RE: CRC32C Parallel Computation Optimization on ARM

From: "Devulapalli, Raghuveer" <raghuveer(dot)devulapalli(at)intel(dot)com>
To: John Naylor <johncnaylorls(at)gmail(dot)com>
Cc: Dmitry Dolgov <9erthalion6(at)gmail(dot)com>, Nathan Bossart <nathandbossart(at)gmail(dot)com>, Xiang Gao <Xiang(dot)Gao(at)arm(dot)com>, Michael Paquier <michael(at)paquier(dot)xyz>, "pgsql-hackers(at)lists(dot)postgresql(dot)org" <pgsql-hackers(at)lists(dot)postgresql(dot)org>
Subject: RE: CRC32C Parallel Computation Optimization on ARM
Date: 2025-03-11 17:46:04
Message-ID: PH8PR11MB82869D313C869B49BAFC5B52FBD12@PH8PR11MB8286.namprd11.prod.outlook.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Hi John,

I am happy to submit a patch with a C fallback version that leverages the specific algorithm/technique mentioned in the white paper to make it clear that Intel has contributed this specific technique to Postgres under Postgres license terms. That should hopefully address any lingering concerns anyone may have w.r.t using this technique for the benefit of Postgres.

Raghuveer

> -----Original Message-----
> From: John Naylor <johncnaylorls(at)gmail(dot)com>
> Sent: Monday, March 10, 2025 6:31 PM
> To: Devulapalli, Raghuveer <raghuveer(dot)devulapalli(at)intel(dot)com>
> Cc: Dmitry Dolgov <9erthalion6(at)gmail(dot)com>; Nathan Bossart
> <nathandbossart(at)gmail(dot)com>; Xiang Gao <Xiang(dot)Gao(at)arm(dot)com>; Michael
> Paquier <michael(at)paquier(dot)xyz>; pgsql-hackers(at)lists(dot)postgresql(dot)org
> Subject: Re: CRC32C Parallel Computation Optimization on ARM
>
> On Tue, Mar 11, 2025 at 3:36 AM Devulapalli, Raghuveer
> <raghuveer(dot)devulapalli(at)intel(dot)com> wrote:
> >
> > Hi John,
> >
> > > On the other hand, looking at Linux kernel sources, it seems a patch
> > > using this technique was contributed by Intel over a decade ago:
> > >
> > > https://github.com/torvalds/linux/blob/master/arch/x86/crypto/crc32c
> > > -pcl-intel-
> > > asm_64.S
> > >
> > > So one more thing to ask our friends at Intel.
> >
> > Intel has contributed SSE4.2 CRC32C [1] and AVX-512 CRC32C [2] based on
> similar techniques to postgres.
> >
> > [1]
> > https://www.postgresql.org/message-
> id/PH8PR11MB8286F844321BA1DEEC51834
> > 8FBFD2(at)PH8PR11MB8286(dot)namprd11(dot)prod(dot)outlook(dot)com
> > [2]
> > https://www.postgresql.org/message-
> id/BL1PR11MB530401FA7E9B1CA432CF9DC
> > 3DC192(at)BL1PR11MB5304(dot)namprd11(dot)prod(dot)outlook(dot)com
>
> No, these are not similar at all. I gave you the paper name and the patents cited
> therein here:
>
> https://www.postgresql.org/message-
> id/CANWCAZbkt89_fVAaCAGBMznwA_xh%3D2Ci5q4GZytZHKjZAEjCRQ%40mail.g
> mail.com
>
> --
> John Naylor
> Amazon Web Services

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Dagfinn Ilmari Mannsåker 2025-03-11 17:52:17 Re: Non-text mode for pg_dumpall
Previous Message Peter Eisentraut 2025-03-11 17:43:48 Re: NOT ENFORCED constraint feature