OSDN Git Service
Bruce Momjian [Tue, 4 Jul 2000 04:45:22 +0000 (04:45 +0000)]
Update TODO list.
Tom Lane [Tue, 4 Jul 2000 02:40:56 +0000 (02:40 +0000)]
DataFill() has no business resetting xact status bits
in the infomask of the provided tuple.
Vadim B. Mikheev [Tue, 4 Jul 2000 01:49:44 +0000 (01:49 +0000)]
misc
Vadim B. Mikheev [Tue, 4 Jul 2000 01:39:24 +0000 (01:39 +0000)]
unlock buffer before releasing in heap_insert
+unlock buffer in heap_fetch under !ItemIdIsUsed().
Tom Lane [Tue, 4 Jul 2000 00:04:03 +0000 (00:04 +0000)]
tuptoaster.h should surely not #include postgres.h.
Tom Lane [Mon, 3 Jul 2000 23:58:32 +0000 (23:58 +0000)]
Add missing #include for TOAST.
Jan Wieck [Mon, 3 Jul 2000 23:19:04 +0000 (23:19 +0000)]
TOAST changes to catalog
Jan
Jan Wieck [Mon, 3 Jul 2000 23:10:14 +0000 (23:10 +0000)]
TOAST
WARNING: This is actually broken - we have self-deadlocks
due to concurrent changes in buffer management.
Vadim and me are working on it.
Jan
Peter Eisentraut [Mon, 3 Jul 2000 20:48:46 +0000 (20:48 +0000)]
Move global internal files to PGDATA/global.
Peter Eisentraut [Mon, 3 Jul 2000 20:46:10 +0000 (20:46 +0000)]
Move PGPORT envar handling to ResetAllOptions(). Improve long options
parsing to not clobber the optarg string -- so that we can bring
SetOptsFile() up to speed.
Peter Eisentraut [Mon, 3 Jul 2000 16:35:39 +0000 (16:35 +0000)]
Have pg_dumpall write CREATE USER and CREATE GROUP commands instead of raw
COPYs of pg_shadow and pg_group.
It also turns out that pg_dumpall was all but broken for multiple servers
running at non-standard port numbers. You might get the users and groups
from one server and the databases from another. Fixed that.
A little user interface and code cleanup along with that. This also takes
care of the portability bug discussed in "[BUGS] pg_dumpall" in March 2000.
Bruce Momjian [Mon, 3 Jul 2000 16:03:27 +0000 (16:03 +0000)]
Very small changes in the contrib's pg_dumplo in the attache.
And:
Note, Bruce I found in the contrib tree any files that we forget
remove during contrib cleaning. Please remove these files:
contrib/lo/test.sql
contrib/pg_dumplo/Makefile.out
contrib/pgbench/pgbench_jis.doc
contrib/spi/new_example.example
contrib/spi/README.MAX
Thanks.
Karel
Bruce Momjian [Mon, 3 Jul 2000 16:01:33 +0000 (16:01 +0000)]
In the attache is new (correct) version. I add TZ (timezone) support
for
to_char() too.
Karel
Hiroshi Inoue [Mon, 3 Jul 2000 04:45:09 +0000 (04:45 +0000)]
Add a check of HEAP_MARKED_FOR_UPDATE in HeapTupleSatisfiesDirty().
This enables unique constraint check in case of SELECT FOR UPDATEd.
Tom Lane [Mon, 3 Jul 2000 03:57:03 +0000 (03:57 +0000)]
Clean up memory-context stuff, other minor infelicities.
Tom Lane [Mon, 3 Jul 2000 03:27:31 +0000 (03:27 +0000)]
Remove unused variable. (Hey Peter, you use gcc, why didn't you
notice this warning?)
Vadim B. Mikheev [Mon, 3 Jul 2000 02:54:21 +0000 (02:54 +0000)]
heap' logging
Bruce Momjian [Sun, 2 Jul 2000 22:01:27 +0000 (22:01 +0000)]
Attached is a new patch which addresses this problem. (oids in
regression tests).
Chris Bitmead
Peter Eisentraut [Sun, 2 Jul 2000 15:21:27 +0000 (15:21 +0000)]
Version number now set in configure, available through Makefile.global
and config.h. Adjusted all referring code.
Scrapped pg_version and changed initdb accordingly. Integrated
src/utils/version.c into src/backend/utils/init/miscinit.c. Changed all
callers.
Set version number to `7.1devel'. (Non-numeric version suffixes now allowed.)
Tom Lane [Sun, 2 Jul 2000 05:38:40 +0000 (05:38 +0000)]
Don't run in CacheMemoryContext more than necessary.
Tom Lane [Sun, 2 Jul 2000 04:46:09 +0000 (04:46 +0000)]
Don't run in CacheMemoryContext longer than necessary.
Tom Lane [Sun, 2 Jul 2000 04:04:09 +0000 (04:04 +0000)]
Remove memory leak from VACUUM parsing.
Tom Lane [Sun, 2 Jul 2000 02:28:38 +0000 (02:28 +0000)]
New memmgr logic in xact.c failed if AbortTransaction() is called when
there is no open transaction.
Bruce Momjian [Sat, 1 Jul 2000 21:27:14 +0000 (21:27 +0000)]
Back out Karels to_char changes.
Peter Eisentraut [Sat, 1 Jul 2000 21:16:44 +0000 (21:16 +0000)]
We don't need config.status substituting anything into Gen_fmgrtab.sh, we
just pass in CPP and AWK via the environment from the Makefile.
Peter Eisentraut [Sat, 1 Jul 2000 15:05:47 +0000 (15:05 +0000)]
Synchronized integrated documentation with latest admin guide change.
Peter Eisentraut [Sat, 1 Jul 2000 15:02:31 +0000 (15:02 +0000)]
A little bit of clean-up/bug fix in Tcl build. Handle gracefully cases
where someone wants to build these but hasn't configured for Tcl.
Bruce Momjian [Sat, 1 Jul 2000 14:10:32 +0000 (14:10 +0000)]
In the attache is new to_date()/to_timestamp() version.
Now the to_timestamp() support WW,W,J,SSSS,DDD conversion from strings and
the am/pm bug is fixed, the to_char() use week-of-year (WW) full compatible
with Oracle.
This patch update relevant regress-tests and docs too.
Karel
~
~
Bruce Momjian [Fri, 30 Jun 2000 18:03:40 +0000 (18:03 +0000)]
Fix select '1--2'; for PEter.
Peter Eisentraut [Fri, 30 Jun 2000 16:14:21 +0000 (16:14 +0000)]
New and revised material for Admin guide, re backup & restore and database
management
Peter Eisentraut [Fri, 30 Jun 2000 16:11:02 +0000 (16:11 +0000)]
Fixes for Solaris/cc suggested by <pgsql-hackers@thewrittenword.com>
Don't use DISABLE_COMPLEX_MACRO on Solaris. Don't define the
replacement function in the header file. Use -KPIC, not -K PIC.
Use CC to link C++ libraries, not ld/ar.
Eliminate file not found warnings in tcl build code.
Tom Lane [Fri, 30 Jun 2000 07:06:05 +0000 (07:06 +0000)]
RemoveView was doing things the hard way...
Tom Lane [Fri, 30 Jun 2000 07:04:23 +0000 (07:04 +0000)]
Use a private memory context to store rule information in each relcache
entry that has rules. This allows us to release the rule parsetrees
on relcache flush without needing a working freeObject() routine.
Formerly, the rule trees were leaked permanently at relcache flush.
Also, clean up handling of rule creation and deletion --- there was
not sufficient locking of the relation being modified, and there was
no reliable notification of other backends that a relcache reload
was needed. Also, clean up relcache.c code so that scans of system
tables needed to load a relcache entry are done in the caller's
memory context, not in CacheMemoryContext. This prevents any
un-pfreed memory from those scans from becoming a permanent memory
leak.
Bruce Momjian [Thu, 29 Jun 2000 16:27:57 +0000 (16:27 +0000)]
Fix quotes in /* */ comments in psql.
Tom Lane [Thu, 29 Jun 2000 07:35:57 +0000 (07:35 +0000)]
Add test code to copy all parse/plan trees. Repair essential omissions
in copyfuncs and equalfuncs exposed by regression tests. We still have
some work to do: these modules really ought to handle most or all of
the utility statement node types. But it's better than it was.
Tom Lane [Thu, 29 Jun 2000 05:50:29 +0000 (05:50 +0000)]
Rearrange config.h entries into a more sensible order, and add some
documentation. Let's try to keep this file a bit neater in future,
hmm? Also (to get back to the original point) update info about
FUNC_MAX_ARGS, and add additional config symbols for debugging
new memory management changes.
Tom Lane [Thu, 29 Jun 2000 02:26:23 +0000 (02:26 +0000)]
Remove freefuncs.c, which hasn't been used in a long time and is not
worth the effort to continue to maintain. Since freeObject() is not
capable of coping with cases like multiple links to a node, it's
unlikely that it ever will be useful again. We now have memory
context management that offers a faster and more reliable way of
getting rid of arbitrary node trees (at the cost of having to know
in advance of building the tree that you'll want to get rid of it).
Tom Lane [Thu, 29 Jun 2000 02:17:42 +0000 (02:17 +0000)]
Add SIGFPE to list of signals never to block. #ifdef all signals in
that list to cope with platforms that don't have them all.
Bruce Momjian [Thu, 29 Jun 2000 01:19:36 +0000 (01:19 +0000)]
Add #ifdef for LC_MESSAGES, because it is not ANSI C standard, and BSDI
does not have it.
Peter Eisentraut [Wed, 28 Jun 2000 18:30:16 +0000 (18:30 +0000)]
Makefile cleanup for interface tree. Now essentially with all the
standard targets and behaviour. Replaced Makefile.in's with
Makefile's and declared the respective variables in Makefile.global.
maintainer-clean target now available at top level, although it does
not work in the backend tree yet.
Cleanup pass over Makefile.shlib, renamed some targets and variables.
The shared library symlink tests are now done by make, not the shell.
ecpg: Remove one warning in sloppy flex output.
PL/Perl and Perl interface: the MakeMaker documentation is confusing,
the realclean target *does* "delete derived files", but it also
uninstalls them. Don't use that.
The submake targets in the various bin directories that update libpq
should `make all', not `make libpq.a'. That is a) unportable, and
b) doesn't build the shared library.
Tom Lane [Wed, 28 Jun 2000 06:05:36 +0000 (06:05 +0000)]
Suppress memory leak in COPY WITH OIDS TO wherever.
Tom Lane [Wed, 28 Jun 2000 05:09:37 +0000 (05:09 +0000)]
Repair incorrect assumption that all versions of 'install' know
how to install multiple files in one invocation.
Tom Lane [Wed, 28 Jun 2000 04:58:36 +0000 (04:58 +0000)]
Call me silly, but I think it would be a good idea if initdb and other
scripts are executable after being installed. Latest changes got the
permissions wrong (non executable).
Tom Lane [Wed, 28 Jun 2000 03:33:33 +0000 (03:33 +0000)]
First phase of memory management rewrite (see backend/utils/mmgr/README
for details). It doesn't really do that much yet, since there are no
short-term memory contexts in the executor, but the infrastructure is
in place and long-term contexts are handled reasonably. A few long-
standing bugs have been fixed, such as 'VACUUM; anything' in a single
query string crashing. Also, out-of-memory is now considered a
recoverable ERROR, not FATAL.
Eliminate a large amount of crufty, now-dead code in and around
memory management.
Fix problem with holding off SIGTRAP, SIGSEGV, etc in postmaster and
backend startup.
Bruce Momjian [Tue, 27 Jun 2000 18:08:49 +0000 (18:08 +0000)]
Tamotsu Nakagawa has posted a fix for this to a local mail list in
Japan. Can someone comment on this? According to him, with the patch
now only the geometry test fails.
Tatsuo Ishii
Peter Eisentraut [Tue, 27 Jun 2000 00:32:06 +0000 (00:32 +0000)]
Makefile cleanup for bin and pl subtrees. They should now support
all the standard semantics. Also get rid of Makefile.in's on the
way and instead declare all variables in Makefile.global.
Bruce Momjian [Mon, 26 Jun 2000 14:16:34 +0000 (14:16 +0000)]
Fix \pset null to use ', not ".
Peter Eisentraut [Sun, 25 Jun 2000 14:25:51 +0000 (14:25 +0000)]
Added new commands and intelligence to psql tab completion.
Peter Eisentraut [Sun, 25 Jun 2000 14:24:59 +0000 (14:24 +0000)]
Relaxed pickiness about user and group commands in transaction blocks.
Bruce Momjian [Sat, 24 Jun 2000 01:51:14 +0000 (01:51 +0000)]
Update FAQ.
Peter Eisentraut [Thu, 22 Jun 2000 22:31:24 +0000 (22:31 +0000)]
Second pass over run-time configuration system. Adjust priorities on some
option settings. Sort out SIGHUP vs BACKEND -- there is no total ordering
here, so make explicit checks. Add comments explaining all of this.
Removed permissions check on SHOW command.
Add examine_subclass to the game, rename to SQL_inheritance to fit the
official data model better. Adjust documentation.
Standalone backend needs to reset all options before it starts. To
facilitate that, have IsUnderPostmaster be set by the postmaster itself,
don't wait for the magic -p switch.
Also make sure that all environment variables and argv's survive
init_ps_display(). Use strdup where necessary.
Have initdb make configuration files (postgresql.conf, pg_hba.conf) mode
0600 -- having configuration files is no fun if you can't edit them.
Peter Eisentraut [Wed, 21 Jun 2000 22:42:32 +0000 (22:42 +0000)]
Fix typo
Bruce Momjian [Tue, 20 Jun 2000 18:04:18 +0000 (18:04 +0000)]
> Great! Please cc: the docs or hackers mailing lists to make sure the
> patches are not lost...
Aggregate doc patches:
The patches are attached. Be great if you could check them over to make
sure all relevant content (and markup) is there...
Isaac Wilcox
Peter Eisentraut [Tue, 20 Jun 2000 16:40:19 +0000 (16:40 +0000)]
Cleaned up PL/pgSQL build. Fixed a couple of copyandpaste'os in the
interfaces and interfaces/odbc make files. Adjusted regression test
driver to start building and installing in the top level directory.
Tom Lane [Tue, 20 Jun 2000 06:41:13 +0000 (06:41 +0000)]
Make renaming a temp table behave sensibly. We don't need to touch
the underlying table at all, just change the mapping entry ... but
that logic was missing.
Tom Lane [Tue, 20 Jun 2000 04:22:21 +0000 (04:22 +0000)]
Make inheritance planning logic a little simpler and clearer,
hopefully even a little faster.
Tom Lane [Tue, 20 Jun 2000 01:41:22 +0000 (01:41 +0000)]
Fix handling of type tuple associated with a temp relation. We have
to apply the tempname->realname mapping to type name lookup as well
as relation name lookup, else the type tuple will not be found when
wanted. This fixes bugs like this one:
create temp table foo (f1 int);
select foo.f2 from foo;
ERROR: Unable to locate type name 'foo' in catalog
Tom Lane [Mon, 19 Jun 2000 23:40:48 +0000 (23:40 +0000)]
Modify index-opening code to guarantee that the indexes of a relation
are opened in a consistent order by different backends (I ordered them
by index OID because that's easy, but any other consistent order would
do as well). This avoids potential deadlock for index types that we
acquire exclusive locks on ... ie, rtree.
Tom Lane [Mon, 19 Jun 2000 23:37:08 +0000 (23:37 +0000)]
Ensure that kernel error code is included in smgr-level error reports.
Tweak mdcreate a little bit so that it returns the right errno.
Peter Eisentraut [Mon, 19 Jun 2000 16:58:48 +0000 (16:58 +0000)]
Integrated ODBC driver into regular build. No more standalone business,
which didn't work anyway. Had to back out unixodbc related patch which
broke the regular ODBC build.
Bruce Momjian [Mon, 19 Jun 2000 14:07:37 +0000 (14:07 +0000)]
Remove old README files.
Bruce Momjian [Mon, 19 Jun 2000 14:02:16 +0000 (14:02 +0000)]
Add missing /contrib files
Bruce Momjian [Mon, 19 Jun 2000 13:54:50 +0000 (13:54 +0000)]
Update to /contrib from Karel.
Bruce Momjian [Mon, 19 Jun 2000 13:40:37 +0000 (13:40 +0000)]
autoconf
Bruce Momjian [Mon, 19 Jun 2000 13:39:03 +0000 (13:39 +0000)]
Rename HAVE_SQLGETPRIVATEPROFILESTRING to something resonable in length
and readability.
Bruce Momjian [Mon, 19 Jun 2000 13:35:30 +0000 (13:35 +0000)]
fix HAVE_SQLGETPRIVATEPROFILESTRING define
Tom Lane [Mon, 19 Jun 2000 03:55:01 +0000 (03:55 +0000)]
Clean up bogosities in pg_opclass, pg_amop, pg_amproc. There are amproc
entries now for int8 and network hash indexes. int24_ops and int42_ops
are gone. pg_opclass no longer contains multiple entries claiming to be
the default opclass for the same datatype. opr_sanity regress test
extended to catch errors like these in the future.
Bruce Momjian [Mon, 19 Jun 2000 00:51:23 +0000 (00:51 +0000)]
Linux/alpha patch.
Tom Lane [Sun, 18 Jun 2000 22:44:35 +0000 (22:44 +0000)]
Reimplement nodeMaterial to use a temporary BufFile (or even memory, if the
materialized tupleset is small enough) instead of a temporary relation.
This was something I was thinking of doing anyway for performance, and Jan
says he needs it for TOAST because he doesn't want to cope with toasting
noname relations. With this change, the 'noname table' support in heap.c
is dead code, and I have accordingly removed it. Also clean up 'noname'
plan handling in planner --- nonames are either sort or materialize plans,
and it seems less confusing to handle them separately under those names.
Peter Eisentraut [Sun, 18 Jun 2000 21:24:54 +0000 (21:24 +0000)]
Separated set constraints and set transaction reference pages, revised set
reference page to new configuration system. Big update to administrator's
guide, chapters Runtime environment, Client authentication, and User
management, the latter two were part of the old Security chapter.
Tom Lane [Sun, 18 Jun 2000 03:11:40 +0000 (03:11 +0000)]
Reinstate BufFileTell().
Tom Lane [Sat, 17 Jun 2000 23:41:51 +0000 (23:41 +0000)]
Get rid of IndexIsUniqueNoCache() kluge by the simple expedient of
passing the index-is-unique flag to index build routines (duh! ...
why wasn't it done this way to begin with?). Aside from eliminating
an eyesore, this should save a few milliseconds in btree index creation
because a full scan of pg_index is not needed any more.
Tom Lane [Sat, 17 Jun 2000 21:49:04 +0000 (21:49 +0000)]
Fix performance problems with pg_index lookups (see, for example,
discussion of 5/19/00). pg_index is now searched for indexes of a
relation using an indexscan. Moreover, this is done once and cached
in the relcache entry for the relation, in the form of a list of OIDs
for the indexes. This list is used by the parser and executor to drive
lookups in the pg_index syscache when they want to know the properties
of the indexes. Net result: index information will be fully cached
for repetitive operations such as inserts.
Tom Lane [Sat, 17 Jun 2000 04:56:39 +0000 (04:56 +0000)]
Clean out another pocket of functions called via nonspecific function
pointers, namely the catcache tuple fetch routines. Also get rid of
the unused and possibly confusing 'size' field in struct cachedesc.
Since it doesn't allow for variable-length fields, anyone who
actually trusted it would likely be making a mistake...
Peter Eisentraut [Sat, 17 Jun 2000 00:10:40 +0000 (00:10 +0000)]
Remove fmgrstamp-h business -- not needed and confusing
Add options to configure to automatically build for Kerberos
support; no more editing of make files.
Bruce Momjian [Fri, 16 Jun 2000 18:59:30 +0000 (18:59 +0000)]
Cleanup README Makefile installs.
Tom Lane [Fri, 16 Jun 2000 05:27:03 +0000 (05:27 +0000)]
Repair unforgivably brain-dead representation of CaseExpr nodes in
stored rules. Bump catversion to reflect incompatibility.
Bruce Momjian [Thu, 15 Jun 2000 19:08:37 +0000 (19:08 +0000)]
Remove old doc files, now README.
Bruce Momjian [Thu, 15 Jun 2000 19:05:22 +0000 (19:05 +0000)]
Add missing /contrib files to CVS.
Bruce Momjian [Thu, 15 Jun 2000 18:55:34 +0000 (18:55 +0000)]
/contrib patch from Karel.
Bruce Momjian [Thu, 15 Jun 2000 13:54:51 +0000 (13:54 +0000)]
Update TODO list.
Tom Lane [Thu, 15 Jun 2000 06:07:34 +0000 (06:07 +0000)]
Repair Large Object bugs demonstrated by Ian Grant's example. inv_write
was inappropriately relying on rel->rd_nblocks to tell if the LO is
empty (apparently a hack to get around a long-dead index bug), causing
misbehavior on a written-but-never-vacuumed LO. Also, inv_read failed
to cope gracefully with 'holes' (unwritten regions) in the object.
Bruce Momjian [Thu, 15 Jun 2000 04:10:30 +0000 (04:10 +0000)]
Final #include cleanup.
Bruce Momjian [Thu, 15 Jun 2000 03:33:12 +0000 (03:33 +0000)]
Clean up #include's.
Bruce Momjian [Thu, 15 Jun 2000 00:52:26 +0000 (00:52 +0000)]
#include cleanups
Bruce Momjian [Wed, 14 Jun 2000 22:33:06 +0000 (22:33 +0000)]
Update include cleaning scripts.
Bruce Momjian [Wed, 14 Jun 2000 22:28:16 +0000 (22:28 +0000)]
Update include scripts
Bruce Momjian [Wed, 14 Jun 2000 18:59:42 +0000 (18:59 +0000)]
Add () around define -1.
Peter Eisentraut [Wed, 14 Jun 2000 18:18:01 +0000 (18:18 +0000)]
Big warnings cleanup for Solaris/GCC. Down to about 40 now, but
we'll get there one day.
Use `cat' to create aclocal.m4, not `aclocal'. Some people don't
have automake installed.
Only run the autoconf rule in the top-level GNUmakefile if the
invoker specified `make configure', don't run it automatically
because of CVS timestamp skew.
Bruce Momjian [Wed, 14 Jun 2000 17:08:57 +0000 (17:08 +0000)]
autoconf
Bruce Momjian [Wed, 14 Jun 2000 17:07:31 +0000 (17:07 +0000)]
Back out libpgtcl tcl/tksh includes.
Bruce Momjian [Wed, 14 Jun 2000 17:01:35 +0000 (17:01 +0000)]
Remove -O0 linux ppc flag
Thomas G. Lockhart [Wed, 14 Jun 2000 13:12:52 +0000 (13:12 +0000)]
Repair markup to allow clean doc builds.
Thomas G. Lockhart [Wed, 14 Jun 2000 13:12:17 +0000 (13:12 +0000)]
Reword to reduce redundant wording in some wording ;)
Thomas G. Lockhart [Wed, 14 Jun 2000 13:10:48 +0000 (13:10 +0000)]
Fix up typos.
Move "query result" tables to left side of page (remove leading spaces).
Tom Lane [Wed, 14 Jun 2000 05:24:50 +0000 (05:24 +0000)]
I had overlooked the fact that some fmgr-callable functions return void
--- ie, they're only called for side-effects. Add a PG_RETURN_VOID()
macro and use it where appropriate. This probably doesn't change the
machine code by a single bit ... it's just for documentation.
Bruce Momjian [Wed, 14 Jun 2000 04:53:44 +0000 (04:53 +0000)]
Fix for memory leak from Denis Perchine
Bruce Momjian [Wed, 14 Jun 2000 03:33:49 +0000 (03:33 +0000)]
Add drop detail item.
Bruce Momjian [Wed, 14 Jun 2000 03:30:49 +0000 (03:30 +0000)]
Add type conversion TODO.detail
Bruce Momjian [Wed, 14 Jun 2000 03:19:24 +0000 (03:19 +0000)]
> If read or write fails. Position will left the same. This
> situation is already tracked in File routines, but a little bit
> incorrectly.
> After small survey in Linux kernel code, I am not sure about
> it. New patch set pos to unknown in the case of read/write
> fails. And do lseek again.
> Here is the full patch for this. This patch reduce amount of
> lseek call ten ti mes for update statement and twenty times for
> select statement. I tested joined up date and count(*) select
> for table with rows > 170000 and 10 indices. I think this is
> worse of trying. Before lseek calls account for more than 5% o
> f time. Now they are 0.89 and 0.15 respectevly.
>
> Due to only one file modification patch should be applied in
> src/backedn/stora ge/file/ dir.
-- Sincerely Yours,
Denis Perchine