Commit Graph

15915 Commits

Author SHA1 Message Date
Bruce Momjian
4b407f6c3c Changes for MinGW/WIN32:
o allow configure to see include/port/win32 include files
        o add matching Win32 accept() prototype
        o allow pg_id to compile with native Win32 API
        o fix invalide mbvalidate() function calls (existing bug)
        o allow /scripts to compile with native Win32 API
        o add win32.c to Win32 compiles (already in *.mak files)
2003-09-07 03:43:57 +00:00
Tom Lane
7703e55c32 Make the default pg_hba.conf include an entry for ::1 only if configure
set HAVE_IPV6.  Per recent discussions.
2003-09-07 03:36:03 +00:00
Tom Lane
f4ed650996 load_file() has to remove pre-existing shlibs that match the new file
on either name or inode; otherwise load_external_function() won't do
anything.  At least on Linux, it appears that recompiling a shlib leads
to a new file with a different inode, so the old code failed to detect
a match.
2003-09-07 02:18:01 +00:00
Peter Eisentraut
3e20a72603 Make all valid uses of the COLLATE clause yield the same result: a syntax
error.  There is no point in providing some kind of forward compatibility
now, because no one can tell what a future implementation will look like.
2003-09-06 14:01:51 +00:00
Bruce Momjian
4060f07578 Rename variables for clarity. 2003-09-06 01:41:56 +00:00
Tom Lane
92aa462247 Second try at IPv4-to-v6 mapping code; avoid assuming that the struct
returned by getaddrinfo_all will have enough room for an IPv6 address.
2003-09-05 23:07:21 +00:00
Tom Lane
3c9bb8886d Allow IPv4-format entries in pg_hba.conf to match IPv6 connections
that have IPv4-embedded-in-IPv6 addresses.  Per idea of Andreas Pflug.
2003-09-05 20:31:36 +00:00
Bruce Momjian
23d07fa357 Done:
< * Allow psql to do table completion for SELECT * FROM schema_part and
> * -Allow psql to do table completion for SELECT * FROM schema_part and
2003-09-05 19:52:10 +00:00
Bruce Momjian
e26af972b5 ifdef out pqGetpwuid() because Win32 doesn't have struct passwd.
Needed for libpq builds, not just for native Win32 port, and probably
for cygwin too.
2003-09-05 17:43:40 +00:00
Bruce Momjian
cec1333bba Add WIN32 URL. 2003-09-05 16:43:13 +00:00
Tom Lane
ba7ee1db3b Add missing space. 2003-09-05 16:13:38 +00:00
Bruce Momjian
702c91b4ff This fixes three minor typos in hba.c.
--
Alvaro Herrera (<alvherre[a]dcc.uchile.cl>)
2003-09-05 03:57:13 +00:00
Bruce Momjian
b0ba70df83 Update information about Linux's overcommit memory behavior.
Andrew Dunstan
2003-09-05 02:54:09 +00:00
Bruce Momjian
5b49f9025a Add:
> * Maintain a map of recently-expired of pages so vacuum can reclaim
>   free space without a sequential scan
2003-09-05 02:51:55 +00:00
Bruce Momjian
04aabb034f Clean up WIN32_CONSOLE code. 2003-09-05 02:31:10 +00:00
Bruce Momjian
248af40d6d Update as done:
> * -Improve concurrency of hash indexes (Tom)
2003-09-05 02:20:50 +00:00
Bruce Momjian
2dc9437160 Enable Win32 to compile libpq again, and enable SSL compiles on that
platform.

Andreas Pflug
2003-09-05 02:08:36 +00:00
Bruce Momjian
1709fde56a Improve prototype casting for thread calls. 2003-09-05 01:53:41 +00:00
Bruce Momjian
8ac3510a4c Add:
> 	o Allow ALTER TABLE ... ALTER CONSTRAINT ... RENAME
2003-09-05 01:10:07 +00:00
Bruce Momjian
8af31863e9 Add:
<
> * Print table names with constraint names in error messages, or make constraint
>   names unique within a schema
2003-09-05 00:50:40 +00:00
Tom Lane
7a3693716d Reimplement hash index locking algorithms, per my recent proposal to
pghackers.  This fixes the problem recently reported by Markus KrÌutner
(hash bucket split corrupts the state of scans being done concurrently),
and I believe it also fixes all the known problems with deadlocks in
hash index operations.  Hash indexes are still not really ready for prime
time (since they aren't WAL-logged), but this is a step forward.
2003-09-04 22:06:27 +00:00
Tom Lane
ca43f71ca5 Avoid consuming unreasonable amounts of memory when GRANT has many
grantees.
2003-09-04 15:53:04 +00:00
Tom Lane
fe055e9280 Looks like plperl has same bug as pltcl. 2003-09-04 15:16:39 +00:00
Tom Lane
9dbfcc2261 Fix some problems with dropped columns in pltcl functions. 2003-09-04 15:10:10 +00:00
Bruce Momjian
6d239ee4b4 Allow non-super users to set log_duration to true, then false, when the
administrator has not turned it on, and fix other PGC_USERLIMIT
variables.
2003-09-04 05:11:20 +00:00
Bruce Momjian
b8245766ab Re-upper case CHECKPOINT_SEGMENTS. 2003-09-04 03:38:55 +00:00
Bruce Momjian
cac43aefd8 Lowercase CHECKPOINT_SEGMENTS. 2003-09-04 00:36:46 +00:00
Bruce Momjian
adb13f231b Add:
> * Allow to_char to print localized month names (Karel)
2003-09-03 23:01:47 +00:00
Bruce Momjian
38eb18092c Tighten casting of function call pointers for threads. 2003-09-03 22:34:08 +00:00
Peter Eisentraut
ee4ae6ceb2 Add note that a for loop where the lower bound is greater than the upper
bound is valid but does nothing.

suggested by Richard Huxton <dev@archonet.com>
2003-09-03 22:17:07 +00:00
Peter Eisentraut
29a20145fd Pass session_authorization to the client and make psql update its prompt
accordingly.
2003-09-03 22:05:09 +00:00
Bruce Momjian
4e85f760e6 Done:
* -Have standalone backend read postgresql.conf (Tom)
2003-09-03 21:27:21 +00:00
Bruce Momjian
4928856fc4 Mark unixware as verified. 2003-09-03 20:54:21 +00:00
Bruce Momjian
e57c9fa6b8 Add Unixware as verified. 2003-09-03 20:51:40 +00:00
Bruce Momjian
1129eab9c7 Update German FAQ.
Ian Barwick
2003-09-03 20:28:14 +00:00
Bruce Momjian
47e6609c54 Mark BSD/OS as verified not needing reentrant function names. 2003-09-03 19:40:56 +00:00
Bruce Momjian
2310dd3af2 Add reporting of specificy threading failure functions. 2003-09-03 19:36:31 +00:00
Bruce Momjian
4cc247f74b Add test for thread-safeness of libc functions. 2003-09-03 19:30:31 +00:00
Bruce Momjian
c351729e1e Mark Unixware as not requiring *_r functions. 2003-09-03 17:55:43 +00:00
Tom Lane
b96e32ebf8 Work around ENOTEMPTY being an alias for EEXIST on some platforms. 2003-09-03 15:49:42 +00:00
Tom Lane
5840b89373 Repair problems with to_char() overrunning its input string.
From Karel Zak.
2003-09-03 14:59:41 +00:00
Tom Lane
5ac2d7c0eb In _bt_check_unique() loop, don't bother applying _bt_isequal() to
killed items; just skip to the next item immediately.  Only check for
key equality when we reach a non-killed item or the end of the index
page.  This saves key comparisons when there are lots of killed items,
as for example in a heavily-updated table that's not been vacuumed lately.
Seems to be a win for pgbench anyway.
2003-09-02 22:10:16 +00:00
Peter Eisentraut
30b4abf5d0 Remove outdated CLI things. 2003-09-02 21:45:37 +00:00
Tom Lane
b916cc435a Cause standalone backend (including bootstrap case) to read the GUC
config file if it exists.  This was already discussed as being a good
idea, and now seems the cleanest way to deal with initdb-time failures
on machines with small SHMMAX.  (The submitted patches instead modified
initdb.sh to pass the correct sizing parameters, but that would still
leave standalone backends prone to failure later.  An admin who needs
to use a standalone backend has enough trouble already, he shouldn't
have to manually configure its shmem settings...)
2003-09-02 19:04:12 +00:00
Tom Lane
d70610c4ee Several fixes for hash indexes that involve changing the on-disk index
layout; therefore, this change forces REINDEX of hash indexes (though
not a full initdb).  Widen hashm_ntuples to double so that hash space
management doesn't get confused by more than 4G entries; enlarge the
allowed number of free-space-bitmap pages; replace the useless bshift
field with a useful bmshift field; eliminate 4 bytes of wasted space
in the per-page special area.
2003-09-02 18:13:32 +00:00
Tom Lane
8b2450c831 Fix a couple typos, add some more comments. 2003-09-02 03:29:01 +00:00
Tom Lane
39673ca47b Rewrite hashbulkdelete() to make it amenable to new bucket locking
scheme.  A pleasant side effect is that it is *much* faster when deleting
a large fraction of the indexed tuples, because of elimination of
redundant hash_step activity induced by hash_adjscans.  Various other
continuing code cleanup.
2003-09-02 02:18:38 +00:00
Peter Eisentraut
5f65345a57 Do not pass server_encoding to the client.
libpq, talking to an old server, should assume SQL_ASCII as the default
client encoding, because that is what the server will actually use (not
the server encoding).
2003-09-01 23:04:49 +00:00
Peter Eisentraut
f10a9033bf Clean up after pygresql removal: adjust/remove documentation and remove
unneeded configure work.
2003-09-01 23:01:49 +00:00
Tom Lane
65c2d427fb Preliminary cleanup for hash index code (doesn't attack the locking problem
yet).  Fix a couple of bugs that would only appear if multiple bitmap pages
are used, including a buffer reference leak and incorrect computation of bit
indexes.  Get rid of 'overflow address' concept, which accomplished nothing
except obfuscating the code and creating a risk of failure due to limited
range of offset field.  Rename some misleadingly-named fields and routines,
and improve documentation.
2003-09-01 20:26:34 +00:00