linux-kernel - Re: [PATCH] introduce CONFIG_BINFMT_ELF

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]

Date:	Sun, 20 Aug 2006 21:16:29 -0300
From:	Marcelo Tosatti <mtosatti@...hat.com>
To:	Solar Designer <solar@...nwall.com>
Cc:	Willy Tarreau <wtarreau@...a.kernel.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] introduce CONFIG_BINFMT_ELF_AOUT

On Sun, Aug 20, 2006 at 03:25:56AM +0400, Solar Designer wrote:
> Willy,
> 
> I propose the attached patch (extracted from 2.4.33-ow1) for inclusion
> into 2.4.34-pre.  (2.6 kernels could benefit from the same change, too.)
> 
> The patch adds a new compile-time option to control the support for
> "ELF binaries with a.out format interpreters or a.out libraries".
> Without this patch, such support is enabled on every system that enables
> the support for ELF binaries - although 99% (100%?) of systems don't
> need this hybrid functionality.  Moreover, this functionality poses a
> security risk - as proven in practice:
> 
> 	http://www.isec.pl/vulnerabilities/isec-0021-uselib.txt
> 
> This uselib() vulnerability did not affect default kernel builds with
> the -ow patch specifically due to separation of the unneeded/risky code
> into CONFIG_BINFMT_ELF_AOUT and having this option disabled by default.
> (Yes, this change in -ow patches pre-dates the discovery of the uselib()
> vulnerability.)
> 
> The patch also changes CONFIG_BINFMT_AOUT to be disabled by default on
> archs that had it default to enabled.  The a.out support is similarly
> risky and not audited/hardened with the same scrutiny that the ELF
> support has received.

I dislike this change. "Make a.out configurable" is a:

- "Hide the problems" trick, making it less likely for any potential bug to
be really fixed.

- Change not suitable for v2.4 inclusion: its not fixing _any_ serious problem.

We had this discussion before, didnt we?
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/