RE: [PATCH] speedup flush_workqueue for singlethread_workqueue

[Anil]

Hi Andrew,
	In the flush_workqueue, can we move lock_cpu_hotplug()/unlock_cpu_hotplug only for _non_single_threaded_ case as shown below
as these locks are not required for single_threaded_workqueue.

Thanks,
Anil	

The attached patch applies on top of your changes.
Signed-off-by: Anil Keshavamurthy <anil.s.keshavamurthy@xxxxxxxxx>
---

 linux-2.6.7-rc2-mm2-root/kernel/workqueue.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff -puN kernel/workqueue.c~andrew2 kernel/workqueue.c
--- linux-2.6.7-rc2-mm2/kernel/workqueue.c~andrew2	2004-06-07 18:45:20.332139681 -0700
+++ linux-2.6.7-rc2-mm2-root/kernel/workqueue.c	2004-06-07 18:46:09.560680511 -0700
@@ -262,7 +262,6 @@ static void flush_cpu_workqueue(struct c
 void fastcall flush_workqueue(struct workqueue_struct *wq)
 {
  	might_sleep();
- 	lock_cpu_hotplug();
 
 	if (is_single_threaded(wq)) {
 		/* Always use cpu 0's area. */
@@ -270,11 +269,12 @@ void fastcall flush_workqueue(struct wor
 	} else {
 		int cpu;
  
+ 		lock_cpu_hotplug();
 		for_each_online_cpu(cpu)
 			flush_cpu_workqueue(wq->cpu_wq + cpu);
+ 		unlock_cpu_hotplug();
  	}
 
- 	unlock_cpu_hotplug();
 }
 
 static struct task_struct *create_workqueue_thread(struct workqueue_struct *wq,

_

 	

>-----Original Message-----
>From: Rusty Russell [mailto:rusty@xxxxxxxxxxxxxxx] 
>Sent: Saturday, June 05, 2004 7:11 PM
>To: Andrew Morton
>Cc: Keshavamurthy, Anil S; lkml - Kernel Mailing List
>Subject: Re: [PATCH] speedup flush_workqueue for singlethread_workqueue
>
>On Sat, 2004-06-05 at 08:30, Andrew Morton wrote:
>> "Anil" <anil.s.keshavamurthy@xxxxxxxxx> wrote:
>> >
>> > 	In the flush_workqueue function for a 
>single_threaded_worqueue case 
>> > the code seemed to loop the same cpu_workqueue_struct for 
>> > each_online_cpu's. The attached patch checks this 
>condition and bails out of for loop there by speeding up the 
>flush_workqueue for a singlethreaded_workqueue.
>> 
>> 
>> OK, thanks.  I think it's a bit clearer to do it as below.  
>I haven't 
>> tested it though.
>
>Me neither, but agree.
>
>Rusty.
>--
>Anyone who quotes me in their signature is an idiot -- Rusty Russell
>
>
>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/