linux-kernel - Re: [PATCH] futex: Ensure get_futex_key

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.DEB.2.11.1410201106260.4121@nanos>
Date:	Mon, 20 Oct 2014 11:11:40 +0200 (CEST)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	Davidlohr Bueso <dave@...olabs.net>
cc:	Linus Torvalds <torvalds@...ux-foundation.org>,
	Catalin Marinas <catalin.marinas@....com>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	Matteo Franchin <Matteo.Franchin@....com>,
	Darren Hart <dvhart@...ux.intel.com>,
	Peter Zijlstra <peterz@...radead.org>,
	Ingo Molnar <mingo@...nel.org>,
	"Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
	Mike Galbraith <umgwanakikbuti@...il.com>
Subject: Re: [PATCH] futex: Ensure get_futex_key_refs() always implies a
 barrier

On Sat, 18 Oct 2014, Davidlohr Bueso wrote:
> On Sat, 2014-10-18 at 13:50 -0700, Linus Torvalds wrote:
> > On Sat, Oct 18, 2014 at 12:58 PM, Davidlohr Bueso <dave@...olabs.net> wrote:
> > >
> > > And [get/put]_futex_keys() shouldn't even be called for private futexes.
> > > The following patch had some very minor testing on a 60 core box last
> > > night, but passes both Darren's and perf's tests. So I *think* this is
> > > right, but lack of sleep and I overall just don't trust them futexes!
> > 
> > Hmm. I don't see the advantage of making the code more complex in
> > order to avoid the functions that are no-ops for the !fshared case?
> > 
> > IOW, as far as I can tell, this patch doesn't actually really *change*
> > anything. Am I missing something?
> 
> Right, all we do is avoid a NOP, but I don't see how this patch makes
> the code more complex. In fact, the whole idea is to make it easier to
> read and makes the key referencing differences between shared and
> private futexes crystal clear, hoping to mitigate future bugs. 

I tend to disagree. The current code is symetric versus get/drop and
you make it unsymetric by avoiding the drop call with a pointless
extra conditional in all call sites.

I really had to look twice to figure out that the patch is correct,
but I really cannot see any value and definitely have a hard time how
this makes the code clearer and would prevent future bugs.

I rather keep it symetric and document the NOP property for private
futexes in both get and drop.

Thanks,

	tglx


--
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/