git.postgresql.org Git - postgresql.git/log
Tom Lane [2008年5月25日 21:51:00 +0000 (21:51 +0000)]
Adjust timestamp regression tests to prevent two low-probability failure
cases. Recent buildfarm experience shows that it is sometimes possible
to execute several SQL commands in less time than the granularity of
Windows' not-very-high-resolution gettimeofday(), leading to a failure
because the tests expect the value of now() to change and it doesn't.
Also, it was recognized some time ago that the same area of the tests
could fail if local midnight passes between the insertion and the checking
of the values for 'yesterday', 'tomorrow', etc. Clean all this up per
ideas from myself and Greg Stark.
There remains a window for failure if the transaction block is entered
exactly at local midnight (so that 'now' and 'today' have the same value),
but that seems low-probability enough to live with.
Since the point of this change is mostly to eliminate buildfarm noise,
back-patch to all versions we are still actively testing.
Link in keywords file instead of copying it.
Use #define/#ifdef instead of sed to fix include files, this should work on Windows too.
Add include directory required after the latest changes to the ecpg build.
Per red Windows buildfarm members.
Fix function headers not matching prototype in header file, per
compiler warnings on msvc.
Use CRITICAL_SECTION instead of Mutexes for thread-locking in libpq on
Windows, for better performance.
Per suggestion from Andrew Chernow, but not his patch since the underlying
code was changed to deal with return values.
ecpg_keywords.o also depends on preproc.h
Synced parser.
Made ecpg parser use backend provided keyword list.
Changed whenever test so exit value is 0.
Tom Lane [2008年5月20日 03:30:22 +0000 (03:30 +0000)]
Adjust -Wl,--asneeded test to avoid using the switch if it breaks
libreadline. What we will do for compatibility :-(
Simplify code in formatting.c now that to upper/lower/initcase do not
modify the passed string.
Done:
< o Use LC_TIME for localized weekday/month names, rather than
> o -Use LC_TIME for localized weekday/month names, rather than
Remove arbitrary 10MB limit on two-phase state file size. It's not that hard
to go beoynd 10MB, as demonstrated by Gavin Sharry's example of dropping a
schema with ~25000 objects. The really bogus thing about the limit was that
it was enforced when a state file file was read in, not when it was written,
so you would end up with a prepared transaction that you can't commit or
abort, and the only recourse was to shut down the server and remove the file
by hand.
Raise the limit to MaxAllocSize, and enforce it also when a state file is
written. We could've removed the limit altogether, but reading in a file
larger than MaxAllocSize would fail anyway because we read it into a
palloc'd buffer.
Backpatch down to 8.1, where 2PC and this issue was introduced.
Tom Lane [2008年5月19日 18:08:16 +0000 (18:08 +0000)]
Make to_char()'s localized month/day names depend on LC_TIME, not LC_MESSAGES.
Euler Taveira de Oliveira
Tom Lane [2008年5月19日 04:14:24 +0000 (04:14 +0000)]
Coercion sanity check in ri_HashCompareOp failed to allow for enums, as per
example from Rod Taylor. On reflection the correct test here is for any
polymorphic type, not specifically ANYARRAY as in the original coding.
Tom Lane [2008年5月18日 20:13:12 +0000 (20:13 +0000)]
Make another try at using -Wl,--as-needed to suppress linking of unnecessary
shared libraries. We've tried this before and had problems with libreadline
not linking properly on some platforms, but that seems to be a libreadline
bug that may have been fixed by now. In any case, it's early enough in the
8.4 devel cycle that we can afford to have some transient breakage while
we work out any portability problems.
On Darwin, we try -Wl,-dead_strip_dylibs, which seems to be the equivalent
incantation there.
Remove old kluge put in to allow Windows regression tests to succeed, and now
found to have been made necessary by our skipping tty detection on Windows. Now
that we are doing tty detection on Windows the kluge is unnecessary and wrong.
Tom Lane [2008年5月18日 03:24:17 +0000 (03:24 +0000)]
okay, looks like ecpg/compatlib needs -lintl too.
Per buildfarm results.
Tom Lane [2008年5月17日 23:36:27 +0000 (23:36 +0000)]
Rewrite the warning about non-transaction-safety of TRUNCATE ... RESTART
IDENTITY to be more explicit about the possible hazards. Per gripe from Neil
and subsequent discussion. Eventually we may be able to get rid of this
warning, but for now it had better be there.
Use isatty() test for pager on Win32; not sure why it was disabled for
that platform.
In psql, test for output to stdout when forcing pager for wide output.
Another fix for compiles on an empty 'po' directory.
Don't call rm with empty file list.
Tom Lane [2008年5月17日 20:23:40 +0000 (20:23 +0000)]
ecpglib needs to link with libintl if it's in use.
Per buildfarm results.
Tom Lane [2008年5月17日 20:02:01 +0000 (20:02 +0000)]
Improve pg_regress so that it reports the fact if any test process
exits with nonzero status. The Windows part of this is untested ...
Tom Lane [2008年5月17日 17:52:14 +0000 (17:52 +0000)]
Fix utterly-bogus code for computing row heights. Per crashes on
spoonbill, though one wonders why it didn't misbehave everywhere.
In passing remove some unnecessary modulo calculations.
Tom Lane [2008年5月17日 17:24:57 +0000 (17:24 +0000)]
Fix a subtle bug exposed by recent wal_sync_method rearrangements.
Formerly, the default value of wal_sync_method was determined inside xlog.c,
but now it is determined inside guc.c. guc.c was reading xlogdefs.h
without having read <fcntl.h>, leading to wrong determination of
DEFAULT_SYNC_METHOD. Obviously xlogdefs.h needs to include <fcntl.h>
for itself to ensure stable results.
Tom Lane [2008年5月17日 16:49:23 +0000 (16:49 +0000)]
Remove DEFAULT_SYNC_FLAGBIT ... not used anymore.
fix ecpg regression tests broken by over-eager adding of $PostgreSQL$ markers
Add $PostgreSQL$ markers to a lot of files that were missing them.
This particular batch was just for *.c and *.h file.
The changes were made with the following 2 commands:
find . \( \( -name 'libstemmer' -o -name 'expected' -o -name 'ppport.h' \) -prune \) -o \( -name '*.[ch]' \) \( -exec grep -q '\$PostgreSQL' {} \; -o -print \) | while read file ; do head -n 1 < $file | grep -q '^/\*' && echo $file; done | xargs -l sed -i -e '1s/^\// /' -e '1i/*\n * $PostgreSQL:$ \n *'
find . \( \( -name 'libstemmer' -o -name 'expected' -o -name 'ppport.h' \) -prune \) -o \( -name '*.[ch]' \) \( -exec grep -q '\$PostgreSQL' {} \; -o -print \) | xargs -l sed -i -e '1i/*\n * $PostgreSQL:$ \n */'
Tom Lane [2008年5月17日 01:20:39 +0000 (01:20 +0000)]
Allow ALTER SEQUENCE START WITH to change the recorded start_value of a
sequence. This seems an obvious extension to the recent patch, and it
makes the code noticeably cleaner and more orthogonal.
Tom Lane [2008年5月16日 23:57:42 +0000 (23:57 +0000)]
Fix declarations of pthread functions, missed in recent commit.
Tom Lane [2008年5月16日 23:36:05 +0000 (23:36 +0000)]
Add a RESTART (without parameter) option to ALTER SEQUENCE, allowing a
sequence to be reset to its original starting value. This requires adding the
original start value to the set of parameters (columns) of a sequence object,
which is a user-visible change with potential compatibility implications;
it also forces initdb.
Also add hopefully-SQL-compatible RESTART/CONTINUE IDENTITY options to
TRUNCATE TABLE. RESTART IDENTITY executes ALTER SEQUENCE RESTART for all
sequences "owned by" any of the truncated relations. CONTINUE IDENTITY is
a no-op option.
Zoltan Boszormenyi
Tom Lane [2008年5月16日 19:15:05 +0000 (19:15 +0000)]
Reduce unnecessary PANIC to ERROR, improve a couple of comments.
Tom Lane [2008年5月16日 18:35:38 +0000 (18:35 +0000)]
Recent patch doesn't compile without <limits.h>
Tom Lane [2008年5月16日 18:34:51 +0000 (18:34 +0000)]
Suppress a possibly-uninitialized-variable warning. (I'm only seeing it
on Apple's gcc and not my other machines, but still it seems worth
getting rid of.)
Implement error checking for pthreads calls in thread-safe mode. They really
should always succeed, but in the likely event of a failure we would
previously fall through *without locking* - the new code will exit(1).
Printing the error message on stderr will not work for all applications, but
it's better than nothing at all - and our API doesn't provide a way to return
the error to the caller.
Tom Lane [2008年5月16日 17:31:17 +0000 (17:31 +0000)]
Suppress compiler warning induced by recent header file refactoring.
Tom Lane [2008年5月16日 17:26:07 +0000 (17:26 +0000)]
Update contrib for tsearch changes.
Update psql startup banner to be shorter, suggest "help" for help.
Add a few "help" entries.
Move \g help entry into "General".
Update psql version mismatch warning text.
Joshua D. Drake
Done in TODO:
> o -Have \l+ show database size, if permissions allow
Add detection of psql pager to trigger on wide output. Also add pager
detection for wrapped lines or lines with newlines that need pager to
display.
Tom Lane [2008年5月16日 16:31:02 +0000 (16:31 +0000)]
Extend GIN to support partial-match searches, and extend tsquery to support
prefix matching using this facility.
Teodor Sigaev and Oleg Bartunov
Add localization support to ecpg.
Author: Euler Taveira de Oliveira <euler@timbira.com>
Done:
PL/pgSQL
> o -Add CASE capability to language (already in SQL)
Tom Lane [2008年5月16日 01:27:06 +0000 (01:27 +0000)]
Persuade GIN to react to control-C in a reasonable amount of time
while building a GIN index.
Tom Lane [2008年5月15日 22:39:49 +0000 (22:39 +0000)]
Support SQL/PSM-compatible CASE statement in plpgsql.
Pavel Stehule
Tom Lane [2008年5月15日 17:37:49 +0000 (17:37 +0000)]
Add code to eval_const_expressions() to support const-simplification of
CoerceViaIO nodes. This improves the ability of the planner to deal with
cases where the node input is a constant. Per bug #4170.
Tom Lane [2008年5月15日 00:17:41 +0000 (00:17 +0000)]
Add support for tracking call counts and elapsed runtime for user-defined
functions.
Note that because this patch changes FmgrInfo, any external C functions
you might be testing with 8.4 will need to be recompiled.
Patch by Martin Pihlak, some editorialization by me (principally, removing
tracking of getrusage() numbers)
Tom Lane [2008年5月14日 19:10:29 +0000 (19:10 +0000)]
Move the "instr_time" typedef and associated macros into a new header
file portability/instr_time.h, and add a couple more macros to eliminate
some abstraction leakage we formerly had. Also update psql to use this
header instead of its own copy of nearly the same code.
This commit in itself is just code cleanup and shouldn't change anything.
It lays some groundwork for the upcoming function-stats patch, though.
Reorgnize psql \? help output, to reduce the size of the "General"
heading at the top; broken into more sections now.
Fix a few warnings that have crept into CVS HEAD.
Remove the special variable for open_sync_bit used in O_SYNC and O_DSYNC
modes, replacing it with a call to a function that derives it from the
sync_method variable, now that it has distinct values for these two cases.
This means that assign_xlog_sync_method() no longer changes any settings,
thus fixing the bug introduced in the change to use a guc enum for
wal_sync_method.
Make the win32 implementation of getrusage() return EINVAL if being
asked for anything other than RUSAGE_SELF, since it's not supported.
This is never called anywhere in the code today, but might be in
the future.
Not backpatching, since it's not called anywhere today.
Remove odd duplicate wording in psql pager section.
Tom Lane [2008年5月13日 22:10:30 +0000 (22:10 +0000)]
Improve plpgsql's RAISE command. It is now possible to attach DETAIL and
HINT fields to a user-thrown error message, and to specify the SQLSTATE
error code to use. The syntax has also been tweaked so that the
Oracle-compatible case "RAISE exception_name" works (though you won't get a
very nice error message if you just write that much). Lastly, support
the Oracle-compatible syntax "RAISE" with no parameters to re-throw
the current error from within an EXCEPTION block.
In passing, allow the syntax SQLSTATE 'nnnnn' within EXCEPTION lists,
so that there is a way to trap errors with custom SQLSTATE codes.
Pavel Stehule and Tom Lane
Don't try to close negative file descriptors, since this can cause
crashes on certain platforms. In particular, the MSVC runtime is known
to do this.
Fixes bug #4162, reported and diagnosed by Javier Pimas
This is the patch replace offnum++ by OffsetNumberNext, to be
consistent. OffsetNumberNext() has some casting that makes it useful.
Fujii Masao
Change \du to use the new printTable API, improving the situation of very wide
tables.
Brendan Jurd, with some help from me.
Fix a bug in the previous patch, which caused the title pointer to be used
before it was actually set.
Improve psql's internal print.c code by introducing an actual print API.
Provides for better code readability, but mainly this is infrastructure changes
to allow further changes such as arbitrary footers on printed tables. Also,
the translation status of each element in the table is more easily customized.
Brendan Jurd, with some editorialization by me.
Todo done:
> * -Improve dead row detection during multi-statement transactions usage
Improve snapshot manager by keeping explicit track of snapshots.
There are two ways to track a snapshot: there's the "registered" list, which
is used for arbitrary long-lived snapshots; and there's the "active stack",
which is used for the snapshot that is considered "active" at any time.
This also allows users of snapshots to stop worrying about snapshot memory
allocation and freeing, and about using PG_TRY blocks around ActiveSnapshot
assignment. This is all done automatically now.
As a consequence, this allows us to reset MyProc->xmin when there are no
more snapshots registered in the current backend, reducing the impact that
long-running transactions have on VACUUM.
Fix breakage by the wal_sync_method patch in installations that use
O_DSYNC (specifically this broke all the Windows buildfarm members)
Check for non-existant connection in prepare statement handling.
Do not close files that weren't opened.
Document that "ROLLBACK TO savepoint" does not un-close cursors.
Put back bufmgr.h in bufpage.h -- it is needed by some macros.
Remove #include bufmgr.h from (most?) source files which already include
bufpage.h.
Report which WAL sync method we are trying to change *to* when it fails,
not which one we had before (that worked, and thus is completley irrelevant)
Convert wal_sync_method to guc enum.
Restructure some header files a bit, in particular heapam.h, by removing some
unnecessary #include lines in it. Also, move some tuple routine prototypes and
macros to htup.h, which allows removal of heapam.h inclusion from some .c
files.
For this to work, a new header file access/sysattr.h needed to be created,
initially containing attribute numbers of system columns, for pg_dump usage.
While at it, make contrib ltree, intarray and hstore header files more
consistent with our header style.
Remaining pieces of fix for contrib makefiles
Tom Lane [2008年5月10日 03:31:58 +0000 (03:31 +0000)]
Restore psql's former behavior that padding spaces to the right of the last
output column are not emitted. (That change already caused more noise in
the regression test output files than I would like.) Provide some needed
editorial help for comments, clean up code formatting.
Add URL for:
o Allow an existing index to be marked as a table's primary key
>
> http://archives.postgresql.org/pgsql-hackers/2008-04/msg00500.php
Add to TODO:
> o Allow an existing index to be marked as a table's primary key
Add URL for:
* Add column to pg_stat_activity that shows the progress of long-running
commands like CREATE INDEX and VACUUM
>
> http://archives.postgresql.org/pgsql-patches/2008-04/msg00203.php
>
Update alternative output regression files for new float8 test of power().
Add TODO item:
> o Clear table counters on TRUNCATE
>
> http://archives.postgresql.org/pgsql-hackers/2008-04/msg00169.php
TODOs done:
> o -Prevent parent tables from altering or dropping constraints
> o -Prevent child tables from altering or dropping constraints
Tom Lane [Fri, 9 May 2008 23:32:05 +0000 (23:32 +0000)]
Change the rules for inherited CHECK constraints to be essentially the same
as those for inherited columns; that is, it's no longer allowed for a child
table to not have a check constraint matching one that exists on a parent.
This satisfies the principle of least surprise (rows selected from the parent
will always appear to meet its check constraints) and eliminates some
longstanding bogosity in pg_dump, which formerly had to guess about whether
check constraints were really inherited or not.
The implementation involves adding conislocal and coninhcount columns to
pg_constraint (paralleling attislocal and attinhcount in pg_attribute)
and refactoring various ALTER TABLE actions to be more like those for
columns.
Alex Hunsaker, Nikhil Sontakke, Tom Lane
Adjust power() error messages to be more descriptive.
Improve logic for finding object files on OBJS lines in contrib Makefiles. If this unbreaks buildfarm mastodon, apply everywhere.
Tom Lane [Fri, 9 May 2008 15:53:07 +0000 (15:53 +0000)]
Give a warning if -s switch is used with a non-custom pgbench test.
Also, clean up the code that assigned the scale into :scale variables.
Greg Smith and Tom Lane
Add "%option noinput" to the scanners to avoid compiler warnings. GCC 4.3
began to realize that the input() function isn't used and printed warnings.
Update C comments to mention SQL:2003 handling of power return values.
Fix Assert introduced in previous patch.
Fix incorrect archive truncation point calculation in the %r recovery_command
parameter. This fixes bug 4137 reported by Wojciech Strzalka, where a WAL
file is deleted too early when starting the recovery of a warm standby server.
Also add a sanity check in pg_standby so that it will refuse to delete anything
earlier than the file being restored, and improve the debug message in case
nothing is deleted.
Simon Riggs. Backpatch to 8.3, which is where %r was introduced.
Tom Lane [Fri, 9 May 2008 05:25:54 +0000 (05:25 +0000)]
Fix bogus expected output that should have made it quite clear that
something was wrong with that tab patch.
Tom Lane [Fri, 9 May 2008 05:25:04 +0000 (05:25 +0000)]
Fix memory stomp that's turning the whole buildfarm pink: you can't hack up
pg_wcsformat without changing pg_wcssize to match. Add some comments to
try to make that clearer, and make a couple other minor editorializations.
Add to TODO:
>
> * Improve the /contrib installation experience
>
> http://archives.postgresql.org/pgsql-hackers/2008-04/msg00132.php
Add regression test for various power expressions with a zero base, and
adjust source code to be more modular.
Have numeric 0 ^ 4.3 return 1, rather than an error, and have 0 ^ 0.0
return 1, rather than error.
This was already the float8 behavior.
Have psql output tab as the proper number of spaces, rather than \x09.
Add psql '\pset format wrapped' mode to wrap output to screen width, or
file/pipe output too if \pset columns' is set.
Bryce Nesbitt
Tom Lane [Thu, 8 May 2008 16:49:37 +0000 (16:49 +0000)]
Fix contrib/xml2 makefile to not override CFLAGS, and in passing make it
auto-configure properly for libxslt present or not.
Tom Lane [Thu, 8 May 2008 16:04:59 +0000 (16:04 +0000)]
Fix rules regression test for recent pg_stat_activity change.
Make the new pg_stat_get_activity use OUT parameters, so you don't have to
specify the column names and types. Also simplifies the view.
Per comments from Tom.
Update TODO wording:
< * Improve detection of shared memory segments being used by other
< FreeBSD jails
> * Improve detection of shared memory segments being used by others
> by checking the SysV shared memory field 'nattch'
> http://archives.postgresql.org/pgsql-hackers/2008-01/msg00673.php
Add to TODO:
>
> o Add CREATE SCHEMA ... LIKE that copies a schema
>
Add to TODO:
> * Add database and transaction-level triggers
>
> http://archives.postgresql.org/pgsql-hackers/2008-03/msg00451.php
Remove OID/foreign key mention in psql variable interpolation docs.
Add Simon Riggs' email address.
Move Client Authentication right after Server Configuration in docs.
Scott Marlowe
This is the main PostgreSQL git repository.
RSS
Atom