Horton

Re: WAL Insertion Lock Improvements

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Cc: Andres Freund <andres@anarazel.de>, Heikki Linnakangas <hlinnaka@iki.fi>, Nathan Bossart <nathandbossart@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Date: 2023-07-27T04:36:43Z
Lists: pgsql-hackers 
On Tue, Jul 25, 2023 at 04:43:16PM +0900, Michael Paquier wrote:
> We really need to do something in terms of documentation with
> something like 0002, so I'll try to look at that next.

I have applied a slightly-tweaked version of 0002 as of 66d86d4 to
improve a bit the documentation of the area, and switched the CF entry
as committed.

(I got interested in what Andres has seen on his latest AIO branch, so
I have a few extra benchmarks running in the background on HEAD, but
nothing able to freeze all the backends yet waiting for a variable
update.  These are still running now.)
--
Michael

Commits

  1. Optimize pg_atomic_exchange_u32 and pg_atomic_exchange_u64.

  2. Document more assumptions of LWLock variable changes with WAL inserts

  3. Optimize WAL insertion lock acquisition and release with some atomics

  4. Avoid the use of a separate spinlock to protect a LWLock's wait queue.