Tom Lane [Tue, 11 Nov 2008 02:42:33 +0000 (02:42 +0000)]
Add support for input and output of interval values formatted per ISO 8601;
specifically, we can input either the "format with designators" or the
"alternative format", and we can output the former when IntervalStyle is set
to iso_8601.
Tom Lane [Mon, 10 Nov 2008 18:02:20 +0000 (18:02 +0000)]
Fix bugs in sqlchar_to_unicode and unicode_to_sqlchar: both were measuring
the length of a UTF8 character with pg_mblen (wrong if DB encoding isn't
UTF8), and the latter was blithely assuming that a static buffer would somehow
revert to all zeroes for each use.
Alvaro Herrera [Mon, 10 Nov 2008 16:25:41 +0000 (16:25 +0000)]
Add a --locale switch to createdb, to ease the creation of databases with
different locales. This is just syntactical sweetener over --lc-collate and
--lc-ctype. Per discussion.
While at it, properly document --lc-ctype and --lc-collate in SGML docs,
which apparently were forgotten (or purposefully ommited?) when they were
created.
Tom Lane [Mon, 10 Nov 2008 14:57:38 +0000 (14:57 +0000)]
Fix old bug in contrib/sslinfo: X509_NAME_to_text freed the BIO_s_mem buffer
it was using too soon. In a situation where pg_do_encoding_conversion is
a no-op, this led to garbage data returned.
In HEAD, also modify the code that's ensuring null termination to make it
a tad more obvious what's happening.
Tom Lane [Sun, 9 Nov 2008 21:24:33 +0000 (21:24 +0000)]
Replace pg_class.reltriggers with relhastriggers, which is just a boolean hint
("there might be triggers") rather than an exact count. This is necessary
catalog infrastructure for the upcoming patch to reduce the strength of
locking needed for trigger addition/removal. Split out and committed
separately for ease of reviewing/testing.
In passing, also get rid of the unused pg_class columns relukeys, relfkeys,
and relrefs, which haven't been maintained in many years and now have no
chance of ever being maintained (because of wishing to avoid locking).
Tom Lane [Sun, 9 Nov 2008 17:51:15 +0000 (17:51 +0000)]
Add a startup check that pg_xlog and pg_xlog/archive_status exist.
If the latter doesn't exist, automatically recreate it. (We don't do
this for pg_xlog, though, per discussion.)
Tom Lane [Sun, 9 Nov 2008 00:28:35 +0000 (00:28 +0000)]
Add a new GUC variable called "IntervalStyle" that decouples interval output
from DateStyle, and create a new interval style that produces output matching
the SQL standard (at least for interval values that fall within the standard's
restrictions). IntervalStyle is also used to resolve the conflict between the
standard and traditional Postgres rules for interpreting negative interval
input.
Tom Lane [Sat, 8 Nov 2008 20:51:49 +0000 (20:51 +0000)]
Fix recently added code for SQL years-months interval syntax so that
it behaves correctly for a leading minus sign, zero year value, and
nonzero month value. Per discussion with Ron Mayer.
Tom Lane [Fri, 7 Nov 2008 23:17:30 +0000 (23:17 +0000)]
Add some more citext test cases, per David Wheeler. Also remove the
citext-to-and-from-xml tests, since those caused variation between
installations with or without libxml without really proving much. Instead
repurpose citext_1.out as the expected results in glibc en_US (and probably
other) locales.
Tom Lane [Thu, 6 Nov 2008 20:51:15 +0000 (20:51 +0000)]
Improve bulk-insert performance by keeping the current target buffer pinned
(but not locked, as that would risk deadlocks). Also, make it work in a small
ring of buffers to avoid having bulk inserts trash the whole buffer arena.
Tom Lane [Thu, 6 Nov 2008 15:18:36 +0000 (15:18 +0000)]
Improve psql's \dC command to take a pattern parameter. Casts are shown
if their source or target types match the pattern (using the same definition
of "match" as \dT does). Per recent discussion.
The logic in systable_beginscan to translate heap attribute numbers to
index column numbers needs to handle the case where you have more than
one scankey on the same index column. toast_fetch_datum_slice() needs it.
Tom Lane [Tue, 4 Nov 2008 15:16:48 +0000 (15:16 +0000)]
Fix bug introduced in recent patch to make plpython cope with OUT arguments:
the proc->argnames array has to be initialized to zero immediately on creation,
since the error recovery path will try to free its elements.
Experimental new support for building man pages via docbook2x, an XSL-based
tool chain. With some polishing, this might help us get rid of our ancient
and crufty man page build mechanism.
Tom Lane [Tue, 4 Nov 2008 00:57:19 +0000 (00:57 +0000)]
Disallow LOCK TABLE outside a transaction block (or function), since this case
almost certainly represents user error. Per a gripe from Sebastian Böhm
and subsequent discussion.
Allow uuid_in() to parse a wider variety of variant input formats for the UUID
data type. This patch takes the approach of allowing an optional hyphen after
each group of four hex digits.
Alvaro Herrera [Mon, 3 Nov 2008 19:24:03 +0000 (19:24 +0000)]
Fix TransactionIdSetStatusBit so that it doesn't try to change a transaction
from COMMITTED to SUBCOMMITTED during recovery. This wasn't previously
possible, but it is now due to the recent changes on clog commit protocol for
subtransactions.
Tom Lane [Mon, 3 Nov 2008 19:08:56 +0000 (19:08 +0000)]
Fix two error-recovery bugs in describeOneTableDetails(), and make the code
to dump sequence values cope with sequences outside the search path and/or
having names that need quoting. No back-patch needed because these are new
problems in 8.4.
Kris Jurka (also a little bit of code beautification by tgl)
Tom Lane [Mon, 3 Nov 2008 01:17:08 +0000 (01:17 +0000)]
Change the pgstat logic so that the stats collector writes the stats file only
upon requests from backends, rather than on a fixed 500msec cycle. (There's
still throttling logic to ensure it writes no more often than once per
500msec, though.) This should result in a significant reduction in stats file
write traffic in typical scenarios where the stats are demanded only
infrequently.
This approach also means that the former difficulty with changing
stats_temp_directory on-the-fly has gone away, so remove the caution about
that as well as the thrashing we did to minimize the trouble window.
In passing, also fix pgstat_report_stat() so that we will send a stats
message if we have function call stats but not table stats to report;
this fixes a bug in the recent patch to support function-call stats.
Tom Lane [Sun, 2 Nov 2008 21:24:52 +0000 (21:24 +0000)]
Remove the last vestiges of the MAKE_PTR/MAKE_OFFSET mechanism. We haven't
allowed different processes to have different addresses for the shmem segment
in quite a long time, but there were still a few places left that used the
old coding convention. Clean them up to reduce confusion and improve the
compiler's ability to detect pointer type mismatches.
Tom Lane [Sun, 2 Nov 2008 01:45:28 +0000 (01:45 +0000)]
Remove all uses of the deprecated functions heap_formtuple, heap_modifytuple,
and heap_deformtuple in favor of the newer functions heap_form_tuple et al
(which do the same things but use bool control flags instead of arbitrary
char values). Eliminate the former duplicate coding of these functions,
reducing the deprecated functions to mere wrappers around the newer ones.
We can't get rid of them entirely because add-on modules probably still
contain many instances of the old coding style.
Tom Lane [Fri, 31 Oct 2008 21:07:55 +0000 (21:07 +0000)]
Simplify ExecutorRun's API and save some trivial number of cycles by having
it just return void instead of sometimes returning a TupleTableSlot. SQL
functions don't need that anymore, and noplace else does either. Eliminating
the return value also means one less hassle for the ExecutorRun hook functions
that will be supported beginning in 8.4.
Update FSM on WAL replay. This is a bit limited; the FSM is only updated
on non-full-page-image WAL records, and quite arbitrarily, only if there's
less than 20% free space on the page after the insert/update (not on HOT
updates, though). The 20% cutoff should avoid most of the overhead, when
replaying a bulk insertion, for example, while ensuring that pages that
are full are marked as full in the FSM.
This is mostly to avoid the nasty worst case scenario, where you replay
from a PITR archive, and the FSM information in the base backup is really
out of date. If there was a lot of pages that the outdated FSM claims to
have free space, but don't actually have any, the first unlucky inserter
after the recovery would traverse through all those pages, just to find
out that they're full. We didn't have this problem with the old FSM
implementation, because we simply threw the FSM information away on a
non-clean shutdown.
Tom Lane [Fri, 31 Oct 2008 19:37:56 +0000 (19:37 +0000)]
Allow SQL-language functions to return the output of an INSERT/UPDATE/DELETE
RETURNING clause, not just a SELECT as formerly.
A side effect of this patch is that when a set-returning SQL function is used
in a FROM clause, performance is improved because the output is collected into
a tuplestore within the function, rather than using the less efficient
value-per-call mechanism.
Unite ReadBufferWithFork, ReadBufferWithStrategy, and ZeroOrReadBuffer
functions into one ReadBufferExtended function, that takes the strategy
and mode as argument. There's three modes, RBM_NORMAL which is the default
used by plain ReadBuffer(), RBM_ZERO, which replaces ZeroOrReadBuffer, and
a new mode RBM_ZERO_ON_ERROR, which allows callers to read corrupt pages
without throwing an error. The FSM needs the new mode to recover from
corrupt pages, which could happend if we crash after extending an FSM file,
and the new page is "torn".
Add fork number to some error messages in bufmgr.c, that still lacked it.
Peter Eisentraut [Fri, 31 Oct 2008 14:35:30 +0000 (14:35 +0000)]
The conversion rule from postgres.sgml to postgres.xml didn't work with
BSD sed. So write it in Perl, which is more portable and a bit faster, too.
We already use Perl for standard documentation builds, so this imposes no
additional requirement.
Tom Lane [Thu, 30 Oct 2008 04:06:16 +0000 (04:06 +0000)]
Fix recoveryLastXTime logic so that it actually does what one would expect.
Per gripe from Kevin Grittner. Backpatch to 8.3, where the bug was introduced.
Peter Eisentraut [Wed, 29 Oct 2008 09:27:24 +0000 (09:27 +0000)]
Use Autoconf provided AS_HELP_STRING macro to automatically format and
align strings in the --help output. Do this through our abstraction layer
to eliminate redundancy and randomness in configure.in.
Tom Lane [Wed, 29 Oct 2008 00:00:39 +0000 (00:00 +0000)]
Be more tense about not creating tuplestores with randomAccess = true unless
backwards scan could actually happen. In particular, pass a flag to
materialize-mode SRFs that tells them whether they need to require random
access. In passing, also suppress unneeded backward-scan overhead for a
Portal's holdStore tuplestore. Per my proposal about reducing I/O costs for
tuplestores.
Tom Lane [Tue, 28 Oct 2008 22:02:06 +0000 (22:02 +0000)]
Extend ExecMakeFunctionResult() to support set-returning functions that return
via a tuplestore instead of value-per-call. Refactor a few things to reduce
ensuing code duplication with nodeFunctionscan.c. This represents the
reasonably noncontroversial part of my proposed patch to switch SQL functions
over to returning tuplestores. For the moment, SQL functions still do things
the old way. However, this change enables PL SRFs to be called in targetlists
(observe changes in plperl regression results).
Tom Lane [Tue, 28 Oct 2008 17:13:51 +0000 (17:13 +0000)]
Change WorkTableScan to not support backward scan. The apparent support
didn't actually work, because nodeRecursiveunion.c creates the underlying
tuplestore with backward scan disabled; which is a decision that we shouldn't
reverse because of performance cost. We could imagine adding signaling from
WorkTableScan to RecursiveUnion about whether backward scan is needed ...
but in practice it'd be a waste of effort, because there simply isn't any
current or plausible future scenario where WorkTableScan would be called on
to scan backward. So just dike out the code that claims to support it.
Tom Lane [Tue, 28 Oct 2008 15:51:03 +0000 (15:51 +0000)]
Arrange to squeeze out the MINIMAL_TUPLE_PADDING in the tuple representation
written to temp files by tuplesort.c and tuplestore.c. This saves 2 bytes per
row for 32-bit machines, and 6 bytes per row for 64-bit machines, which seems
worth the slight additional uglification of the tuple read/write routines.
Tom Lane [Mon, 27 Oct 2008 19:37:22 +0000 (19:37 +0000)]
Install a more robust solution for the problem of infinite error-processing
recursion when we are unable to convert a localized error message to the
client's encoding. We've been over this ground before, but as reported by
Ibrar Ahmed, it still didn't work in the case of conversion failures for
the conversion-failure message itself :-(. Fix by installing a "circuit
breaker" that disables attempts to localize this message once we get into
recursion trouble.
Patch all supported branches, because it is in fact broken in all of them;
though I had to add some missing translations to the older branches in
order to expose the failure in the particular test case I was using.
Magnus Hagander [Mon, 27 Oct 2008 09:42:31 +0000 (09:42 +0000)]
Add support for multiple error messages from libpq, by simply appending them
after each other (since we already add a newline on each, this makes them
multiline).
Previously a new error would just overwrite the old one, so for example any
error caused when trying to connect with SSL enabled would be overwritten
by the error message form the non-SSL connection when using sslmode=prefer.
Tom Lane [Sun, 26 Oct 2008 02:46:25 +0000 (02:46 +0000)]
Better solution to the IN-list issue: instead of having an arbitrary cutoff,
treat Var and non-Var IN-list items differently. Only non-Var items are
candidates to go into an ANY(ARRAY) construct --- we put all Vars as separate
OR conditions on the grounds that that leaves more scope for optimization.
Per suggestion from Robert Haas.
Tom Lane [Sat, 25 Oct 2008 19:51:32 +0000 (19:51 +0000)]
Be a little smarter about qual handling for semi-joins: a qual that mentions
only the outer side can be pushed down rather than having to be evaluated
at the join.
Tom Lane [Sat, 25 Oct 2008 17:19:09 +0000 (17:19 +0000)]
Add a heuristic to transformAExprIn() to make it prefer expanding "x IN (list)"
into an OR of equality comparisons, rather than x = ANY(ARRAY[...]), when there
are Vars in the right-hand side. This avoids a performance regression compared
to pre-8.2 releases, in cases where the OR form can be optimized into scans
of multiple indexes. Limit the possible downside by preferring this form only
when the list isn't very long (I set the cutoff at 32 elements, which is a
bit arbitrary but in the right ballpark). Per discussion with Jim Nasby.
In passing, also make it try the OR form if it cannot select a common type
for the array elements; we've seen a complaint or two about how the OR form
worked for such cases and ARRAY doesn't.
Tom Lane [Fri, 24 Oct 2008 23:42:35 +0000 (23:42 +0000)]
Reduce the memory footprint of large pending-trigger-event lists, as per my
recent proposal. In typical cases, we now need 12 bytes per insert or delete
event and 16 bytes per update event; previously we needed 40 bytes per
event on 32-bit hardware and 80 bytes per event on 64-bit hardware. Even
in the worst case usage pattern with a large number of distinct triggers being
fired in one query, usage is at most 32 bytes per event. It seems to be a
bit faster than the old code as well, due to reduction of palloc overhead.
This commit doesn't address the TODO item of allowing the event list to spill
to disk; rather it's trying to stave off the need for that. However, it
probably makes that task a bit easier by reducing the data structure's
dependency on pointers. It would now be practical to dump an event list to
disk by "chunks" instead of individual events.