linux-kernel - Re: [PATCH] ext4: fix OOB read when checking dotdot dir

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAHk-=wgzYVZ0ZNvcqC+yToX6nFx+SNZqTcyEvzm2RMP-TU-Dqw@mail.gmail.com>
Date: Wed, 19 Mar 2025 09:28:56 -0700
From: Linus Torvalds <torvalds@...uxfoundation.org>
To: "Theodore Ts'o" <tytso@....edu>
Cc: Jakub Acs <acsjakub@...zon.com>, linux-kernel@...r.kernel.org, 
	linux-ext4@...r.kernel.org, Andreas Dilger <adilger.kernel@...ger.ca>, 
	Mahmoud Adam <mngyadam@...zon.com>, stable@...r.kernel.org, security@...nel.org
Subject: Re: [PATCH] ext4: fix OOB read when checking dotdot dir

On Wed, 19 Mar 2025 at 06:06, Theodore Ts'o <tytso@....edu> wrote:
>
> I'd change the check to:
>
>         else if (unlikely(next_offset == size && de->name_len == 1 &&
>                           strcmp(".", de->name) == 0))
>
> which is a bit more optimized.

Why would you use 'strcmp()' when you just checked that the length is one?

IOW, if you are talking about "a bit more optimized", please just check

        de->name[0] == '.'

after you've checked that the length is 1.

No?

             Linus