linux-kernel - RE: [PATCH 1/4] fs, afs: convert afs_cell.usage from atomic_t to refcount

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:   Wed, 22 Feb 2017 17:19:49 +0000
From:   "Reshetova, Elena" <elena.reshetova@...el.com>
To:     David Howells <dhowells@...hat.com>
CC:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-afs@...ts.infradead.org" <linux-afs@...ts.infradead.org>,
        "peterz@...radead.org" <peterz@...radead.org>,
        "gregkh@...uxfoundation.org" <gregkh@...uxfoundation.org>,
        Hans Liljestrand <ishkamiel@...il.com>,
        "Kees Cook" <keescook@...omium.org>,
        David Windsor <dwindsor@...il.com>
Subject: RE: [PATCH 1/4] fs, afs: convert afs_cell.usage from atomic_t to
 refcount_t

> Elena Reshetova <elena.reshetova@...el.com> wrote:
> 
> > refcount_t type and corresponding API should be
> > used instead of atomic_t when the variable is used as
> > a reference counter. This allows to avoid accidental
> > refcounter overflows that might lead to use-after-free
> > situations.
> 
> This causes an assertion failure because cells aren't immediately destroyed
> when their refcount reaches 0, but may be resurrected provided the cache lock
> is held.  However, attempting to increment the 0 refcount does nothing, not
> even giving a warning.

This is strange, it is supposed to give a warning I think now when it is even not inlined. 
Peter, am I confusing smth?

> 
> So please place a hold on this patch.  I will check the other AFS patches also.
> 

Thank you very much David for testing the patches! 
I guess for this one and other two patches it means that if we want to do the atomic_t --> refcount_t conversions, 
we need to do +1 on the whole counting scheme to avoid issues around reaching zero.  
Do you see this approach reasonable? I can give it a try, if it makes sense in your opinion. 

Best Regards,
Elena.