[PATCH 1/2] v2 PF_DEAD: cleanup usage

[Oleg Nesterov]

schedule() checks PF_DEAD on every context switch, and sets ->state = EXIT_DEAD
to ensure that exited task will be deactivated.

It is possible to set new ->state value in do_exit(), along with PF_DEAD flag,
and remove the check from schedule()'s hot path.

To avoid mixing ->state/->exit_state values, this patch adds new TASK_DEAD state.

Signed-off-by: Oleg Nesterov <oleg@xxxxxxxxxx>

--- 2.6.15-rc2/include/linux/sched.h~1_CLEAN	2005-11-25 20:58:09.000000000 +0300
+++ 2.6.15-rc2/include/linux/sched.h	2005-11-25 21:02:22.000000000 +0300
@@ -127,6 +127,7 @@ extern unsigned long nr_iowait(void);
 #define EXIT_DEAD		32
 /* in tsk->state again */
 #define TASK_NONINTERACTIVE	64
+#define TASK_DEAD		128
 
 #define __set_task_state(tsk, state_value)		\
 	do { (tsk)->state = (state_value); } while (0)
--- 2.6.15-rc2/kernel/exit.c~1_CLEAN	2005-11-25 20:58:14.000000000 +0300
+++ 2.6.15-rc2/kernel/exit.c	2005-11-25 21:04:26.000000000 +0300
@@ -875,6 +875,7 @@ fastcall NORET_TYPE void do_exit(long co
 	preempt_disable();
 	BUG_ON(tsk->flags & PF_DEAD);
 	tsk->flags |= PF_DEAD;
+	tsk->state = TASK_DEAD;
 
 	schedule();
 	BUG();
--- 2.6.15-rc2/kernel/sched.c~1_CLEAN	2005-11-25 20:58:14.000000000 +0300
+++ 2.6.15-rc2/kernel/sched.c	2005-11-25 21:04:48.000000000 +0300
@@ -3000,9 +3000,6 @@ need_resched_nonpreemptible:
 
 	spin_lock_irq(&rq->lock);
 
-	if (unlikely(prev->flags & PF_DEAD))
-		prev->state = EXIT_DEAD;
-
 	switch_count = &prev->nivcsw;
 	if (prev->state && !(preempt_count() & PREEMPT_ACTIVE)) {
 		switch_count = &prev->nvcsw;
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/