pgsql: Properly fix AVX-512 CRC calculation bug

From: John Naylor <john(dot)naylor(at)postgresql(dot)org>
To: pgsql-committers(at)lists(dot)postgresql(dot)org
Subject: pgsql: Properly fix AVX-512 CRC calculation bug
Date: 2025-06-23 11:05:59
Message-ID: E1uTezv-002yZk-0B@gemulon.postgresql.org
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-committers

Properly fix AVX-512 CRC calculation bug

The problem that led to the workaround in f83f14881c7 was not in fact
a compiler bug, but a failure to zero the upper bits of the vector
register containing the initial scalar CRC value. Fix that and revert
the workaround.

Diagnosed-by: Nathan Bossart <nathandbossart(at)gmail(dot)com>
Diagnosed-by: Raghuveer Devulapalli <raghuveer(dot)devulapalli(at)intel(dot)com>
Tested-by: Andy Fan <zhihuifan1213(at)163(dot)com>
Tested-by: Soumyadeep Chakraborty <soumyadeep2007(at)gmail(dot)com>
Reviewed-by: Nathan Bossart <nathandbossart(at)gmail(dot)com>
Reviewed-by: Raghuveer Devulapalli <raghuveer(dot)devulapalli(at)intel(dot)com>
Discussion: https://postgr.es/m/PH8PR11MB82866B07AA6758D12F699C00FB70A@PH8PR11MB8286.namprd11.prod.outlook.com

Branch
------
master

Details
-------
https://git.postgresql.org/pg/commitdiff/43da394304fba820830da2cef2c0214fe292c037

Modified Files
--------------
src/port/pg_crc32c_sse42.c | 2 +-
src/port/pg_crc32c_sse42_choose.c | 4 +---
2 files changed, 2 insertions(+), 4 deletions(-)

Browse pgsql-committers by date

  From Date Subject
Next Message Jelte Fennema-Nio 2025-06-23 11:59:47 Re: pgsql: Improve runtime and output of tests for replication slots checkp
Previous Message Alexander Korotkov 2025-06-23 10:24:30 Re: pgsql: Improve runtime and output of tests for replication slots checkp