sched: Add a preempt count base offset to __might_sleep()

Add a preempt count base offset to compare against the current
preempt level count. It prepares to pull up the might_sleep
check from cond_resched() to cond_resched_lock() and
cond_resched_bh().

For these two helpers, we need to respectively ensure that once
we'll unlock the given spinlock / reenable local softirqs, we
will reach a sleepable state.

Signed-off-by: Frederic Weisbecker <[e-mail]>
[ Move and rename preempt_count_equals() ]
Signed-off-by: Peter Zijlstra <[e-mail]>
LKML-Reference: <[e-mail]>
Signed-off-by: Ingo Molnar <[e-mail]>

include/linux/kernel.h

···
 #endif
 
 #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP
-  void __might_sleep(char *file, int line);
+  void __might_sleep(char *file, int line, int preempt_offset);
 /**
  * might_sleep - annotation for functions that can sleep
  *
···
  * supposed to.
  */
 # define might_sleep() \
-	do { __might_sleep(__FILE__, __LINE__); might_resched(); } while (0)
+	do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0)
 #else
-  static inline void __might_sleep(char *file, int line) { }
+  static inline void __might_sleep(char *file, int line, int preempt_offset) { }
 # define might_sleep() do { might_resched(); } while (0)
 #endif
 

kernel/sched.c

···
 
 static void __cond_resched(void)
 {
-	__might_sleep(__FILE__, __LINE__);
+	__might_sleep(__FILE__, __LINE__, 0);
 
 	add_preempt_count(PREEMPT_ACTIVE);
 	schedule();
···
 }
 
 #ifdef CONFIG_DEBUG_SPINLOCK_SLEEP
-void __might_sleep(char *file, int line)
+static inline int preempt_count_equals(int preempt_offset)
+{
+	int nested = preempt_count() & ~PREEMPT_ACTIVE;
+
+	return (nested == PREEMPT_INATOMIC_BASE + preempt_offset);
+}
+
+void __might_sleep(char *file, int line, int preempt_offset)
 {
 #ifdef in_atomic
 	static unsigned long prev_jiffy;	/* ratelimiting */
 
-	if ((!in_atomic() && !irqs_disabled()) ||
-		    system_state != SYSTEM_RUNNING || oops_in_progress)
+	if ((preempt_count_equals(preempt_offset) && !irqs_disabled()) ||
+	    system_state != SYSTEM_RUNNING || oops_in_progress)
 		return;
 	if (time_before(jiffies, prev_jiffy + HZ) && prev_jiffy)
 		return;