[COMMITTERS] pgsql: Fix excessive memory consumption in the new sort pre-reading cod

2016-10-05 Thread Heikki Linnakangas
Fix excessive memory consumption in the new sort pre-reading code. LogicalTapeRewind() should not allocate large read buffer, if the tape is completely empty. The calling code relies on that, for its calculation of how much memory to allocate for the read buffers. That lead to massive overallocati

Re: [COMMITTERS] pgsql: Change the way pre-reading in external sort's merge phase works.

2016-10-05 Thread Heikki Linnakangas
On 10/06/2016 04:23 AM, Peter Geoghegan wrote: I noticed a simple oversight in this patch. It looks like you missed one place where state->maxTapes ought to be replaced with numInputTapes -- the loop that calls LogicalTapeAssignReadBufferSize() needs that changed too, in order to continue to resp

[COMMITTERS] pgsql: Remove -Wl,-undefined,dynamic_lookup in macOS build.

2016-10-05 Thread Tom Lane
Remove -Wl,-undefined,dynamic_lookup in macOS build. We don't need this anymore, and it prevents build-time error checking that's usually good to have, so remove it. Undoes one change of commit cac765820. Unfortunately, it's much harder to get a similar effect on other common platforms, because

[COMMITTERS] pgsql: Try to fix python shlib probe for MinGW.

2016-10-05 Thread Tom Lane
Try to fix python shlib probe for MinGW. Per discussion with Andrew Dunstan, it seems there are three peculiarities of the Python installation on MinGW (or at least, of the instance on buildfarm member frogmouth). First, the library name doesn't contain "2.7" but just "27". It looks like we can

Re: [COMMITTERS] pgsql: Change the way pre-reading in external sort's merge phase works.

2016-10-05 Thread Peter Geoghegan
On Mon, Oct 3, 2016 at 3:38 AM, Heikki Linnakangas wrote: > Change the way pre-reading in external sort's merge phase works. I noticed a simple oversight in this patch. It looks like you missed one place where state->maxTapes ought to be replaced with numInputTapes -- the loop that calls LogicalT

[COMMITTERS] pgsql: Update obsolete comments and perldoc.

2016-10-05 Thread Robert Haas
Update obsolete comments and perldoc. Loose ends from commit 2a0f89cd717ce6d49cdc47850577823682167e87. Daniel Gustafsson Branch -- master Details --- http://git.postgresql.org/pg/commitdiff/61f9e7ba3c4e4bde887f980b9316fb818ede59b6 Modified Files -- src/test/perl/PostgresNod

[COMMITTERS] pgsql: Update obsolete comments and perldoc.

2016-10-05 Thread Robert Haas
Update obsolete comments and perldoc. Loose ends from commit 2a0f89cd717ce6d49cdc47850577823682167e87. Daniel Gustafsson Branch -- REL9_6_STABLE Details --- http://git.postgresql.org/pg/commitdiff/bfcd07bd8fdf1ddaba375aa5b3aea419a898 Modified Files -- src/test/perl/Post

[COMMITTERS] pgsql: In python shlib probe, cater for OpenBSD, which omits the .so sy

2016-10-05 Thread Tom Lane
In python shlib probe, cater for OpenBSD, which omits the .so symlink. Most Unix-oid platforms provide a symlink "libfoo.so" -> "libfoo.so.n.n" to allow the linker to resolve a reference "-lfoo" to a version-numbered shared library. OpenBSD has apparently hacked ld(1) to do this internally, becau

Re: [COMMITTERS] pgsql: Huh, we do need to look in $python_configdir for the Python shli

2016-10-05 Thread Tom Lane
I wrote: > Andrew Dunstan writes: >> Looks like there are still problems on Windows/mingw and OpenBSD - see >> frogmouth and curculio on the buildfarm. > Yeah. I just sent you an offlist request to look into what's happening > on frogmouth, and I'm talking to curculio's owner as well. The answ

Re: [COMMITTERS] pgsql: Huh, we do need to look in $python_configdir for the Python shli

2016-10-05 Thread Tom Lane
Andrew Dunstan writes: > Looks like there are still problems on Windows/mingw and OpenBSD - see > frogmouth and curculio on the buildfarm. Yeah. I just sent you an offlist request to look into what's happening on frogmouth, and I'm talking to curculio's owner as well. > Do we need to fall back

Re: [COMMITTERS] pgsql: Huh, we do need to look in $python_configdir for the Python shli

2016-10-05 Thread Andrew Dunstan
On 10/04/2016 04:38 PM, Tom Lane wrote: Huh, we do need to look in $python_configdir for the Python shlib. Debian does it that way, for no doubt what seems to them a good reason. Thanks to Aidan Van Dyk for confirmation. Looks like there are still problems on Windows/mingw and OpenBSD - s

Re: [COMMITTERS] pgsql: Extend framework from commit 53be0b1ad to report latch waits.

2016-10-05 Thread Amit Kapila
On Wed, Oct 5, 2016 at 5:53 PM, Amit Kapila wrote: > On Tue, Oct 4, 2016 at 8:32 PM, Robert Haas wrote: >> Extend framework from commit 53be0b1ad to report latch waits. >> > > I am facing windows below compilation error: > > 1>pgstat.h(726): warning C4005: 'WAIT_TIMEOUT' : macro redefinition > 1>

[COMMITTERS] pgsql: Re-alphabetize #include directives.

2016-10-05 Thread Robert Haas
Re-alphabetize #include directives. Thomas Munro Branch -- master Details --- http://git.postgresql.org/pg/commitdiff/eb3bc0bd1a68e30f09f36937da5aa5338b3b994c Modified Files -- src/backend/postmaster/bgworker.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) -- S

Re: [COMMITTERS] pgsql: Extend framework from commit 53be0b1ad to report latch waits.

2016-10-05 Thread Amit Kapila
On Tue, Oct 4, 2016 at 8:32 PM, Robert Haas wrote: > Extend framework from commit 53be0b1ad to report latch waits. > I am facing windows below compilation error: 1>pgstat.h(726): warning C4005: 'WAIT_TIMEOUT' : macro redefinition 1>c:\Program Files (x86)\Microsoft SDKs\Windows\v7.0A\include\wine

[COMMITTERS] pgsql: Rename WAIT_* constants to PG_WAIT_*.

2016-10-05 Thread Robert Haas
Rename WAIT_* constants to PG_WAIT_*. Windows apparently has a constant named WAIT_TIMEOUT, and some of these other names are pretty generic, too. Insert "PG_" at the front of each name in order to disambiguate. Michael Paquier Branch -- master Details --- http://git.postgresql.org/pg/