linux-kernel - Re: refcount_t: underflow; use-after-free with CIFS umount after scsi-misc commit ef2cc88e2a205b8a11a19e78db63a70d3728cdf5

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: <CAHk-=whyQ-Ets2v3bg7eP8GAGpzA7f2pqyiw5tPj8HFoxFcEjw@mail.gmail.com>
Date:   Sun, 8 Dec 2019 19:15:58 -0800
From:   Linus Torvalds <torvalds@...ux-foundation.org>
To:     Al Viro <viro@...iv.linux.org.uk>
Cc:     Arthur Marsh <arthur.marsh@...ernode.on.net>,
        SCSI development list <linux-scsi@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        CIFS <linux-cifs@...r.kernel.org>,
        "James E.J. Bottomley" <jejb@...ux.ibm.com>
Subject: Re: refcount_t: underflow; use-after-free with CIFS umount after
 scsi-misc commit ef2cc88e2a205b8a11a19e78db63a70d3728cdf5

On Sun, Dec 8, 2019 at 7:10 PM Linus Torvalds
<torvalds@...ux-foundation.org> wrote:
>
> You're right that at least the CI bots might want to disable it for
> bisecting. Or force a particular seed for RANDSTRUCT - I seem to
> recall that there was some way to make it be at least repeatable for
> any particular structure.

Yeah, if you build in the same directory and don't do a "distclean" or
"git clean -dqfx", the seed remains in
scripts/gcc-plgins/randomize_layout_seed.h across builds, and the
result then _should_ be repeatable.

But yes, it's the kind of noise that is likely not worth fighting for
bisection (or any random bug hunting) at all, and turning off
RANDSTRUCT is probably the right thing to do.

But I don't think Arthur had RANDSTRUCT enabled, so that should be fine.

            Linus