a cleanup and a bugfix in the preemptive kernel:

- the PREEMPT_ACTIVE trick is not needed

- schedule() should check for need_resched, we might miss a
  reschedule otherwise.

the cleanup also fixes the bug. The only reason why i kept
preempt_schedule() was to fix up p->state to TASK_RUNNING,
to make it possible to preempt from places that mark the
task TASK_UNINTERRUPTIBLE before adding the task to a waitqueue,
and thus a preemption in that small window could cause the
task to be removed from the runqueue erroneously.

include/linux/sched.h

@@ -92,7 +92,6 @@ extern unsigned long nr_running(void);
 #define TASK_UNINTERRUPTIBLE	2
 #define TASK_ZOMBIE		4
 #define TASK_STOPPED		8
-#define PREEMPT_ACTIVE		0x4000000
 
 #define __set_task_state(tsk, state_value)		\
 	do { (tsk)->state = (state_value); } while (0)

kernel/sched.c

@@ -761,6 +761,7 @@ asmlinkage void schedule(void)
 
 	if (unlikely(in_interrupt()))
 		BUG();
+need_resched:
 	preempt_disable();
 	prev = current;
 	rq = this_rq();
@@ -769,15 +770,6 @@ asmlinkage void schedule(void)
 	prev->sleep_timestamp = jiffies;
 	spin_lock_irq(&rq->lock);
 
-#ifdef CONFIG_PREEMPT
-	/*
-	 * if entering from preempt_schedule, off a kernel preemption,
-	 * go straight to picking the next task.
-	 */
-	if (unlikely(preempt_get_count() & PREEMPT_ACTIVE))
-		goto pick_next_task;
-#endif
-
 	switch (prev->state) {
 	case TASK_INTERRUPTIBLE:
 		if (unlikely(signal_pending(prev))) {
@@ -789,7 +781,7 @@ asmlinkage void schedule(void)
 	case TASK_RUNNING:
 		;
 	}
-#if CONFIG_SMP || CONFIG_PREEMPT
+#if CONFIG_SMP
 pick_next_task:
 #endif
 	if (unlikely(!rq->nr_running)) {
@@ -838,6 +830,8 @@ asmlinkage void schedule(void)
 
 	reacquire_kernel_lock(current);
 	preempt_enable_no_resched();
+	if (test_thread_flag(TIF_NEED_RESCHED))
+		goto need_resched;
 	return;
 }
 
@@ -847,12 +841,8 @@ asmlinkage void schedule(void)
  */
 asmlinkage void preempt_schedule(void)
 {
-	do {
-		current_thread_info()->preempt_count += PREEMPT_ACTIVE;
-		schedule();
-		current_thread_info()->preempt_count -= PREEMPT_ACTIVE;
-		barrier();
-	} while (test_thread_flag(TIF_NEED_RESCHED));
+	current->state = TASK_RUNNING;
+	schedule();
 }
 #endif /* CONFIG_PREEMPT */