36 +0000 (19:28 +0000)]
Standardize on using the Min, Max, and Abs macros that are in our c.h file,
getting rid of numerous ad-hoc versions that have popped up in various
places. Shortens code and avoids conflict with Windows min() and max()
macros.
Tom Lane [Wed, 20 Oct 2004 16:42:46 +0000 (16:42 +0000)]
Use temp files in current directory, not /tmp, to reduce security risk
while running this script.
Tom Lane [Wed, 20 Oct 2004 16:04:50 +0000 (16:04 +0000)]
Allow functions returning void or cstring to appear in FROM clause,
to make life cushy for the JDBC driver. Centralize the decision-making
that affects this by inventing a get_type_func_class() function, rather
than adding special cases in half a dozen places.
Neil Conway [Wed, 20 Oct 2004 02:12:07 +0000 (02:12 +0000)]
When using GCC, change the default CFLAGS to:
-O2 -Wall -Wmissing-prototypes -Wpointer-arith
Check whether the version of GCC we are using supports any of:
-Wdeclaration-after-statement
-Wendif-labels
-Wold-style-definition
And add the supported flags to CFLAGS.
Tom Lane [Tue, 19 Oct 2004 15:04:17 +0000 (15:04 +0000)]
Add a HINT about the likely reason for 'invalid multibyte character for locale' failure.
Peter Eisentraut [Tue, 19 Oct 2004 13:39:27 +0000 (13:39 +0000)]
New translation
Peter Eisentraut [Tue, 19 Oct 2004 13:38:53 +0000 (13:38 +0000)]
Organize the help output a little better. Improve clarity of some messages.
Tom Lane [Mon, 18 Oct 2004 23:23:19 +0000 (23:23 +0000)]
Add some code to ensure that we don't lose communication sync due to
an oversize message, per suggestion from Oliver Jowett. I'm a bit
dubious that this is a real problem, since the client likely doesn't
have any more space available than the server, but it's not hard to
make it behave according to the protocol intention.
Peter Eisentraut [Mon, 18 Oct 2004 22:19:15 +0000 (22:19 +0000)]
New translation
Peter Eisentraut [Mon, 18 Oct 2004 22:19:00 +0000 (22:19 +0000)]
Might want to initialize the gettext library, if we're going to do all that
translation work.
Peter Eisentraut [Mon, 18 Oct 2004 22:18:03 +0000 (22:18 +0000)]
Translation update
Tom Lane [Mon, 18 Oct 2004 22:00:42 +0000 (22:00 +0000)]
Add PQprepare/PQsendPrepared functions to libpq to support preparing
statements without necessarily specifying the datatypes of their parameters.
Abhijit Menon-Sen with some help from Tom Lane.
Bruce Momjian [Mon, 18 Oct 2004 19:08:58 +0000 (19:08 +0000)]
Remove use of "miscadmin.h" in port C files and remove test for frontend
in miscadmin.h.
Peter Eisentraut [Mon, 18 Oct 2004 17:58:54 +0000 (17:58 +0000)]
New translation
Peter Eisentraut [Mon, 18 Oct 2004 17:56:46 +0000 (17:56 +0000)]
Translation updates
Peter Eisentraut [Mon, 18 Oct 2004 17:53:40 +0000 (17:53 +0000)]
New translation
Peter Eisentraut [Mon, 18 Oct 2004 17:09:03 +0000 (17:09 +0000)]
Updated SQL key words table
Bruce Momjian [Mon, 18 Oct 2004 16:24:40 +0000 (16:24 +0000)]
The attached patch fixes psql's win32 frontend-only build, by using
pg_strcasecmp in variables.c, and #ifdef'ing out PostmasterPid in
miscadmin.h.
Dave Page
Bruce Momjian [Mon, 18 Oct 2004 16:16:20 +0000 (16:16 +0000)]
Force rebuild.
Bruce Momjian [Mon, 18 Oct 2004 16:13:43 +0000 (16:13 +0000)]
Improve indentation of sublists:
< The proper solution to this will probably the use of a master/slave
< replication solution like Sloney and a connection pooling tool like
< pgpool.
> The proper solution to this will probably the use of a master/slave
> replication solution like Sloney and a connection pooling tool like
> pgpool.
114,
116c114,116
< You can use any of the master/slave replication servers to use a
< standby server for data warehousing. To allow read/write queries to
< multiple servers, you need multi-master replication like pgcluster.
> You can use any of the master/slave replication servers to use a
> standby server for data warehousing. To allow read/write queries to
> multiple servers, you need multi-master replication like pgcluster.
166,
167c166,167
< Currently large objects entries do not have owners. Permissions can
< only be set at the pg_largeobject table level.
> Currently large objects entries do not have owners. Permissions can
> only be set at the pg_largeobject table level.
173c173
< This requires the TOAST column to be stored EXTERNAL.
> This requires the TOAST column to be stored EXTERNAL.
359,
360c359,360
< One complexity is whether moving a schema should move all existing
< schema objects or just define the location for future object creation.
> One complexity is whether moving a schema should move all existing
> schema objects or just define the location for future object creation.
364,
365c364,365
< Currently non-global system tables must be in the default database
< schema. Global system tables can never be moved.
> Currently non-global system tables must be in the default database
> schema. Global system tables can never be moved.
371,
375c371,375
< This might require some background daemon to maintain clustering
< during periods of low usage. It might also require tables to be only
< paritally filled for easier reorganization. Another idea would
< be to create a merged heap/index data file so an index lookup would
< automatically access the heap data too.
> This might require some background daemon to maintain clustering
> during periods of low usage. It might also require tables to be only
> paritally filled for easier reorganization. Another idea would
> be to create a merged heap/index data file so an index lookup would
> automatically access the heap data too.
379,
380c379,380
< To do this, determine the ideal cluster index for each system
< table and set the cluster setting during initdb.
> To do this, determine the ideal cluster index for each system
> table and set the cluster setting during initdb.
385,
386c385,386
< This requires the use of a savepoint before each COPY line is
< processed, with ROLLBACK on COPY failure.
> This requires the use of a savepoint before each COPY line is
> processed, with ROLLBACK on COPY failure.
395,
398c395,398
< This requires using the row ctid to map cursor rows back to the
< original heap row. This become more complicated if WITH HOLD cursors
< are to be supported because WITH HOLD cursors have a copy of the row
< and no FOR UPDATE lock.
> This requires using the row ctid to map cursor rows back to the
> original heap row. This become more complicated if WITH HOLD cursors
> are to be supported because WITH HOLD cursors have a copy of the row
> and no FOR UPDATE lock.
405,
406c405,406
< Because WITH HOLD cursors exist outside transactions, this allows
< them to be listed so they can be closed.
> Because WITH HOLD cursors exist outside transactions, this allows
> them to be listed so they can be closed.
413,
415c413,415
< This is useful for returning the auto-generated key for an INSERT.
< One complication is how to handle rules that run as part of
< the insert.
> This is useful for returning the auto-generated key for an INSERT.
> One complication is how to handle rules that run as part of
> the insert.
422c422
< This is basically the same as SET search_path.
> This is basically the same as SET search_path.
426,
427c426,427
< This requires a checking function to be called after the server
< configuration file is read.
> This requires a checking function to be called after the server
> configuration file is read.
432c432
< Currently only constants are supported.
> Currently only constants are supported.
438,
439c438,439
< This requires the cached PL/PgSQL byte code to be invalidated when
< an object referenced in the function is changed.
> This requires the cached PL/PgSQL byte code to be invalidated when
> an object referenced in the function is changed.
512,
513c512,513
< Document differences between ecpg and the SQL standard and
< information about the Informix-compatibility module.
> Document differences between ecpg and the SQL standard and
> information about the Informix-compatibility module.
Michael Meskes [Mon, 18 Oct 2004 13:36:23 +0000 (13:36 +0000)]
Synced parser
Bruce Momjian [Mon, 18 Oct 2004 11:45:26 +0000 (11:45 +0000)]
Update Japanese FAQ.
Jun Kuwamura
Bruce Momjian [Mon, 18 Oct 2004 03:27:14 +0000 (03:27 +0000)]
Update wording:
* Allow a database in tablespace t1 with tables created in tablespace t2
to be used as a template for a new database created with default
tablespace t2
All objects in the default database tablespace must have default tablespace
specifications. This is because new databases are created by copying
directories. If you mix default tablespace tables and tablespace-specified
tables in the same directory, creating a new database from such a mixed
directory would create a new database with tables that had incorrect
explicit tablespaces. To fix this would require modifying pg_class in the
newly copied database, which we don't currently do.
Bruce Momjian [Mon, 18 Oct 2004 02:56:42 +0000 (02:56 +0000)]
Add:
>
> * Allow a database in tablespace t1 with tables created in tablespace t2
> to be used as a template for a new database created with default
> tablespace t2
>
> All objects in the default database tablespace must have default tablespace
> specifications. This is because new databases are created by copying
> directories. If you mix default tablespace tables and tablespace-specified
> tables in the same directory, creating a new database from such a mixed
> directory would create a new database with tables that had incorrect
> explicit tablespaces. To fix this would require modifying pg_class in the
> newly copied database, which we don't currently do.
Neil Conway [Mon, 18 Oct 2004 01:45:38 +0000 (01:45 +0000)]
Trivial fix: remove a pointless cast.
Tom Lane [Mon, 18 Oct 2004 00:20:41 +0000 (00:20 +0000)]
Tweak pg_dump to handle default tablespaces correctly --- same logic
as the corrected pg_get_indexdef code.
Bruce Momjian [Sun, 17 Oct 2004 23:53:30 +0000 (23:53 +0000)]
Better document win32_open and its ability to allow concurrent
unlink/rename().
Bruce Momjian [Sun, 17 Oct 2004 23:39:22 +0000 (23:39 +0000)]
Fix some typos.
Tom Lane [Sun, 17 Oct 2004 23:09:31 +0000 (23:09 +0000)]
Make the standard stopword files be sought relative to share_dir, so
that a tsearch2 installation can be relocatable.
Tom Lane [Sun, 17 Oct 2004 22:01:51 +0000 (22:01 +0000)]
Allow background writing to be shut down by setting limit values to zero.
This does not disable the bgwriter process: it still has to wake up often
enough to collect fsync requests from backends in a timely fashion. But
it responds to the recent gripe about not being able to prevent the disk
from being spun up constantly.
Tom Lane [Sun, 17 Oct 2004 21:17:27 +0000 (21:17 +0000)]
pg_get_indexdef() didn't do quite the right thing with identifying
an index's tablespace.
Peter Eisentraut [Sun, 17 Oct 2004 21:12:55 +0000 (21:12 +0000)]
New translation
Peter Eisentraut [Sun, 17 Oct 2004 21:10:25 +0000 (21:10 +0000)]
New translation
Peter Eisentraut [Sun, 17 Oct 2004 21:08:22 +0000 (21:08 +0000)]
Translation update
Peter Eisentraut [Sun, 17 Oct 2004 21:05:03 +0000 (21:05 +0000)]
Translation updates
Peter Eisentraut [Sun, 17 Oct 2004 21:04:41 +0000 (21:04 +0000)]
Adjust message
Tom Lane [Sun, 17 Oct 2004 20:47:21 +0000 (20:47 +0000)]
Give a more user-friendly error message in situation where CREATE DATABASE
specifies a new default tablespace and the template database already has
some tables in that tablespace. There isn't any way to solve this fully
without modifying the clone database's pg_class contents, so for now the
best we can do is issue a better error message.
Tom Lane [Sun, 17 Oct 2004 20:02:26 +0000 (20:02 +0000)]
Make locale_messages_assign() really work on Windows; the prior hack
only covered the case of assigning "", and failed to recognize that
actually setlocale(LC_MESSAGES,...) does not work at all on this platform.
Magnus Hagander, some code prettification by Tom Lane.
Tom Lane [Sat, 16 Oct 2004 22:52:55 +0000 (22:52 +0000)]
Remove dllist.c from libpq. It's overkill for what libpq needs; we can
just stick a list-link into struct PGnotify instead. Result is a smaller
faster and more robust library (mainly because we reduce the number of
malloc's and free's involved in notify processing), plus less pollution
of application link-symbol namespace.
Tom Lane [Sat, 16 Oct 2004 21:50:02 +0000 (21:50 +0000)]
Include dllist.c directly instead of assuming that libpq will provide it.
Whack some semblance of project-conventions-conformance into pg_autovacuum.h.
Tom Lane [Sat, 16 Oct 2004 21:16:36 +0000 (21:16 +0000)]
Give a more user-friendly error message in case where a table is created
in a schema whose default tablespace has been dropped.
Tom Lane [Sat, 16 Oct 2004 20:10:57 +0000 (20:10 +0000)]
Generate the various .def files for libpq DLLs from a single list of
exportable functions. The .def files are removed from CVS, but will
still be present in distribution tarballs, since we can't assume that
Windows boxes will have sed.
Tom Lane [Sat, 16 Oct 2004 19:08:38 +0000 (19:08 +0000)]
Limit NBuffers and some related values to INT_MAX / BLCKSZ, to prevent
arithmetic overflow during initial sizing calculations. This is not
water-tight but it should avoid the grossest sorts of failures.
Tom Lane [Sat, 16 Oct 2004 18:57:26 +0000 (18:57 +0000)]
Give the ResourceOwner mechanism full responsibility for releasing buffer
pins at end of transaction, and reduce AtEOXact_Buffers to an Assert
cross-check that this was done correctly. When not USE_ASSERT_CHECKING,
AtEOXact_Buffers is a complete no-op. This gets rid of an O(NBuffers)
bottleneck during transaction commit/abort, which recent testing has shown
becomes significant above a few tens of thousands of shared buffers.
Tom Lane [Sat, 16 Oct 2004 18:05:07 +0000 (18:05 +0000)]
Remove BufferLocks[] array in favor of a single pointer to the buffer
(if any) currently waited for by LockBufferForCleanup(), which is all
that we were using it for anymore. Saves some space and eliminates
proportional-to-NBuffers slowdown in UnlockBuffers().
Bruce Momjian [Sat, 16 Oct 2004 03:32:08 +0000 (03:32 +0000)]
Fix Cygwin using of pg_ctl by not usig START.
Reini Urban
Bruce Momjian [Sat, 16 Oct 2004 03:26:43 +0000 (03:26 +0000)]
Export only required symbols in libpq on Win32.
Magnus Hagander
Bruce Momjian [Sat, 16 Oct 2004 03:10:17 +0000 (03:10 +0000)]
Allow pg_ctl to determine the server is up when getting a request for a
password.
Make password error message a #define and use it consistently.
Sean Chittenden