MIME-Version: 1.0
References: <be0339cc-1ae1-4892-9445-8e6d8995a44d@eisentraut.org> <47AEACEE-E5CC-40E4-8611-01F3A465AF16@yandex-team.ru>
In-Reply-To: <47AEACEE-E5CC-40E4-8611-01F3A465AF16@yandex-team.ru>
From: Hannu Krosing <hannuk@google.com>
Date: Tue, 18 Jun 2024 17:08:57 +0200
Message-ID: <CAMT0RQQ=2u5GATAMWqnfsW7baTZT_6QgRVrZXd8oFS8bk1kVbQ@mail.gmail.com>
Subject: Re: What is a typical precision of gettimeofday()?
To: "Andrey M. Borodin" <x4mmm@yandex-team.ru>
Cc: Peter Eisentraut <peter@eisentraut.org>, Ants Aasma <ants@cybertec.at>, gregsmithpgsql@gmail.com, 
	pgsql-hackers <pgsql-hackers@postgresql.org>
Content-Type: multipart/alternative; boundary="000000000000d1a092061b2b7673"
Archived-At: <https://www.postgresql.org/message-id/CAMT0RQQ%3D2u5GATAMWqnfsW7baTZT_6QgRVrZXd8oFS8bk1kVbQ%40mail.gmail.com>
Precedence: bulk

--000000000000d1a092061b2b7673
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

I plan to send patch to pg_test_timing in a day or two

the underlying time precision on modern linux seems to be

2 ns for some Intel CPUs
10 ns for Zen4
40 ns for ARM (Ampere)

---
Hannu


|


On Tue, Jun 18, 2024 at 7:48=E2=80=AFAM Andrey M. Borodin <x4mmm@yandex-tea=
m.ru>
wrote:

>
>
> > On 19 Mar 2024, at 13:28, Peter Eisentraut <peter@eisentraut.org> wrote=
:
> >
> > I feel that we don't actually have any information about this
> portability concern.  Does anyone know what precision we can expect from
> gettimeofday()?  Can we expect the full microsecond precision usually?
>
> At PGConf.dev Hannu Krossing draw attention to pg_test_timing module. I=
=E2=80=99ve
> tried this module(slightly modified to measure nanoseconds) on some
> systems, and everywhere I found ~100ns resolution (95% of ticks fall into
> 64ns and 128ns buckets).
>
> I=E2=80=99ll add cc Hannu, and also pg_test_timing module authors Ants an=
g Greg.
> Maybe they can add some context.
>
>
> Best regards, Andrey Borodin.

--000000000000d1a092061b2b7673
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">I plan to send patch to pg_test_timing in a day or two=C2=
=A0<br><br>the underlying time precision on modern linux seems to be<br><br=
>2 ns for some Intel CPUs<br>10 ns for Zen4<br>40 ns for ARM (Ampere)<br><b=
r>---<div>Hannu<br><br><br><br>|<br><br><br>=C2=A0</div></div><br><div clas=
s=3D"gmail_quote"><div dir=3D"ltr" class=3D"gmail_attr">On Tue, Jun 18, 202=
4 at 7:48=E2=80=AFAM Andrey M. Borodin &lt;<a href=3D"mailto:x4mmm@yandex-t=
eam.ru">x4mmm@yandex-team.ru</a>&gt; wrote:<br></div><blockquote class=3D"g=
mail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204=
,204,204);padding-left:1ex"><br>
<br>
&gt; On 19 Mar 2024, at 13:28, Peter Eisentraut &lt;<a href=3D"mailto:peter=
@eisentraut.org" target=3D"_blank">peter@eisentraut.org</a>&gt; wrote:<br>
&gt; <br>
&gt; I feel that we don&#39;t actually have any information about this port=
ability concern.=C2=A0 Does anyone know what precision we can expect from g=
ettimeofday()?=C2=A0 Can we expect the full microsecond precision usually?<=
br>
<br>
At PGConf.dev Hannu Krossing draw attention to pg_test_timing module. I=E2=
=80=99ve tried this module(slightly modified to measure nanoseconds) on som=
e systems, and everywhere I found ~100ns resolution (95% of ticks fall into=
 64ns and 128ns buckets).<br>
<br>
I=E2=80=99ll add cc Hannu, and also pg_test_timing module authors Ants ang =
Greg. Maybe they can add some context.<br>
<br>
<br>
Best regards, Andrey Borodin.</blockquote></div>

--000000000000d1a092061b2b7673--