Re: rt14: strace -> migrate_disable_atomic imbalance
From: Peter Zijlstra <peterz@infradead.org>
Date: 2011-09-22 08:39:20
Also in:
lkml
From: Peter Zijlstra <peterz@infradead.org>
Date: 2011-09-22 08:39:20
Also in:
lkml
On Wed, 2011-09-21 at 20:50 +0200, Peter Zijlstra wrote:
+static void wait_task_inactive_sched_out(struct preempt_notifier *n,
+ struct task_struct *next)
+{
+ struct task_struct *p;
+ struct wait_task_inactive_blocked *blocked =
+ container_of(n, struct wait_task_inactive_blocked, notifier);
+
+ if (current->on_rq) /* we're not inactive yet */
+ return;
+
+ hlist_del(&n->link);
+
+ p = ACCESS_ONCE(blocked->waiter);
+ blocked->waiter = NULL;
+ wake_up_process(p);
+} Trying a wakeup from there isn't going to actually ever work of-course.. Duh!