Horton

Re: refactor architecture-specific popcount code

John Naylor <johncnaylorls@gmail.com>

View thread View on postgr.es
From: John Naylor <johncnaylorls@gmail.com>
To: Nathan Bossart <nathandbossart@gmail.com>
Cc: Greg Burd <greg@burd.me>, Heikki Linnakangas <hlinnaka@iki.fi>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2026-02-05T07:48:44Z
Lists: pgsql-hackers 
On Thu, Feb 5, 2026 at 4:43 AM Nathan Bossart <nathandbossart@gmail.com> wrote:
> Sure.  I'm tempted to suggest that we only use the plain C version here,
> too.  The SSE4.2 bms_num_members() test I did yesterday used it and showed
> improvement at one word.  If we do that, we can rip out even more code
> since we no longer need the popcount built-ins.

Unlike the 32-bit case, people do run production on 64-bit platforms
that are not Arm/x86, so that would require effort to see if the
builtins are worth it for them. That seems like a separate effort. I
can help with that, but let's get the tested stuff in first.

> * tests plain C version on an Apple M3 *
>
> Yeah, the plain C version might be marginally slower than the built-in
> version for that test, but it still seems quite a bit faster than HEAD.
>
>     HEAD  v8  v10
>       40  25   29

That's good to know, and maybe it'll be true elsewhere.

--
John Naylor
Amazon Web Services

Commits

  1. Make use of pg_popcount() in more places.

  2. Remove uses of popcount builtins.

  3. Remove some unnecessary optimizations in popcount code.

  4. Remove specialized word-length popcount implementations.

  5. Move x86-64-specific popcount code to pg_popcount_x86.c.

  6. Refactor some SIMD and popcount macros.

  7. Rename "fast" and "slow" popcount functions.

  8. Rename pg_popcount_avx512.c to pg_popcount_x86.c.

  9. Remove trailing zero words from Bitmapsets

  10. Don't use _BitScanForward64/_BitScanReverse64 on 32-bit MSVC builds