Horton

Re: refactor architecture-specific popcount code

Nathan Bossart <nathandbossart@gmail.com>

View thread View on postgr.es
From: Nathan Bossart <nathandbossart@gmail.com>
To: John Naylor <johncnaylorls@gmail.com>
Cc: Heikki Linnakangas <hlinnaka@iki.fi>, pgsql-hackers@postgresql.org
Date: 2026-01-30T21:33:19Z
Lists: pgsql-hackers

Attachments

On Fri, Jan 30, 2026 at 03:22:45PM +0700, John Naylor wrote:
> 0001 - I'm pretty sure this is comparable to HEAD if the optimized
> function is pg_popcount_sse42(). Has the AVX512 version been tested
> with 8-byte inputs? That seems to have a lot of pre- and
> post-processing involved. The inline wrapper only bypasses for 7 or
> less bytes.

Here [0] is the latest perf data I see for the AVX-512 popcount patch,
although that's comparing to v16, which IIRC lacks a few other inlining
tricks.  There's a chance the SSE4.2 version is faster at that particular
length.  I'm not sure we need to worry about that, but I can do a bit of
testing if you'd like.

> 0002
> - I tried running this on x86-64 with alignment sanitizer and no
> alarms went off during "make check", but adding
> pg_attribute_no_sanitize_alignment() would prevent surprises in the
> future.

Done.

> - I imagine that the old SIZEOF_VOID_P check is superfluous now, since
> the whole file is gated by HAVE_X86_64_POPCNTQ.

I think you're right.  There was some concern about this when I was first
adding the SSE4.2-specific pg_popcount() [1], but all the configure-time
checks for HAVE_X86_64_POPCNTQ are restricted to 64-bit x86, so I bet we
could safely assume SIZEOF_VOID_P == 8 in that file.

> - Maybe we can remove the aligned 32-bit path in
> pg_popcount_(masked_)portable(), since that's on-topic for this patch
> and would simplify things further.

IMHO that's a reasonable thing for us to do.

[0] https://postgr.es/m/20240404171828.GA3866970%40nathanxps13
[1] https://postgr.es/m/CAApHDvojPyh6dLKooqjXSZE%3D0Ed590Lq1BxF7WQ9knSggyuJEA%40mail.gmail.com

-- 
nathan

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