Bruce Momjian [Wed, 11 May 2005 18:05:37 +0000 (18:05 +0000)]
This patch will ensure that the hash table iteration performed by
AtCommit_Portals is restarted when a portal is deleted. This is
necessary since the deletion of a portal may cause the deletion of
another which on rare occations may cause the iterator to return a
deleted portal an thus a renewed attempt delete.
Neil Conway [Wed, 11 May 2005 06:24:55 +0000 (06:24 +0000)]
This patch refactors away some duplicated code in the index AM build
methods: they all invoke UpdateStats() since they have computed the
number of heap tuples, so I created a function in catalog/index.c that
each AM now calls.
Neil Conway [Wed, 11 May 2005 01:41:41 +0000 (01:41 +0000)]
This patch reduces the size of the message header used by statistics
collector messages, per recent discussion on pgsql-patches. This
actually required quite a few changes -- for example,
"databaseid != InvalidOid" was used to check whether a slot in the
backend entry table was initialized, but that no longer works since
the slot might be initialized prior to receiving the BESTART message
which contains the database id. We now use procpid > 0 to indicate
that a slot is non-empty.
Other changes:
- various comment improvements and cleanups
- there's no need to zero-out the entire activity buffer in
pgstat_add_backend(), we can just set activity[0] to '\0'.
- remove the counting of the # of connections to a database; this
was not used anywhere
One change in behavior I wasn't sure about: previously, the code
would create a hash table entry for a database as soon as any message
was received whose header referenced that database. Now, we only
create hash table entries as needed (so for example BESTART won't
create a database hash table entry, since it doesn't need to
access anything in the per-db hash table). It would be easy enough
to retain the old behavior, but AFAICS it is not required.
Neil Conway [Wed, 11 May 2005 01:26:02 +0000 (01:26 +0000)]
Code cleanup: in C89, there is no point casting the first argument to
memset() or MemSet() to a char *. For one, memset()'s first argument is
a void *, and further void * can be implicitly coerced to/from any other
pointer type.
Bruce Momjian [Tue, 10 May 2005 02:16:15 +0000 (02:16 +0000)]
Done:
< * Add session start time and last statement time to pg_stat_activity
> * -Add session start time and last statement time to pg_stat_activity 134c134
< * Add the client IP address and port to pg_stat_activity
> * -Add the client IP address and port to pg_stat_activity
Neil Conway [Mon, 9 May 2005 11:31:34 +0000 (11:31 +0000)]
Complete the following TODO items:
* Add session start time to pg_stat_activity
* Add the client IP address and port to pg_stat_activity
Original patch from Magnus Hagander, code review by Neil Conway. Catalog
version bumped. This patch sends the client IP address and port number in
every statistics message; that's not ideal, but will be fixed up shortly.
Tom Lane [Sat, 7 May 2005 21:32:24 +0000 (21:32 +0000)]
Repair very-low-probability race condition between relation extension
and VACUUM: in the interval between adding a new page to the relation
and formatting it, it was possible for VACUUM to come along and decide
it should format the page too. Though not harmful in itself, this would
cause data loss if a third transaction were able to insert tuples into
the vacuumed page before the original extender got control back.
Tom Lane [Sat, 7 May 2005 21:22:01 +0000 (21:22 +0000)]
Adjust time qual checking code so that we always check TransactionIdIsInProgress
before we check commit/abort status. Formerly this was done in some paths
but not all, with the result that a transaction might be considered
committed for some purposes before it became committed for others.
Per example found by Jan Wieck.
Bruce Momjian [Sat, 7 May 2005 15:45:23 +0000 (15:45 +0000)]
Add description:
< Currently locale can only be set during initdb.
> Currently locale can only be set during initdb. No global tables have
> locale-aware columns. However, the database template used during
> database creation might have locale-aware indexes. The indexes would
> need to be reindexed to match the new locale.
Neil Conway [Sat, 7 May 2005 05:48:50 +0000 (05:48 +0000)]
Revert the ld --as-needed patch. This breaks Fedora Core 3, due to a strange
interaction between ld, readline, termcap, and psql. The symptom is psql
failing with this error on startup:
symbol lookup error: /usr/lib64/libreadline.so.4: undefined symbol: BC
I'm still trying to find the best way to solve this, but in the mean time
I'm reverting the patch in order to unbreak FC3.
Bruce Momjian [Sat, 7 May 2005 04:24:34 +0000 (04:24 +0000)]
Add items:
> * Prevent to_char() on interval from returning meaningless values
>
> For example, to_char('1 month', 'mon') is meaningless. Basically,
> most date-related parameters to to_char() are meaningless for
> intervals because interval is not anchored to a date.
>
> * Allow to_char() on interval values to accumulate the highest unit
> requested
>
> o to_char(INTERVAL '1 hour 5 minutes', 'MI') => 65
> o to_char(INTERVAL '43 hours 20 minutes', 'MI' ) => 2600
> o to_char(INTERVAL '43 hours 20 minutes', 'WK:DD:HR:MI') => 0:1:19:20
> o to_char(INTERVAL '3 years 5 months','MM') => 41
>
> Some special format flag would be required to request such
> accumulation. Such functionality could also be added to EXTRACT.
> Prevent accumulation that crosses the month/day boundary because of
> the uneven number of days in a month.
>
Tom Lane [Fri, 6 May 2005 17:24:55 +0000 (17:24 +0000)]
For some reason access/tupmacs.h has been #including utils/memutils.h,
which is neither needed by nor related to that header. Remove the bogus
inclusion and instead include the header in those C files that actually
need it. Also fix unnecessary inclusions and bad inclusion order in
tsearch2 files.
Tom Lane [Fri, 6 May 2005 00:19:14 +0000 (00:19 +0000)]
Marginal performance improvements in dynahash: make sure that everything
associated with a hashtable is allocated in that hashtable's private
context, so that hash_destroy only has to destroy the context and not
do any retail pfree's; and tighten the inner loop of hash_seq_search.
Bruce Momjian [Thu, 5 May 2005 19:15:54 +0000 (19:15 +0000)]
On Win32, libintl replaces snprintf() with its own version that
understands arg control, so we don't need our own. In fact, it
also uses macros that conflict with ours, so we _can't_ use
our own.
Neil Conway [Thu, 5 May 2005 11:50:18 +0000 (11:50 +0000)]
The issue has been raised in the past that our build system links each
executable against the maximal set of libraries it might need. So for
example, if one executable requires `libreadline', all executables are
linked against it.
The easiest fix is to make use of GNU ld's --as-needed flag, which
ignores linker arguments that are not actually needed by the specified
object files. The attached patch modifies configure to check for this
flag (when using GNU ld), and if ld supports it, adds the flag to
LDFLAGS (we need to do the check since only relatively recent versions
of GNU ld support this capability). Currently only GNU ld is supported;
I'm not aware of any other linkers that support this functionality.
Tom Lane [Thu, 5 May 2005 03:37:23 +0000 (03:37 +0000)]
Adjust nodeBitmapIndexscan to keep the target index opened from plan
startup to end, rather than re-opening it in each MultiExecBitmapIndexScan
call. I had foolishly thought that opening/closing wouldn't be much
more expensive than a rescan call, but that was sheer brain fade.
This seems to fix about half of the performance lossage reported by
Sergey Koposov. I'm still not sure where the other half went.
Tom Lane [Thu, 5 May 2005 00:19:47 +0000 (00:19 +0000)]
Allow implicit cast from any named composite type to RECORD. At the
moment this has no particular use except to allow table rows to be
passed to record_out(), but that case seems to be useful in itself
per recent example from Elein. Further down the road we could look
at letting PL functions be declared to accept RECORD parameters.
Tom Lane [Tue, 3 May 2005 19:42:41 +0000 (19:42 +0000)]
Clean up MultiXactIdExpand's API by separating out the case where we
are creating a new MultiXactId from two regular XIDs. The original
coding was unnecessarily complicated and didn't save any code anyway.
Tom Lane [Tue, 3 May 2005 19:17:59 +0000 (19:17 +0000)]
Alter the signature for encoding conversion functions to declare the
output area as INTERNAL not CSTRING. This is to prevent people from
calling the functions by hand. This is a permanent solution for the
back branches but I hope it is just a stopgap for HEAD.
Tom Lane [Tue, 3 May 2005 16:51:00 +0000 (16:51 +0000)]
Change tsearch2 to not use the unsafe practice of creating functions
that return INTERNAL without also having INTERNAL arguments. Since the
functions in question aren't meant to be called by hand anyway, I just
redeclared them to take 'internal' instead of 'text'. Also add code
to ProcedureCreate() to enforce the restriction, as I should have done
to start with :-(
Neil Conway [Mon, 2 May 2005 00:37:07 +0000 (00:37 +0000)]
Change SPI functions to use a `long' when specifying the number of tuples
to produce when running the executor. This is consistent with the internal
executor APIs (such as ExecutorRun), which also use a long for this purpose.
It also allows FETCH_ALL to be passed -- since FETCH_ALL is defined as
LONG_MAX, this wouldn't have worked on platforms where int and long are of
different sizes. Per report from Tzahi Fadida.
Tom Lane [Sun, 1 May 2005 18:56:19 +0000 (18:56 +0000)]
Change CREATE TYPE to require datatype output and send functions to have
only one argument. (Per recent discussion, the option to accept multiple
arguments is pretty useless for user-defined types, and would be a likely
source of security holes if it was used.) Simplify call sites of
output/send functions to not bother passing more than one argument.
Tom Lane [Sat, 30 Apr 2005 20:31:39 +0000 (20:31 +0000)]
Change catalog entries for record_out and record_send to show only one
argument, since that's all they are using now. Adjust type_sanity
regression test so that it will complain if anyone tries to define
multiple-argument output functions in future.
Tom Lane [Sat, 30 Apr 2005 20:04:33 +0000 (20:04 +0000)]
Make record_out and record_send extract type information from the passed
record object itself, rather than relying on a second OID argument to be
correct. This patch just changes the function behavior and not the
catalogs, so it's OK to back-patch to 8.0. Will remove the now-redundant
second argument in pg_proc in a separate patch in HEAD only.
Tom Lane [Sat, 30 Apr 2005 19:03:33 +0000 (19:03 +0000)]
Use the standard lock manager to establish priority order when there
is contention for a tuple-level lock. This solves the problem of a
would-be exclusive locker being starved out by an indefinite succession
of share-lockers. Per recent discussion with Alvaro.
Neil Conway [Sat, 30 Apr 2005 08:08:51 +0000 (08:08 +0000)]
GCC 4.0 includes a new warning option, -Wformat-literal, that emits
a warning when a variable is used as a format string for printf()
and similar functions (if the variable is derived from untrusted
data, it could include unexpected formatting sequences). This
emits too many warnings to be enabled by default, but it does
flag a few dubious constructs in the Postgres tree. This patch
fixes up the obvious variants: functions that are passed a variable
format string but no additional arguments.
Most of these are harmless (e.g. the ruleutils stuff), but there
is at least one actual bug here: if you create a trigger named
"%sfoo", pg_dump will read uninitialized memory and fail to dump
the trigger correctly.
Tom Lane [Fri, 29 Apr 2005 22:28:24 +0000 (22:28 +0000)]
Restructure LOCKTAG as per discussions of a couple months ago.
Essentially, we shoehorn in a lockable-object-type field by taking
a byte away from the lockmethodid, which can surely fit in one byte
instead of two. This allows less artificial definitions of all the
other fields of LOCKTAG; we can get rid of the special pg_xactlock
pseudo-relation, and also support locks on individual tuples and
general database objects (including shared objects). None of those
possibilities are actually exploited just yet, however.
I removed pg_xactlock from pg_class, but did not force initdb for
that change. At this point, relkind 's' (SPECIAL) is unused and
could be removed entirely.
Neil Conway [Fri, 29 Apr 2005 07:08:06 +0000 (07:08 +0000)]
This patch fixes a bug in the error message emitted by pg_restore on an
incorrect -F argument: write_msg() expects its first parameter to be a
"module name", not the format string.
Tom Lane [Thu, 28 Apr 2005 21:47:18 +0000 (21:47 +0000)]
Implement sharable row-level locks, and use them for foreign key references
to eliminate unnecessary deadlocks. This commit adds SELECT ... FOR SHARE
paralleling SELECT ... FOR UPDATE. The implementation uses a new SLRU
data structure (managed much like pg_subtrans) to represent multiple-
transaction-ID sets. When more than one transaction is holding a shared
lock on a particular row, we create a MultiXactId representing that set
of transactions and store its ID in the row's XMAX. This scheme allows
an effectively unlimited number of row locks, just as we did before,
while not costing any extra overhead except when a shared lock actually
has to be shared. Still TODO: use the regular lock manager to control
the grant order when multiple backends are waiting for a row lock.
Bruce Momjian [Thu, 28 Apr 2005 13:09:59 +0000 (13:09 +0000)]
Add psql \set ON_ERROR_ROLLBACK to allow statements in a transaction to
error without affecting the entire transaction. Valid values are
"on|interactive|off".