linux-kernel - Re: [patch] mm: fix anon

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: <alpine.LFD.2.00.0810171717470.3438@nehalem.linux-foundation.org>
Date:	Fri, 17 Oct 2008 17:25:26 -0700 (PDT)
From:	Linus Torvalds <torvalds@...ux-foundation.org>
To:	Hugh Dickins <hugh@...itas.com>
cc:	Nick Piggin <npiggin@...e.de>, linux-kernel@...r.kernel.org,
	linux-mm@...ck.org
Subject: Re: [patch] mm: fix anon_vma races

On Sat, 18 Oct 2008, Hugh Dickins wrote:
> 
> But they're doing it to make the page's ptes accessible to
> memory reclaim, and the CPU doing memory reclaim will not
> (unless by coincidence) have done that anon_vma_prepare() -
> it's just reading the links which the faulters are providing.

Ahh. Good point. Then yes, those ones would really need the 
smp_read_barrier_depends() too.

Very annoying.

Of course, we _could_ just admit that the situation is really *really* 
unlikely, and there are probably something like five people running 
Linux/alpha, and that we don't care enough. With just the smp_wmb(), we 
cover all non-alpha people, since this is all through a pointer, so all 
the readers will inevitably be of the smp_read_barrier_depends kind.

If we want to guarantee the proper smp_read_barrier_depends() behaviour, 
then we'd need to find them all. Possibly by renaming the whole field 
(prepend an underscore like we usually do) and forcing all readers to use 
some "get_anon_vma(vma)" helper function, aka

  static inline struct anon_vma *get_anon_vma(struct vm_area_struct *vma)
  {
	struct anon_vma *anon_vma = vma->_anon_vma;
	smp_read_barrier_depends();
	return anon_vma;
  }

which would find them all.

Ugh. I really would have preferred not having something like that. 
Especially considering how limited that issue really is. Hmm. 

		Linus
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/