Messages in this thread | | | Date | Wed, 15 Dec 2010 19:59:29 +0100 | From | Oleg Nesterov <> | Subject | Re: [PATCH RFC] reduce runqueue lock contention |
| |
On 12/13, Frank Rowand wrote: > > I have not been able to make sense of the task_running() check in > try_to_wake_up(), even with that clue. The try_to_wake_up() code in > question is: > ... > > What am I missing, or is the task_running() test not needed?
I am afraid I can misuderstood this all, including the question ;)
But, with __ARCH_WANT_UNLOCKED_CTXSW task_running() checks ->oncpu. However, schedule() drops rq->lock after prev was deactivated but before it clears prev->oncpu.
Oleg.
| |