< * Allow server log information to be output as INSERT statements
> * -Allow server log information to be output as CSV format

redirect_stderr to logging_collector.
Original patch from Arul Shaji, subsequently modified by Greg Smith, and then
heavily modified by me.

>
> * Reduce XID consumption of read-only queries
>
>    http://archives.postgresql.org/pgsql-hackers/2007-08/msg00516.php
>

are not one of the query's defined result relations, but nonetheless have
triggers fired against them while the query is active.  This was formerly
impossible but can now occur because of my recent patch to fix the firing
order for RI triggers.  Caching a ResultRelInfo avoids duplicating work by
repeatedly opening and closing the same relation, and also allows EXPLAIN
ANALYZE to "see" and report on these extra triggers.  Use the same mechanism
to cache open relations when firing deferred triggers at transaction shutdown;
this replaces the former one-element-cache strategy used in that case, and
should improve performance a bit when there are deferred triggers on a number
of relations.

row within one query: we were firing check triggers before all the updates
were done, leading to bogus failures.  Fix by making the triggers queued by
an RI update go at the end of the outer query's trigger event list, thereby
effectively making the processing "breadth-first".  This was indeed how it
worked pre-8.0, so the bug does not occur in the 7.x branches.
Per report from Pavel Stehule.

>   A third idea would be for a heap scan to check if all rows are visible
>   and if so set a per-table flag which can be checked by index scans.
>   Any change to the table would have to clear the flag.  To detect
>   changes during the heap scan a counter could be set at the start and
>   checked at the end --- if it is the same, the table has not been
>   modified --- any table change would increment the counter.

that still thought they could set HEAP_XMAX_COMMITTED immediately after
seeing the other transaction commit.  Make them use the same logic as
tqual.c does to determine if the hint bit can be set yet.

< 	o Use backend PREPARE/EXECUTE facility for ecpg where possible

- Really prepare statements
- Added more regression tests
- Added auto-prepare mode
- Use '$n' for positional variables, '?' is still possible via ecpg option
- Cleaned up the sources a little bit

thorough testing of async-commit mode from the buildfarm.  This patch MUST
get reverted before 8.3 release!

First, we cannot assume that XLogAsyncCommitFlush guarantees hint bits will be
settable, because clog.c's inexact LSN bookkeeping results in windows where a
previously flushed transaction is considered unhintable because it shares an
LSN slot with a later unflushed transaction.  But repair_frag requires
XMIN_COMMITTED to be correct so that it can distinguish tuples moved by the
current vacuum.  Since not being able to set the bit is an uncommon corner
case, the most practical way of dealing with it seems to be to abandon
shrinking (ie, don't invoke repair_frag) when we find a non-dead tuple whose
XMIN_COMMITTED bit couldn't be set.

Second, it is possible for the same reason that a RECENTLY_DEAD tuple does not
get its XMAX_COMMITTED bit set during scan_heap.  But by the time repair_frag
examines the tuple it might be possible to set the bit.  We therefore must
take buffer content lock when calling HeapTupleSatisfiesVacuum a second time,
else we can get an Assert failure in SetBufferCommitInfoNeedsSave.  This
latter bug is latent in existing releases, but I think it cannot actually
occur without async commit, since the first HeapTupleSatisfiesVacuum call
should always have set the bit.  So I'm not going to back-patch it.

In passing, reduce the existing "cannot shrink relation" messages from NOTICE
to LOG level.  The new message must be no higher than LOG if we don't want
unpredictable regression test failures, and consistency seems like a good
idea.  Also arrange that only one such message is reported per VACUUM FULL;
in typical scenarios you could get spammed with many such messages, which
seems a bit useless.

certain corner cases.  Per discussion, the code does what we want, but
it really needs to be documented that these functions act differently
from regexp_matches.

enlarge the memory chunk in-place when it was feasible to do so.  This turns
out to not work well at all for scenarios involving repeated cycles of
palloc/repalloc/pfree: the eventually freed chunks go into the wrong freelist
for the next initial palloc request, and so we consume memory indefinitely.
While that could be defended against, the number of cases where the
optimization can still be applied drops significantly, and adjusting the
initial sizes of StringInfo buffers makes it drop to almost nothing.
Seems better to just remove the extra complexity.
Per recent discussion and testing.

likewise increase the initial size of the scanner's literal buffer to 1024
(from 128).  Instrumentation of the regression tests suggests that this
saves a useful amount of repalloc() traffic --- the number of calls occurring
during one set of tests drops from about 6900 to about 3900.  The old sizes
were chosen in the late 90's with an eye to machines much smaller than
are common today.

regexp_split_to_table() within a single query.  This is only a partial
solution, as it turns out that with enough matches per string these
functions can also tickle a repalloc() misbehavior.  But fixing that
is a topic for a separate patch.

that cached compiled patterns will still be there when the function is next
called.  Clean up looping logic, thereby fixing bug identified by Pavel
Stehule.  Share setup code between the two functions, add some comments, and
avoid risky mixing of int and size_t variables.  Clean up the documentation a
tad, and accept all the flag characters mentioned in table 9-19 rather than
just a subset.

these uses of printQuery() in FETCH_COUNT patch a year ago :-(.  Per report
from Tomoaki Sato.

Brendan Jurd

constant flow of new connection requests could prevent the postmaster from
completing a shutdown or crash restart.  This is done by labeling child
processes that are "dead ends", that is, we know that they were launched only
to tell a client that it can't connect.  These processes are managed
separately so that they don't confuse us into thinking that we can't advance
to the next stage of a shutdown or restart sequence, until the very end
where we must wait for them to drain out so we can delete the shmem segment.
Per discussion of a misbehavior reported by Keaton Adams.

Since this code was baroque already, and my first attempt at fixing the
problem made it entirely impenetrable, I took the opportunity to rewrite it
in a state-machine style.  That eliminates some duplicated code sections and
hopefully makes everything a bit clearer.

hash table is allocated in a child context of the agg node's memory
context, MemoryContextReset() will reset but *not* delete the child
context. Since ExecReScanAgg() proceeds to build a new hash table
from scratch (in a new sub-context), this results in leaking the
header for the previous memory context. Therefore, use
MemoryContextResetAndDeleteChildren() instead.

Credit: My colleague Sailesh Krishnamurthy at Truviso for isolating
the cause of the leak.

GUC assign hooks are supposed to be made whether doit is true or not.

child memory contexts is indented two spaces to the right of its
parent context.  This should make it easier to deduce the memory
context hierarchy from the output of MemoryContextStats().

Per report and partial patch from Chad Wagner.

even if the compiler is not defining __GNUC__.  Per report from
Dirk Tilger that it is possible for icc to not do that.

check for __INTEL_COMPILER.  Per report from Dirk Tilger.
Not back-patched since I don't fully trust it yet ...

improve grammar a tad.  Per Greg Stark.

expected, exit with an error, rather than complaining about the error
on stderr but continuing onward.

between the setting of log_line_prefix and the setting of log_timezone.  We
can't realistically set log_timezone any earlier than we do now, so the best
behavior seems to be to use GMT zone if any timestamps are to be logged during
early startup.  Create a dummy zone variable with a minimal definition of GMT
(in particular it will never know about leap seconds), so that we can set it
up without reference to any external files.

as well as regular backends: if no regular backend launches before the autovac
launcher tries to start an autovac worker, the postmaster would get an Assert
fault due to calling PostmasterRandom before random_seed was initialized.
Cleanest solution seems to be to take the initialization of random_seed out
of ServerLoop and let PostmasterRandom do it for itself.

displayed in the postmaster log.  This avoids Windows-specific problems with
localized time zone names that are in the wrong encoding, and generally seems
like a good idea to forestall other potential platform-dependent issues.
To preserve the existing behavior that all backends will log in the same time
zone, create a new GUC variable log_timezone that can only be changed on a
system-wide basis, and reference log-related calculations to that zone instead
of the TimeZone variable.

This fixes the issue reported by Hiroshi Saito that timestamps printed by
xlog.c startup could be improperly localized on Windows.  We still need a
simpler patch for that problem in the back branches, however.

not bothering to initialize is_autovacuum for regular backends, meaning there
was a significant chance of the postmaster prematurely sending them SIGTERM
during database shutdown.  Also, leaving the cancel key unset for an autovac
worker meant that any client could send it SIGINT, which doesn't sound
especially good either.