servers. AFAICT it's harmless at the moment because nothing can depend on
either, but as soon as we introduce an object type with such dependencies,
tableoid needs to be set or pg_dump will fail to interpret the dependencies
correctly. In theory, I guess the uninitialized garbage in tableoid could
cause the object to be mistaken for some other object with same OID as well.

This was unintentionally broken in 8.4 while tightening up checking of
ordinary non-Julian date inputs to forbid references to "year zero".
Per bug #5672 from Benjamin Gigot.

The previous patches failed to cover a lot of symlinks that are only
added in platform-specific cases.  Make the lists match what's in the
Makefile for each branch.

Josh Kupershmidt

These are just cosmetic and don't seem worth back-patching far.
I put them into 9.0 just because it was trivial to do so.

Also do some further work in the back branches, where quite a bit wasn't
covered by Magnus' original back-patch.

Backpatch to 8.2 as that's how far the structure looks the same.

for git. Change other references from cvs to git as well.

See git_topo_order instead.

Poking around for remaining occurrences of CVS keyword strings, I came
across one that apparently reflects the use of a $Revision: ...$ string
in the original input data.  Dunno why anybody would be using that in
an MTA's Received: lines, but there it is.  Put it back to the way that
it was originally, according to inspection of the CVS repo.

Not sure why these symlinks are removed here and not in the port/
Makefile, but I won't second-guess that choice right now.

This script is intended to substitute for cvs2cl in generating release
notes and scrutinizing what got back-patched to which branches.

Script by me.  Support for --since by Alex Hunsaker.

with Magnus's script to remove these.

The previous coding just terminated the COPY immediately after seeing
the EOF marker (-1 where a row field count is expected).  The expected
CopyDone or CopyFail message just got thrown away later, since we weren't
in COPY mode anymore.  This behavior complicated matters for the JDBC
driver, and arguably was the wrong thing in any case since a CopyFail
message after the marker wouldn't be honored.

Note that there is a behavioral change here: extra data after the EOF
marker was silently ignored before, but now it will cause an error.
Hence not back-patching, although this is arguably a bug.

Per report and patch by Kris Jurka.

the same number of columns expected by the insert.  This suggests that there
were extra parentheses that converted the intended column list into a row
expression.

Original patch by Marko Tiikkaja, rather heavily editorialized by me.

These checks are also present in objectaddress.c, so there's no need to
recheck here.

Per a question from Robert Haas.

since it can happen when a process fails to start when the system
is under high load.

Per several bug reports and many peoples investigation.

Back-patch to 8.4, which is as far back as the "deadman-switch"
for shared memory access exists.

copy-editing.

There was an incorrect Assert in hstoreValidOldFormat(), which would cause
immediate core dumps when attempting to work with pre-9.0 hstore data,
but of course only in an assert-enabled build.

Also, ghstore_decompress() incorrectly applied DatumGetHStoreP() to a datum
that wasn't actually an hstore, but rather a ghstore (ie, a gist signature
bitstring).  That used to be harmless, but could now result in misbehavior
if the hstore format conversion code happened to trigger.  In reality,
since ghstore is not marked toastable (and doesn't need to be), this
function is useless anyway; we can lobotomize it down to returning the
passed-in pointer.

Both bugs found by Andrew Gierth, though this isn't exactly his proposed
patch.

when fld is of composite type.  Per discussion of bug #5644 from Valentine
Gogichashvili.

new WAL arrives via streaming replication. This reduces the latency, and
also allows us to use a longer polling interval, which is good for energy
efficiency.

We still need to poll to check for the appearance of a trigger file, but
the interval is now 5 seconds (instead of 100ms), like when waiting for
a new WAL segment to appear in WAL archive.

dynamic pool of event handles, we can permanently assign one for each
shared latch. Thanks to that, we no longer need a separate shared memory
block for latches, and we don't need to know in advance how many shared
latches there is, so you no longer need to remember to update
NumSharedLatches when you introduce a new latch to the system.

some "can't happen" scenarios, and spinlocks should only be held for
a few instructions anyway. As pointed out by Fujii Masao.

In these cases a qual can get marked with the removable rel in its
required_relids, but this is just to schedule its evaluation correctly, not
because it really depends on the rel.  We were assuming that, in effect,
we could throw away *all* quals so marked, which is nonsense.  Tighten up
the logic to be a little more paranoid about which quals belong to the
outer join being considered for removal, and arrange for all quals that
don't belong to be updated so they will still get evaluated correctly.

Also fix another problem that happened to be exposed by this test case,
which was that make_join_rel() was failing to notice some cases where
a constant-false qual could be used to prove a join relation empty.  If it's
a pushed-down constant false, then the relation is empty even if it's an
outer join, because the qual applies after the outer join expansion.

Per report from Nathan Grange.  Back-patch into 9.0.