Re: add AVX2 support to simd.h

Поиск
Список
Период
Сортировка
От John Naylor
Тема Re: add AVX2 support to simd.h
Дата
Msg-id CANWCAZbY8fxmz6TRXmPz_s1VOZQQgTkfvayaXvWbqQnq2D849g@mail.gmail.com
обсуждение исходный текст
Ответ на Re: add AVX2 support to simd.h  (Nathan Bossart <nathandbossart@gmail.com>)
Ответы Re: add AVX2 support to simd.h  (Nathan Bossart <nathandbossart@gmail.com>)
Список pgsql-hackers
On Tue, Mar 19, 2024 at 9:03 AM Nathan Bossart <nathandbossart@gmail.com> wrote:
>
> On Sun, Mar 17, 2024 at 09:47:33AM +0700, John Naylor wrote:
> > I haven't looked at the patches, but the graphs look good.
>
> I spent some more time on these patches.  Specifically, I reordered them to
> demonstrate the effects on systems without AVX2 support.  I've also added a
> shortcut to jump to the one-by-one approach when there aren't many
> elements, as the overhead becomes quite noticeable otherwise.  Finally, I
> ran the same benchmarks again on x86 and Arm out to 128 elements.
>
> Overall, I think 0001 and 0002 are in decent shape, although I'm wondering
> if it's possible to improve the style a bit.

I took a brief look, and 0001 isn't quite what I had in mind. I can't
quite tell what it's doing with the additional branches and "goto
retry", but I meant something pretty simple:

- if short, do one element at a time and return
- if long, do one block unconditionally, then round the start pointer
up so that "end - start" is an exact multiple of blocks, and loop over
them



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Ajin Cherian
Дата:
Сообщение: Re: Proposal: Filter irrelevant change before reassemble transactions during logical decoding
Следующее
От: Nathan Bossart
Дата:
Сообщение: Re: add AVX2 support to simd.h