Slow functions in index expressions might cause this loop to take long
enough to make it worth being cancellable.  Probably it would be enough
to call CHECK_FOR_INTERRUPTS here, but for consistency with other
per-sample-row loops in this file, let's use vacuum_delay_point.

Report and patch by Jeff Janes.  Back-patch to all supported branches.

Previously the funcCtx was a child of the tmpCtx, but that was broken
by commit eaa5808e, which made
MemoryContextReset() delete, not reset, child contexts.  The behavior of
having a tmpCtx reset also clear the other context seems rather dubious
anyway, so let's just disentangle them.  Per report from Erik Rijkers.

In passing, fix badly-inaccurate comments about these contexts.

These were workarounds for a long-gone flex bug; all supported versions
of flex emit an extern declaration as expected.

Get rid of unnecessary expr_yylex declaration (we haven't supported
flex 2.5.4 in a long time, and even if we still did, the declaration in
pgbench.h makes this one unnecessary and inappropriate).  Fix copyright
dates, improve some layout choices, etc.

We can use that macro as long as we put the value into a local variable.
Commit 735cd612 was not wrong on its own terms, but I think this way
looks nicer, and it should save a few cycles on 32-bit machines.

If set, the pager will not be used unless this many lines are to be
displayed, even if that is more than the screen depth. Default is zero,
meaning it's disabled.

There is probably more work to be done in giving the user control over
when the pager is used, particularly when wide output forces use of the
pager regardless of how many lines there are, but this is a start.

The stddev calculation included a faster but unportable sqrt function.
This is not worth the extra effort, and won't work everywhere. If the
standard library function is good enough for the SQL function it
should be good enough here too.

Andreas Karlsson

inet, cidr, and timetz indexes still cannot support index-only scans,
because they don't store the original unmodified value in the index, but a
derived approximate value.

Stddev is calculated on the fly, and the code in commit 717f7095 was
using Float8GetDatumFast() inappropriately to convert the result to a
Datum. Mea culpa. It now uses Float8GetDatum().

The new fields are min_time, max_time, mean_time and stddev_time.

Based on an original patch from Mitsumasa KONDO, modified by me. Reviewed by Petr Jelínek.

The gbt_var_key_copy function was doing two different things depending on
the boolean argument. Seems cleaner to have two separate functions.

Remove unused argument from gbt_num_compress.

We cannot use the index's tuple descriptor directly to describe the index
tuples returned in an index-only scan. That's because the index might use
a different datatype for the values stored on disk than the type originally
indexed. As long as they were both pass-by-ref, it worked, but will not work
for pass-by-value types of different sizes. I noticed this as a crash when I
started hacking a patch to add fetch methods to btree_gist.

This improves on commit bbfd7eda by
making two simple changes:

* pg_attribute_noreturn now takes parentheses, ie pg_attribute_noreturn().
Likewise pg_attribute_unused(), pg_attribute_packed().  This reduces
pgindent's tendency to misformat declarations involving them.

* attributes are now always attached to function declarations, not
definitions.  Previously some places were taking creative shortcuts,
which were not merely candidates for bad misformatting by pgindent
but often were outright wrong anyway.  (It does little good to put a
noreturn annotation where callers can't see it.)  In any case, if
we would like to believe that these macros can be used with non-gcc
compilers, we should avoid gratuitous variance in usage patterns.

I also went through and manually improved the formatting of a lot of
declarations, and got rid of excessively repetitive (and now obsolete
anyway) comments informing the reader what pg_attribute_printf is for.

This adds a new GiST opclass method, 'fetch', which is used to reconstruct
the original Datum from the value stored in the index. Also, the 'canreturn'
index AM interface function gains a new 'attno' argument. That makes it
possible to use index-only scans on a multi-column index where some of the
opclasses support index-only scans but some do not.

This patch adds support in the box and point opclasses. Other opclasses
can added later as follow-on patches (btree_gist would be particularly
interesting).

Anastasia Lubennikova, with additional fixes and modifications by me.

Remove the gistcentryinit function, inlining the relevant part of it into
the only caller.

Jeff Janes

It is only used in src/backend/replication/syncrep.c.

Back-patch to all supported branches except 9.1 which declares the
function as static.

This is meant to make it easier to insert simple debugging cross-checks
in plpgsql functions.

Pavel Stehule, reviewed by Jim Nasby

Several submitted and even committed patches have run into the problem
that C89, our baseline, does not provide minimum/maximum values for
various integer datatypes. C99's stdint.h does, but we can't rely on
it.

Several parts of the code defined limits locally, so instead centralize
the definitions to c.h.

This patch also changes the more obvious usages of literal limit values;
there's more places that could be changed, but it's less clear whether
it's beneficial to change those.

Author: Andrew Gierth
Discussion: 87619tc5wc.fsf@news-spur.riddles.org.uk

Since commit a2e35b53, most CREATE and ALTER commands return the
ObjectAddress of the affected object.  This is useful for event triggers
to try to figure out exactly what happened.  This patch extends this
idea a bit further to cover ALTER TABLE as well: an auxiliary
ObjectAddress is returned for each of several subcommands of ALTER
TABLE.  This makes it possible to decode with precision what happened
during execution of any ALTER TABLE command; for instance, which
constraint was added by ALTER TABLE ADD CONSTRAINT, or which parent got
dropped from the parents list by ALTER TABLE NO INHERIT.

As with the previous patch, there is no immediate user-visible change
here.

This is all really just continuing what c504513f started.

Reviewed by Stephen Frost.

The POSIX spec says that rint() rounds halfway cases to nearest even.
Our substitute implementation failed to do that, rather rounding halfway
cases away from zero; and it also got some other cases (such as minus
zero) wrong.  This led to observable cross-platform differences, as
reported in bug #12885 from Rich Schaaf; in particular, casting from
float to int didn't honor round-to-nearest-even on builds using rint.c.

Implement something that attempts to cover all cases per spec, and add
some simple regression tests so that we'll notice if any platforms still
get this wrong.

Although this is a bug fix, no back-patch, as a behavioral change in
the back branches was agreed not to be a good idea.

Pedro Gimeno Fortea, reviewed by Michael Paquier and myself

Even though the main benefit of the Lehman and Yao algorithm for
btrees is that no locks need be held between page reads in an
index search, we were holding a buffer pin on each leaf page after
it was read until we were ready to read the next one.  The reason
was so that we could treat this as a weak lock to create an
"interlock" with vacuum's deletion of heap line pointers, even
though our README file pointed out that this was not necessary for
a scan using an MVCC snapshot.

The main goal of this patch is to reduce the blocking of vacuum
processes by in-progress btree index scans (including a cursor
which is idle), but the code rearrangement also allows for one
less buffer content lock to be taken when a forward scan steps from
one page to the next, which results in a small but consistent
performance improvement in many workloads.

This patch leaves behavior unchanged for some cases, which can be
addressed separately so that each case can be evaluated on its own
merits.  These unchanged cases are when a scan uses a non-MVCC
snapshot, an index-only scan, and a scan of a btree index for which
modifications are not WAL-logged.  If later patches allow  all of
these cases to drop the buffer pin after reading a leaf page, then
the btree vacuum process can be simplified; it will no longer need
the "super-exclusive" lock to delete tuples from a page.

Reviewed by Heikki Linnakangas and Kyotaro Horiguchi

... which is set to the OID of a copied text search config, whenever the
COPY clause is used.

This is in the spirit of commit a2e35b53.

I failed to realize that server names reported in the object args array
would get quoted, which is wrong; remove that, making sure that it's
only quoted in the string-formatted identity.

This bug was introduced by my commit cf34e373, which was backpatched,
but since object name/args arrays are new in commit a6762014, there
is no need to backpatch this any further.

There are other comments in there that don't precisely match what's
implemented, but this one confused me enough to be worth fixing.

Report by Tom Lane

Previously only concatenation was recommended.

Report by Pavel Stehule

Also, fix pager enable selection for such cases, and other cleanups for
zero-column output.

Report by Thom Brown

Report by Álvaro Hernández Tortosa

Previously we called DirectFunctionCall3() with dummy arguments.

Patch by Jon Nelson

Report by Tom Lane

Also add regression test.  Previously this was documented to work, but
didn't.

ExecOpenScanRelation assumed that any relation listed in the ExecRowMark
list has been locked by InitPlan; but this is not true if the rel's
markType is ROW_MARK_COPY, which is possible if it's a foreign table.

In most (possibly all) cases, failure to acquire a lock here isn't really
problematic because the parser, planner, or plancache would have taken the
appropriate lock already.  In principle though it might leave us vulnerable
to working with a relation that we hold no lock on, and in any case if the
executor isn't depending on previously-taken locks otherwise then it should
not do so for ROW_MARK_COPY relations.

Noted by Etsuro Fujita.  Back-patch to all active versions, since the
inconsistency has been there a long time.  (It's almost certainly
irrelevant in 9.0, since that predates foreign tables, but the code's
still wrong on its own terms.)

Previously, CHECK constraints of the same scope were checked in whatever
order they happened to be read from pg_constraint.  (Usually, but not
reliably, this would be creation order for domain constraints and reverse
creation order for table constraints, because of differing implementation
details.)  Nondeterministic results of this sort are problematic at least
for testing purposes, and in discussion it was agreed to be a violation of
the principle of least astonishment.  Therefore, borrow the principle
already established for triggers, and apply such checks in name order
(using strcmp() sort rules).  This lets users control the check order
if they have a mind to.

Domain CHECK constraints still follow the rule of checking lower nested
domains' constraints first; the name sort only applies to multiple
constraints attached to the same domain.

In passing, I failed to resist the temptation to wordsmith a bit in
create_domain.sgml.

Apply to HEAD only, since this could result in a behavioral change in
existing applications, and the potential regression test failures have
not actually been observed in our buildfarm.

Noticed by Coverity

It worked in my Windows VM with VS2013, but buildfarm animal mastodon,
running MSVC 2005, was not happy. Amit Kapila also reported a similar error
earlier in his environment. Let's see if this helps.

Earlier versions of this tool were available (and still are) on github.

Thanks to Michael Paquier, Alvaro Herrera, Peter Eisentraut, Amit Kapila,
and Satoshi Nagayasu for review.

Recovery delays are implemented by waiting on a latch, and latches take
milliseconds as a parameter. The required amount of waiting was computed
using microsecond resolution though and the wait loop's abort condition
was checking the delay in microseconds as well.  This could lead to
short spurts of busy looping when the overall wait time was below a
millisecond, but above 0 microseconds.

Instead just formulate the wait loop's abort condition in millisecond
granularity as well. Given that that's recovery_min_apply_delay
resolution, it seems harmless to not wait for less than a millisecond.

Backpatch to 9.4 where recovery_min_apply_delay was introduced.

Discussion: 20150323141819.GH26995@alap3.anarazel.de