sched, idle: Fix the idle polling state logic

commit ea8117478918a4734586d35ff530721b682425be upstream.

Mike reported that commit 7d1a9417 ("x86: Use generic idle loop")
regressed several workloads and caused excessive reschedule
interrupts.

The patch in question failed to notice that the x86 code had an
inverted sense of the polling state versus the new generic code (x86:
default polling, generic: default !polling).

Fix the two prominent x86 mwait based idle drivers and introduce a few
new generic polling helpers (fixing the wrong smp_mb__after_clear_bit
usage).

Also switch the idle routines to using tif_need_resched() which is an
immediate TIF_NEED_RESCHED test as opposed to need_resched which will
end up being slightly different.

Reported-by: Mike Galbraith <bitbucket@online.de>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Cc: lenb@kernel.org
Cc: tglx@linutronix.de
Link: http://lkml.kernel.org/n/tip-nc03imb0etuefmzybzj7sprf@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

2 files changed, 73 insertions, 7 deletions

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 178a8d909f14..7b64cd4e3e56 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2469,34 +2469,98 @@ static inline int tsk_is_polling(struct task_struct *p)
 {
 	return task_thread_info(p)->status & TS_POLLING;
 }
-static inline void current_set_polling(void)
+static inline void __current_set_polling(void)
 {
 	current_thread_info()->status |= TS_POLLING;
 }
 
-static inline void current_clr_polling(void)
+static inline bool __must_check current_set_polling_and_test(void)
+{
+	__current_set_polling();
+
+	/*
+	 * Polling state must be visible before we test NEED_RESCHED,
+	 * paired by resched_task()
+	 */
+	smp_mb();
+
+	return unlikely(tif_need_resched());
+}
+
+static inline void __current_clr_polling(void)
 {
 	current_thread_info()->status &= ~TS_POLLING;
-	smp_mb__after_clear_bit();
+}
+
+static inline bool __must_check current_clr_polling_and_test(void)
+{
+	__current_clr_polling();
+
+	/*
+	 * Polling state must be visible before we test NEED_RESCHED,
+	 * paired by resched_task()
+	 */
+	smp_mb();
+
+	return unlikely(tif_need_resched());
 }
 #elif defined(TIF_POLLING_NRFLAG)
 static inline int tsk_is_polling(struct task_struct *p)
 {
 	return test_tsk_thread_flag(p, TIF_POLLING_NRFLAG);
 }
-static inline void current_set_polling(void)
+
+static inline void __current_set_polling(void)
 {
 	set_thread_flag(TIF_POLLING_NRFLAG);
 }
 
-static inline void current_clr_polling(void)
+static inline bool __must_check current_set_polling_and_test(void)
+{
+	__current_set_polling();
+
+	/*
+	 * Polling state must be visible before we test NEED_RESCHED,
+	 * paired by resched_task()
+	 *
+	 * XXX: assumes set/clear bit are identical barrier wise.
+	 */
+	smp_mb__after_clear_bit();
+
+	return unlikely(tif_need_resched());
+}
+
+static inline void __current_clr_polling(void)
 {
 	clear_thread_flag(TIF_POLLING_NRFLAG);
 }
+
+static inline bool __must_check current_clr_polling_and_test(void)
+{
+	__current_clr_polling();
+
+	/*
+	 * Polling state must be visible before we test NEED_RESCHED,
+	 * paired by resched_task()
+	 */
+	smp_mb__after_clear_bit();
+
+	return unlikely(tif_need_resched());
+}
+
 #else
 static inline int tsk_is_polling(struct task_struct *p) { return 0; }
-static inline void current_set_polling(void) { }
-static inline void current_clr_polling(void) { }
+static inline void __current_set_polling(void) { }
+static inline void __current_clr_polling(void) { }
+
+static inline bool __must_check current_set_polling_and_test(void)
+{
+	return unlikely(tif_need_resched());
+}
+static inline bool __must_check current_clr_polling_and_test(void)
+{
+	return unlikely(tif_need_resched());
+}
 #endif
 
 /*
diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h
index e7e04736802f..4ae6f32c8033 100644
--- a/include/linux/thread_info.h
+++ b/include/linux/thread_info.h
@@ -107,6 +107,8 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag)
 #define set_need_resched()	set_thread_flag(TIF_NEED_RESCHED)
 #define clear_need_resched()	clear_thread_flag(TIF_NEED_RESCHED)
 
+#define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED)
+
 #if defined TIF_RESTORE_SIGMASK && !defined HAVE_SET_RESTORE_SIGMASK
 /*
  * An arch can define its own version of set_restore_sigmask() to get the