Horton

Re: PGPROC alignment (was Re: pgsql: Separate RecoveryConflictReasons from procsignals)

Bertrand Drouvot <bertranddrouvot.pg@gmail.com>

View thread View on postgr.es
From: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
To: Heikki Linnakangas <hlinnaka@iki.fi>
Cc: Andres Freund <andres@anarazel.de>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2026-02-11T10:58:41Z
Lists: pgsql-hackers 
Hi,

On Wed, Feb 11, 2026 at 12:03:51PM +0200, Heikki Linnakangas wrote:
> On 11/02/2026 06:40, Bertrand Drouvot wrote:
> > > With this, sizeof(PGPROC) == 864 without the explicit alignment to
> > > PG_CACHE_LINE_SIZE, and 896 with it.
> > 
> > I can see 876 -> 896 on my side:
> > 
> > /*    872      |       4 */    uint32 wait_event_info;
> > /* XXX 20-byte padding   */
> > 
> >                                 /* total size (bytes):  896 */
> >                               }
> 
> Interesting. I've attached 'pahole bin/postgres' output from my laptop. It's
> Linux on arm64.

Thanks!

Got it: I was using "-DCACHEDEBUG -DWAL_DEBUG -DLOCK_DEBUG -DDEBUG_DEADLOCK"
so that with LOCK_DEBUG in place then LWLock size is 32 bytes (vs 16 in your case).

Regards,

-- 
Bertrand Drouvot
PostgreSQL Contributors Team
RDS Open Source Databases
Amazon Web Services: https://aws.amazon.com

Commits

  1. Use C11 alignas in typedef definitions

  2. Align PGPROC to cache line boundary

  3. Rearrange fields in PGPROC, for clarity

  4. Split PGPROC 'links' field into two, for clarity

  5. Remove useless store to local variable

  6. Separate RecoveryConflictReasons from procsignals