Messages in this thread | | | Date | Wed, 9 Dec 2009 09:06:42 +0100 (CET) | From | Thomas Gleixner <> | Subject | Re: [patch 2/2] sched: Protect sched_rr_get_param access to task->sched_class |
| |
On Wed, 9 Dec 2009, Peter Zijlstra wrote:
> On Wed, 2009-12-09 at 09:00 +0100, Peter Zijlstra wrote: > > On Tue, 2009-12-08 at 20:24 +0000, Thomas Gleixner wrote: > > > plain text document attachment > > > (sched-fix-sched-get-rr-param-race.patch) > > > sched_rr_get_param calls task->sched_class->get_rr_interval(task) > > > without protection against a concurrent sched_setscheduler() call > > > which modifies task->sched_class. > > > > > > Serialize the access with task_rq_lock(task); > > > > > > Signed-off-by: Thomas Gleixner <tglx@linutronix.de> > > > --- > > > kernel/sched.c | 4 ++++ > > > 1 file changed, 4 insertions(+) > > > > > > Index: linux-2.6-tip/kernel/sched.c > > > =================================================================== > > > --- linux-2.6-tip.orig/kernel/sched.c > > > +++ linux-2.6-tip/kernel/sched.c > > > @@ -6887,6 +6887,8 @@ SYSCALL_DEFINE2(sched_rr_get_interval, p > > > { > > > struct task_struct *p; > > > unsigned int time_slice; > > > + unsigned long flags; > > > + struct rq *rq; > > > int retval; > > > struct timespec t; > > > > > > @@ -6903,7 +6905,9 @@ SYSCALL_DEFINE2(sched_rr_get_interval, p > > > if (retval) > > > goto out_unlock; > > > > > > + rq = task_rq_lock(p, &flags); > > > time_slice = p->sched_class->get_rr_interval(p); > > > + task_rq_unlock(rq, &flags); > > > > > > read_unlock(&tasklist_lock); > > > jiffies_to_timespec(time_slice, &t); > > > > > > > Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl> > > Oh, that was too quick, get_rr_interval_fair() actually takes the > rq->lock, so lets add this: > > unsigned int get_rr_interval_fair(struct task_struct *task) > { > struct sched_entity *se = &task->se; > - unsigned long flags; > - struct rq *rq; > unsigned int rr_interval = 0; > > /* > * Time slice is 0 for SCHED_OTHER tasks that are on an otherwise > * idle runqueue: > */ > - rq = task_rq_lock(task, &flags); > if (rq->cfs.load.weight) > rr_interval = NS_TO_JIFFIES(sched_slice(&rq->cfs, se)); > - task_rq_unlock(rq, &flags); > - > return rr_interval; > }
Duh, yes. I looked at that actually and made a mental note to remove it. :(
Thanks for catching that,
tglx
| |