[PATCH] workqueue.c neatening II

Rusty Russell (rusty@rustcorp.com.au)
Mon, 23 Jun 2003 13:20:25 +1000


Linus applied my patch before Jeff's critique. Here's the fix.

Linus, please apply.
Rusty.

Name: Workqueue Exit Neatening
Author: Rusty Russell
Status: Tested on 2.5.73

D: Jeff Garzik points out the initializing the exit completion at
D: exit time is foolish: we should just initialize it at creation time
D: live everything else in that structure, and avoid the memory barrier.

diff -urpN --exclude TAGS -X /home/rusty/devel/kernel/kernel-patches/current-dontdiff --minimal .22396-linux-2.5.73/kernel/workqueue.c .22396-linux-2.5.73.updated/kernel/workqueue.c
--- .22396-linux-2.5.73/kernel/workqueue.c 2003-06-23 10:52:59.000000000 +1000
+++ .22396-linux-2.5.73.updated/kernel/workqueue.c 2003-06-23 12:01:18.000000000 +1000
@@ -275,6 +275,7 @@ static int create_workqueue_thread(struc
INIT_LIST_HEAD(&cwq->worklist);
init_waitqueue_head(&cwq->more_work);
init_waitqueue_head(&cwq->work_done);
+ init_completion(&cwq->exit);

init_completion(&startup.done);
startup.cwq = cwq;
@@ -320,10 +321,7 @@ static void cleanup_workqueue_thread(str

cwq = wq->cpu_wq + cpu;
if (cwq->thread) {
- printk("Cleaning up workqueue thread for %i\n", cpu);
- /* Initiate an exit and wait for it: */
- init_completion(&cwq->exit);
- wmb(); /* Thread must see !cwq->thread after completion init */
+ /* Tell thread to exit and wait for it. */
cwq->thread = NULL;
wake_up(&cwq->more_work);

--
  Anyone who quotes me in their sig is an idiot. -- Rusty Russell.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/