Content-Type: text/plain;
	charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 16.0 \(3864.300.41.1.7\))
Subject: Re: Fix uninitialized xl_running_xacts padding
From: Chao Li <li.evan.chao@gmail.com>
In-Reply-To: <aY74IKOgGLF5nj5A@ip-10-97-1-34.eu-west-3.compute.internal>
Date: Mon, 16 Feb 2026 07:39:09 +0800
Cc: Michael Paquier <michael@paquier.xyz>,
 Anthonin Bonnefoy <anthonin.bonnefoy@datadoghq.com>,
 PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Content-Transfer-Encoding: quoted-printable
Message-Id: <4AECB8A0-A04A-4CE5-9161-B8C28799D4CB@gmail.com>
References: <CAO6_Xqoxp7C+y0L==xZXH5V=9PjpBx4T9vJYs87EbxFp_9nwOA@mail.gmail.com>
 <aY7z0NH83hnaGR7k@paquier.xyz>
 <aY74IKOgGLF5nj5A@ip-10-97-1-34.eu-west-3.compute.internal>
To: Bertrand Drouvot <bertranddrouvot.pg@gmail.com>
Archived-At: <https://www.postgresql.org/message-id/4AECB8A0-A04A-4CE5-9161-B8C28799D4CB%40gmail.com>
Precedence: bulk


> On Feb 13, 2026, at 18:08, Bertrand Drouvot =
<bertranddrouvot.pg@gmail.com> wrote:
>=20
> Hi,
>=20
> On Fri, Feb 13, 2026 at 06:50:08PM +0900, Michael Paquier wrote:
>> On Fri, Feb 13, 2026 at 10:39:14AM +0100, Anthonin Bonnefoy wrote:
>>> The 3 bytes of padding after subxid_overflow were left =
uninitialized,
>>> leading to the random 'ca ce 9b' data being written in the WAL. The
>>> attached patch fixes the issue by zeroing the xl_running_xacts
>>> structure in LogCurrentRunningXacts using MemSet.
>>=20
>> This uninitialized padding exists for as long as this code exists,
>> down to efc16ea52067.  No objection here to clean up that on HEAD.
>=20
> It's not as important as when a struct which is used as an hash key =
has padding
> bytes uninitialized (and byte comparisons are done on the key) but I'm =
also
> +1 to make it "cleaner".
>=20
> Regards,
>=20
> --=20
> Bertrand Drouvot
> PostgreSQL Contributors Team
> RDS Open Source Databases
> Amazon Web Services: https://aws.amazon.com
>=20

I have no objection on cleanup the padding bytes. As the structure is =
small, maybe we can use {0} initializer:
```
xl_running_xacts xlrec =3D {0};
```
That will allow compilers to optimize the initialization. Anyway, =
that=E2=80=99s not a big deal, no strong opinion here.

Best regards,
--
Chao Li (Evan)
HighGo Software Co., Ltd.
https://www.highgo.com/