linux-kernel - Re: [PATCH v10 0/5] Introduce mseal

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: <56001.1715726927@cvs.openbsd.org>
Date: Tue, 14 May 2024 16:48:47 -0600
From: "Theo de Raadt" <deraadt@...nbsd.org>
To: Matthew Wilcox <willy@...radead.org>
cc: Jonathan Corbet <corbet@....net>,
    Andrew Morton <akpm@...ux-foundation.org>, jeffxu@...omium.org,
    keescook@...omium.org, jannh@...gle.com, sroettger@...gle.com,
    gregkh@...uxfoundation.org, torvalds@...ux-foundation.org,
    usama.anjum@...labora.com, Liam.Howlett@...cle.com,
    surenb@...gle.com, merimus@...gle.com, rdunlap@...radead.org,
    jeffxu@...gle.com, jorgelo@...omium.org, groeck@...omium.org,
    linux-kernel@...r.kernel.org, linux-kselftest@...r.kernel.org,
    linux-mm@...ck.org, pedro.falcato@...il.com, dave.hansen@...el.com,
    linux-hardening@...r.kernel.org
Subject: Re: [PATCH v10 0/5] Introduce mseal

Matthew Wilcox <willy@...radead.org> wrote:

> > Not taking a position on merging, but I have to ask: are we convinced at
> > this point that mseal() isn't a chrome-only system call?  Did we ever
> > see the glibc patches that were promised?
> 
> I think _this_ version of mseal() is OpenBSD's mimmutable() with a
> basically unused extra 'flags' argument.  As such, we have an existance
> proof that it's useful beyond Chrome.

Yes, it is close enough.

> I think Liam still had concerns around the
> walk-the-vmas-twice-to-error-out-early part of the implementation?
> Although we can always fix the implementation later; changing the API
> is hard.

Yes I am a bit worried about the point Liam brings up -- we've discussed
it privately at length.  Matthew, to keep it short I have a different
viewpoint:

Some of the Linux m* system calls have non-conforming, partial-work-then-return-error
behaviour.  I cannot find anything like this in any system call in any other
operating system, and I believe there is a defacto rule against doing this, and
Linux has an optimization which violating this, and I think it could be fixed
with fairly minor expense, and can't imagine it affecting a single application.

I worry that the non-atomicity will one day be used by an attacker.