public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH 0/2] introduce wait_event_common()
@ 2013-06-06 20:02 Oleg Nesterov
  2013-06-06 20:03 ` [PATCH 1/2] wait: introduce wait_event_common(wq, condition, state, timeout) Oleg Nesterov
  2013-06-06 20:03 ` [PATCH 2/2] wait: introduce prepare_to_wait_event() Oleg Nesterov
  0 siblings, 2 replies; 7+ messages in thread
From: Oleg Nesterov @ 2013-06-06 20:02 UTC (permalink / raw)
  To: Andrew Morton, Daniel Vetter, Dave Jones, David Howells,
	Imre Deak, Jens Axboe, Linus Torvalds, Lukas Czerner,
	Paul E. McKenney, Tejun Heo
  Cc: linux-kernel

Hello.

To remind, I think that 4c663cfc "wait: fix false timeouts when using
wait_event_timeout()" is not enough, wait(wq, true, 0) still returns
zero.

But to me the main problem is that wait_event* macros duplicate the
same code again and again. Imho it would be nice to create a single
helper. To simplify the review, this is the code after 1/2:

	#define __wait_no_timeout(tout)	\
		(__builtin_constant_p(tout) && (tout) == MAX_SCHEDULE_TIMEOUT)

	/* uglified signal_pending_state() optimized for constant state */
	#define __wait_signal_pending(state)					\
		((state == TASK_INTERRUPTIBLE) ? signal_pending(current) :	\
		 (state == TASK_KILLABLE) ? fatal_signal_pending(current) :	\
		  0)

	#define __wait_event_common(wq, condition, state, tout)			\
	({									\
		DEFINE_WAIT(__wait);						\
		long __ret = 0, __tout = tout;					\
										\
		for (;;) {							\
			prepare_to_wait(&wq, &__wait, state);			\
			if (condition) {					\
				__ret = __wait_no_timeout(tout) ?: __tout ?: 1;	\
				break;						\
			}							\
										\
			if (__wait_signal_pending(state)) {			\
				__ret = -ERESTARTSYS;				\
				break;						\
			}							\
										\
			if (__wait_no_timeout(tout))				\
				schedule();					\
			else if (__tout)					\
				__tout = schedule_timeout(__tout);		\
			else							\
				break;						\
		}								\
		finish_wait(&wq, &__wait);					\
		__ret;								\
	})

	#define wait_event_common(wq, condition, state, tout)			\
	({									\
		long __ret;							\
		if (condition)							\
			__ret = __wait_no_timeout(tout) ?: (tout) ?: 1;		\
		else								\
			__ret = __wait_event_common(wq, condition, state, tout);\
		__ret;								\
	})

2/2 doesn't look like a cleanup. But personally I think that it makes
sense to shrink .text,

	-	4977769 2930984 10104832        18013585        112dd91 vmlinux
	+	4976847	2930984	10104832	18012663	112d9f7	vmlinux

on my build.

Please comment.

Oleg.


^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2013-06-19 16:20 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2013-06-06 20:02 [PATCH 0/2] introduce wait_event_common() Oleg Nesterov
2013-06-06 20:03 ` [PATCH 1/2] wait: introduce wait_event_common(wq, condition, state, timeout) Oleg Nesterov
2013-06-18 22:06   ` Andrew Morton
2013-06-19 16:14     ` Oleg Nesterov
2013-06-06 20:03 ` [PATCH 2/2] wait: introduce prepare_to_wait_event() Oleg Nesterov
2013-06-06 21:36   ` Tejun Heo
2013-06-07 13:07     ` Oleg Nesterov

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox