]> granicus.if.org Git - postgresql/log
postgresql
18 years agoFix incorrect treatment of RL_PROMPT_START_IGNORE/RL_PROMPT_END_IGNORE,
Tom Lane [Tue, 3 Jan 2006 23:32:30 +0000 (23:32 +0000)]
Fix incorrect treatment of RL_PROMPT_START_IGNORE/RL_PROMPT_END_IGNORE,
per http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=343616 via Martin Pitt.

18 years agoAdd checks to verify that a plpgsql function returning a rowtype is actually
Tom Lane [Tue, 3 Jan 2006 22:48:10 +0000 (22:48 +0000)]
Add checks to verify that a plpgsql function returning a rowtype is actually
returning the rowtype it's supposed to return.  Per reports from David Niblett
and Michael Fuhr.

18 years agoAssume select() might modify struct timeout, so remove previous
Bruce Momjian [Tue, 3 Jan 2006 19:54:08 +0000 (19:54 +0000)]
Assume select() might modify struct timeout, so remove previous
optimization.

18 years agoUse setitimer() for stats file write, rather than do a gettimeofday()
Bruce Momjian [Tue, 3 Jan 2006 16:42:17 +0000 (16:42 +0000)]
Use setitimer() for stats file write, rather than do a gettimeofday()
call for every stats packet read to adjust select() timeout.  Other
stylistic improvements.

18 years agoDon't assume that LC_MESSAGES is always available on WIN32. Per Magnus.
Tom Lane [Mon, 2 Jan 2006 20:25:45 +0000 (20:25 +0000)]
Don't assume that LC_MESSAGES is always available on WIN32.  Per Magnus.

18 years agoReset flex state in a way that works for both flex 2.5.4 and 2.5.31.
Tom Lane [Mon, 2 Jan 2006 19:55:25 +0000 (19:55 +0000)]
Reset flex state in a way that works for both flex 2.5.4 and 2.5.31.

18 years agoRemove the nexus between trial_buffs and trial_conns, and don't test shared buffers...
Andrew Dunstan [Mon, 2 Jan 2006 16:45:12 +0000 (16:45 +0000)]
Remove the nexus between trial_buffs and trial_conns, and don't test shared buffers lower than or equal to a value we already know is good, but use that value instead. This will make it easier to adjust the trial values and/or formulae in future if necessary.

18 years agoRemove unnecessary assignment, per Tom.
Bruce Momjian [Mon, 2 Jan 2006 03:30:41 +0000 (03:30 +0000)]
Remove unnecessary assignment, per Tom.

18 years agoChange if (!(x || y)) construct with if (!x && !y), for clarity.
Bruce Momjian [Mon, 2 Jan 2006 00:58:00 +0000 (00:58 +0000)]
Change if (!(x || y)) construct with if (!x && !y), for clarity.

18 years agoRewrite ProcessConfigFile() to avoid misbehavior at EOF, as per report
Tom Lane [Sun, 1 Jan 2006 19:52:40 +0000 (19:52 +0000)]
Rewrite ProcessConfigFile() to avoid misbehavior at EOF, as per report
from Andrus Moor.  The former state-machine-style coding wasn't actually
doing much except obscuring the control flow, and it didn't extend
readily to fix this case, so I just took it out.  Also, add a
YY_FLUSH_BUFFER call to ensure the lexer is reset correctly if the
previous scan failed partway through the file.

18 years agoRemove DOS line endings ("\r\n") from several .po files. DOS line endings
Neil Conway [Sun, 1 Jan 2006 10:13:56 +0000 (10:13 +0000)]
Remove DOS line endings ("\r\n") from several .po files. DOS line endings
are inconsistent with the rest of the .po files, and apparently cause
problems for Sun's cc. Per report on IRC from "bitvector2".

18 years agoTrim an obsolete comment: it is no longer relevant that "SELECT" was
Neil Conway [Sun, 1 Jan 2006 01:41:42 +0000 (01:41 +0000)]
Trim an obsolete comment: it is no longer relevant that "SELECT" was
formerly "RETRIEVE", and so forth.

18 years agoClean up initdb's code for selecting max_connections and shared_buffers
Tom Lane [Sat, 31 Dec 2005 23:50:59 +0000 (23:50 +0000)]
Clean up initdb's code for selecting max_connections and shared_buffers
a little bit, and set the minimum buffers-per-connection ratio to 10 not
5.  I folded the two test routines into one to counteract the illusion
that the tests can be twiddled independently, and added some documentation
pointing out the necessary connection between the sets of values tested.
Fixes strange choices of parameters that I noticed CVS tip making on
Darwin with Apple's undersized default SHMMAX.

18 years agoRename variable to cmd_str.
Bruce Momjian [Sat, 31 Dec 2005 19:39:11 +0000 (19:39 +0000)]
Rename variable to cmd_str.

18 years agoRename variable 'what' to 'stat_msg'.
Bruce Momjian [Sat, 31 Dec 2005 17:46:19 +0000 (17:46 +0000)]
Rename variable 'what' to 'stat_msg'.

18 years agoPrefix client-side prepare with '[protocol]' rather than '[client]'.
Bruce Momjian [Sat, 31 Dec 2005 16:50:44 +0000 (16:50 +0000)]
Prefix client-side prepare with '[protocol]' rather than '[client]'.

18 years agoRename send_rfq to send_ready_for_query.
Bruce Momjian [Fri, 30 Dec 2005 23:49:48 +0000 (23:49 +0000)]
Rename send_rfq to send_ready_for_query.

18 years agoMmark client-side prepare/bind/execute statements with "[client]" so
Bruce Momjian [Fri, 30 Dec 2005 22:55:20 +0000 (22:55 +0000)]
Mmark client-side prepare/bind/execute statements with "[client]" so
they can be easily distinguished from SQL commands.

18 years agoAdd support for Solaris x86_64 using Sun's compiler.
Bruce Momjian [Fri, 30 Dec 2005 21:43:41 +0000 (21:43 +0000)]
Add support for Solaris x86_64 using Sun's compiler.

Pierre Girard

18 years agoRepair EXPLAIN failure when trying to display a plan condition that involves
Tom Lane [Fri, 30 Dec 2005 18:34:22 +0000 (18:34 +0000)]
Repair EXPLAIN failure when trying to display a plan condition that involves
selection of a field from the result of a function returning RECORD.
I believe this case is new in 8.1; it's due to the addition of OUT parameters.
Per example from Michael Fuhr.

18 years agoAdd:
Bruce Momjian [Fri, 30 Dec 2005 03:24:58 +0000 (03:24 +0000)]
Add:

> * Make CLUSTER preserve recently-dead tuples per MVCC requirements

18 years agoDisallow setting ssl = on if SSL is not compiled in.
Peter Eisentraut [Fri, 30 Dec 2005 00:13:50 +0000 (00:13 +0000)]
Disallow setting ssl = on if SSL is not compiled in.

18 years agoIndex: src/pl/plpython/plpython.c
Neil Conway [Thu, 29 Dec 2005 21:47:32 +0000 (21:47 +0000)]
Index: src/pl/plpython/plpython.c
===================================================================
RCS file: /Users/neilc/postgres/cvs_root/pgsql/src/pl/plpython/plpython.c,v
retrieving revision 1.67
diff -c -r1.67 plpython.c
*** src/pl/plpython/plpython.c 26 Dec 2005 04:28:48 -0000 1.67
--- src/pl/plpython/plpython.c 29 Dec 2005 16:54:57 -0000
***************
*** 2,8 ****
   * plpython.c - python as a procedural language for PostgreSQL
   *
   * This software is copyright by Andrew Bosma
!  * but is really shameless cribbed from pltcl.c by Jan Weick, and
   * plperl.c by Mark Hollomon.
   *
   * The author hereby grants permission to use, copy, modify,
--- 2,8 ----
   * plpython.c - python as a procedural language for PostgreSQL
   *
   * This software is copyright by Andrew Bosma
!  * but is really shamelessly cribbed from pltcl.c by Jan Wieck, and
   * plperl.c by Mark Hollomon.
   *
   * The author hereby grants permission to use, copy, modify,
***************
*** 1996,2002 ****
   int i,
   rv;
   PLyPlanObject *plan;
-  char    *nulls;
   MemoryContext oldcontext;

   if (list != NULL)
--- 1996,2001 ----
***************
*** 2018,2024 ****
   if (nargs != plan->nargs)
   {
   char    *sv;
-
   PyObject   *so = PyObject_Str(list);

   if (!so)
--- 2017,2022 ----
***************
*** 2036,2048 ****
   oldcontext = CurrentMemoryContext;
   PG_TRY();
   {
!  nulls = palloc(nargs * sizeof(char));

   for (i = 0; i < nargs; i++)
   {
   PyObject   *elem,
      *so;
-  char    *sv;

   elem = PySequence_GetItem(list, i);
   if (elem != Py_None)
--- 2034,2045 ----
   oldcontext = CurrentMemoryContext;
   PG_TRY();
   {
!  char    *nulls = palloc(nargs * sizeof(char));

   for (i = 0; i < nargs; i++)
   {
   PyObject   *elem,
      *so;

   elem = PySequence_GetItem(list, i);
   if (elem != Py_None)
***************
*** 2051,2070 ****
   if (!so)
   PLy_elog(ERROR, "function \"%s\" could not execute plan",
    PLy_procedure_name(PLy_curr_procedure));
!  sv = PyString_AsString(so);

!  /*
!   * FIXME -- if this elogs, we have Python reference leak
!   */
!  plan->values[i] =
!  FunctionCall3(&(plan->args[i].out.d.typfunc),
!    CStringGetDatum(sv),
!  ObjectIdGetDatum(plan->args[i].out.d.typioparam),
!    Int32GetDatum(-1));

!  Py_DECREF(so);
!  Py_DECREF(elem);

   nulls[i] = ' ';
   }
   else
--- 2048,2073 ----
   if (!so)
   PLy_elog(ERROR, "function \"%s\" could not execute plan",
    PLy_procedure_name(PLy_curr_procedure));
!  Py_DECREF(elem);

!  PG_TRY();
!  {
!  char *sv = PyString_AsString(so);

!  plan->values[i] =
!  FunctionCall3(&(plan->args[i].out.d.typfunc),
!    CStringGetDatum(sv),
!  ObjectIdGetDatum(plan->args[i].out.d.typioparam),
!    Int32GetDatum(-1));
!  }
!  PG_CATCH();
!  {
!  Py_DECREF(so);
!  PG_RE_THROW();
!  }
!  PG_END_TRY();

+  Py_DECREF(so);
   nulls[i] = ' ';
   }
   else

18 years agoGet rid of the SpinLockAcquire/SpinLockAcquire_NoHoldoff distinction
Tom Lane [Thu, 29 Dec 2005 18:08:05 +0000 (18:08 +0000)]
Get rid of the SpinLockAcquire/SpinLockAcquire_NoHoldoff distinction
in favor of having just one set of macros that don't do HOLD/RESUME_INTERRUPTS
(hence, these correspond to the old SpinLockAcquire_NoHoldoff case).
Given our coding rules for spinlock use, there is no reason to allow
CHECK_FOR_INTERRUPTS to be done while holding a spinlock, and also there
is no situation where ImmediateInterruptOK will be true while holding a
spinlock.  Therefore doing HOLD/RESUME_INTERRUPTS while taking/releasing a
spinlock is just a waste of cycles.  Qingqing Zhou and Tom Lane.

18 years agoUpdate:
Bruce Momjian [Thu, 29 Dec 2005 16:45:11 +0000 (16:45 +0000)]
Update:

< * Add missing rtree optimizer selectivity
> * Improve selectivity functions for geometric operators

18 years agoMove declaration of check_function_bodies to where the perl headers
Andrew Dunstan [Thu, 29 Dec 2005 14:28:31 +0000 (14:28 +0000)]
Move declaration of check_function_bodies to where the perl headers
haven't had a chance to mangle the definition of DLLIMPORT (thanks again, perl guys).

18 years agoRevert some careless search-and-replace: "ADD" in comment text should
Neil Conway [Thu, 29 Dec 2005 04:53:18 +0000 (04:53 +0000)]
Revert some careless search-and-replace: "ADD" in comment text should
not be replaced with "ADD_P".

18 years agoRemove a space that incorrectly precedes a comma in the PL/pgSQL
Bruce Momjian [Thu, 29 Dec 2005 04:02:32 +0000 (04:02 +0000)]
Remove a space that incorrectly precedes a comma in the PL/pgSQL
documentation.

Michael Fuhr

18 years agoArrange to set the LC_XXX environment variables to match our locale
Tom Lane [Wed, 28 Dec 2005 23:22:51 +0000 (23:22 +0000)]
Arrange to set the LC_XXX environment variables to match our locale
setup.  This protects against undesired changes in locale behavior
if someone carelessly does setlocale(LC_ALL, "") (and we know who
you are, perl guys).

18 years agoUpdate regression tests for new referential integrity error message
Bruce Momjian [Wed, 28 Dec 2005 23:17:09 +0000 (23:17 +0000)]
Update regression tests for new referential integrity error message
wording ("table" added).

18 years agoFix plperl validator to honor check_function_bodies: when that is OFF,
Tom Lane [Wed, 28 Dec 2005 18:34:16 +0000 (18:34 +0000)]
Fix plperl validator to honor check_function_bodies: when that is OFF,
we want it to check the argument/result data types and no more.  In
particular, libperl shouldn't get initialized in this case.

18 years agoMove plpgsql's fetchArgInfo() into funcapi.c, and rename to
Tom Lane [Wed, 28 Dec 2005 18:11:25 +0000 (18:11 +0000)]
Move plpgsql's fetchArgInfo() into funcapi.c, and rename to
get_func_arg_info() for consistency with other names there.
This code will probably be useful to other PLs when they start to
support OUT parameters, so better to have it in the main backend.
Also, fix plpgsql validator to detect bogus OUT parameters even when
check_function_bodies is off.

18 years agoMention "table" in "violates foreign key constraint" message that was
Bruce Momjian [Wed, 28 Dec 2005 16:47:21 +0000 (16:47 +0000)]
Mention "table" in "violates foreign key constraint" message that was
lacking it.  Perhaps it was suppressed because of line length
considerations, but "table" should appear.

18 years agoRemove silly use of DLLIMPORT.
Tom Lane [Wed, 28 Dec 2005 16:38:38 +0000 (16:38 +0000)]
Remove silly use of DLLIMPORT.

18 years agoAdd COPY CSV test that tests CSV output of \.
Bruce Momjian [Wed, 28 Dec 2005 15:07:02 +0000 (15:07 +0000)]
Add COPY CSV test that tests CSV output of \.

18 years agoUpdate COPY CSV \. wording for externally created files.
Bruce Momjian [Wed, 28 Dec 2005 14:38:32 +0000 (14:38 +0000)]
Update COPY CSV \. wording for externally created files.

18 years agoUpdate:
Bruce Momjian [Wed, 28 Dec 2005 03:59:39 +0000 (03:59 +0000)]
Update:

< * Add missing optimizer selectivities for date, r-tree, etc
> * Add missing rtree optimizer selectivity

18 years agoAdd regression tests for CSV and \., and add automatic quoting of a
Bruce Momjian [Wed, 28 Dec 2005 03:25:32 +0000 (03:25 +0000)]
Add regression tests for CSV and \., and add automatic quoting of a
single column dump that has a \. value, so the load works properly.  I
also added documentation describing this issue.

18 years agoDone:
Bruce Momjian [Wed, 28 Dec 2005 02:12:25 +0000 (02:12 +0000)]
Done:

< * %Make row-wise comparisons work per SQL spec
> * -Make row-wise comparisons work per SQL spec

18 years agoImplement SQL-compliant treatment of row comparisons for < <= > >= cases
Tom Lane [Wed, 28 Dec 2005 01:30:02 +0000 (01:30 +0000)]
Implement SQL-compliant treatment of row comparisons for < <= > >= cases
(previously we only did = and <> correctly).  Also, allow row comparisons
with any operators that are in btree opclasses, not only those with these
specific names.  This gets rid of a whole lot of indefensible assumptions
about the behavior of particular operators based on their names ... though
it's still true that IN and NOT IN expand to "= ANY".  The patch adds a
RowCompareExpr expression node type, and makes some changes in the
representation of ANY/ALL/ROWCOMPARE SubLinks so that they can share code
with RowCompareExpr.

I have not yet done anything about making RowCompareExpr an indexable
operator, but will look at that soon.

initdb forced due to changes in stored rules.

18 years agoIncrease amount of shared buffers initdb tries to allocate to 4000,
Andrew Dunstan [Tue, 27 Dec 2005 23:54:01 +0000 (23:54 +0000)]
Increase amount of shared buffers initdb tries to allocate to 4000,
and add logic to try max_fsm_pages up to 200000, plus accompanying minor
docs changes.

18 years agoOur code had:
Bruce Momjian [Tue, 27 Dec 2005 18:10:48 +0000 (18:10 +0000)]
Our code had:

                if (c == '\\' && cstate->line_buf.len == 0)

The problem with that is the because of the input and _output_
buffering, cstate->line_buf.len could be zero even if we are not on the
first character of a line.  In fact, for a typical line, it is zero for
all characters on the line.  The proper solution is to introduce a
boolean, first_char_in_line, that we set as we enter the loop and clear
once we process a character.

I have restructured the line-reading code in copy.c by:

        o  merging the CSV/non-CSV functions into a single function
        o  used macros to centralize and clarify the buffering code
        o  updated comments
        o  renamed client_encoding_only to encoding_embeds_ascii
        o  added a high-bit test to the encoding_embeds_ascii test for
           performance
        o  in CSV mode, allow a backslash followed by a non-period to
           continue being processed as a data value

There should be no performance impact from this patch because it is
functionally equivalent.  If you apply the patch you will see copy.c is
much clearer in this area now and might suggest additional
optimizations.

I have also attached a 8.1-only patch to fix the CSV \. handling bug
with no code restructuring.

18 years agoProtect ADD and HEADER symbols from conflicting with MIPS includes.
Bruce Momjian [Tue, 27 Dec 2005 04:00:08 +0000 (04:00 +0000)]
Protect ADD and HEADER symbols from conflicting with MIPS includes.

18 years agoMore uses of IS_HIGHBIT_SET() macro.
Bruce Momjian [Mon, 26 Dec 2005 19:30:45 +0000 (19:30 +0000)]
More uses of IS_HIGHBIT_SET() macro.

18 years agoRename pg_make_encrypted_password to PQencryptPassword.
Peter Eisentraut [Mon, 26 Dec 2005 14:58:06 +0000 (14:58 +0000)]
Rename pg_make_encrypted_password to PQencryptPassword.

18 years agoAdd:
Bruce Momjian [Mon, 26 Dec 2005 05:14:52 +0000 (05:14 +0000)]
Add:

* %Make row-wise comparisons work per SQL spec

  Right now, '(a, b) < (1, 2)' is processed as 'a < 1 and b < 2', but
  the SQL standard requires it to be processed as a column-by-column
  comparison, so the proper comparison is '(a < 1) OR (a = 1 AND b < 2)'.

18 years agoVarious cosmetic code cleanup for PL/Python:
Neil Conway [Mon, 26 Dec 2005 04:28:48 +0000 (04:28 +0000)]
Various cosmetic code cleanup for PL/Python:

- use "bool" rather than "int" for boolean variables

- use "PLy_malloc" rather than "malloc" in two places

- define "PLy_strdup", and use it rather than malloc() + strcpy() in
  two places (which should have been memcpy(), anyway).

- remove a bunch of redundant parentheses from expressions that do not
  need the parentheses for code clarity

18 years agoI have added these macros to c.h:
Bruce Momjian [Sun, 25 Dec 2005 02:14:19 +0000 (02:14 +0000)]
I  have added these macros to c.h:

        #define HIGHBIT                 (0x80)
        #define IS_HIGHBIT_SET(ch)      ((unsigned char)(ch) & HIGHBIT)

and removed CSIGNBIT and mapped it uses to HIGHBIT.  I have also added
uses for IS_HIGHBIT_SET where appropriate.  This change is
purely for code clarity.

18 years agoMinor doc tweak: "NOT NULL" is redundant with "SERIAL" in example.
Neil Conway [Sun, 25 Dec 2005 01:41:15 +0000 (01:41 +0000)]
Minor doc tweak: "NOT NULL" is redundant with "SERIAL" in example.

18 years agoSpell fix. Andrew.
Bruce Momjian [Sat, 24 Dec 2005 19:29:38 +0000 (19:29 +0000)]
Spell fix.  Andrew.

18 years agoUpdate why unified diff is _sometimes_ better.
Bruce Momjian [Sat, 24 Dec 2005 18:37:17 +0000 (18:37 +0000)]
Update why unified diff is _sometimes_ better.

18 years agoPrevious commit message should have been:
Bruce Momjian [Sat, 24 Dec 2005 18:23:02 +0000 (18:23 +0000)]
Previous commit message should have been:

Add comment marker for PG_ENCODING_BE_LAST.

18 years agoAdd
Bruce Momjian [Sat, 24 Dec 2005 18:21:34 +0000 (18:21 +0000)]
Add

18 years agoAlignment cleanup.
Bruce Momjian [Sat, 24 Dec 2005 18:11:30 +0000 (18:11 +0000)]
Alignment cleanup.

18 years agoFormatting cleanups.
Bruce Momjian [Sat, 24 Dec 2005 17:19:40 +0000 (17:19 +0000)]
Formatting cleanups.

18 years agoFormatting cleanup.
Bruce Momjian [Sat, 24 Dec 2005 16:49:48 +0000 (16:49 +0000)]
Formatting cleanup.

18 years agoFix long standing Asian multibyte charsets bug.
Tatsuo Ishii [Sat, 24 Dec 2005 09:35:36 +0000 (09:35 +0000)]
Fix long standing Asian multibyte charsets bug.
See:

Subject: [HACKERS] bugs with certain Asian multibyte charsets
From: Tatsuo Ishii <ishii@sraoss.co.jp>
To: pgsql-hackers@postgresql.org
Date: Sat, 24 Dec 2005 18:25:33 +0900 (JST)

for more details/

18 years agoWups, fat-fingered the calculation the first time. Update comment in
Tom Lane [Fri, 23 Dec 2005 23:28:35 +0000 (23:28 +0000)]
Wups, fat-fingered the calculation the first time.  Update comment in
postgresql.conf.sample too.

18 years agoUpdate info about shared memory space calculation to match CVS tip's behavior.
Tom Lane [Fri, 23 Dec 2005 23:02:26 +0000 (23:02 +0000)]
Update info about shared memory space calculation to match CVS tip's behavior.

18 years agoFix make_relative_path() to support cases where target_path and bin_path
Tom Lane [Fri, 23 Dec 2005 22:34:22 +0000 (22:34 +0000)]
Fix make_relative_path() to support cases where target_path and bin_path
differ by more than the last directory component.  Instead of insisting
that they match up to the last component, accept whatever common prefix
they have, and try to replace the non-matching part of bin_path with
the non-matching part of target_path in the actual executable's path.
In one way this is tighter than the old code, because it insists on
a match to the part of bin_path we want to substitute for, rather than
blindly stripping one directory component from the executable's path.
Per gripe from Martin Pitt and subsequent discussion.

18 years agoAllow CREATE/ALTER ROLE PASSWORD NULL to allow restoring the default state
Peter Eisentraut [Fri, 23 Dec 2005 16:46:39 +0000 (16:46 +0000)]
Allow CREATE/ALTER ROLE PASSWORD NULL to allow restoring the default state
of having no password.

18 years agoFix for rearranging encoding id ISO-8859-5 to ISO-8859-8.
Tatsuo Ishii [Fri, 23 Dec 2005 02:11:02 +0000 (02:11 +0000)]
Fix for rearranging encoding id ISO-8859-5 to ISO-8859-8.
Also make the code more robust by searching for target encoding
in the internal charset map.

Problem reported by Sagi Bashari on 2005/12/21.
See "[BUGS] BUG #2120: Crash when doing UTF8<->ISO_8859_8 encoding conversion"
on pgsql-bugs list for more details.

18 years agoAdd an officially exported libpq function to encrypt passwords, and
Tom Lane [Fri, 23 Dec 2005 01:16:38 +0000 (01:16 +0000)]
Add an officially exported libpq function to encrypt passwords, and
modify the previous \password patch to use it instead of depending
on a not-officially-exported function.  Per discussion.

18 years agoAdd quotes around search_path "$user" so that SHOW output can be used in
Bruce Momjian [Fri, 23 Dec 2005 00:38:04 +0000 (00:38 +0000)]
Add quotes around search_path "$user" so that SHOW output can be used in
SET.

18 years agoUpdate item:
Bruce Momjian [Thu, 22 Dec 2005 23:05:32 +0000 (23:05 +0000)]
Update item:

>
>   A more complex solution would be to save multiple plans for different
>   cardinality and use the appropriate plan based on the EXECUTE values.
>

18 years agoAdjust string comparison so that only bitwise-equal strings are considered
Tom Lane [Thu, 22 Dec 2005 22:50:00 +0000 (22:50 +0000)]
Adjust string comparison so that only bitwise-equal strings are considered
equal: if strcoll claims two strings are equal, check it with strcmp, and
sort according to strcmp if not identical.  This fixes inconsistent
behavior under glibc's hu_HU locale, and probably under some other locales
as well.  Also, take advantage of the now-well-defined behavior to speed up
texteq, textne, bpchareq, bpcharne: they may as well just do a bitwise
comparison and not bother with strcoll at all.

NOTE: affected databases may need to REINDEX indexes on text columns to be
sure they are self-consistent.

18 years agoUpdate interval documenation to mention the storage system used.
Bruce Momjian [Thu, 22 Dec 2005 21:45:19 +0000 (21:45 +0000)]
Update interval documenation to mention the storage system used.

18 years agoAdd documentation example of using interval multiplication with 'days'.
Bruce Momjian [Wed, 21 Dec 2005 23:22:55 +0000 (23:22 +0000)]
Add documentation example of using interval multiplication with 'days'.

Backpatch to 8.1.X.

18 years agoAdd additional example for interval multiplication.
Bruce Momjian [Wed, 21 Dec 2005 16:02:24 +0000 (16:02 +0000)]
Add additional example for interval multiplication.

Fix example for day and hours interval subtraction for new computation
method.

Update interval examples to display zero seconds, which is our default.

Backpatch to 8.1.X.

18 years agoMultibyte encodings support for ISpell dictionary
Teodor Sigaev [Wed, 21 Dec 2005 13:05:49 +0000 (13:05 +0000)]
Multibyte encodings support for ISpell dictionary

18 years agoTeach planner how to rearrange join order for some classes of OUTER JOIN.
Tom Lane [Tue, 20 Dec 2005 02:30:36 +0000 (02:30 +0000)]
Teach planner how to rearrange join order for some classes of OUTER JOIN.
Per my recent proposal.  I ended up basing the implementation on the
existing mechanism for enforcing valid join orders of IN joins --- the
rules for valid outer-join orders are somewhat similar.

18 years agoFix broken markup.
Tom Lane [Tue, 20 Dec 2005 00:51:45 +0000 (00:51 +0000)]
Fix broken markup.

18 years agoAdd new psql command \password for changing role password with client-side
Peter Eisentraut [Sun, 18 Dec 2005 02:17:16 +0000 (02:17 +0000)]
Add new psql command \password for changing role password with client-side
password encryption.  Also alter createuser command to the same effect.

18 years agoFix typo.
Bruce Momjian [Sat, 17 Dec 2005 21:08:24 +0000 (21:08 +0000)]
Fix typo.

18 years agoUpdate s_lock.c comments.
Bruce Momjian [Sat, 17 Dec 2005 20:39:16 +0000 (20:39 +0000)]
Update s_lock.c comments.

18 years agoUpdate ASM comments.
Bruce Momjian [Sat, 17 Dec 2005 20:15:43 +0000 (20:15 +0000)]
Update ASM comments.

18 years agoRemove item:
Bruce Momjian [Sat, 17 Dec 2005 19:03:24 +0000 (19:03 +0000)]
Remove item:

< * Allow star join optimizations
<
<   While our bitmap scan allows multiple indexes to be joined to get
<   to heap rows, a star joins allows multiple dimension _tables_ to
<   be joined to index into a larger main fact table.  The join is
<   usually performed by either creating a cartesian product of all
<   the dimmension tables and doing a single join on that product or
<   using subselects to create bitmaps of each dimmension table match
<   and merge the bitmaps to perform the join on the fact table.  Some
<   of these algorithms might be patented.

18 years agoUpdate:
Bruce Momjian [Sat, 17 Dec 2005 18:07:01 +0000 (18:07 +0000)]
Update:

< * Flush cached query plans when the dependent objects change or
<   when the cardinality of parameters changes dramatically
> * Flush cached query plans when the dependent objects change,
>   when the cardinality of parameters changes dramatically, or
>   when new ANALYZE statistics are available

18 years agoAdd mention of possible patent problems with star joins, per Joshua
Bruce Momjian [Sat, 17 Dec 2005 17:04:41 +0000 (17:04 +0000)]
Add mention of possible patent problems with star joins, per Joshua
Drake:

<   and merge the bitmaps to perform the join on the fact table.
>   and merge the bitmaps to perform the join on the fact table.  Some
>   of these algorithms might be patented.

18 years agoAdd:
Bruce Momjian [Sat, 17 Dec 2005 16:43:11 +0000 (16:43 +0000)]
Add:

* Allow star join optimizations

  While our bitmap scan allows multiple indexes to be joined to get
  to heap rows, a star joins allows multiple dimension _tables_ to
  be joined to index into a larger main fact table.  The join is
  usually performed by either creating a cartesian product of all
  the dimmension tables and doing a single join on that product or
  using subselects to create bitmaps of each dimmension table match
  and merge the bitmaps to perform the join on the fact table.

18 years agoAdd more function to TODO:
Bruce Momjian [Sat, 17 Dec 2005 14:18:33 +0000 (14:18 +0000)]
Add more function to TODO:

< * %Add pg_get_acldef(), pg_get_typedefault(), and pg_get_attrdef()
> * %Add pg_get_acldef(), pg_get_typedefault(), pg_get_attrdef(),
>   pg_get_tabledef(), pg_get_domaindef(), pg_get_functiondef()

18 years agoUpdate:
Bruce Momjian [Sat, 17 Dec 2005 04:10:47 +0000 (04:10 +0000)]
Update:

< * Flush cached query plans when the dependent objects change
> * Flush cached query plans when the dependent objects change or
>   when the cardinality of parameters changes dramatically

18 years agoUpdate FAQ, second version.
Bruce Momjian [Sat, 17 Dec 2005 03:21:00 +0000 (03:21 +0000)]
Update FAQ, second version.

J.Kuwamura

18 years agoUpdate Japanese FAQ.
Bruce Momjian [Sat, 17 Dec 2005 03:20:12 +0000 (03:20 +0000)]
Update Japanese FAQ.

J.Kuwamura

18 years agoAdd:
Bruce Momjian [Sat, 17 Dec 2005 01:48:37 +0000 (01:48 +0000)]
Add:

> * Allow the count returned by SELECT, etc to be to represent as an int64
>   to allow a higher range of values

18 years agoAdd URL for Solaris qsort() bug.
Bruce Momjian [Sat, 17 Dec 2005 00:35:50 +0000 (00:35 +0000)]
Add URL for Solaris qsort() bug.

18 years agoUpdate doc mention that Solaris qsort() might be fixed.
Bruce Momjian [Sat, 17 Dec 2005 00:33:33 +0000 (00:33 +0000)]
Update doc mention that Solaris qsort() might be fixed.

18 years agoAdd a note to Win32 gettimeofday() emulation, per Qingqing Zhou.
Alvaro Herrera [Fri, 16 Dec 2005 21:55:27 +0000 (21:55 +0000)]
Add a note to Win32 gettimeofday() emulation, per Qingqing Zhou.

18 years agoForce update.
Bruce Momjian [Fri, 16 Dec 2005 18:56:55 +0000 (18:56 +0000)]
Force update.

18 years agoFix some typos in the TODO list.
Neil Conway [Fri, 16 Dec 2005 18:44:07 +0000 (18:44 +0000)]
Fix some typos in the TODO list.

18 years agoRethink prior patch to filter out dead backend entries from the pgstats
Tom Lane [Fri, 16 Dec 2005 04:03:40 +0000 (04:03 +0000)]
Rethink prior patch to filter out dead backend entries from the pgstats
file.  The original code probed the PGPROC array separately for each PID,
which was not good for large numbers of backends: not only is the runtime
O(N^2) but most of it is spent holding ProcArrayLock.  Instead, take the
lock just once and copy the active PIDs into an array, then use qsort
and bsearch so that the lookup time is more like O(N log N).

18 years agoDefend against crash while processing Describe Statement or Describe Portal
Tom Lane [Wed, 14 Dec 2005 17:06:28 +0000 (17:06 +0000)]
Defend against crash while processing Describe Statement or Describe Portal
messages, when client attempts to execute these outside a transaction (start
one) or in a failed transaction (reject message, except for COMMIT/ROLLBACK
statements which we can handle).  Per report from Francisco Figueiredo Jr.

18 years agoFix problem with whole-row Vars referencing sub-select outputs, per
Tom Lane [Wed, 14 Dec 2005 16:28:32 +0000 (16:28 +0000)]
Fix problem with whole-row Vars referencing sub-select outputs, per
example from Jim Dew.  Add some simple regression tests, since this is
an area we seem to break regularly :-(

18 years agoUse a proper enum for tri-valued variables.
Bruce Momjian [Mon, 12 Dec 2005 15:48:04 +0000 (15:48 +0000)]
Use a proper enum for tri-valued variables.

18 years agoDocument tri-valued variables in createdb, +1, -1, 0.
Bruce Momjian [Mon, 12 Dec 2005 15:41:52 +0000 (15:41 +0000)]
Document tri-valued variables in createdb, +1, -1, 0.

18 years agoUpdate wording, query -> statement:
Bruce Momjian [Mon, 12 Dec 2005 13:36:36 +0000 (13:36 +0000)]
Update wording, query -> statement:

< * %Allow pooled connections to list all prepared queries
> * %Allow pooled connections to list all prepared statements
28c28
<   the queries prepared in the current session.
>   the statements prepared in the current session.
143c143
<    o Allow a warm standby system to also allow read-only queries
>    o Allow a warm standby system to also allow read-only statements
404c404
< * Add GUC to issue notice about queries that use unjoined tables
> * Add GUC to issue notice about statements that use unjoined tables
490c490
<    Another idea would be to allow actual SELECT queries in a COPY.
>    Another idea would be to allow actual SELECT statements in a COPY.
554c554
<  o Allow function argument names to be queries from PL/PgSQL
>  o Allow function argument names to be statements from PL/PgSQL
591c591
<  o Improve psql's handling of multi-line queries
>  o Improve psql's handling of multi-line statements
<    Currently, while \e saves a single query as one entry, interactive
<    queries are saved one line at a time.  Ideally all queries
>    Currently, while \e saves a single statement as one entry, interactive
>    statements are saved one line at a time.  Ideally all statements
665c665
<  o Allow query results to be automatically batched to the client
>  o Allow statement results to be automatically batched to the client
667c667
<    Currently, all query results are transfered to the libpq
>    Currently, all statement results are transfered to the libpq
672c672
<    One complexity is that a query like SELECT 1/col could error
>    One complexity is that a statement like SELECT 1/col could error
739c739
< * Allow queries across databases or servers with transaction
> * Allow statements across databases or servers with transaction
<   inheritance, allow it to work for UPDATE and DELETE queries, and allow
<   it to be used for all queries with little performance impact
>   inheritance, allow it to work for UPDATE and DELETE statements, and allow
>   it to be used for all statements with little performance impact
876c876
< * Consider automatic caching of queries at various levels:
> * Consider automatic caching of statements at various levels:
947c947
<   a single session using multiple threads to execute a query faster.
>   a single session using multiple threads to execute a statement faster.
1025c1025
< * Log queries where the optimizer row estimates were dramatically
> * Log statements where the optimizer row estimates were dramatically
1146c1146
<    of result sets using new query protocol
>    of result sets using new statement protocol

18 years agoImprove support of multibyte encoding:
Teodor Sigaev [Mon, 12 Dec 2005 11:10:12 +0000 (11:10 +0000)]
Improve support of multibyte encoding:
- tsvector_(in|out)
- tsquery_(in|out)
- to_tsvector
- to_tsquery, plainto_tsquery
- 'simple' dictionary

18 years agoDivide the lock manager's shared state into 'partitions', so as to
Tom Lane [Sun, 11 Dec 2005 21:02:18 +0000 (21:02 +0000)]
Divide the lock manager's shared state into 'partitions', so as to
reduce contention for the former single LockMgrLock.  Per my recent
proposal.  I set it up for 16 partitions, but on a pgbench test this
gives only a marginal further improvement over 4 partitions --- we need
to test more scenarios to choose the number of partitions.

18 years agoImplement IS NOT DISTINCT FROM, update the regression tests and docs.
Neil Conway [Sun, 11 Dec 2005 10:54:28 +0000 (10:54 +0000)]
Implement IS NOT DISTINCT FROM, update the regression tests and docs.
Patch from Pavel Stehule, minor fixups by myself.

18 years agoRemove item:
Bruce Momjian [Sun, 11 Dec 2005 02:16:25 +0000 (02:16 +0000)]
Remove item:

< * Add GUC variable to run a command on database panic or
<   smart/fast/immediate shutdown