This is more secure, and saves a redirect since we no longer accept
plain HTTP connections on the website.

References in code comments should probably be updated too, but
that doesn't seem to need back-patching, whereas this does.

Also, in the 9.2 branch, remove suggestion that you can get the
source code via FTP, since that service will be shut down soon.

Daniel Gustafsson, with a few additional changes by me

Discussion: https://postgr.es/m/9A2C89A7-0BB8-41A8-B288-8B7BD09D7D44@yesql.se

Using flex's -i switch to achieve case-insensitivity is not a very safe
practice, because the scanner's behavior may then depend on the locale
that flex was invoked in.  In the particular example at hand, that's
not academic: the possible matches for "FIRST" will be different in a
Turkish locale than elsewhere.  Do it the hard way instead, as our
other scanners do.

Also, drop use of -b -CF -p, because this scanner is only used when
parsing the contents of a GUC variable.  That's not done often, and
the amount of text to be parsed can be expected to be trivial, so
prioritizing scanner speed over code size seems like quite the wrong
tradeoff.  Using flex's default optimization options reduces the
size of syncrep_gram.o by more than 50%.

The case-insensitivity problem is new in HEAD (cf commit 3901fd70).
The poor choice of optimization flags exists also in 9.6, but it doesn't
seem important enough to back-patch.

Discussion: https://postgr.es/m/24403.1495225931@sss.pgh.pa.us

Mark any old hash indexes as invalid so that they don't get used, and
create a script to run REINDEX on all of them.  Without this, we'd
still try to use any upgraded hash indexes, but it would fail.

Amit Kapila, reviewed by me.  Per a suggestion from Tom Lane.

Discussion: http://postgr.es/m/CAA4eK1Jidtagm7Q81q-WoegOVgkotv0OxvHOjFxcvFRP4X=mSw@mail.gmail.com

Reported-by: Ntushar <tushar.ahuja@enterprisedb.com>
Author: Dilip Kumar <dilipbalaut@gmail.com>

If one host in a multi-host connection string times out, move on to
the next specified host instead of giving up entirely.

Takayuki Tsunakawa, reviewed by Michael Paquier.  I added
a minor adjustment to the documentation.

Discussion: http://postgr.es/m/0A3221C70F24FB45833433255569204D1F6F42F5@G01JPEXMBYT05

This makes it also work for replication connections.

Report and patch by Daisuke Higuchi.

Discussion: http://postgr.es/m/1803D792815FC24D871C00D17AE95905B1A34A@g01jpexmbkw24

Otherwise, set_plan_refs() can get applied to the same list
multiple times through different references, leading to chaos.

Amit Langote, Dilip Kumar, and Robert Haas, reviewed by Ashutosh
Bapat.  Original report by Sveinn Sveinsson.

Discussion: http://postgr.es/m/20170517141151.1435.79890@wrigleys.postgresql.org

This was evidently intended to match the struct's typedef name,
but it didn't quite.  Noted while testing find_typedefs.

Since commit e7b3349a, MergeAttributes
destructively modifies the input List, to which the caller's
CreateStmt still points.  One may wonder whether this was already a
bug, but commit f0e44751 made things
noticeably worse by adding additional destructive modifications so
that the caller's List might, in the case of creation a partitioned
table, no longer even be structurally valid.  Restore the status quo
ante by assigning the return value of MergeAttributes back to
stmt->tableElts in the caller.

In most of the places where DefineRelation is called, it doesn't
matter what stmt->tableElts points to here or whether it's valid or
not, because the caller doesn't use the statement for anything after
DefineRelation returns anyway.  However, ProcessUtilitySlow passes it
to EventTriggerCollectSimpleCommand, and that function tries to invoke
copyObject on it.  If any of the CreateStmt's substructure is invalid
at that point, undefined behavior will result.

One might wonder whether this whole area needs further revision -
perhaps DefineRelation() ought not to be destructively modifying the
caller-provided CreateStmt at all.  However, that would be a behavior
change for any event triggers using C code to inspect the CreateStmt,
so for now, just fix the crash.

Report by Amit Langote, who provided a somewhat different patch for it.

Discussion: http://postgr.es/m/bf6a39a7-100a-74bd-1156-3c16a1429d88@lab.ntt.co.jp

Different names were used between function declaration and definition.

Reported-by: Daniel Gustafsson

Fix for hot_standby=on change.

Reported-by: Huong Dangminh

Author: Huong Dangminh

Commit 8d3b9cce added extra arguments to the sendAuthRequest function,
but neglected this caller inside #ifdef USE_BSD_AUTH.

Per report from Pierre-Emmanuel André.

Discussion: https://www.postgresql.org/message-id/20170519090336.whzmjzrsap6ktbgg@digipea.digitick.local

Author: Masahiko Sawada <sawada.mshk@gmail.com>

Notably, m68k only needs 2-byte alignment. Per report from Christoph Berg.

Discussion: https://www.postgresql.org/message-id/20170517193957.fwntkgi6epuso5l2@msg.df7cb.de

This seemed like a good idea originally because there's no way to mark
a range partition as accepting NULL, but that now seems more like a
current limitation than something we want to lock down for all time.
For example, there's a proposal to add the notion of a default
partition which accepts all rows not otherwise routed, which directly
conflicts with the idea that a range-partitioned table should never
allow nulls anywhere.  So let's change this while we still can, by
putting the NOT NULL test into the partition constraint instead of
changing the column properties.

Amit Langote and Robert Haas, reviewed by Amit Kapila

Discussion: http://postgr.es/m/8e2dd63d-c6fb-bb74-3c2b-ed6d63629c9d@lab.ntt.co.jp

Daniel Gustafsson

It previously wrote out slot_name = '', which was incorrect.
Reported-by: NMasahiko Sawada <sawada.mshk@gmail.com>

When creating a subscription with slot_name = NONE, we failed to check
that also create_slot = false and enabled = false were set.  This
created an invalid subscription and could later lead to a crash if a
NULL slot name was accessed.  Add more checks around that for
robustness.
Reported-by: Ntushar <tushar.ahuja@enterprisedb.com>

perltidy run not included.

It's better to use the actual keynum here rather than 0, because
someday someone might try to make list partitioning work with
multiple partitioning columns.

Jeevan Ladhe

Discussion: http://postgr.es/m/CAOgcT0M6-mx+dSX47JGJuJP1CKr4XssBFVmKNETt0OZYWpFr+w@mail.gmail.com

Partially revert commit c079673d.
There were complaints that splitting switch descriptions would
complicate translation efforts.  There are probably ways to resolve
the formatting problem without doing that, but undo it while we're
discussing.

Jeevan Ladhe, with some changes by me.

Discussion: http://postgr.es/m/CAOgcT0NZ_30-pjBpW2OgneV1ammArHkZDZ8B_KFC3q+_Xb2H9A@mail.gmail.com

Add some tests for parsing different option combinations.  Fix some of
the resulting error messages for recent changes in option naming.

Author: Masahiko Sawada <sawada.mshk@gmail.com>

When stdin is a terminal, it's possible to end a COPY FROM STDIN with
a keyboard EOF signal (typically control-D), and then keep on issuing
SQL commands.  One would expect another COPY FROM STDIN to work as well,
but on some platforms it did not.  This turns out to be because we were
not resetting the stream's feof() flag, and BSD-ish versions of fread()
and fgets() won't attempt to read more data if that's set.

The misbehavior is observed on BSDen (including macOS), but not Linux,
Windows, or SysV-ish Unixen, which makes this a portability bug not
just a missing feature.

Add a clearerr() call to fix the behavior, and improve the prompt that's
issued when copying from a TTY to mention that EOF signals work.

It's been like this forever, so back-patch to all supported branches.

Thomas Munro

Discussion: https://postgr.es/m/CAEepm=0MCGfYf=JAMiYhO6JPtv9-3ZfBo8fcGeCZ8oMzaw+Z+Q@mail.gmail.com

We used to only check for a supported relkind on the subscriber during
replication, which is needed to ensure that the setup is valid and we
don't crash.  But it's also useful to tell the user immediately when
CREATE or ALTER SUBSCRIPTION is executed that the relation being added
to the subscription is not of a supported relkind.

Author: Petr Jelinek <petr.jelinek@2ndquadrant.com>
Reported-by: Ntushar <tushar.ahuja@enterprisedb.com>

Reformat various places in which pgindent will make a mess, and
fix a few small violations of coding style that I happened to notice
while perusing the diffs from a pgindent dry run.

There is one actual bug fix here: the need-to-enlarge-the-buffer code
path in icu_convert_case was obviously broken.  Perhaps it's unreachable
in our usage?  Or maybe this is just sadly undertested.

Commit 827d6f97 contained the same misunderstanding of hash_create's API
as commit 090010f2.  As in 5d00b764, remove the unnecessary layer of
memory context.  (This bug is less significant than the other one, since
the extra context would be under a relatively short-lived context, but
it's still a bug.)

This reverts commit 4a03f935.

Since commit 4e37b3e1, buildfarm member frogmouth has been failing
occasionally with symptoms indicating that some expected stats data is
getting dropped.  The reason that that commit changed the behavior seems
probably to be that more data is getting shoved at the collector in a short
span of time.  In current sources, the stats test's first session sends
about 9KB of data while exiting, which is probably the same as what was
sent just before wait_for_stats() in the previous test design.  But now,
the test's second session is starting up concurrently, and it sends another
2KB (presumably reflecting its initial catalog accesses).  Since frogmouth
is running on Windows XP, which reputedly has a default socket receive
buffer size of only 8KB, it is not very surprising if this has put us over
the threshold where the receive buffer can overflow and drop messages.

The same mechanism could very easily explain the intermittent stats test
failures we've been seeing for years, since background processes such
as the bgwriter will sometimes send data concurrently with all this, and
could thus cause occasional buffer overflows.

Hence, insert some code into pgstat_init() to increase the stats socket's
receive buffer size to 100KB if it's less than that.  (On failure, emit a
LOG message, but keep going.)  Modern systems seem to have default sizes
in the range of 100KB-250KB, but older platforms don't.  I couldn't find
any platforms that wouldn't accept 100KB, so in theory this won't cause
any portability problems.

If this is successful at reducing the buildfarm failure rate in HEAD,
we should back-patch it, because it's certain that similar buffer overflows
happen in the field on platforms with small buffer sizes.  Going forward,
there might be an argument for trying to increase the buffer size even
more, but let's take a baby step first.

Discussion: https://postgr.es/m/22173.1494788088@sss.pgh.pa.us

Thomas Munro, reviewed by Amit Langote

Discussion: http://postgr.es/m/CAEepm=15Jss-yhFApuKzxcoCuFnb8TR8iQiWMjG=CLYPx48QLw@mail.gmail.com

Paul Jungwirth, reviewed by Ashutosh Bapat.

Discussion: http://postgr.es/m/e24a6a6d-5670-739b-00f3-41a226a80f25@illuminatedcomputing.com

Modifying the permissions of a persistent file isn't really much nicer
than modifying its contents, even if git doesn't currently notice it.
Adjust the test script to make a copy and set the permissions of that
instead.

Michael Paquier, per a gripe from me.  Back-patch to 9.5 where these
tests were introduced.

Discussion: https://postgr.es/m/14836.1494885946@sss.pgh.pa.us

Masahiko Sawada