MIME-Version: 1.0
References: 
 <CAOzEurSW8cNr6TPKsjrstnPfhf4QyQqB4tnPXGGe8N4e_v7Jig@mail.gmail.com>
 <CAN55FZ247JdiT8Sd1SRiyOJxk3Ei=pDCL4kpdP=HqLRjOhKf1Q@mail.gmail.com>
 <CAN55FZ2AxiwSah7TiQoMB==r=JKT0bOtooCB7ov4xRrGkVmJ1A@mail.gmail.com>
 <CAOzEurR5nFt=-SijfU7y0BHVcrT6RG9ovvdVfKt_uBZfEQew9w@mail.gmail.com>
 <CAOzEurSqgA69er9SzhPnXwmsVpO7-piUOuOy3dXcHOi__nSQcg@mail.gmail.com>
 <CA+K2RumC79NwWxBdofHOYo8SCSs0YCJic05Du=xOszRmoPf9FA@mail.gmail.com>
 <CAN55FZ0houfWHn8_MEEefhprZvc33jr07GrBYo+Bp2yw=TVnKA@mail.gmail.com>
 <CA+K2Ru=jHuz_Wpgar4Sobtxeb33qxx=o59ToOhZ=vpmkMqErnA@mail.gmail.com>
 <CAN55FZ1J+6eM=F5GreWEBMJcNV_gifYyYY1b6xpYzun=nWPhMQ@mail.gmail.com>
In-Reply-To: 
 <CAN55FZ1J+6eM=F5GreWEBMJcNV_gifYyYY1b6xpYzun=nWPhMQ@mail.gmail.com>
From: KAZAR Ayoub <ma_kazar@esi.dz>
Date: Thu, 21 Aug 2025 20:36:42 +0100
Message-ID: 
 <CA+K2RukEWfNAp821Fy1LYWCoE_fOKMU8efsP2VLb5ZM8OEETWA@mail.gmail.com>
Subject: Re: Speed up COPY FROM text/CSV parsing using SIMD
To: Nazir Bilal Yavuz <byavuz81@gmail.com>
Cc: Shinya Kato <shinya11.kato@gmail.com>, pgsql-hackers@postgresql.org
Content-Type: multipart/alternative; boundary="0000000000003b79de063ce536ec"
Archived-At: 
 <https://www.postgresql.org/message-id/CA%2BK2RukEWfNAp821Fy1LYWCoE_fOKMU8efsP2VLb5ZM8OEETWA%40mail.gmail.com>
Precedence: bulk

--0000000000003b79de063ce536ec
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

> On Thu, 14 Aug 2025 at 18:00, KAZAR Ayoub <ma_kazar@esi.dz> wrote:
> >> Thanks for running that benchmark! Would you mind sharing a reproducer
> >> for the regression you observed?
> >
> > Of course, I attached the sql to generate the text and csv test files.
> > If having a 1/3 of line length of special characters can be an
> exaggeration, something lower might still reproduce some regressions of
> course for the same idea.
>
> Thank you so much!
>
> I am able to reproduce the regression you mentioned but both
> regressions are %20 on my end. I found that (by experimenting) SIMD
> causes a regression if it advances less than 5 characters.
>
> So, I implemented a small heuristic. It works like that:
>
> - If advance < 5 -> insert a sleep penalty (n cycles).
> - Each time advance < 5, n is doubled.
> - Each time advance =E2=89=A5 5, n is halved.
>
> I am sharing a POC patch to show heuristic, it can be applied on top
> of v1-0001. Heuristic version has the same performance improvements
> with the v1-0001 but the regression is %5 instead of %20 compared to
> the master.
>
> --
> Regards,
> Nazir Bilal Yavuz
> Microsoft

Yes this is good, i'm also getting about 5% regression only now.


Regards,
Ayoub Kazar

--0000000000003b79de063ce536ec
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><div dir=3D"ltr"><br></div><div class=3D"=
gmail_quote"><blockquote class=3D"gmail_quote" style=3D"margin:0px 0px 0px =
0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Thu, 14 A=
ug 2025 at 18:00, KAZAR Ayoub &lt;<a href=3D"mailto:ma_kazar@esi.dz" target=
=3D"_blank">ma_kazar@esi.dz</a>&gt; wrote:<br>
&gt;&gt; Thanks for running that benchmark! Would you mind sharing a reprod=
ucer<br>
&gt;&gt; for the regression you observed?<br>
&gt;<br>
&gt; Of course, I attached the sql to generate the text and csv test files.=
<br>
&gt; If having a 1/3 of line length of special characters can be an exagger=
ation, something lower might still reproduce some regressions of course for=
 the same idea.<br>
<br>
Thank you so much!<br>
<br>
I am able to reproduce the regression you mentioned but both<br>
regressions are %20 on my end. I found that (by experimenting) SIMD<br>
causes a regression if it advances less than 5 characters.<br>
<br>
So, I implemented a small heuristic. It works like that:<br>
<br>
- If advance &lt; 5 -&gt; insert a sleep penalty (n cycles).<br>
- Each time advance &lt; 5, n is doubled.<br>
- Each time advance =E2=89=A5 5, n is halved.<br>
<br>
I am sharing a POC patch to show heuristic, it can be applied on top<br>
of v1-0001. Heuristic version has the same performance improvements<br>
with the v1-0001 but the regression is %5 instead of %20 compared to<br>
the master.<br>
<br>
--<br>
Regards,<br>
Nazir Bilal Yavuz<br>
Microsoft</blockquote><div>Yes this is good, i&#39;m also getting about 5% =
regression only now.</div><div><br><br><br></div><div>Regards,<br></div><di=
v>Ayoub Kazar</div></div></div>
</div>

--0000000000003b79de063ce536ec--