Bruce Momjian [Sat, 17 Feb 2007 03:11:32 +0000 (03:11 +0000)]
Remove rint() for to_char MS and US output. We can't us rint() because
we can't overflow to the next higher units, and we might print the lower
units for MS.
Bruce Momjian [Sat, 17 Feb 2007 01:35:41 +0000 (01:35 +0000)]
Add:
>
> o Allow row and record variables to be set to NULL constants,
> and allow NULL tests on such variables
>
> Because a row is not scalar, do not allow assignment
> from NULL-valued scalars.
Tom Lane [Fri, 16 Feb 2007 23:32:08 +0000 (23:32 +0000)]
Teach find_nonnullable_rels to handle OR cases: if every arm of an OR
forces a particular relation nonnullable, then we can say that the OR does.
This is worth a little extra trouble since it may allow reduction of
outer joins to plain joins.
Tom Lane [Fri, 16 Feb 2007 22:04:02 +0000 (22:04 +0000)]
Fix new RI operator selection code to do the right thing when working with
an opclass for a generic type such as ANYARRAY. The original coding failed
to check that PK and FK columns were of the same array type. Per discussion
with Tom Dunstan. Also, make the code a shade more readable by not trying
to economize on variables.
Bruce Momjian [Fri, 16 Feb 2007 21:34:04 +0000 (21:34 +0000)]
Reduce the amount of memory "clobbered" for every process title change,
on platforms that need this. This is done by only writing past the
previously stored message, if it was longer.
Tom Lane [Fri, 16 Feb 2007 20:57:19 +0000 (20:57 +0000)]
Adjust the definition of is_pushed_down so that it's always true for INNER
JOIN quals, just like WHERE quals, even if they reference every one of the
join's relations. Now that we can reorder outer and inner joins, it's
possible for such a qual to end up being assigned to an outer join plan node,
and we mustn't have it treated as a join qual rather than a filter qual for
the node. (If it were, the join could produce null-extended rows that it
shouldn't.) Per bug report from Pelle Johansson.
Alvaro Herrera [Fri, 16 Feb 2007 17:49:15 +0000 (17:49 +0000)]
Install a more correct fix in the timestamp and timestamptz regression tests:
remove duplicated tests in timestamp, and complete timestamptz with the tests
that were missing to more closely mirror timestamp.
Alvaro Herrera [Fri, 16 Feb 2007 15:42:42 +0000 (15:42 +0000)]
Fix the timestamptz test problem, by moving the tests that use the
timestamp_tbl table into the timestamp test. Also, restore a test that
used to exist as a valid test in the timestamptz test.
Tom Lane [Fri, 16 Feb 2007 03:49:04 +0000 (03:49 +0000)]
Fix another problem in 8.2 changes that allowed "one-time" qual conditions to
be checked at plan levels below the top; namely, we have to allow for Result
nodes inserted just above a nestloop inner indexscan. Should think about
using the general Param mechanism to pass down outer-relation variables, but
for the moment we need a back-patchable solution. Per report from Phil Frost.
Bruce Momjian [Fri, 16 Feb 2007 02:59:41 +0000 (02:59 +0000)]
SSL improvements:
o read global SSL configuration file
o add GUC "ssl_ciphers" to control allowed ciphers
o add libpq environment variable PGSSLKEY to control SSL hardware keys
Tom Lane [Fri, 16 Feb 2007 00:14:01 +0000 (00:14 +0000)]
Restructure code that is responsible for ensuring that clauseless joins are
considered when it is necessary to do so because of a join-order restriction
(that is, an outer-join or IN-subselect construct). The former coding was a
bit ad-hoc and inconsistent, and it missed some cases, as exposed by Mario
Weilguni's recent bug report. His specific problem was that an IN could be
turned into a "clauseless" join due to constant-propagation removing the IN's
joinclause, and if the IN's subselect involved more than one relation and
there was more than one such IN linking to the same upper relation, then the
only valid join orders involve "bushy" plans but we would fail to consider the
specific paths needed to get there. (See the example case added to the join
regression test.) On examining the code I wonder if there weren't some other
problem cases too; in particular it seems that GEQO was defending against a
different set of corner cases than the main planner was. There was also an
efficiency problem, in that when we did realize we needed a clauseless join
because of an IN, we'd consider clauseless joins against every other relation
whether this was sensible or not. It seems a better design is to use the
outer-join and in-clause lists as a backup heuristic, just as the rule of
joining only where there are joinclauses is a heuristic: we'll join two
relations if they have a usable joinclause *or* this might be necessary to
satisfy an outer-join or IN-clause join order restriction. I refactored the
code to have just one place considering this instead of three, and made sure
that it covered all the cases that any of them had been considering.
Backpatch as far as 8.1 (which has only the IN-clause form of the disease).
By rights 8.0 and 7.4 should have the bug too, but they accidentally fail
to fail, because the joininfo structure used in those releases preserves some
memory of there having once been a joinclause between the inner and outer
sides of an IN, and so it leads the code in the right direction anyway.
I'll be conservative and not touch them.
Alvaro Herrera [Thu, 15 Feb 2007 23:23:23 +0000 (23:23 +0000)]
Restructure autovacuum in two processes: a dummy process, which runs
continuously, and requests vacuum runs of "autovacuum workers" to postmaster.
The workers do the actual vacuum work. This allows for future improvements,
like allowing multiple autovacuum jobs running in parallel.
For now, the code keeps the original behavior of having a single autovac
process at any time by sleeping until the previous worker has finished.
Tom Lane [Thu, 15 Feb 2007 03:07:13 +0000 (03:07 +0000)]
Repair oversight in 8.2 change that improved the handling of "pseudoconstant"
WHERE clauses. createplan.c is now willing to stick a gating Result node
almost anywhere in the plan tree, and in particular one can wind up directly
underneath a MergeJoin node. This means it had better be willing to handle
Mark/Restore. Fortunately, that's trivial in such cases, since we can just
pass off the call to the input node (which the planner has previously ensured
can handle Mark/Restore). Per report from Phil Frost.
Tom Lane [Wed, 14 Feb 2007 01:58:58 +0000 (01:58 +0000)]
Fix up foreign-key mechanism so that there is a sound semantic basis for the
equality checks it applies, instead of a random dependence on whatever
operators might be named "=". The equality operators will now be selected
from the opfamily of the unique index that the FK constraint depends on to
enforce uniqueness of the referenced columns; therefore they are certain to be
consistent with that index's notion of equality. Among other things this
should fix the problem noted awhile back that pg_dump may fail for foreign-key
constraints on user-defined types when the required operators aren't in the
search path. This also means that the former warning condition about "foreign
key constraint will require costly sequential scans" is gone: if the
comparison condition isn't indexable then we'll reject the constraint
entirely. All per past discussions.
Along the way, make the RI triggers look into pg_constraint for their
information, instead of using pg_trigger.tgargs; and get rid of the always
error-prone fixed-size string buffers in ri_triggers.c in favor of building up
the RI queries in StringInfo buffers.
initdb forced due to columns added to pg_constraint and pg_trigger.
Tom Lane [Tue, 13 Feb 2007 19:39:42 +0000 (19:39 +0000)]
Disallow committing a prepared transaction unless we are in the same database
it was executed in. Someday it might be nice to allow cross-DB commits, but
work would be needed in NOTIFY and perhaps other places. Per Heikki.
Tom Lane [Tue, 13 Feb 2007 19:18:54 +0000 (19:18 +0000)]
Improve postmaster's behavior if an accept() call fails. Because the server
socket is still read-ready, the code was a tight loop, wasting lots of CPU.
We can't do anything to clear the failure, other than wait, but we should give
other processes more chance to finish and release FDs; so insert a small sleep.
Also, avoid bogus "close(-1)" in this case. Per report from Jim Nasby.
Tom Lane [Tue, 13 Feb 2007 02:31:03 +0000 (02:31 +0000)]
Repair bug in 8.2's new logic for planning outer joins: we have to allow joins
that overlap an outer join's min_righthand but aren't fully contained in it,
to support joining within the RHS after having performed an outer join that
can commute with this one. Aside from the direct fix in make_join_rel(),
fix has_join_restriction() and GEQO's desirable_join() to consider this
possibility. Per report from Ian Harding.
Tom Lane [Fri, 9 Feb 2007 20:17:59 +0000 (20:17 +0000)]
Minor tweak to make rowtypes regression test run faster. We don't
currently have any better strategy for this query than re-running the
sub-select over and over; it seems unlikely that doing so 10000 times
is a more useful test than doing it a few dozen times.
Tom Lane [Fri, 9 Feb 2007 16:12:19 +0000 (16:12 +0000)]
Call pgstat_drop_database during DROP DATABASE, so that any stats file
entries for the victim database go away sooner rather than later. We already
did the equivalent thing at the per-relation level, not sure why it's not
been done for whole databases. With this change, pgstat_vacuum_tabstat
should usually not find anything to do; though we still need it as a backstop
in case DROPDB or TABPURGE messages get lost under load.
Bruce Momjian [Fri, 9 Feb 2007 03:37:45 +0000 (03:37 +0000)]
Done!
< * Merge xmin/xmax/cmin/cmax back into three header fields
<
< Before subtransactions, there used to be only three fields needed to
< store these four values. This was possible because only the current
< transaction looks at the cmin/cmax values. If the current transaction
< created and expired the row the fields stored where xmin (same as
< xmax), cmin, cmax, and if the transaction was expiring a row from a
< another transaction, the fields stored were xmin (cmin was not
< needed), xmax, and cmax. Such a system worked because a transaction
< could only see rows from another completed transaction. However,
< subtransactions can see rows from outer transactions, and once the
< subtransaction completes, the outer transaction continues, requiring
< the storage of all four fields. With subtransactions, an outer
< transaction can create a row, a subtransaction expire it, and when the
< subtransaction completes, the outer transaction still has to have
< proper visibility of the row's cmin, for example, for cursors.
<
< One possible solution is to create a phantom cid which represents a
< cmin/cmax pair and is stored in local memory. Another idea is to
< store both cmin and cmax only in local memory.
<
> * -Merge xmin/xmax/cmin/cmax back into three header fields
Tom Lane [Fri, 9 Feb 2007 03:35:35 +0000 (03:35 +0000)]
Combine cmin and cmax fields of HeapTupleHeaders into a single field, by
keeping private state in each backend that has inserted and deleted the same
tuple during its current top-level transaction. This is sufficient since
there is no need to be able to determine the cmin/cmax from any other
transaction. This gets us back down to 23-byte headers, removing a penalty
paid in 8.0 to support subtransactions. Patch by Heikki Linnakangas, with
minor revisions by moi, following a design hashed out awhile back on the
pghackers list.
Bruce Momjian [Fri, 9 Feb 2007 00:34:31 +0000 (00:34 +0000)]
Update:
< * Consider placing all sequences in a single table, now that system
< tables are full transactional
> * Consider placing all sequences in a single table
Tom Lane [Thu, 8 Feb 2007 18:37:30 +0000 (18:37 +0000)]
Fix an ancient logic error in plpgsql's exec_stmt_block: it thought it could
get away with not (re)initializing a local variable if the variable is marked
"isconst" and not "isnull". Unfortunately it makes this decision after having
already freed the old value, meaning that something like
for i in 1..10 loop
declare c constant text := 'hi there';
leads to subsequent accesses to freed memory, and hence probably crashes.
(In particular, this is why Asif Ali Rehman's bug leads to crash and not
just an unexpectedly-NULL value for SQLERRM: SQLERRM is marked CONSTANT
and so triggers this error.)
The whole thing seems wrong on its face anyway: CONSTANT means that you can't
change the variable inside the block, not that the initializer expression is
guaranteed not to change value across successive block entries. Hence,
remove the "optimization" instead of trying to fix it.
Tom Lane [Thu, 8 Feb 2007 18:37:14 +0000 (18:37 +0000)]
Rearrange use of plpgsql_add_initdatums() so that only the parsing of a
DECLARE section needs to know about it. Formerly, everyplace besides DECLARE
that created variables needed to do "plpgsql_add_initdatums(NULL)" to prevent
those variables from being sucked up as part of a subsequent DECLARE block.
This is obviously error-prone, and in fact the SQLSTATE/SQLERRM patch had
failed to do it for those two variables, leading to the bug recently exhibited
by Asif Ali Rehman: a DECLARE within an exception handler tried to reinitialize
SQLERRM.
Although the SQLSTATE/SQLERRM patch isn't in any pre-8.1 branches, and so
I can't point to a demonstrable failure there, it seems wise to back-patch
this into the older branches anyway, just to keep the logic similar to HEAD.
Bruce Momjian [Thu, 8 Feb 2007 15:28:58 +0000 (15:28 +0000)]
Win32 regression test fixes:
For win32 in general, this makes it possible to run the regression tests
as an admin user by using the same restricted token method that's used
by pg_ctl and initdb.
For vc++, it adds building of pg_regress.exe, adds a resultmap, and
fixes how it runs the install.
Bruce Momjian [Thu, 8 Feb 2007 15:16:19 +0000 (15:16 +0000)]
Add /contrib/pg_standby:
pg_standby is a production-ready program that can be used to
create a Warm Standby server. Other configuration is required
as well, all of which is described in the main server manual.
Bruce Momjian [Thu, 8 Feb 2007 15:09:47 +0000 (15:09 +0000)]
Add /contrib/pg_standby:
pg_standby is a production-ready program that can be used to
create a Warm Standby server. Other configuration is required
as well, all of which is described in the main server manual.
Normalize fgets() calls to use sizeof() for calculating the buffer size
where possible, and fix some sites that apparently thought that fgets()
will overwrite the buffer by one byte.
Also add some strlcpy() to eliminate some weird memory handling.
Bruce Momjian [Thu, 8 Feb 2007 05:05:53 +0000 (05:05 +0000)]
Reduce WAL activity for page splits:
> Currently, an index split writes all the data on the split page to
> WAL. That's a lot of WAL traffic. The tuples that are copied to the
> right page need to be WAL logged, but the tuples that stay on the
> original page don't.