[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 02 Feb 2011 07:22:01 +0100
From: Mike Galbraith <efault@....de>
To: Milton Miller <miltonm@....com>
Cc: "Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
Peter Zijlstra <peterz@...radead.org>,
akpm@...ux-foundation.org, Anton Blanchard <anton@...ba.org>,
xiaoguangrong@...fujitsu.com, mingo@...e.hu, jaxboe@...ionio.com,
npiggin@...il.com, JBeulich@...ell.com, rusty@...tcorp.com.au,
torvalds@...ux-foundation.org, benh@...nel.crashing.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/3 v2] call_function_many: fix list delete vs add race
On Tue, 2011-02-01 at 14:00 -0800, Milton Miller wrote:
> On Tue, 1 Feb 2011 about 14:00:26 -0800, "Paul E. McKenney" wrote:
> > Starting with smp_call_function_many():
> >
> > o The check for refs is redundant:
> >
> > /* some callers might race with other cpus changing the mask */
> > if (unlikely(!refs)) {
> > csd_unlock(&data->csd);
> > return;
> > }
> >
> > The memory barriers and atomic functions in
> > generic_smp_call_function_interrupt() prevent the callback from
> > being reused before the cpumask bits have all been cleared, right?
>
> The issue is not the cpumask in the csd, but the mask passed in from the
> caller. If other cpus clear the mask between the cpumask_first and and
> cpumask_next above (where we established there were at least two cpus not
> ourself) and the cpumask_copy, then this can happen. Both Mike Galbraith
> and Jan Beulich saw this in practice (Mikes case was mm_cpumask(mm)).
Mine (and Jan's) is a flavor of one hit and fixed via copy in ia64.
http://git2.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=75c1c91cb92806f960fcd6e53d2a0c21f343081c
-Mike
--
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/
Powered by blists - more mailing lists