[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 10 Oct 2007 13:46:36 -0700
From: "Paul Menage" <menage@...gle.com>
To: "David Rientjes" <rientjes@...gle.com>
Cc: "Paul Jackson" <pj@....com>,
"Andrew Morton" <akpm@...ux-foundation.org>,
nickpiggin@...oo.com.au, a.p.zijlstra@...llo.nl, serue@...ibm.com,
clg@...ibm.com, linux-kernel@...r.kernel.org,
ebiederm@...ssion.com, svaidy@...ux.vnet.ibm.com, xemul@...nvz.org,
containers@...ts.osdl.org, balbir@...ux.vnet.ibm.com
Subject: Re: [PATCH] task containersv11 add tasks file interface fix for cpusets
On 10/6/07, David Rientjes <rientjes@...gle.com> wrote:
>
> It can race with sched_setaffinity(). It has to give up tasklist_lock as
> well to call set_cpus_allowed() and can race
>
> cpus_allowed = cpuset_cpus_allowed(p);
> cpus_and(new_mask, new_mask, cpus_allowed);
> retval = set_cpus_allowed(p, new_mask);
>
> and allow a task to have a cpu outside of the cpuset's new cpus_allowed if
> you've taken it away between cpuset_cpus_allowed() and set_cpus_allowed().
cpuset_cpus_allowed() takes callback_mutex, which is held by
update_cpumask() when it updates cs->cpus_allowed. So if we continue
to hold callback_mutex across the task update loop this wouldn't be a
race. Having said that, holding callback mutex for that long might not
be a good idea.
A cleaner solution might be to drop callback_mutex after updating
cs->cpus_allowed in update_cpumask() and then make sched_setaffinity()
do a post-check:
cpus_allowed = cpuset_cpus_allowed(p);
again:
cpus_and(new_mask, new_mask, cpus_allowed);
retval = set_cpus_allowed(p, new_mask);
if (!retval) {
/* Check for races with cpuset updates */
cpus_allowed = cpuset_cpus_allowed(p);
if (!cpus_subset(new_mask, cpus_allowed)) {
/*
* We raced with a change to cpuset update,
* and our cpumask is now outside the
* permitted cpumask for the cpuset. Since a
* change to the cpuset's cpus resets the
* cpumask for each task, do the same thing
* here.
*/
new_mask = cpus_allowed;
goto again;
}
}
Paul
-
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