]> granicus.if.org Git - postgresql/log
postgresql
24 years agoAdded patch by Christof Petig <christof.petig@wtal.de> that fixes some bugs in preproc.y.
Michael Meskes [Tue, 31 Oct 2000 15:11:38 +0000 (15:11 +0000)]
Added patch by Christof Petig <christof.petig@wtal.de> that fixes some bugs in preproc.y.

24 years agoUse $(filter ...), not $(findstring ...).
Peter Eisentraut [Tue, 31 Oct 2000 14:37:25 +0000 (14:37 +0000)]
Use $(filter ...), not $(findstring ...).

24 years agoAdded long-standing transaction when restoring BLOBS (uses commit every BLOB_BATCH_SIZE)
Philip Warner [Tue, 31 Oct 2000 14:20:30 +0000 (14:20 +0000)]
Added long-standing transaction when restoring BLOBS (uses commit every BLOB_BATCH_SIZE)
Prevent dumping of languages from template1.

24 years agoChange internal string representation of BitString node to include a
Peter Eisentraut [Tue, 31 Oct 2000 13:59:53 +0000 (13:59 +0000)]
Change internal string representation of BitString node to include a
leading 'b', as it appears to be more convenient this way for the input
and node functions.

24 years agoHashing functions from Marko Kreen <marko@l-t.ee>
Peter Eisentraut [Tue, 31 Oct 2000 13:11:28 +0000 (13:11 +0000)]
Hashing functions from Marko Kreen <marko@l-t.ee>

24 years agoChange the parser to convert SQL "position" and "substring" syntax to
Peter Eisentraut [Tue, 31 Oct 2000 10:22:13 +0000 (10:22 +0000)]
Change the parser to convert SQL "position" and "substring" syntax to
position() and substring() functions, so that it works transparently for
bit types as well.  Alias the text functions appropriately.

Add position() for bit types.

Add new constant node T_BitString that represents literals of the form
B'1001 and pass those to zpbit type.

24 years agoDisallow zero-length delimited identifier (per SQL).
Peter Eisentraut [Mon, 30 Oct 2000 17:54:16 +0000 (17:54 +0000)]
Disallow zero-length delimited identifier (per SQL).

24 years agoSync preproc.y with gram.y.
Michael Meskes [Mon, 30 Oct 2000 14:43:40 +0000 (14:43 +0000)]
Sync preproc.y with gram.y.

24 years agoAdd -enable-uniconv
Tatsuo Ishii [Mon, 30 Oct 2000 13:27:13 +0000 (13:27 +0000)]
Add -enable-uniconv

24 years agoAdd support for code conversion between Unicode and other encodings.
Tatsuo Ishii [Mon, 30 Oct 2000 10:41:05 +0000 (10:41 +0000)]
Add support for code conversion between Unicode and other encodings.
Supported encodings are: EUC_JP, EUC_CN, EUC_KR, EUC_TW, Shift JIS,
Big5, ISO8859-[1-5].
TODO: testings! and documentations...

24 years agoAvoid dependency on backend's multibyte module as possible.
Tatsuo Ishii [Mon, 30 Oct 2000 10:31:46 +0000 (10:31 +0000)]
Avoid dependency on backend's multibyte module as possible.
Now frontend/libpq has its own version of pg_encoding_to_char
and pg_char_to_encoding.

24 years agoAdd new configure option "--enable-uniconv" that enables automatic
Tatsuo Ishii [Mon, 30 Oct 2000 07:17:31 +0000 (07:17 +0000)]
Add new configure option "--enable-uniconv" that enables automatic
code conversion between Unicode and other encodings. Note that
this option requires --enable-multibyte also.
The reason why this is optional is that the feature requires huge
mapping tables and I don't think every user need the feature.

24 years agoinclude pg_wchar.h to import a fucntion prototype of pg_mbcliplen
Tatsuo Ishii [Mon, 30 Oct 2000 06:48:36 +0000 (06:48 +0000)]
include pg_wchar.h to import a fucntion prototype of pg_mbcliplen

24 years agoWAL fixes.
Vadim B. Mikheev [Sun, 29 Oct 2000 18:33:41 +0000 (18:33 +0000)]
WAL fixes.

24 years agoupdate flags.
Bruce Momjian [Sun, 29 Oct 2000 18:13:47 +0000 (18:13 +0000)]
update flags.

24 years agoAutomatic dependency tracking for C++ (GCC only)
Peter Eisentraut [Sun, 29 Oct 2000 16:13:28 +0000 (16:13 +0000)]
Automatic dependency tracking for C++ (GCC only)

24 years agoRemove special treatment of '|' operator, in the spirit of "sane" binary
Peter Eisentraut [Sun, 29 Oct 2000 16:11:33 +0000 (16:11 +0000)]
Remove special treatment of '|' operator, in the spirit of "sane" binary
operators.

24 years agoUSE_POSIX_TIME replaced by HAVE_TM_ZONE || HAVE_INT_TIMEZONE, which are
Peter Eisentraut [Sun, 29 Oct 2000 13:17:34 +0000 (13:17 +0000)]
USE_POSIX_TIME replaced by HAVE_TM_ZONE || HAVE_INT_TIMEZONE, which are
equivalent.

In linux.h there were some #undef HAVE_INT_TIMEZONE, which are useless
because HAVE_TM_ZONE overrides it anyway, and messing with configure
results isn't cool.

24 years agoOmit perl and python interfaces when making temporary installation.
Peter Eisentraut [Sun, 29 Oct 2000 12:49:49 +0000 (12:49 +0000)]
Omit perl and python interfaces when making temporary installation.

24 years agoShould be "test =", not "test ==".
Peter Eisentraut [Sun, 29 Oct 2000 11:36:44 +0000 (11:36 +0000)]
Should be "test =", not "test ==".

24 years agoRemoved multibyte stuff since client does not know about encoding in the backendFixed...
Michael Meskes [Sun, 29 Oct 2000 09:44:58 +0000 (09:44 +0000)]
Removed multibyte stuff since client does not know about encoding in the backendFixed quoting bug reported by Sascha Demetrio (sd@b-comp.de).

24 years ago#define JMP_BUF has been unnecessary since the arrival of the sigsetjmp
Peter Eisentraut [Sat, 28 Oct 2000 23:53:01 +0000 (23:53 +0000)]
#define JMP_BUF has been unnecessary since the arrival of the sigsetjmp
test.

24 years agoAs we don't use struct dirent, the #define SYSV_DIRENT is useless. If we
Peter Eisentraut [Sat, 28 Oct 2000 22:53:17 +0000 (22:53 +0000)]
As we don't use struct dirent, the #define SYSV_DIRENT is useless.  If we
every do need it we'd better use AC_HEADER_DIRENT from Autoconf.

#define NO_EMPTY_STMTS is completely unused.

24 years agoMake initdb safe against using
Peter Eisentraut [Sat, 28 Oct 2000 22:14:14 +0000 (22:14 +0000)]
Make initdb safe against using
a) mismatching backend program, by checking --version output
b) mismatching bki files, by putting a version-identifying comment atop
   those files.

24 years agoBack out change to gram.y for parens.
Bruce Momjian [Sat, 28 Oct 2000 19:41:00 +0000 (19:41 +0000)]
Back out change to gram.y for parens.

24 years agobeos fixes from Cyril VELTER
Bruce Momjian [Sat, 28 Oct 2000 18:27:57 +0000 (18:27 +0000)]
beos fixes from Cyril VELTER

24 years agoWAL
Vadim B. Mikheev [Sat, 28 Oct 2000 16:21:00 +0000 (16:21 +0000)]
WAL

24 years agoOkay, here's my attempt at fixing the problems with parentheses in
Bruce Momjian [Sat, 28 Oct 2000 15:44:04 +0000 (15:44 +0000)]
Okay, here's my attempt at fixing the problems with parentheses in
subqueries.  It passes the normal 'runcheck' tests, and I've tried
a few simple things like
  select 1 as foo union (((((select 2))))) order by foo;

There are a few things that it doesn't do that have been talked
about here at least a little:

1) It doesn't allow things like "IN(((select 1)))" -- the select
here has to be at the top level.  This is not new.
2) It does NOT preserve the odd syntax I found when I started looking
at this, where a SELECT statement could begin with parentheses.  Thus,
  (SELECT a from foo) order by a;
fails.

I have preserved the ability, used in the regression tests, to
have a single select statement in what appears to be a RuleActionMulti
(but wasn't -- the parens were part of select_clause syntax).
In my version, this is a special form.

This may cause some discussion: I have differentiated the two kinds
of RuleActionMulti.  Perhaps nobody knew there were two kinds, because
I don't think the second form appears in the regression tests. This
one uses square brackets instead of parentheses, but originally was
otherwise the same as the one in parentheses.  In this version of
gram.y, the square bracket form treats SELECT statements the same
as the other allowed statements.  As discussed before on this list,
psql cannot make sense out of the results of such a thing, but an
application might.  And I have designs on just such an application.

++ kevin o'gorman

24 years agoRemove compiler warnings (add missing include files). Only link against
Peter Eisentraut [Sat, 28 Oct 2000 15:10:36 +0000 (15:10 +0000)]
Remove compiler warnings (add missing include files).  Only link against
-lm if it's actually there.

24 years agoAdd --version and --help options to "postgres".
Peter Eisentraut [Sat, 28 Oct 2000 01:07:00 +0000 (01:07 +0000)]
Add --version and --help options to "postgres".

24 years agoMake use of system-specific linker option to embed shared library search
Peter Eisentraut [Fri, 27 Oct 2000 23:59:39 +0000 (23:59 +0000)]
Make use of system-specific linker option to embed shared library search
path into executables and shared libraries (-rpath or -R for most).  Can be
disabled with --disable-rpath, since some binary packaging standards do not
like this option.

24 years agoFix java driver example.
Bruce Momjian [Fri, 27 Oct 2000 22:28:23 +0000 (22:28 +0000)]
Fix java driver example.

24 years agoUse compiler driver, not linker, to link shared libraries on Unixware.
Peter Eisentraut [Fri, 27 Oct 2000 20:09:48 +0000 (20:09 +0000)]
Use compiler driver, not linker, to link shared libraries on Unixware.

24 years agoMake regression tests work with VPATH builds.
Peter Eisentraut [Fri, 27 Oct 2000 20:00:22 +0000 (20:00 +0000)]
Make regression tests work with VPATH builds.

24 years agoFix for drop db message format changes
Tatsuo Ishii [Fri, 27 Oct 2000 02:25:34 +0000 (02:25 +0000)]
Fix for drop db message format changes

24 years agoRemove gcc-only macro definition
Tatsuo Ishii [Fri, 27 Oct 2000 02:23:51 +0000 (02:23 +0000)]
Remove gcc-only macro definition

24 years agoDisallow bits beyond the mask length for CIDR values, per discussion
Tom Lane [Fri, 27 Oct 2000 01:55:23 +0000 (01:55 +0000)]
Disallow bits beyond the mask length for CIDR values, per discussion
on pghackers.  Arrange for the sort ordering of general INET values
to be network part as major sort key, host part as minor sort key.
I did not force an initdb for this change, but anyone who's running
indexes on general INET values may need to recreate those indexes.

24 years agoRe-implement LIMIT/OFFSET as a plan node type, instead of a hack in
Tom Lane [Thu, 26 Oct 2000 21:38:24 +0000 (21:38 +0000)]
Re-implement LIMIT/OFFSET as a plan node type, instead of a hack in
ExecutorRun.  This allows LIMIT to work in a view.  Also, LIMIT in a
cursor declaration will behave in a reasonable fashion, whereas before
it was overridden by the FETCH count.

24 years agoUpdate TODO list.
Bruce Momjian [Thu, 26 Oct 2000 19:55:10 +0000 (19:55 +0000)]
Update TODO list.

24 years agoFix breakage I introduced yesterday in MULTIBYTE compilations.
Tom Lane [Thu, 26 Oct 2000 17:31:35 +0000 (17:31 +0000)]
Fix breakage I introduced yesterday in MULTIBYTE compilations.
Sorry 'bout that, chief...

24 years agoClean up gcc warnings in MULTIBYTE mode.
Tom Lane [Thu, 26 Oct 2000 17:04:12 +0000 (17:04 +0000)]
Clean up gcc warnings in MULTIBYTE mode.

24 years agoIn flex --version test, redirect stdin to /dev/null, because some lex' hang
Peter Eisentraut [Thu, 26 Oct 2000 16:28:01 +0000 (16:28 +0000)]
In flex --version test, redirect stdin to /dev/null, because some lex' hang
waiting for input.  From Pete Forman <gsez020@kryten.bedford.waii.com>.

24 years agoUpdate TODO list.
Bruce Momjian [Thu, 26 Oct 2000 11:41:55 +0000 (11:41 +0000)]
Update TODO list.

24 years agoClean up broken test for whether to wait for input in SSL case.
Tom Lane [Wed, 25 Oct 2000 22:27:25 +0000 (22:27 +0000)]
Clean up broken test for whether to wait for input in SSL case.
Per discussion with Magnus Hagander.

24 years agoautoconf
Bruce Momjian [Wed, 25 Oct 2000 21:55:13 +0000 (21:55 +0000)]
autoconf

24 years agoEnsure clause_selectivity() behaves sanely when examining an uplevel Var
Tom Lane [Wed, 25 Oct 2000 21:48:12 +0000 (21:48 +0000)]
Ensure clause_selectivity() behaves sanely when examining an uplevel Var
or a Var that references a subquery output.

24 years agoTeach psql about new relkind for views.
Tom Lane [Wed, 25 Oct 2000 20:36:52 +0000 (20:36 +0000)]
Teach psql about new relkind for views.

24 years agoSupport SET/SHOW/RESET client_encoding and server_encoding even when
Tom Lane [Wed, 25 Oct 2000 19:44:44 +0000 (19:44 +0000)]
Support SET/SHOW/RESET client_encoding and server_encoding even when
MULTIBYTE support is not compiled (you just can't set them to anything
but SQL_ASCII).  This should reduce interoperability problems between
MB-enabled clients and non-MB-enabled servers.

24 years agoMinor cleanup.
Tom Lane [Wed, 25 Oct 2000 19:36:03 +0000 (19:36 +0000)]
Minor cleanup.

24 years agoSynced preproc.y with gram.y.
Michael Meskes [Wed, 25 Oct 2000 19:28:09 +0000 (19:28 +0000)]
Synced preproc.y with gram.y.

24 years agoAccept CREATE DATABASE WITH ENCODING 'SQL_ASCII' even when MULTIBYTE
Tom Lane [Wed, 25 Oct 2000 18:56:16 +0000 (18:56 +0000)]
Accept CREATE DATABASE WITH ENCODING 'SQL_ASCII' even when MULTIBYTE
support is not present.  This allows a non-MB server to load a pg_dumpall
script produced by an MB-enabled server, so long as only ASCII encoding
was used.

24 years agoRearrange handling of -L linker options so that they are always before all
Peter Eisentraut [Wed, 25 Oct 2000 16:13:52 +0000 (16:13 +0000)]
Rearrange handling of -L linker options so that they are always before all
the -l options.  (This was not the case when using the  OpenSSL or Kerberos
options.)  Also make sure that shared library links get to see all the -L
options.  Get Kerberos 5 support to compile on Redhat 7.0.  Add OpenSSL and
-lsocket (if used/found) to libpq link.

24 years agoFrom Zoltan Kovacs back in April (sorry for the delay Zoltan!):
Thomas G. Lockhart [Wed, 25 Oct 2000 14:15:49 +0000 (14:15 +0000)]
From Zoltan Kovacs back in April (sorry for the delay Zoltan!):

I modified the current ODBC driver for

* referential integrity error reporting,
* SELECT in transactions and
* disabling autocommit.

I tested these changes with Borland C++ Builder -> ODBCExpress ->
WinODBC driver (DLL) -> Postgres 7.0beta1 and Borland C++ Builder -> BDE ->
WinODBC driver (DLL) -> Postgres 7.0beta1. The patch is based on snapshot of
22th April (I don't think that someone has modified it since that: Byron
hasn't gave any sign of living for about a month and I didn't find any
comments about the ODBC driver on the list).

24 years agoFinal (?) fix for tar (null block at end)
Philip Warner [Wed, 25 Oct 2000 10:21:38 +0000 (10:21 +0000)]
Final (?) fix for tar (null block at end)
Dump template db in dumpall

24 years agoAdded some more C constructs to the parser.
Michael Meskes [Wed, 25 Oct 2000 07:00:33 +0000 (07:00 +0000)]
Added some more C constructs to the parser.

24 years agoWAL utils defs
Vadim B. Mikheev [Wed, 25 Oct 2000 00:49:14 +0000 (00:49 +0000)]
WAL utils defs

24 years agoCheck for SIGHUP and process config file updates just after waiting
Tom Lane [Tue, 24 Oct 2000 21:33:52 +0000 (21:33 +0000)]
Check for SIGHUP and process config file updates just after waiting
for input, not just before.

24 years agoDo not execute fastpath function calls if in transaction ABORT state.
Tom Lane [Tue, 24 Oct 2000 20:59:35 +0000 (20:59 +0000)]
Do not execute fastpath function calls if in transaction ABORT state.
Just like queries, doing nothing is better than possibly getting weird
error messages.  Also, improve comments.

24 years agoInteger binary operators, from Marko Kreen <marko@l-t.ee>. Renamed bitxor
Peter Eisentraut [Tue, 24 Oct 2000 20:16:48 +0000 (20:16 +0000)]
Integer binary operators, from Marko Kreen <marko@l-t.ee>.  Renamed bitxor
operator to '#' for consistency.  Parser still needs work.

24 years agoFix AbortOutOfAnyTransaction logic to avoid notice about
Tom Lane [Tue, 24 Oct 2000 20:06:39 +0000 (20:06 +0000)]
Fix AbortOutOfAnyTransaction logic to avoid notice about
'AbortTransaction and not in in-progress state' when client disconnects
just after an error.  Notice seems pretty harmless, so I'm not going
to worry about back-patching this into 7.0.* ...

24 years agoWups, messed up the comment markers on that last change.
Tom Lane [Tue, 24 Oct 2000 19:31:13 +0000 (19:31 +0000)]
Wups, messed up the comment markers on that last change.

24 years agoIgnore missing postmaster.opts.default file.
Peter Eisentraut [Tue, 24 Oct 2000 19:11:15 +0000 (19:11 +0000)]
Ignore missing postmaster.opts.default file.

24 years agoUpdate TODO list.
Bruce Momjian [Tue, 24 Oct 2000 19:09:20 +0000 (19:09 +0000)]
Update TODO list.

24 years agoNew config.guess and config.sub with Darwin support (from official GNU
Peter Eisentraut [Tue, 24 Oct 2000 18:11:48 +0000 (18:11 +0000)]
New config.guess and config.sub with Darwin support (from official GNU
source).  Adjust resultmap accordingly and fix some other regexps.

24 years agoCreate empty file so that CVS sources compile (Vadim can fill in real
Tom Lane [Tue, 24 Oct 2000 18:05:14 +0000 (18:05 +0000)]
Create empty file so that CVS sources compile (Vadim can fill in real
definition later...)

24 years agoUpdate plperl README per suggestions from Edward Bridges.
Tom Lane [Tue, 24 Oct 2000 17:47:32 +0000 (17:47 +0000)]
Update plperl README per suggestions from Edward Bridges.

24 years agoDon't call AC_SUBST (in libtool.m4) before AC_INIT. The m4 diversions get
Peter Eisentraut [Tue, 24 Oct 2000 17:41:50 +0000 (17:41 +0000)]
Don't call AC_SUBST (in libtool.m4) before AC_INIT.  The m4 diversions get
messed up and we end up with the comments from included files in the output
"configure".

24 years agoEnable plperl to be built when --with-perl is selected. Dunno whether
Tom Lane [Tue, 24 Oct 2000 17:03:46 +0000 (17:03 +0000)]
Enable plperl to be built when --with-perl is selected.  Dunno whether
latest fixes make it safe or not, but we won't find out if no one builds
it, eh?

24 years agoEliminate POLLUTE=1 hack for cross-Perl-version compatibility by using
Tom Lane [Tue, 24 Oct 2000 17:01:06 +0000 (17:01 +0000)]
Eliminate POLLUTE=1 hack for cross-Perl-version compatibility by using
Devel::PPPort instead.  Thanks to Gilles Darold for doing the legwork.

24 years agoMake test for <netinet/tcp.h> IRIX-proof.
Tom Lane [Tue, 24 Oct 2000 14:55:28 +0000 (14:55 +0000)]
Make test for <netinet/tcp.h> IRIX-proof.

24 years agoVarious fixes to TAR header format
Philip Warner [Tue, 24 Oct 2000 13:24:30 +0000 (13:24 +0000)]
Various fixes to TAR header format
Fix for endian bug in TAR output
Nicer error messages in pg_dump

24 years agoWAL misc
Vadim B. Mikheev [Tue, 24 Oct 2000 09:56:23 +0000 (09:56 +0000)]
WAL misc

24 years agoRejigger do_lo_list's query so that obj_description() is evaluated only
Tom Lane [Tue, 24 Oct 2000 04:13:28 +0000 (04:13 +0000)]
Rejigger do_lo_list's query so that obj_description() is evaluated only
once per distinct LO, not once per pg_largeobject tuple.

24 years agoReconsider page size for large objects: rather than stuffing disk pages
Tom Lane [Tue, 24 Oct 2000 03:34:53 +0000 (03:34 +0000)]
Reconsider page size for large objects: rather than stuffing disk pages
as full as possible, seems better to use a tuple size around BLCKSZ/4
so that less space is wasted when a LO tuple is updated.  Also, this
lets us use a logical page size that's an exact power of two, avoiding
partial-page writes when client is sending us stuff in power-of-2
buffer chunks.

24 years agoIncrease buffer size to improve performance of client-side lo_import.
Tom Lane [Tue, 24 Oct 2000 03:15:14 +0000 (03:15 +0000)]
Increase buffer size to improve performance of client-side lo_import.

24 years agoRemove arbitrary limit on number of simultaneously open large objects.
Tom Lane [Tue, 24 Oct 2000 03:14:08 +0000 (03:14 +0000)]
Remove arbitrary limit on number of simultaneously open large objects.
This used to make some sense under the old implementation, but now an
open LO is pretty darn cheap, so why restrict it?

24 years agoUpdate TODO list.
Bruce Momjian [Tue, 24 Oct 2000 01:59:22 +0000 (01:59 +0000)]
Update TODO list.

24 years agoMajor overhaul of large-object implementation, by Denis Perchine with
Tom Lane [Tue, 24 Oct 2000 01:38:44 +0000 (01:38 +0000)]
Major overhaul of large-object implementation, by Denis Perchine with
kibitzing from Tom Lane.  Large objects are now all stored in a single
system relation "pg_largeobject" --- no more xinv or xinx files, no more
relkind 'l'.  This should offer substantial performance improvement for
large numbers of LOs, since there won't be directory bloat anymore.
It'll also fix problems like running out of locktable space when you
access thousands of LOs in one transaction.
Also clean up cruft in read/write routines.  LOs with "holes" in them
(never-written byte ranges) now work just like Unix files with holes do:
a hole reads as zeroes but doesn't occupy storage space.
INITDB forced!

24 years agoAdd some modicum of error detection...
Tom Lane [Tue, 24 Oct 2000 00:08:02 +0000 (00:08 +0000)]
Add some modicum of error detection...

24 years agoRemove testlo2.c --- there seems to be no difference between this and
Tom Lane [Tue, 24 Oct 2000 00:04:11 +0000 (00:04 +0000)]
Remove testlo2.c --- there seems to be no difference between this and
testlo.c, except it's even skimpier on error checking :-(

24 years agoIf a field is incompressible ('compressed' data is actually larger than
Tom Lane [Mon, 23 Oct 2000 23:42:04 +0000 (23:42 +0000)]
If a field is incompressible ('compressed' data is actually larger than
source, due to addition of header overhead), store it as plain data
rather than pseudo-compressed data.  This saves a few microseconds when
reading it out, but much more importantly guarantees that the toaster
won't actually expand tuples that contain incompressible data.  That's
essential to avoid 'Tuple too big' failures with large objects.

24 years agoUpdate TODO list.
Bruce Momjian [Mon, 23 Oct 2000 22:15:36 +0000 (22:15 +0000)]
Update TODO list.

24 years agoPolish shared library build to reduce number of special hacks. In
Peter Eisentraut [Mon, 23 Oct 2000 21:44:12 +0000 (21:44 +0000)]
Polish shared library build to reduce number of special hacks.  In
particular, allow linking with arbitrary commands rather than only $(AR) or
$(LD), and treat C++ without hacks.

Add option to disable shared libraries.  This takes the place of the
BSD_SHLIB variable.  The regression test driver ignores the plpgsql test
if there are no shared libraries available.

24 years agoautoconf
Bruce Momjian [Mon, 23 Oct 2000 16:47:48 +0000 (16:47 +0000)]
autoconf

24 years agoBack out xti.h portion of the patch.
Bruce Momjian [Mon, 23 Oct 2000 14:50:44 +0000 (14:50 +0000)]
Back out xti.h portion of the patch.

24 years agoNew relcache hash table with RelFileNode as key to be used
Vadim B. Mikheev [Mon, 23 Oct 2000 04:10:24 +0000 (04:10 +0000)]
New relcache hash table with RelFileNode as key to be used
from bufmgr - it would be nice to have separate hash in smgr
for node <--> fd mappings, but for the moment it's easy to
add new hash to relcache.
Fixed small bug in xlog.c:ReadRecord.

24 years agoMinor updates/corrections for CREATE/DROP FUNCTION/AGGREGATE/OPERATOR.
Tom Lane [Mon, 23 Oct 2000 00:46:07 +0000 (00:46 +0000)]
Minor updates/corrections for CREATE/DROP FUNCTION/AGGREGATE/OPERATOR.

24 years agoSome small polishing of Mark Hollomon's cleanup of DROP command: might
Tom Lane [Sun, 22 Oct 2000 23:32:48 +0000 (23:32 +0000)]
Some small polishing of Mark Hollomon's cleanup of DROP command: might
as well allow DROP multiple INDEX, RULE, TYPE as well.  Add missing
CommandCounterIncrement to DROP loop, which could cause trouble otherwise
with multiple DROP of items affecting same catalog entries.  Try to
bring a little consistency to various error messages using 'does not exist',
'nonexistent', etc --- I standardized on 'does not exist' since that's
what the vast majority of the existing uses seem to be.

24 years agoplpgsql regress tests seem a tad out of date ... repair bit rot.
Tom Lane [Sun, 22 Oct 2000 23:25:11 +0000 (23:25 +0000)]
plpgsql regress tests seem a tad out of date ... repair bit rot.

24 years agoCleanup error messages in findLastBuiltinOid
Philip Warner [Sun, 22 Oct 2000 23:16:55 +0000 (23:16 +0000)]
Cleanup error messages in findLastBuiltinOid

24 years agoMakeover for Unixware 7.1.1
Peter Eisentraut [Sun, 22 Oct 2000 22:15:13 +0000 (22:15 +0000)]
Makeover for Unixware 7.1.1

* Makefile: Add more standard targets.  Improve shell redirection in GNU
make detection.
* src/backend/access/transam/rmgr.c: Fix incorrect(?) C.
* src/backend/libpq/pqcomm.c (StreamConnection): Work around accept() bug.
* src/include/port/unixware.h: ...with help from here.
* src/backend/nodes/print.c (plannode_type): Remove some "break"s after
"return"s.
* src/backend/tcop/dest.c (DestToFunction): ditto.
* src/backend/nodes/readfuncs.c: Add proper prototypes.
* src/backend/utils/adt/numutils.c (pg_atoi): Cope specially with strtol()
setting EINVAL. This saves us from creating an extra set of regression test
output for the affected systems.
* src/include/storage/s_lock.h (tas): Correct prototype.
* src/interfaces/libpq/fe-connect.c (parseServiceInfo): Don't use variable
as dimension in array definition.
* src/makefiles/Makefile.unixware: Add support for GCC.
* src/template/unixware: same here
* src/test/regress/expected/abstime-solaris-1947.out: Adjust whitespace.
* src/test/regress/expected/horology-solaris-1947.out: Part of this file
was evidently missing.
* src/test/regress/pg_regress.sh: Fix shell.  mkdir -p returns non-zero if
the directory exists.
* src/test/regress/resultmap: Add entries for Unixware.

24 years agoFix incorrect logic for clearing BufferDirtiedByMe in ReleaseRelationBuffers
Tom Lane [Sun, 22 Oct 2000 20:20:49 +0000 (20:20 +0000)]
Fix incorrect logic for clearing BufferDirtiedByMe in ReleaseRelationBuffers
and DropBuffers.  Formerly we cleared the flag for each buffer currently
belonging to the target rel or database, but that's completely wrong!
Must look at BufferTagLastDirtied to see whether the BufferDirtiedByMe
flag is relevant to target rel or not; this is *independent* of the
current contents of the buffer.  Vadim spotted this problem, but his
fix was only partially correct...

24 years agoPatch VACUUM problem with moving chain of update tuples when source
Tom Lane [Sun, 22 Oct 2000 19:49:43 +0000 (19:49 +0000)]
Patch VACUUM problem with moving chain of update tuples when source
and destination of a tuple lie on the same page.
(Previously fixed in REL7_0 branch, now apply to current.)

24 years agoFix typo.
Tom Lane [Sun, 22 Oct 2000 19:11:05 +0000 (19:11 +0000)]
Fix typo.

24 years agodatlastsysoid is an Oid, not an int4 ...
Tom Lane [Sun, 22 Oct 2000 18:29:58 +0000 (18:29 +0000)]
datlastsysoid is an Oid, not an int4 ...

24 years agoUse new datlastsysoid field in pg_database + some cleanups & fixes
Philip Warner [Sun, 22 Oct 2000 18:13:09 +0000 (18:13 +0000)]
Use new datlastsysoid field in pg_database + some cleanups & fixes

24 years agoAdd new datlastsysoid to pg_database.
Philip Warner [Sun, 22 Oct 2000 17:55:49 +0000 (17:55 +0000)]
Add new datlastsysoid to pg_database.
This field stores the last allocated OID after the database was created.
Used by pg_dump in deciding what is user-defined vs. system-defined.

24 years agoFixed mixing of two enum datatypes.
Michael Meskes [Sun, 22 Oct 2000 13:36:19 +0000 (13:36 +0000)]
Fixed mixing of two enum datatypes.

24 years agoBACKED OUT.
Bruce Momjian [Sun, 22 Oct 2000 05:27:23 +0000 (05:27 +0000)]
BACKED OUT.

> Regression tests opr_sanity and sanity_check are now failing.

Um, Bruce, I've said several times that I didn't think Perchine's large
object changes should be applied until someone had actually reviewed
them.

24 years agoCrash on NULL again in system table check.
Bruce Momjian [Sun, 22 Oct 2000 05:14:01 +0000 (05:14 +0000)]
Crash on NULL again in system table check.