Re: [PATCH v3] sched/deadline: Fix bad accounting of nr_running

2014-02-19 Thread Steven Rostedt
On Wed, 19 Feb 2014 14:14:54 +0100 Juri Lelli wrote: > Steven, could you test it? OK, I tested it on my x86_64 box that was also causing troubles. Please fix up the whitespace issues and send a formal patch to Peter. Reported-by: Steven Rostedt Reviewed-by: Steven Rostedt Tested-by: Steven R

Re: [PATCH v3] sched/deadline: Fix bad accounting of nr_running

2014-02-19 Thread Juri Lelli
On 02/19/2014 11:32 AM, Juri Lelli wrote: > On 02/19/2014 09:46 AM, Peter Zijlstra wrote: >> On Tue, Feb 18, 2014 at 09:50:12PM -0500, Steven Rostedt wrote: >>> Rationale for this odd behavior is that, when a task is throttled, it is removed only from the dl_rq, but we keep it on_rq (as

Re: [PATCH v3] sched/deadline: Fix bad accounting of nr_running

2014-02-19 Thread Juri Lelli
On 02/19/2014 09:46 AM, Peter Zijlstra wrote: > On Tue, Feb 18, 2014 at 09:50:12PM -0500, Steven Rostedt wrote: >> >>> Rationale for this odd behavior is that, when a task is throttled, it >>> is removed only from the dl_rq, but we keep it on_rq (as this is not >>> a "full dequeue", that is the t

Re: [PATCH v3] sched/deadline: Fix bad accounting of nr_running

2014-02-19 Thread Peter Zijlstra
On Tue, Feb 18, 2014 at 09:50:12PM -0500, Steven Rostedt wrote: > > > Rationale for this odd behavior is that, when a task is throttled, it > > is removed only from the dl_rq, but we keep it on_rq (as this is not > > a "full dequeue", that is the task is not actually sleeping). But, it > > is als