linux-kernel - Re: [GIT PULL] execve updates for v6.13-rc1

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] [day] [month] [year] [list]

Message-ID: <CAHk-=wjRE5S_vpQdRH-ZH2Q6SU1cmX0HhwzmfpjgYtoQAtok=Q@mail.gmail.com>
Date: Thu, 21 Nov 2024 14:38:15 -0800
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Kees Cook <kees@...nel.org>
Cc: Zbigniew Jędrzejewski-Szmek <zbyszek@...waw.pl>, 
	"Eric W. Biederman" <ebiederm@...ssion.com>, linux-kernel@...r.kernel.org, 
	Alexander Viro <viro@...iv.linux.org.uk>, 
	Christophe JAILLET <christophe.jaillet@...adoo.fr>, Dan Carpenter <dan.carpenter@...aro.org>, 
	Nir Lichtman <nir@...htman.org>, syzbot+03e1af5c332f7e0eb84b@...kaller.appspotmail.com, 
	Tycho Andersen <tandersen@...flix.com>, Vegard Nossum <vegard.nossum@...cle.com>
Subject: Re: [GIT PULL] execve updates for v6.13-rc1

On Thu, 21 Nov 2024 at 14:06, Kees Cook <kees@...nel.org> wrote:
>
> I think I finally figured out why you keep saying this. I think you mean
> to imply "ps -e" (or similar), not "ps". Asking for more process details
> ("ps a", "ps -f", "ps -e", etc) uses cmdline.

Ah. I never use plain 'ps'. The output is too useless.

> Since comm is mutable anyway, I feel like the "friendlier" default for
> userspace would be option 2.

The thing is, I still violently disagree.

I don't see what's "friendlier" in being (a) slower and (b) giving the
wrong output.

argv[0] isn't what we *normally* use.

And I've seen lots of cases where argv[0] is actually plain made-up garbage.

Christ, I went and looked at OUR OWN TEST-CASES, and they just happily
lie about "argv[0]".

Just go check tools/testing/selftests/exec/execveat.c, and see.

So  no. THERE IS NO WAY I WILL ACCEPT THE GARBAGE THAT IS ARGV[0].

What is so hard to understand about the fact that argv[0] has never
*EVER* been meaningful? We're not making it so now.

            Linus