linux-hardening - Re: [PATCH] vfs: Replace all non-returning strlcpy with strscpy

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 for Android: free password hash cracker in your pocket

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

Date:   Tue, 16 May 2023 13:37:16 -0700
From:   Kees Cook <keescook@...omium.org>
To:     Christian Brauner <brauner@...nel.org>
Cc:     Azeem Shaikh <azeemshaikh38@...il.com>,
        linux-hardening@...r.kernel.org, linux-fsdevel@...r.kernel.org,
        linux-kernel@...r.kernel.org,
        Alexander Viro <viro@...iv.linux.org.uk>
Subject: Re: [PATCH] vfs: Replace all non-returning strlcpy with strscpy

On Mon, May 15, 2023 at 09:50:25AM +0200, Christian Brauner wrote:
> On Wed, 10 May 2023 22:11:19 +0000, Azeem Shaikh wrote:
> > strlcpy() reads the entire source buffer first.
> > This read may exceed the destination size limit.
> > This is both inefficient and can lead to linear read
> > overflows if a source string is not NUL-terminated [1].
> > In an effort to remove strlcpy() completely [2], replace
> > strlcpy() here with strscpy().
> > No return values were used, so direct replacement is safe.
> > 
> > [...]
> 
> I sincerely hope we'll be done with swapping out various string
> functions for one another at some point. Such patches always seems
> benign and straightforward but the potential for subtle bugs is
> feels rather high...

Agreed. The long-term goal is to remove "strlcpy"[1] and "strncpy"[2]
completely from the kernel, leaving only "strscpy". From there, I hope
to do a treewide change to scrub the kernel of the pattern:
	strscpy(fixed-sized-dest, fixed-sized-source, sizeof(fixed-size-dest))
and replace it with a much stricter "strcpy" that refuses to work on
dynamically sized arguments. This will get us away from the pointless
exercise of duplicating sizeof() arguments when the compiler can very
happily do it itself.

But doing the return value transitions (and padding checks) for strlcpy
and strncpy need to happen first. It's a long road.

> Applied to the vfs.misc branch of the vfs/vfs.git tree.
> Patches in the vfs.misc branch should appear in linux-next soon.

Thanks!

-Kees

[1] https://github.com/KSPP/linux/issues/89
[2] https://github.com/KSPP/linux/issues/90

-- 
Kees Cook