[COMMITTERS] pgsql: Use O_DIRECT in O_SYNC test of different size. Restructure O_DI

2011-01-15 Thread Bruce Momjian
Use O_DIRECT in O_SYNC test of different size. Restructure O_DIRECT error reporting to be more consistent. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=6dc15e3befaa6a3ff72633a2084ad1e1466edcde Modified Files -- src/tools/fsync

[COMMITTERS] pgsql: Reverse number of stars used for test_fsync details.

2011-01-15 Thread Bruce Momjian
Reverse number of stars used for test_fsync details. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=3eebb33dddcfe4ac0719b697c1ebd3694038054e Modified Files -- src/tools/fsync/test_fsync.c | 24 1 files c

[COMMITTERS] pgsql: In test_fsync, warn about options without o_direct that are not

2011-01-15 Thread Bruce Momjian
In test_fsync, warn about options without o_direct that are not used by Postgres, and cases where o_direct does not work with certain file systems. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=431605f666cfb223cd615ec8c63cbdea07295550 Modif

[COMMITTERS] pgsql: Move a couple of declarations to reflect where the routines real

2011-01-15 Thread Tom Lane
Move a couple of declarations to reflect where the routines really are. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=6ca452ba7fca14dad16425a56ffa1c8a93496b5f Modified Files -- src/include/commands/defrem.h |6 +++--- 1 files

[COMMITTERS] pgsql: Add .gitignore to silence git complaints about parser/scanner ou

2011-01-15 Thread Tom Lane
Add .gitignore to silence git complaints about parser/scanner output files. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=36750dcef58550c652cfff861f9aad057a391fb9 Modified Files -- src/backend/replication/.gitignore |3 +++ 1

[COMMITTERS] pgsql: Have test_fsync output details that fdatasync is the default

2011-01-15 Thread Bruce Momjian
Have test_fsync output details that fdatasync is the default wal_sync_method on Linux. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=001d3664e32c0d156215bbfeccea3272aaf17722 Modified Files -- src/tools/fsync/test_fsync.c |3

[COMMITTERS] pgsql: Restructure test_fync to use modular C so there is less duplicat

2011-01-15 Thread Bruce Momjian
Restructure test_fync to use modular C so there is less duplicate code and it can be enhanced easier. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=169516ad9395e91d206cbf5bf32c5d2fa34d4111 Modified Files -- src/tools/fsync/READM

[COMMITTERS] pgsql: Enumerate available tablespaces after starting the backup

2011-01-15 Thread Magnus Hagander
Enumerate available tablespaces after starting the backup This closes a race condition where if a tablespace was created after the enumeration happened but before the do_pg_start_backup() was called, the backup would be incomplete. Now that it's done while we are in backup mode, WAL replay will re

[COMMITTERS] pgsql: Improve output display of test_fsync.

2011-01-15 Thread Bruce Momjian
Improve output display of test_fsync. Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=3ab80cfe031b616638eb6956010dcc9cb6426631 Modified Files -- src/tools/fsync/test_fsync.c | 35 ++- 1 files chang

[COMMITTERS] pgsql: Apply patch for test_fsync to add tests for O_DIRECT.

2011-01-15 Thread Bruce Momjian
Apply patch for test_fsync to add tests for O_DIRECT. Adjusted patch by Josh Berkus Branch -- master Details --- http://git.postgresql.org/gitweb?p=postgresql.git;a=commitdiff;h=677b06ca462ec6fd98da9369a2eae6085c9d7fed Modified Files -- src/tools/fsync/Makefile |4 +-

[COMMITTERS] pgsql: Treat a WAL sender process that hasn't started streaming yet as

2011-01-15 Thread Heikki Linnakangas
Treat a WAL sender process that hasn't started streaming yet as a regular backend, as far as the postmaster shutdown logic is concerned. That means, fast shutdown will wait for WAL sender processes to exit before signaling bgwriter to finish. This avoids race conditions between a base backup stoppi