MIME-Version: 1.0
References: <CACG=ezaWg7_nt-8ey4aKv2w9LcuLthHknwCawmBgEeTnJrJTcw@mail.gmail.com>
 <ff143b24-a093-40da-9833-d36b83726bdf@iki.fi> <CAGjGUAKO1GCzG5wBMt5RosWo0PatgFpYY=Gjgt77tN2brNe=Bg@mail.gmail.com>
 <CACG=ezYokoiumOFnqUfg_ffHD5s8T+6iHYfzKLfa=QQ-1pNrBg@mail.gmail.com>
 <CACG=ezY9xq73jcX_EjVqx5-f90nbQ9PyhFCTW2fwFCS2wmNiFw@mail.gmail.com>
 <CACG=eza+27CfLBobJJccRhXrA3He6c1irAnoyTtSC1-z9UXLrg@mail.gmail.com>
 <CAPpHfduczcop9s6gKUpLGgFUe2y4ERGMJx6SS6Kp+s-kQPwMjg@mail.gmail.com>
 <CACG=ezbye4g_ERNqE=gBcvQ0YypRaVENhNUu8xrs4PL12UdnUA@mail.gmail.com>
 <CACG=ezaMncd0-BcGHBgsSR2eqHfrz9WznHGLKX8biz6zu-azGw@mail.gmail.com> <CALT9ZEHm138m2rTi6-N_Xcm0ibfbK0b7yysdPHj1v144YUOFQg@mail.gmail.com>
In-Reply-To: <CALT9ZEHm138m2rTi6-N_Xcm0ibfbK0b7yysdPHj1v144YUOFQg@mail.gmail.com>
From: Pavel Borisov <pashkin.elfe@gmail.com>
Date: Thu, 12 Sep 2024 16:25:53 +0400
Message-ID: <CALT9ZEG4OoQk8OUBCNY8cT3GuUvcQ0NrphHZ9H8--VjwM5VWUw@mail.gmail.com>
Subject: Re: POC: make mxidoff 64 bits
To: Maxim Orlov <orlovmg@gmail.com>
Cc: Alexander Korotkov <aekorotkov@gmail.com>, wenhui qiu <qiuwenhuifx@gmail.com>, 
	Heikki Linnakangas <hlinnaka@iki.fi>, Postgres hackers <pgsql-hackers@lists.postgresql.org>
Content-Type: multipart/alternative; boundary="000000000000d3b1180621eb35e2"
Archived-At: <https://www.postgresql.org/message-id/CALT9ZEG4OoQk8OUBCNY8cT3GuUvcQ0NrphHZ9H8--VjwM5VWUw%40mail.gmail.com>
Precedence: bulk

--000000000000d3b1180621eb35e2
Content-Type: text/plain; charset="UTF-8"

On Thu, 12 Sept 2024 at 16:09, Pavel Borisov <pashkin.elfe@gmail.com> wrote:

> Hi, Maxim!
>
> Previously we accessed offsets in shared MultiXactState without locks as
> 32-bit read is always atomic. But I'm not sure it's so when offset become
> 64-bit.
> E.g. GetNewMultiXactId():
>
> nextOffset = MultiXactState->nextOffset;
> is outside lock.
>
> There might be other places we do the same as well.
>
I think the replacement of plain assignments by
pg_atomic_read_u64/pg_atomic_write_u64 would be sufficient.

(The same I think is needed for the patchset [1])
[1]
https://www.postgresql.org/message-id/flat/CAJ7c6TMvPz8q+nC=JoKniy7yxPzQYcCTnNFYmsDP-nnWsAOJ2g@mail.gmail.com

Regards,
Pavel Borisov

--000000000000d3b1180621eb35e2
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"></div><br><div class=3D"gmail_quote"><div=
 dir=3D"ltr" class=3D"gmail_attr">On Thu, 12 Sept 2024 at 16:09, Pavel Bori=
sov &lt;<a href=3D"mailto:pashkin.elfe@gmail.com">pashkin.elfe@gmail.com</a=
>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" style=3D"margin:0px=
 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><di=
v dir=3D"ltr"><div dir=3D"ltr">Hi, Maxim!</div><div class=3D"gmail_quote"><=
div><br></div><div>Previously we accessed offsets in shared MultiXactState =
without locks as 32-bit read is always atomic. But I&#39;m not sure it&#39;=
s so when offset become 64-bit.</div><div>E.g.=C2=A0GetNewMultiXactId():</d=
iv><div><br></div><div>nextOffset =3D MultiXactState-&gt;nextOffset;</div><=
div>is outside lock.=C2=A0</div><div><br></div><div>There might be other pl=
aces we do the same as well.=C2=A0</div></div></div></blockquote><div><div>=
I think the replacement=C2=A0of plain assignments by pg_atomic_read_u64/pg_=
atomic_write_u64 would be sufficient.</div><div><br></div><div>(The same I =
think is needed for the patchset [1])</div></div><div>[1] <a href=3D"https:=
//www.postgresql.org/message-id/flat/CAJ7c6TMvPz8q+nC=3DJoKniy7yxPzQYcCTnNF=
YmsDP-nnWsAOJ2g@mail.gmail.com">https://www.postgresql.org/message-id/flat/=
CAJ7c6TMvPz8q+nC=3DJoKniy7yxPzQYcCTnNFYmsDP-nnWsAOJ2g@mail.gmail.com</a><br=
></div><div><br></div><div>Regards,</div><div>Pavel Borisov</div></div></di=
v>

--000000000000d3b1180621eb35e2--