Skip to contentSkip to conversation

question on macros in wait.h

DISCUSSION David Brownell 2 participants 2 messages Mar 21, 2003 Scheduler

lore
Share & tools
lore
#0 lore raw
Quoted 4 lines

is there some deeper reason that there's no macro for waiting uninterruptablely with a timeout? Or did just nobody feel a need as yet?

Those macros seem to have moved out of <linux/sched.h> (2.4) and wait_event_interruptible_timeout() was added about 6 months ago; the changelog entry says it was for smbfs. So I'd guess "no need yet".

Here's an updated version of your patch, now using the same calling convention that the other two "can return 'early'" calls there provide. It's behaved in my testing, to replace the chaos in the usb synchronous call wrappers.

- Dave

include/linux/wait.h +26 -0
@@ -173,6 +173,32 @@
 	__ret;								\
 })
 
+#define __wait_event_timeout(wq, condition, ret)			\
+do {									\
+	wait_queue_t __wait;						\
+	init_waitqueue_entry(&__wait, current);				\
+									\
+	add_wait_queue(&wq, &__wait);					\
+	for (;;) {							\
+		set_current_state(TASK_UNINTERRUPTIBLE);		\
+		if (condition)						\
+			break;						\
+		ret = schedule_timeout(ret);				\
+		if (!ret)						\
+			break;						\
+	}								\
+	current->state = TASK_RUNNING;					\
+	remove_wait_queue(&wq, &__wait);				\
+} while (0)
+
+#define wait_event_timeout(wq, condition, timeout)			\
+({									\
+	long __ret = timeout;						\
+	if (!(condition))						\
+		__wait_event_timeout(wq, condition, __ret);		\
+	__ret;								\
+})
+	
 #define __wait_event_interruptible_timeout(wq, condition, ret)		\
 do {									\
 	wait_queue_t __wait;						\
#1 lore raw

Am Freitag, 21. März 2003 07:52 schrieb David Brownell:

Quoted 9 lines
Quoted 3 lines

is there some deeper reason that there's no macro for waiting uninterruptablely with a timeout? Or did just nobody feel a need as yet?

Those macros seem to have moved out of <linux/sched.h> (2.4) and wait_event_interruptible_timeout() was added about 6 months ago; the changelog entry says it was for smbfs. So I'd guess "no need yet".

Quoted 4 lines

Here's an updated version of your patch, now using the same calling convention that the other two "can return 'early'" calls there provide. It's behaved in my testing, to replace the chaos in the usb synchronous call wrappers.

Much better than my version.

Regards
Oliver