Re: scheduler: IRQs disabled over context switches

[Ingo Molnar <mingo@elte.hu>]

* Davide Libenzi <davidel@xmailserver.org> wrote:

> We used to do it in 2.4. What changed to make it fragile? The
> threading (TLS) thing?

it _should_ work, but in the past we only had trouble from such changes
(at least in the O(1) tree of scheduling - 2.4 scheduler is OK.). We
could try the patch below. It certainly boots on SMP x86. But it causes
a 3.5% slowdown in lat_ctx so i'd not do it unless there are some really
good reasons.

	Ingo

--- linux/kernel/sched.c.orig	
+++ linux/kernel/sched.c	
@@ -247,9 +247,15 @@ static DEFINE_PER_CPU(struct runqueue, r
  * Default context-switch locking:
  */
 #ifndef prepare_arch_switch
-# define prepare_arch_switch(rq, next)	do { } while (0)
-# define finish_arch_switch(rq, next)	spin_unlock_irq(&(rq)->lock)
-# define task_running(rq, p)		((rq)->curr == (p))
+# define prepare_arch_switch(rq, next)				\
+		do {						\
+			spin_lock(&(next)->switch_lock);	\
+			spin_unlock(&(rq)->lock);		\
+		} while (0)
+# define finish_arch_switch(rq, prev) \
+		spin_unlock_irq(&(prev)->switch_lock)
+# define task_running(rq, p) \
+		((rq)->curr == (p) || spin_is_locked(&(p)->switch_lock))
 #endif
 
 /*