pgsql: Inline initial comparisons in TestForOldSnapshot()

From: Kevin Grittner <kgrittn(at)postgresql(dot)org>
To: pgsql-committers(at)postgresql(dot)org
Subject: pgsql: Inline initial comparisons in TestForOldSnapshot()
Date: 2016-04-21 13:47:58
Message-ID: E1atExa-0001od-Dk@gemulon.postgresql.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers pgsql-hackers

Inline initial comparisons in TestForOldSnapshot()

Even with old_snapshot_threshold = -1 (which disables the "snapshot
too old" feature), performance regressions were seen at moderate to
high concurrency. For example, a one-socket, four-core system
running 200 connections at saturation could see up to a 2.3%
regression, with larger regressions possible on NUMA machines.
By inlining the early (smaller, faster) tests in the
TestForOldSnapshot() function, the i7 case dropped to a 0.2%
regression, which could easily just be noise, and is clearly an
improvement. Further testing will show whether more is needed.

Branch
------
master

Details
-------
http://git.postgresql.org/pg/commitdiff/11e178d0dc4bc2328ae4759090b3c48b07023fab

Modified Files
--------------
src/backend/storage/buffer/bufmgr.c | 28 +++++-----------------------
src/include/storage/bufmgr.h | 32 +++++++++++++++++++++++++++++++-
2 files changed, 36 insertions(+), 24 deletions(-)

Responses

Browse pgsql-committers by date

  From Date Subject
Next Message Kevin Grittner 2016-04-21 14:16:18 Re: Re: [COMMITTERS] pgsql: Avoid extra locks in GetSnapshotData if old_snapshot_threshold <
Previous Message Robert Haas 2016-04-21 03:54:30 pgsql: postgres_fdw: Don't push down certain full joins.

Browse pgsql-hackers by date

  From Date Subject
Next Message Craig Ringer 2016-04-21 13:57:11 Re: Wire protocol compression
Previous Message Andres Freund 2016-04-21 13:45:05 Re: [BUGS] Breakage with VACUUM ANALYSE + partitions