| From: | John Naylor <john(dot)naylor(at)enterprisedb(dot)com> |
|---|---|
| To: | Nathan Bossart <nathandbossart(at)gmail(dot)com> |
| Cc: | Thomas Munro <thomas(dot)munro(at)gmail(dot)com>, Andres Freund <andres(at)anarazel(dot)de>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: use ARM intrinsics in pg_lfind32() where available |
| Date: | 2022-08-29 08:19:22 |
| Message-ID: | CAFBsxsEouaTwbmpqV+EW2=wFbhw2vHRe26NQTRcd0=NaOFDy7A@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Mon, Aug 29, 2022 at 11:25 AM John Naylor
<john(dot)naylor(at)enterprisedb(dot)com> wrote:
> +static inline bool
> +vector32_is_highbit_set(const Vector32 v)
> +{
> +#ifdef USE_SSE2
> + return (_mm_movemask_epi8(v) & 0x8888) != 0;
> +#endif
> +}
>
> I'm not sure why we need this function -- AFAICS it just adds more
> work on x86 for zero benefit. For our present application, can we just
> cast to Vector8 (for Arm's sake) and call the 8-bit version?
It turns out MSVC animal drongo doesn't like this cast -- on x86 they
are the same underlying type. Will look into that as more results come
in.
--
John Naylor
EDB: http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Julien Rouhaud | 2022-08-29 09:00:12 | Re: Schema variables - new implementation for Postgres 15 |
| Previous Message | John Naylor | 2022-08-29 07:51:03 | Re: use ARM intrinsics in pg_lfind32() where available |