Bruce Momjian [Sat, 7 Oct 2006 23:28:04 +0000 (23:28 +0000)]
Allow HISTORY/history.html to be generated by removing <link> elements,
per suggestion from Peter. Without this, the main doc links prevent
those files from being built.
Tom Lane [Sat, 7 Oct 2006 22:21:38 +0000 (22:21 +0000)]
Fix ancient oversight in psql's \d pattern processing code: when seeing two
quote chars inside quote marks, should emit one quote *and stay in inquotes
mode*. No doubt the lack of reports of this have something to do with the
poor documentation of the feature ...
The -X option in pg_dump was supposed to be a workaround for the lack of
portable long options. But we have had portable long options for a long
time now, so this is obsolete. Now people have added options which *only*
work with -X but not as regular long option, so I'm putting a stop to this:
-X is deprecated; it still works, but it has been removed from the
documentation, and please don't add more of them.
Tom Lane [Sat, 7 Oct 2006 20:16:57 +0000 (20:16 +0000)]
When planning a query at Bind time, be careful to pass the correct
query_list into the Portal, ie, the one seen and possibly modified by
the planner. My fault :-( Per report from Sergey Koposov.
Tom Lane [Sat, 7 Oct 2006 19:25:29 +0000 (19:25 +0000)]
On platforms that have getrlimit(RLIMIT_STACK), use it to ensure that
max_stack_depth is not set to an unsafe value.
This commit also provides configure-time checking for <sys/resource.h>,
and cleans up some perhaps-unportable code associated with use of that
include file and getrlimit().
Tom Lane [Sat, 7 Oct 2006 16:43:28 +0000 (16:43 +0000)]
Adjust HINT for stack depth limit to mention checking the underlying
platform limit, rather than just blindly raising max_stack_depth.
Also, tweak the code to work properly if someone sets max_stack_depth
to more than 2Gb, which guc.c will allow on a 64-bit machine.
Tom Lane [Sat, 7 Oct 2006 00:11:53 +0000 (00:11 +0000)]
Fix string_to_array() to correctly handle the case where there are
overlapping possible matches for the separator string, such as
string_to_array('123xx456xxx789', 'xx').
Also, revise the logic of replace(), split_part(), and string_to_array()
to avoid O(N^2) work from redundant searches and conversions to pg_wchar
format when there are N matches to the separator string.
Backpatched the full patch as far as 8.0. 7.4 also has the bug, but the
code has diverged a lot, so I just went for a quick-and-dirty fix of the
bug itself in that branch.
Tom Lane [Fri, 6 Oct 2006 18:23:35 +0000 (18:23 +0000)]
Fix SysCacheGetAttr() to handle the case where the specified syscache has not
been initialized yet. This can happen because there are code paths that call
SysCacheGetAttr() on a tuple originally fetched from a different syscache
(hopefully on the same catalog) than the one specified in the call. It
doesn't seem useful or robust to try to prevent that from happening, so just
improve the function to cope instead. Per bug#2678 from Jeff Trout. The
specific example shown by Jeff is new in 8.1, but to be on the safe side
I'm backpatching 8.0 as well. We could patch 7.x similarly but I think
that's probably overkill, given the lack of evidence of old bugs of this ilk.
Tom Lane [Thu, 5 Oct 2006 23:33:33 +0000 (23:33 +0000)]
Cleanup for pglz_compress code: remove dead code, const-ify API of
remaining functions, simplify pglz_compress's API to not require a useless
data copy when compression fails. Also add a check in pglz_decompress that
the expected amount of data was decompressed.
Tom Lane [Thu, 5 Oct 2006 17:57:40 +0000 (17:57 +0000)]
Make use of qsort_arg in several places that were formerly using klugy
static variables. This avoids any risk of potential non-reentrancy,
and in particular offers a much cleaner workaround for the Intel compiler
bug that was affecting ginutil.c.
Tom Lane [Thu, 5 Oct 2006 00:07:45 +0000 (00:07 +0000)]
Second try at fixing libedit configuration for Bob Friesenhahn.
Buildfarm results from 'gazelle' show that there are indeed libedit
versions for which history.h is a needed header, even though it's
apparently been dropped entirely in other versions. Grumble.
Tom Lane [Wed, 4 Oct 2006 22:31:14 +0000 (22:31 +0000)]
Adjust configure's handling of libedit to not look for a history.h file.
Per Bob Friesenhahn's report, this file is not supplied by some versions
of libedit, and even when it is supplied it seems to be just a link to
readline.h, so we don't need to include it anyway.
Also, ensure that we won't try to use a too-old version of Bison.
The previous coding would bleat but then use it anyway; better to invoke
the 'missing' script if any grammar files need to be rebuilt.
Tom Lane [Wed, 4 Oct 2006 18:58:08 +0000 (18:58 +0000)]
Fix initdb's logic for picking shared_buffers/max_fsm_pages, which had
gotten rather thoroughly whacked out by careless recent changes: the
intended ratio between the two was off by a lot, and the minimum number
of shared buffers tried had increased by a lot. Problem exposed by
failures on buildfarm members with smaller SHMMAX values.
Tom Lane [Wed, 4 Oct 2006 17:52:52 +0000 (17:52 +0000)]
Improve error messages from to_tsquery per yesterday's discussion:
provide the bad input, and be sure to mention that we are talking about
a tsearch query.
Tom Lane [Tue, 3 Oct 2006 22:18:23 +0000 (22:18 +0000)]
Switch over to using our own qsort() all the time, as has been proposed
repeatedly. Now that we don't have to worry about memory leaks from
glibc's qsort, we can safely put CHECK_FOR_INTERRUPTS into the tuplesort
comparators, as was requested a couple months ago. Also, get rid of
non-reentrancy and an extra level of function call in tuplesort.c by
providing a variant qsort_arg() API that passes an extra void * argument
through to the comparison routine. (We might want to use that in other
places too, I didn't look yet.)
Bruce Momjian [Tue, 3 Oct 2006 22:12:14 +0000 (22:12 +0000)]
pgevent fixes:
1) Make vcbuild actually build the pgevent dll.
2) Change the pgevent DLL file so it doens't specify ordinal for the
functions. You're not supposed to do that. You're actually supposed to
declare them as PRIVATE as well, but mingw doesn't support that. VC++
will throw a warning and not an error though, so we can live with it.
Tom Lane [Sun, 1 Oct 2006 22:25:48 +0000 (22:25 +0000)]
Make some marginal performance improvements in reportErrorPosition(),
which turns out to be a dominant part of the runtime in scenarios
involving lots of parse-time warnings (such as Stephen Frost's example
of an INSERT with a lot of backslash-containing strings). There's not
a whole lot we can do about the character-at-a-time scanning, but we
can at least avoid traversing the query twice.
Tom Lane [Sun, 1 Oct 2006 22:08:18 +0000 (22:08 +0000)]
Do a CHECK_FOR_INTERRUPTS after emitting a message of less than ERROR
severity. This is to ensure the user can cancel a query that's spitting
out lots of notice/warning messages, even if they're coming from a loop
that doesn't otherwise contain a CHECK_FOR_INTERRUPTS. Per gripe from
Stephen Frost.
Tom Lane [Sun, 1 Oct 2006 18:37:53 +0000 (18:37 +0000)]
Remove accented characters in comments, to avoid failures when this
file is read with an incompatible client_encoding setting. Per report
from Tim N. van der Leeuw.
Tom Lane [Sun, 1 Oct 2006 17:23:38 +0000 (17:23 +0000)]
Fix overly enthusiastic Assert introduced in 8.1: it's expecting a
CaseTestExpr, but forgot that the optimizer is sometimes able to replace
CaseTestExpr by Const.
Bruce Momjian [Sat, 30 Sep 2006 03:55:32 +0000 (03:55 +0000)]
Add:
> * Allow more complex user/database default GUC settings
> Currently, ALTER USER and ALTER DATABASE support per-user and
> per-database defaults. Consider adding per-user-and-database
> defaults so things like search_path can be defaulted for a
> specific user connecting to a specific database.
>
>
Tom Lane [Fri, 29 Sep 2006 21:22:21 +0000 (21:22 +0000)]
Allow assignment to array elements not contiguous with those already
present; intervening positions are filled with nulls. This behavior
is required by SQL99 but was not implementable before 8.2 due to lack
of support for nulls in arrays. I have only made it work for the
one-dimensional case, which is all that SQL99 requires. It seems quite
complex to get it right in higher dimensions, and since we never allowed
extension at all in higher dimensions, I think that must count as a
future feature addition not a bug fix.
Tom Lane [Thu, 28 Sep 2006 20:51:43 +0000 (20:51 +0000)]
Fix IS NULL and IS NOT NULL tests on row-valued expressions to conform to
the SQL spec, viz IS NULL is true if all the row's fields are null, IS NOT
NULL is true if all the row's fields are not null. The former coding got
this right for a limited number of cases with IS NULL (ie, those where it
could disassemble a ROW constructor at parse time), but was entirely wrong
for IS NOT NULL. Per report from Teodor.
I desisted from changing the behavior for arrays, since on closer inspection
it's not clear that there's any support for that in the SQL spec. This
probably needs more consideration.
Tom Lane [Wed, 27 Sep 2006 18:40:10 +0000 (18:40 +0000)]
Replace strncpy with strlcpy in selected places that seem possibly relevant
to performance. (A wholesale effort to get rid of strncpy should be
undertaken sometime, but not during beta.) This commit also fixes dynahash.c
to correctly truncate overlength string keys for hashtables, so that its
callers don't have to anymore.
Tom Lane [Wed, 27 Sep 2006 16:29:46 +0000 (16:29 +0000)]
Add strlcpy() to the set of functions supported by src/port/ when not
available directly on the platform. Per discussion, this function is
sufficiently widely recognized to be treated as standard.
Bruce Momjian [Tue, 26 Sep 2006 11:18:29 +0000 (11:18 +0000)]
Done:
< o Add command to archive partially filled write-ahead logs? [pitr]
<
< Currently only full WAL files are archived. This means that the
< most recent transactions aren't available for recovery in case
< of a disk failure.
<
Reorder:
< o %Create dump tool for write-ahead logs for use in determining
< transaction id for point-in-time recovery
> o %Create dump tool for write-ahead logs for use in determining
> transaction id for point-in-time recovery
Tom Lane [Mon, 25 Sep 2006 22:01:10 +0000 (22:01 +0000)]
Add a check to prevent overwriting valid data if smgrnblocks() gives a
wrong answer, as has been seen to occur with a buggy Linux kernel. Not
really our bug, but it's a simple test in a seldom-used control path,
so might as well have a defense.