From: | Nathan Bossart <nathandbossart(at)gmail(dot)com> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | John Naylor <johncnaylorls(at)gmail(dot)com>, Ants Aasma <ants(at)cybertec(dot)at>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: add AVX2 support to simd.h |
Date: | 2024-01-02 22:00:18 |
Message-ID: | 20240102220018.GB1063425@nathanxps13 |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On Tue, Jan 02, 2024 at 12:50:04PM -0500, Tom Lane wrote:
> The patch needs better comments (as in, more than "none whatsoever").
Yes, will do.
> Also, do you really want to structure the header so that USE_SSE2
> doesn't get defined? In that case you are committing to provide
> an AVX2 replacement every single place that there's USE_SSE2, which
> doesn't seem like a great thing to require. OTOH, maybe there's
> no choice given than we need a different definition for Vector8 and
> Vector32?
Yeah, the precedent is to use these abstracted types elsewhere so that any
SIMD-related improvements aren't limited to one architecture. There are a
couple of places that do explicitly check for USE_NO_SIMD, though. Maybe
there's an eventual use-case for using SSE2 intrinsics even when you have
AVX2 support, but for now, ensuring we have an AVX2 replacement for
everything doesn't seem particularly burdensome.
--
Nathan Bossart
Amazon Web Services: https://aws.amazon.com
From | Date | Subject | |
---|---|---|---|
Next Message | Michał Kłeczek | 2024-01-02 22:23:23 | Re: SET ROLE x NO RESET |
Previous Message | Nathan Bossart | 2024-01-02 21:45:44 | Re: verify predefined LWLocks have entries in wait_event_names.txt |