Re: [PATCH -v2] balance RT task resched only on runqueue

[Peter Zijlstra <a.p.zijlstra@chello.nl>]

On Wed, 2008-03-05 at 10:00 -0500, Steven Rostedt wrote:
> [
>   Update: I didn't like the task_rq(p), so I changed the test to be
>   rq->curr == p. No need to resched if the p isn't running. And
>   I think this is a bit cleaner.
> ]
> 
> In the RT kernel we discovered a place that can reschedule a task
> without holding the tasks runqueue lock.  This was caused by the RT
> balancing code that pulls RT tasks to the current run queue and will
> reschedule the current task.
> 
> There's a slight chance that the pulling of the RT tasks will release
> the current runqueue's lock and retake it (in the double_lock_balance).
> During this time that the runqueue is released, the current task can
> migrate to another runqueue.
> 
> In the prio_changed_rt code, after the pull, if the current task is of
> lesser priority than one of the RT tasks pulled, resched_task is called
> on the current task. If the current task had migrated in that small
> window, resched_task will be called without holding the runqueue lock
> for the runqueue that the task is on.
> 
> This race condition also exists in the mainline kernel and this patch
> adds a check to make sure the task hasn't migrated before calling
> resched_task.
> 
> Signed-off-by: Steven Rostedt <srostedt@redhat.com>

Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>

> ---
>  kernel/sched_rt.c |    6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
> 
> Index: linux-compile.git/kernel/sched_rt.c
> ===================================================================
> --- linux-compile.git.orig/kernel/sched_rt.c	2008-03-05 09:31:22.000000000 -0500
> +++ linux-compile.git/kernel/sched_rt.c	2008-03-05 09:53:00.000000000 -0500
> @@ -1107,9 +1107,11 @@ static void prio_changed_rt(struct rq *r
>  			pull_rt_task(rq);
>  		/*
>  		 * If there's a higher priority task waiting to run
> -		 * then reschedule.
> +		 * then reschedule. Note, the above pull_rt_task
> +		 * can release the rq lock and p could migrate.
> +		 * Only reschedule if p is still on the same runqueue.
>  		 */
> -		if (p->prio > rq->rt.highest_prio)
> +		if (p->prio > rq->rt.highest_prio && rq->curr == p)
>  			resched_task(p);
>  #else
>  		/* For UP simply resched on drop of prio */
> 
>