]> granicus.if.org Git - postgresql/log
postgresql
21 years agoDocument threading status.
Bruce Momjian [Thu, 14 Aug 2003 05:27:18 +0000 (05:27 +0000)]
Document threading status.

Update to POSIX getpwuid_r() function.

21 years agoAdjust for compiler options for compiler bug.
Bruce Momjian [Wed, 13 Aug 2003 23:59:49 +0000 (23:59 +0000)]
Adjust for compiler options for compiler bug.

Larry Rosenman

21 years agolibpq failed to cope with COPY FROM STDIN if the command was issued
Tom Lane [Wed, 13 Aug 2003 18:56:21 +0000 (18:56 +0000)]
libpq failed to cope with COPY FROM STDIN if the command was issued
via extended query protocol, because it sends Sync right after Execute
without realizing that the command to be executed is COPY.  There seems
to be no reasonable way for it to realize that, either, so the best fix
seems to be to make the backend ignore Sync during copy-in mode.  Bit of
a wart on the protocol, but little alternative.  Also, libpq must send
another Sync after terminating the COPY, if the command was issued via
Execute.

21 years agoAdd PQexecPrepared() and PQsendQueryPrepared() functions, to allow
Tom Lane [Wed, 13 Aug 2003 16:29:03 +0000 (16:29 +0000)]
Add PQexecPrepared() and PQsendQueryPrepared() functions, to allow
libpq users to perform Bind/Execute of previously prepared statements.
Per yesterday's discussion, this offers enough performance improvement
to justify bending the 'no new features during beta' rule.

21 years agoMove MemoryContextCheck() call from bottom of PostgresMain loop to
Tom Lane [Wed, 13 Aug 2003 16:16:23 +0000 (16:16 +0000)]
Move MemoryContextCheck() call from bottom of PostgresMain loop to
just before CommitTransactionCommand().  This is a more sensible place
to put it since commit discards a lot of contexts, and we'd not find
out about stomps affecting only transaction-local contexts.

21 years agoCorrect error messages
Teodor Sigaev [Wed, 13 Aug 2003 14:41:59 +0000 (14:41 +0000)]
Correct error messages

21 years agoMore accuracy work with MinWords option of headline function
Teodor Sigaev [Wed, 13 Aug 2003 14:37:55 +0000 (14:37 +0000)]
More accuracy work with MinWords option of headline function

21 years agoI'm quite fond of doing VPATH builds, i.e. building outside the source
Bruce Momjian [Wed, 13 Aug 2003 03:12:04 +0000 (03:12 +0000)]
I'm quite fond of doing VPATH builds, i.e. building outside the source
tree. This also catches lots of little Makefile bugs, so here's a small
patch for one of them (replacing an explicit reference to thread.c with
a reference to it as the first prerequsite of the rule makes make look
for it in the place where it was found (the source tree) rather than in
the build tree. (using GNU make 3.79.1)

John Gray

21 years agoInclude 'IPv4', 'IPv6', or 'Unix' in socket-creation failure messages,
Tom Lane [Tue, 12 Aug 2003 22:42:01 +0000 (22:42 +0000)]
Include 'IPv4', 'IPv6', or 'Unix' in socket-creation failure messages,
in hopes of soothing fears of those with partial IPv6 support.  Still an
open question whether we should report EAFNOSUPPORT errors at all,
though.

21 years agoCope with NoData message from backend. Needed for case where
Tom Lane [Tue, 12 Aug 2003 21:34:44 +0000 (21:34 +0000)]
Cope with NoData message from backend.  Needed for case where
PQexecParams is used with a statement that doesn't return data.

21 years agoMarginal hacks to move some processing out of the per-client-message
Tom Lane [Tue, 12 Aug 2003 18:52:38 +0000 (18:52 +0000)]
Marginal hacks to move some processing out of the per-client-message
processing loop; avoids extra overhead when using parse/bind/execute
messages instead of single Query message.

21 years agoChange some frequently-reached elog(DEBUG...) calls to ereport(DEBUG...)
Tom Lane [Tue, 12 Aug 2003 18:23:21 +0000 (18:23 +0000)]
Change some frequently-reached elog(DEBUG...) calls to ereport(DEBUG...)
for speed reasons.  (ereport falls out much more quickly when no output
is needed than elog does.)

21 years agoAvoid unnecessary work when stats collection is disabled. Tighten
Tom Lane [Tue, 12 Aug 2003 16:21:18 +0000 (16:21 +0000)]
Avoid unnecessary work when stats collection is disabled.  Tighten
search loop in pgstat_initstats.  Per report from Gavin Sherry.

21 years agoApplied patch from Oliver Jowett to clean up the jdbc regression test build
Barry Lind [Mon, 11 Aug 2003 23:42:04 +0000 (23:42 +0000)]
Applied patch from Oliver Jowett to clean up the jdbc regression test build

 Modified Files:
  jdbc/build.xml
  jdbc/org/postgresql/test/jdbc3/Jdbc3TestSuite.java

21 years agoRewriter and planner should use only resno, not resname, to identify
Tom Lane [Mon, 11 Aug 2003 23:04:50 +0000 (23:04 +0000)]
Rewriter and planner should use only resno, not resname, to identify
target columns in INSERT and UPDATE targetlists.  Don't rely on resname
to be accurate in ruleutils, either.  This fixes bug reported by
Donald Fraser, in which renaming a column referenced in a rule did not
work very well.

21 years agoAdd:
Bruce Momjian [Mon, 11 Aug 2003 22:39:48 +0000 (22:39 +0000)]
Add:

> * Prevent unneeded quoting in psql \d output using fmtId()

21 years agoApplied patch from Oliver Jowett to better handle invalid input for getArray
Barry Lind [Mon, 11 Aug 2003 21:33:50 +0000 (21:33 +0000)]
Applied patch from Oliver Jowett to better handle invalid input for getArray
(no longer throw an index out of range exception)

 Modified Files:
  jdbc/org/postgresql/jdbc2/AbstractJdbc2ResultSet.java

21 years agoApplied patch from Oliver Jowett to clean up some aditional warning messages
Barry Lind [Mon, 11 Aug 2003 21:27:52 +0000 (21:27 +0000)]
Applied patch from Oliver Jowett to clean up some aditional warning messages
from ant.

 Modified Files:
  jdbc/build.xml

21 years agoApplied patch from Oliver Jewett to fix a deprecation in newer versions of ant
Barry Lind [Mon, 11 Aug 2003 21:25:01 +0000 (21:25 +0000)]
Applied patch from Oliver Jewett to fix a deprecation in newer versions of ant

 Modified Files:
  jdbc/build.xml

21 years agoApplied patch from Oliver Jewett to clean up the testing README file
Barry Lind [Mon, 11 Aug 2003 21:20:50 +0000 (21:20 +0000)]
Applied patch from Oliver Jewett to clean up the testing README file

 Modified Files:
  jdbc/org/postgresql/test/README

21 years agoApplied patch by Oliver Jowett to clean up some exception handling
Barry Lind [Mon, 11 Aug 2003 21:18:47 +0000 (21:18 +0000)]
Applied patch by Oliver Jowett to clean up some exception handling

 Modified Files:
  jdbc/org/postgresql/core/QueryExecutor.java
  jdbc/org/postgresql/util/PSQLException.java

21 years agoApplied patch from Oliver Jowett to improve a buffer sizing.
Barry Lind [Mon, 11 Aug 2003 21:12:00 +0000 (21:12 +0000)]
Applied patch from Oliver Jowett to improve a buffer sizing.

 Modified Files:
  jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java

21 years agoApplied patch from Kim Ho to fix a regression against a 7.4 server. The result
Barry Lind [Mon, 11 Aug 2003 20:54:55 +0000 (20:54 +0000)]
Applied patch from Kim Ho to fix a regression against a 7.4 server.  The result
of transaction isolation level changed from uppercase to lower case between 7.3 and 7.4.  In testing, a regression was also fixed in this area when talking to
a 7.2 server due to changes in how notice messages are processed in the current
code.

 Modified Files:
  jdbc/build.xml jdbc/org/postgresql/core/BaseStatement.java
  jdbc/org/postgresql/jdbc1/AbstractJdbc1Connection.java

21 years agoCode cleanup inspired by recent resname bug report (doesn't fix the bug
Tom Lane [Mon, 11 Aug 2003 20:46:47 +0000 (20:46 +0000)]
Code cleanup inspired by recent resname bug report (doesn't fix the bug
yet, though).  Avoid using nth() to fetch tlist entries; provide a
common routine get_tle_by_resno() to search a tlist for a particular
resno.  This replaces a couple uses of nth() and a dozen hand-coded
search loops.  Also, replace a few uses of nth(length-1, list) with
llast().

21 years agoDo not link in libwsock32 on non-win32 platforms. Improve grammar
Tom Lane [Mon, 11 Aug 2003 18:07:38 +0000 (18:07 +0000)]
Do not link in libwsock32 on non-win32 platforms.  Improve grammar
of thread-safety comments.

21 years agoTranslation updates
Peter Eisentraut [Mon, 11 Aug 2003 15:19:58 +0000 (15:19 +0000)]
Translation updates

21 years agoRepair potential deadlock created by recent changes to recycle btree
Tom Lane [Sun, 10 Aug 2003 19:48:08 +0000 (19:48 +0000)]
Repair potential deadlock created by recent changes to recycle btree
index pages: when _bt_getbuf asks the FSM for a free index page, it is
possible (and, in some cases, even moderately likely) that the answer
will be the same page that _bt_split is trying to split.  _bt_getbuf
already knew that the returned page might not be free, but it wasn't
prepared for the possibility that even trying to lock the page could
be problematic.  Fix by doing a conditional rather than unconditional
grab of the page lock.

21 years agoMake various links point to specific places instead of entire parts
Tom Lane [Sun, 10 Aug 2003 01:20:34 +0000 (01:20 +0000)]
Make various links point to specific places instead of entire parts
(doubtless these are hangovers from the old separate-books days).

21 years agoRequest verbose error message output in bug reports.
Tom Lane [Sun, 10 Aug 2003 01:19:18 +0000 (01:19 +0000)]
Request verbose error message output in bug reports.

21 years agoUpdate docs for 7.4 array features and polymorphic functions.
Tom Lane [Sat, 9 Aug 2003 22:50:22 +0000 (22:50 +0000)]
Update docs for 7.4 array features and polymorphic functions.
This is Joe Conway's patch of 7-Aug plus further editorializing
of my own.

21 years agoI get the following warning when I compile src/bin/pg_id/pg_id.c:
Bruce Momjian [Sat, 9 Aug 2003 01:25:46 +0000 (01:25 +0000)]
I get the following warning when I compile src/bin/pg_id/pg_id.c:

gcc -pipe -g -Wall -Wmissing-prototypes -Wmissing-declarations
-I../../../src/include   -c -o pg_id.o pg_id.c -MMD
pg_id.c: In function `main':
pg_id.c:35: warning: unused variable `optarg'

The attached trivial patch fixes the warning by removing the variable.

Neil Conway

21 years agoThis makes psql use the new pg_get_viewdef function. It also puts one
Bruce Momjian [Sat, 9 Aug 2003 01:21:54 +0000 (01:21 +0000)]
This makes psql use the new pg_get_viewdef function.  It also puts one
extra line break in to improve formatting.

Christopher Kings-Lynne

21 years agoAnother pgindent run with updated typedefs.
Bruce Momjian [Fri, 8 Aug 2003 21:42:59 +0000 (21:42 +0000)]
Another pgindent run with updated typedefs.

21 years agopgindent fix for new typedefs.
Bruce Momjian [Fri, 8 Aug 2003 21:25:06 +0000 (21:25 +0000)]
pgindent fix for new typedefs.

21 years agoAdd:
Bruce Momjian [Fri, 8 Aug 2003 20:31:26 +0000 (20:31 +0000)]
Add:

> * Allow clients to query WITH HOLD cursors and prepared statements
503d503
<

21 years agoRemove simple_prompt from /contrib C files, now that it is in /port.
Bruce Momjian [Fri, 8 Aug 2003 20:20:49 +0000 (20:20 +0000)]
Remove simple_prompt from /contrib C files, now that it is in /port.

They had the old versions anyway.

21 years agoAvoid malloc(0). Although standard mallocs seem not to object,
Tom Lane [Fri, 8 Aug 2003 19:19:32 +0000 (19:19 +0000)]
Avoid malloc(0).  Although standard mallocs seem not to object,
some debugging malloc packages do.

21 years agoFix nasty little order-of-operations bug in _SPI_cursor_operation.
Tom Lane [Fri, 8 Aug 2003 19:18:21 +0000 (19:18 +0000)]
Fix nasty little order-of-operations bug in _SPI_cursor_operation.
Per report from Mendola Gaetano.

21 years agoUpdate:
Bruce Momjian [Fri, 8 Aug 2003 18:55:35 +0000 (18:55 +0000)]
Update:

<  o -Implement SQLDA
>  o Implement SQLDA

21 years agoRemove, no in /port.
Bruce Momjian [Fri, 8 Aug 2003 18:43:49 +0000 (18:43 +0000)]
Remove, no in /port.

21 years agoUpdate.
Bruce Momjian [Fri, 8 Aug 2003 18:36:34 +0000 (18:36 +0000)]
Update.

21 years agoUpdate items.
Bruce Momjian [Fri, 8 Aug 2003 17:17:17 +0000 (17:17 +0000)]
Update items.

21 years agoUpdate as done:
Bruce Momjian [Fri, 8 Aug 2003 17:13:02 +0000 (17:13 +0000)]
Update as done:

< * Allow easy display of usernames in a group
> * -Allow easy display of usernames in a group
88,89d87
< * -Delay resolution of array expression type so assignment coercion
<   can be performed on empty array expressions (Joe)
94c92,94
<  o Support construction of array result values in expressions
>  o -Support construction of array result values in expressions (Joe)
>  o Delay resolution of array expression type so assignment coercion
>    can be performed on empty array expressions (Joe)
148c148
< * Allow LIKE indexing optimization for non-ASCII locales
> * -Allow LIKE indexing optimization for non-ASCII locales using special index
173c173
< * Return proper effected tuple count from complex commands [return]
> * -Return proper effected tuple count from complex commands [return]
236c236
<  o Allow SHOW of non-modifiable variables, like pg_controldata
>   o -Allow SHOW of some non-modifiable variables, like pg_controldata
257a258
>  o Add capability to create and call PROCEDURES
272c273
<
> * Allow psql \du to show groups, and add \dg for groups
424c425
< * Improve Subplan list handling
> * -Improve Subplan list handling

21 years agoNot needed anymore.
Bruce Momjian [Fri, 8 Aug 2003 17:05:11 +0000 (17:05 +0000)]
Not needed anymore.

21 years agoDone:
Bruce Momjian [Fri, 8 Aug 2003 16:34:40 +0000 (16:34 +0000)]
Done:

<  o Allow array declarations and other data types in PL/PgSQL DECLARE
>  o -Allow array declarations and other data types in PL/PgSQL DECLARE
254c254
<  o Allow PL/PgSQL to support array element assignment
>  o -Allow PL/PgSQL to support array element assignment (Joe)

21 years agoRemove:
Bruce Momjian [Fri, 8 Aug 2003 16:33:35 +0000 (16:33 +0000)]
Remove:

<  o Add PL/PgSQL PROCEDURES that can return multiple values

21 years agosprompt not needed in libpq.
Bruce Momjian [Fri, 8 Aug 2003 16:32:45 +0000 (16:32 +0000)]
sprompt not needed in libpq.

21 years agoSuppress unused-variable warnings when building without Asserts.
Tom Lane [Fri, 8 Aug 2003 14:39:45 +0000 (14:39 +0000)]
Suppress unused-variable warnings when building without Asserts.

21 years agoConversion functions must be STRICT to prevent them from getting null inputs.
Tom Lane [Fri, 8 Aug 2003 14:31:12 +0000 (14:31 +0000)]
Conversion functions must be STRICT to prevent them from getting null inputs.

21 years agoFix null checking of type decimal datums without indicator in Informix
Peter Eisentraut [Fri, 8 Aug 2003 13:17:58 +0000 (13:17 +0000)]
Fix null checking of type decimal datums without indicator in Informix
mode.

from Dave Cramer

21 years agoBe a little bit more careful about using sqlstate, in case libpq returned
Peter Eisentraut [Fri, 8 Aug 2003 13:16:20 +0000 (13:16 +0000)]
Be a little bit more careful about using sqlstate, in case libpq returned
NULL for it.

21 years agoTranslation updates by Dennis Björklund
Peter Eisentraut [Fri, 8 Aug 2003 11:03:29 +0000 (11:03 +0000)]
Translation updates by Dennis Björklund

21 years agoRemove postgres_ext.h inclusion -- not necessary.
Peter Eisentraut [Fri, 8 Aug 2003 10:43:54 +0000 (10:43 +0000)]
Remove postgres_ext.h inclusion -- not necessary.

21 years agoMove simple_prompt()/sprompt.c into /port.
Bruce Momjian [Fri, 8 Aug 2003 04:52:22 +0000 (04:52 +0000)]
Move simple_prompt()/sprompt.c into /port.

21 years agoMakefile cleanup.
Bruce Momjian [Fri, 8 Aug 2003 03:22:24 +0000 (03:22 +0000)]
Makefile cleanup.

21 years agoCleanup.
Bruce Momjian [Fri, 8 Aug 2003 03:18:40 +0000 (03:18 +0000)]
Cleanup.

21 years agoTypo cleanup.
Bruce Momjian [Fri, 8 Aug 2003 03:09:56 +0000 (03:09 +0000)]
Typo cleanup.

21 years agoMakefile cleanup.
Bruce Momjian [Fri, 8 Aug 2003 03:03:54 +0000 (03:03 +0000)]
Makefile cleanup.

21 years agoMore thread.c cleanup.
Bruce Momjian [Fri, 8 Aug 2003 03:00:31 +0000 (03:00 +0000)]
More thread.c cleanup.

21 years agoMore threading cleanups.
Bruce Momjian [Fri, 8 Aug 2003 02:55:08 +0000 (02:55 +0000)]
More threading cleanups.

21 years agomore thread.c consistency.
Bruce Momjian [Fri, 8 Aug 2003 02:48:24 +0000 (02:48 +0000)]
more thread.c consistency.

21 years agothreads.c -> thread.c, be consistent.
Bruce Momjian [Fri, 8 Aug 2003 02:46:40 +0000 (02:46 +0000)]
threads.c -> thread.c, be consistent.

21 years agoUpdate done items:
Bruce Momjian [Fri, 8 Aug 2003 02:00:20 +0000 (02:00 +0000)]
Update done items:

< * Allow elog() to return error codes, module name, file name, line
<   number, not just messages (Peter E)
< * Add error codes (Peter E)
< * Make error messages more consistent [error]
> * -Allow elog() to return error codes, module name, file name, line
>   number, not just messages (Tom)
> * -Add error codes (Tom)
> * -Make error messages more consistent
40c40
< * Add GUC log_statement_and_duration to print statement and >= min duration
> * -Add GUC log_statement_and_duration to print statement and >= min duration
84c84
< * Allow current datestyle to restrict dates;  prevent month/day swapping
> * -Allow current datestyle to restrict dates;  prevent month/day swapping
86c86
< * Prevent month/day swapping of ISO dates to make invalid dates valid
> * -Prevent month/day swapping of ISO dates to make invalid dates valid
88c88
< * Delay resolution of array expression type so assignment coercion
> * -Delay resolution of array expression type so assignment coercion
93c93
<  o Allow arrays to be ORDER'ed
>  o -Allow arrays to be ORDER'ed
116c116
< * Remove Cyrillic recode support
> * -Remove Cyrillic recode support
144c144
< * Certain indexes will not shrink, e.g. indexes on ever-increasing
> * -Certain indexes will not shrink, e.g. indexes on ever-increasing
185c185
< * Have SELECT '13 minutes'::interval display zero seconds in ISO datestyle
> * -Have SELECT '13 minutes'::interval display zero seconds in ISO datestyle
196c196
<  o -Add ALTER TABLE tab SET WITHOUT OIDS (Rod)
>  o --Add ALTER TABLE tab SET WITHOUT OIDS (Rod)
221c221
<    stored in the backend
>    stored in the backend (Gavin)
235c235
<  o Allow EXPLAIN EXECUTE to see prepared plans
>  o -Allow EXPLAIN EXECUTE to see prepared plans
241d240
<  o Add untrusted version of plpython
265c264
< * Allow psql to show transaction status if backend protocol changes made
> * -Allow psql to show transaction status if backend protocol changes made
272,273c271,272
< * Modify pg_get_triggerdef() to take a boolean to pretty-print,
<  and use that as part of pg_dump along with psql
> * -Modify pg_get_triggerdef() to take a boolean to pretty-print,
>   and use that as part of pg_dump along with psql
292c291
<  o Add SQLSTATE
>  o -Add SQLSTATE
296c295
<  o Implement SQLDA (do we really need this?)
>  o -Implement SQLDA
364d362
< * Allow binding query args over FE/BE protocol
378c376,377
< * Provide automatic running of vacuum in the background (Tom) [vacuum]
> * Provide automatic running of vacuum in the background in backend
>   rather than in /contrib [vacuum]
427c426
< * Allow Subplans to use efficient joins(hash, merge) with upper variable
> * -Allow Subplans to use efficient joins(hash, merge) with upper variable
429c428
< * Allow merge and hash joins on expressions not just simple variables (Tom)
> * -Allow merge and hash joins on expressions not just simple variables (Tom)
474c473
< * Remove memory/file descriptor freeing befor elog(ERROR)  (Bruce)
> * Remove memory/file descriptor freeing before ereport(ERROR)  (Bruce)
489,490c488,489
<  o Show transaction status in psql
<  o Allow binding of query parameters, support for prepared queries
>  o -Show transaction status in psql
>  o -Allow binding of query parameters, support for prepared queries
492,494c491,493
<  o Remove hard-coded limits on user/db/password names
<  o Remove unused elements of startup packet (unused, tty, passlength)
<  o Fix COPY/fastpath protocol?
>  o -Remove hard-coded limits on user/db/password names
>  o -Remove unused elements of startup packet (unused, tty, passlength)
>  o -Fix COPY/fastpath protocol
496,497c495
<  o Replication support?
<  o Error codes
>  o -Error codes
500d497
<  o ecpg improvements?
503c500
<  o Report server version number, database encoding, client encoding
>  o -Report server version number, database encoding, client encoding

21 years agoRemove space between function name and opening paren in pg_dump, for
Bruce Momjian [Fri, 8 Aug 2003 01:21:02 +0000 (01:21 +0000)]
Remove space between function name and opening paren in pg_dump, for
clarity.

21 years agoNo need for TODO.detail error.
Bruce Momjian [Fri, 8 Aug 2003 00:28:26 +0000 (00:28 +0000)]
No need for TODO.detail error.

21 years agoFix floating-point timestamp comparisons to not go nuts if NaN is
Tom Lane [Fri, 8 Aug 2003 00:10:31 +0000 (00:10 +0000)]
Fix floating-point timestamp comparisons to not go nuts if NaN is
encountered; per bug report from Christian van der Leeden 8/7/03.
Also, adjust larger/smaller routines (MAX/MIN) to share code with
comparisons for timestamp, interval, timetz.

21 years agoClean up function header.
Bruce Momjian [Thu, 7 Aug 2003 23:43:32 +0000 (23:43 +0000)]
Clean up function header.

21 years agoFix typo.
Tom Lane [Thu, 7 Aug 2003 21:38:55 +0000 (21:38 +0000)]
Fix typo.

21 years agoDon't assume that struct option is available just because we can find a
Tom Lane [Thu, 7 Aug 2003 21:11:58 +0000 (21:11 +0000)]
Don't assume that struct option is available just because we can find a
getopt_long().  This is more or less the same problem as we saw earlier
with getaddrinfo() and struct addrinfo, and for the same reason: random
user-added libraries might contain the subroutine, but there's no
guarantee we will find the matching header files.

21 years agoSCO_ACCEPT_BUG code didn't get updated for new SockAddr struct definition.
Tom Lane [Thu, 7 Aug 2003 19:37:13 +0000 (19:37 +0000)]
SCO_ACCEPT_BUG code didn't get updated for new SockAddr struct definition.

21 years agocreate_unique_plan() should not discard existing output columns of the
Tom Lane [Thu, 7 Aug 2003 19:20:24 +0000 (19:20 +0000)]
create_unique_plan() should not discard existing output columns of the
subplan it starts with, as they may be needed at upper join levels.
See comments added to code for the non-obvious reason why.  Per bug report
from Robert Creager.

21 years agoDon't use HAVE_STRUCT_ADDRINFO as a guide to whether netdb.h defines
Tom Lane [Thu, 7 Aug 2003 16:45:21 +0000 (16:45 +0000)]
Don't use HAVE_STRUCT_ADDRINFO as a guide to whether netdb.h defines
macros like AI_NUMERICHOST; instead, test the macros individually.
Should fix recent reports of trouble on AIX and Unixware.

21 years agoMake table column type TEXT.
Bruce Momjian [Thu, 7 Aug 2003 16:37:31 +0000 (16:37 +0000)]
Make table column type TEXT.

21 years agoAdd iteration option to thread test program.
Bruce Momjian [Thu, 7 Aug 2003 16:14:03 +0000 (16:14 +0000)]
Add iteration option to thread test program.

21 years agoUpdate pgindent readme.
Bruce Momjian [Thu, 7 Aug 2003 15:02:43 +0000 (15:02 +0000)]
Update pgindent readme.

21 years agoRemove unnecessary use of multiple cat processes to feed psql; this
Tom Lane [Thu, 7 Aug 2003 14:36:31 +0000 (14:36 +0000)]
Remove unnecessary use of multiple cat processes to feed psql; this
reduces the number of concurrent processes launched during parallel
regression tests, possibly avoiding failures such as exceeding a user's
max number of processes.  This essentially completes the reversion
of 1.27->1.28.

21 years agoMove dollar sign items to object section.
Bruce Momjian [Thu, 7 Aug 2003 14:12:50 +0000 (14:12 +0000)]
Move dollar sign items to object section.

21 years agoUpdate typedef names for pgindent 7.4.
Bruce Momjian [Thu, 7 Aug 2003 05:18:14 +0000 (05:18 +0000)]
Update typedef names for pgindent 7.4.

21 years agoUpdate ecpg thread testing program to be more automated.
Bruce Momjian [Thu, 7 Aug 2003 05:12:00 +0000 (05:12 +0000)]
Update ecpg thread testing program to be more automated.

21 years agomarkup fixes
Barry Lind [Thu, 7 Aug 2003 05:06:40 +0000 (05:06 +0000)]
markup fixes

21 years agoInclude postgres_ext.h instead of postgres_fe.h. This allows ecpg
Bruce Momjian [Thu, 7 Aug 2003 04:44:26 +0000 (04:44 +0000)]
Include postgres_ext.h instead of postgres_fe.h.  This allows ecpg
output C files to proper compile again.

21 years agoNoticed broken markup in ecpg.sgml. Attached patch fixes it. Please
Bruce Momjian [Thu, 7 Aug 2003 04:17:21 +0000 (04:17 +0000)]
Noticed broken markup in ecpg.sgml. Attached patch fixes it. Please
apply.

Joe Conway

21 years agoAdd -lm for ecpg/pgtypeslib/Makefile link so -lm isn't required for ecpg
Bruce Momjian [Thu, 7 Aug 2003 04:03:18 +0000 (04:03 +0000)]
Add -lm for ecpg/pgtypeslib/Makefile link so -lm isn't required for ecpg
compiles.

21 years agoSometimes the third time is the charm. Third try to fix the sql injection
Barry Lind [Thu, 7 Aug 2003 04:03:13 +0000 (04:03 +0000)]
Sometimes the third time is the charm.  Third try to fix the sql injection
vulnerability.  This fix completely removes the ability (hack) of being able
to bind a list of values in an in clause.  It was demonstrated that by allowing
that functionality you open up the possibility for certain types of
sql injection attacks.  The previous fix attempts all focused on preventing
the insertion of additional sql statements (the semi-colon problem:
xxx; any new sql statement here).  But that still left the ability to
change the where clause on the current statement or perform a subselect
which can circumvent applicaiton security logic and/or allow you to call
any stored function.

 Modified Files:
  jdbc/org/postgresql/jdbc1/AbstractJdbc1Statement.java

21 years agoUpdate release items from Tom.
Bruce Momjian [Thu, 7 Aug 2003 03:59:25 +0000 (03:59 +0000)]
Update release items from Tom.

21 years agoApplied doc patch for the jdbc docs submitted by Nic Ferrier for functionality
Barry Lind [Wed, 6 Aug 2003 23:50:19 +0000 (23:50 +0000)]
Applied doc patch for the jdbc docs submitted by Nic Ferrier for functionality
he supplied a few months ago, but didn't get around to docing until now.  And
he also added some doc for calling stored functions in general from jdbc that was missing.

 Modified Files:
  sgml/jdbc.sgml

21 years agoAdd:
Bruce Momjian [Wed, 6 Aug 2003 23:36:02 +0000 (23:36 +0000)]
Add:

>  o Add ALTER DATABASE ... OWNER TO newowner

21 years agoRename fields of DestReceiver to avoid collisions with (ill-considered)
Tom Lane [Wed, 6 Aug 2003 17:46:46 +0000 (17:46 +0000)]
Rename fields of DestReceiver to avoid collisions with (ill-considered)
macros in some platforms' sys/socket.h.

21 years agoAdopt a random backoff algorithm for sleep delays when waiting for a
Tom Lane [Wed, 6 Aug 2003 16:43:43 +0000 (16:43 +0000)]
Adopt a random backoff algorithm for sleep delays when waiting for a
spinlock.  Per recent pghackers discussion.

21 years agoFix compiler-detected problem for Alphas: it seems strlen returns
Tom Lane [Wed, 6 Aug 2003 15:54:06 +0000 (15:54 +0000)]
Fix compiler-detected problem for Alphas: it seems strlen returns
something wider than int on that platform.  Also, remove bogus
assumption that sizeof("INT_MAX") has something to do with the maximum
number of digits in an int.

21 years agoOpenFTS vs Tsearch2 comment
Teodor Sigaev [Wed, 6 Aug 2003 14:53:01 +0000 (14:53 +0000)]
OpenFTS vs Tsearch2 comment

21 years agoAdd documentation about ts_debug
Teodor Sigaev [Wed, 6 Aug 2003 09:41:13 +0000 (09:41 +0000)]
Add documentation about ts_debug

21 years agoAdd ts_debug function for debugging configurations
Teodor Sigaev [Wed, 6 Aug 2003 09:19:21 +0000 (09:19 +0000)]
Add ts_debug function for debugging  configurations

21 years agoApplied patch from kho@redhat.com to fix a problem with trying to use a fetch
Barry Lind [Wed, 6 Aug 2003 05:53:13 +0000 (05:53 +0000)]
Applied patch from kho@redhat.com to fix a problem with trying to use a fetch
when a cursor wasn't being used.

 Modified Files:
  jdbc/org/postgresql/jdbc1/AbstractJdbc1ResultSet.java

21 years agoAdd ecpg thread testing file.
Bruce Momjian [Wed, 6 Aug 2003 02:19:51 +0000 (02:19 +0000)]
Add ecpg thread testing file.

21 years agoImprove documentation of ParseDateTime(). Reorder tests to prevent
Tom Lane [Tue, 5 Aug 2003 18:30:21 +0000 (18:30 +0000)]
Improve documentation of ParseDateTime().  Reorder tests to prevent
writing one more value into return arrays than will fit.  This is
potentially a stack smash, though I do not think it is a problem in
current uses of the routine, since a failure return causes elog anyway.

21 years agoFix several places where fractional-second inputs were misprocessed
Tom Lane [Tue, 5 Aug 2003 17:39:19 +0000 (17:39 +0000)]
Fix several places where fractional-second inputs were misprocessed
in HAVE_INT64_TIMESTAMP cases, including two potential stack smashes
when more than six fractional digits were supplied.  Per bug report
from Philipp Reisner.

21 years agoModify:
Bruce Momjian [Tue, 5 Aug 2003 15:31:33 +0000 (15:31 +0000)]
Modify:

Prevent interval from supressing ':00' seconds display

21 years agoRemove Tom mention on array items.
Bruce Momjian [Tue, 5 Aug 2003 15:04:16 +0000 (15:04 +0000)]
Remove Tom mention on array items.

21 years agoAdd mention of :00 seconds.
Bruce Momjian [Tue, 5 Aug 2003 14:37:49 +0000 (14:37 +0000)]
Add mention of :00 seconds.