linux-kernel - Re: [PATCH 1/1] futex: do not fail on invalid op

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: <CA+55aFw3kfhAh9G-UJfkAbyDeufqLikQ+9qxLh0ga+uuC_fa5g@mail.gmail.com>
Date:   Mon, 23 Oct 2017 06:48:46 -0400
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     Jiri Slaby <jslaby@...e.cz>
Cc:     Ingo Molnar <mingo@...hat.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Darren Hart <dvhart@...radead.org>
Subject: Re: [PATCH 1/1] futex: do not fail on invalid op

On Mon, Oct 23, 2017 at 5:31 AM, Jiri Slaby <jslaby@...e.cz> wrote:
>
> So let us soften the failure to print only a (ratelimited) message and
> return 0 silently in these cases until userspace keeps up.

No, please make it do that ratelimited warning, but then continue as
if things were ok - trying to approximate what the code used to do.

Don't make it return success without actually doing anything.

So just make it do

        oparg = 1 << (oparg & 31);

or something. That's what that sequence would have done at least on 32-bit x86.

Maybe that was what it happened to expect by random luck. And maybe it
didn't matter. But don't just say  "it worked" without doing anything.

              Linus