While the space is optional, it seems nicer to be consistent with what
you get if you do "SET search_path=...". SET always normalizes the
separator to be comma+space.

Christoph Martin

This reverts commit 083d29c6.

The commit changed the code so that it causes an errors when
IDENTIFY_SYSTEM returns three columns. But which prevents us
from using the replication-related utilities against the server
with older version. This is not what we want. For that
compatibility, we allow the utilities to receive three columns
as the result of IDENTIFY_SYSTEM eventhough it actually returns
four columns in 9.4 or later.

Pointed out by Andres Freund.

5a991ef8 added new column into
the result of IDENTIFY_SYSTEM command. But it was not reflected into
several codes checking that result. Specifically though the number of
columns in the result was increased to 4, it was still compared with 3
in some replication codes.

Back-patch to 9.4 where the number of columns in IDENTIFY_SYSTEM
result was increased.

Report from Michael Paquier

Programs need execute permission on a DLL file to load it.  MSYS
"install" ignores the mode argument, and our Cygwin build statically
links libpq into programs.  That explains the lack of buildfarm trouble.
Back-patch to 9.0 (all supported versions).

In support of this, have the MSVC build follow GNU make in preferring
GNUmakefile over Makefile when a directory contains both.

Michael Paquier, reviewed by MauMau.

strncmp() is a specialized API unsuited for routine copying into
fixed-size buffers.  On a system where the length of a single filename
can exceed MAXPGPATH, the pg_archivecleanup change prevents a simple
crash in the subsequent strlen().  Few filesystems support names that
long, and calling pg_archivecleanup with untrusted input is still not a
credible use case.  Therefore, no back-patch.

David Rowley

The aboriginal sample placed connection parameters in
groupOfUniqueNames/uniqueMember.  OpenLDAP, at least as early as version
2.4.23, rejects uniqueMember entries that do not conform to the syntax
for a distinguished name.  Use device/description, which is free-form.
Back-patch to 9.4 for web site visibility.

Oversight in commit 7cbe57c3.
Back-patch to 9.4, where that commit first appeared.  In passing,
release-note the FDW API postcondition change from the same commit.

Adjust Release Notes to reflect holding off wrapped expanded mode until 9.5 (thanks Michael Paquier)

collateral damage on other formats, by Sergey Muraviov.

Move the functions within the file so that public interface functions come
first, followed by internal functions. Previously, be_tls_write was first,
then internal stuff, and finally the rest of the public interface, which
clearly didn't make much sense.

Per Andres Freund's complaint.

The old text explained what happened if we didn't have working int64
arithmetic.  Since that case has been explicitly rejected by configure
since 8.4.3, documenting it in the 9.x branches can only produce confusion.

Commit f30015b6 made this happen for
timestamp and timestamptz, but it seems pretty inconsistent to not
do it for simple dates as well.

(In passing, I re-pgindent'd json.c.)

Previously only CREATE_REPLICATION_SLOT was exposed as an index term.
That's odd and there is no reason not to add index terms for other
replication commands.

Back-patch to 9.4.

Update the notes to include commits through today, and do a lot of
wordsmithing and markup adjustment.  Notably, don't use <link> where <xref>
will do; since we got rid of the text-format HISTORY file, there is no
longer a reason to avoid <xref>.

The new DISCARD SEQUENCES option was inadequately described, and hadn't
been mentioned at all in the initial Description paragraph.  Rather than
rectifying the latter the hard way, it seemed better to rewrite the
description as a summary, instead of having it basically duplicate
statements made under Parameters.  Be more consistent about the ordering
of the options, too.

The upstream stylesheets for man output insert a *roff comment for an
occurrence of an indexterm, for reasons that have apparently been lost
in history.  This, however, is done incorrectly and causes some
formatting problems.  This hasn't been an issue until now, but the
reorganization of indexterm elements inside variablelists has triggered
this issue.

The upstream fix (http://sourceforge.net/p/docbook/bugs/1340/) is to
drop indexterms altogether in man output, and so we'll do the same here.

PG_RETURN_BOOL() should only be used in functions following the V1 SQL
function API.  This coding accidentally fails to fail since letting the
compiler coerce the Datum representation of bool back to plain bool
does give the right answer; but that doesn't make it a good idea.

Back-patch to older branches just to avoid unnecessary code divergence.

Make lists of the names of all operators that are claimed to be commutator
pairs or negator pairs.  This is analogous to the existing queries that
make lists of all operator names appearing in particular opclass strategy
slots.  Unexpected additions to these lists are likely to be mistakes; had
we had these queries in place before, bug #11178 might've been prevented.

<@ and @> are each other's commutators, but they were incorrectly marked
as being each other's negators instead.  (This was actually questioned
in a comment in the original commit, but nobody followed through :-(.)
Per bug #11178 from Christian Pronovost.

In passing, fix some JSONB operator descriptions that were randomly
different from the phrasing of every other similar description.

catversion bump for pg_catalog contents change.

from commit e4c3c99a

Print a clear error message in this case, rather than wait for initdb
--sync-only to generate a "Permission denied" error.

The inheritance example in the manual implies that the capital of New York
is New York City, but in reality it's Albany.

George Hartzell

I removed the flag earlier, but missed a few references in jsonb.h.

When the TAP tests are run in-tree (make check), set the shared library
path using the appropriate environment variable, using a logic similar
to pg_regress, so that the right libraries are used.

See 3a9d430a.  A new one snuck in.

FreeBSD hasn't made any use of kern.ipc.semmap since 1.1, and newer
releases reject attempts to set it altogether; so stop recommending
that it be adjusted.  Per bug #11161.

Back-patch to all supported branches.  Before 9.3, also incorporate
commit 7a42dff4, which touches the same text and for some reason
was not back-patched at the time.

Fabien COELHO and Robert Haas

This provides a small but worthwhile speedup when sorting text, at least
in cases to which the sortsupport machinery applies.

Robert Haas and Peter Geoghegan

Previously the help message described that -m is an option for
"stop", "restart" and "promote" commands in pg_ctl. But actually
that's not an option for "promote". So this commit fixes that
incorrect description in the help message.

Back-patch to 9.3 where the incorrect description was added.

We don't use set elements (collection of books) anymore, so this is just
dead code.

parseRelOptions() tended to leak memory in the caller's context.  Most
of the time this doesn't really matter since the caller's context is
at most query-lifespan, and the function won't be invoked very many times.
However, when testing with CLOBBER_CACHE_RECURSIVELY, the same relcache
entry can get rebuilt a *lot* of times in one query, leading to significant
intraquery memory bloat if it has any reloptions.  Noted while
investigating a related report from Tomas Vondra.

In passing, get rid of some Asserts that are redundant with the one
done by deconstruct_array().

As with other patches to avoid leaks in CLOBBER_CACHE testing, it doesn't
really seem worth back-patching this.

When replacing rd_indexlist, rd_indexattr, etc, we neglected to pfree any
old value of these fields.  Under ordinary circumstances, the old value
would always be NULL, so this seemed reasonable enough.  However, in cases
where we're rebuilding a system catalog's relcache entry and another cache
flush occurs on that same catalog meanwhile, it's possible for the field to
not be NULL when we return to the outer level, because we already refilled
it while recovering from the inner flush.  This leads to a fairly small
session-lifespan leak in CacheMemoryContext.  In real-world usage the leak
would be too small to notice; but in testing with CLOBBER_CACHE_RECURSIVELY
the leakage can add up to the point of causing OOM failures, as reported by
Tomas Vondra.

The issue has been there a long time, but it only seems worth fixing in
HEAD, like the previous fix in this area (commit 078b2ed2).

This option is equivalent to --slot option which pg_receivexlog has
already supported, which specifies the replication slot to use for
WAL streaming. pg_recvlogical has already supported both options,
and this commit makes pg_receivexlog consistent with pg_recvlogical
regarding the slot option.

Back-patch to 9.4 where the slot option was added.

Michael Paquier

Some error messages complained about --init and --stop being used
whereas the --create and --drop are the correct verbs. Fix that.

Also a XLogRecPtr was tested in a boolean fashion instead of being
compared to InvalidXLogRecPtr.

Backpatch to 9.4 where pg_recvlogical was introduced.

Michael Paquier

Fabien COELHO

When doing logical decoding using START_LOGICAL_REPLICATION in a
walsender process the walsender sometimes was sending out keepalive
messages too frequently. Asking for feedback every time.

WalSndWaitForWal() sends out keepalive messages when it's waiting for
new WAL to be generated locally when it sees that the remote side
hasn't yet flushed WAL up to the local position. That generally is
good but causes problems if the remote side only writes but doesn't
flush changes yet. So check for both remote write and flush position.

Additionally we've asked for feedback to the keepalive message which
isn't warranted when waiting for WAL in contrast to preventing
timeouts because of wal_sender_timeout.

Complaint and patch by Steve Singer.