On Fri, 20 Jul 2007 14:24:07 -0500
Bob Nelson <[EMAIL PROTECTED]> wrote:
> From: Maynard Johnson <[EMAIL PROTECTED]>
>
> This patch adds to the capability of spu_switch_event_register so that
> the caller is also notified of currently active SPU tasks.
> Exports spu_switch_event_register and spu_switch_event_unregister so
> that OProfile can get access to the notifications provided.
>
> Signed-off-by: Maynard Johnson <[EMAIL PROTECTED]>
> Signed-off-by: Carl Love <[EMAIL PROTECTED]>
> Signed-off-by: Bob Nelson <[EMAIL PROTECTED]>
> Acked-by: Arnd Bergmann <[EMAIL PROTECTED]>
> Acked-by: Paul Mackerras <[EMAIL PROTECTED]>
>
> ---
>
> We would like this patch included in -mm and 2.6.23
>
> Changed "for (node = 0; node < MAX_NUMNODES; node++)" loop to
> for_each_online_node(node).
> Added comment to memory barrier.
> Better info in changelog.
here it is:
---
a/arch/powerpc/platforms/cell/spufs/sched.c~oprofile-enable-spu-switch-notification-to-detect-currently-active-spu-tasks-update
+++ a/arch/powerpc/platforms/cell/spufs/sched.c
@@ -220,13 +220,14 @@ static void notify_spus_active(void)
* When the awakened processes see their "notify_active" flag is set,
* they will call spu_switch_notify();
*/
- for (node = 0; node < MAX_NUMNODES; node++) {
+ for_each_online_node(node) {
struct spu *spu;
mutex_lock(&spu_prio->active_mutex[node]);
list_for_each_entry(spu, &spu_prio->active_list[node], list) {
struct spu_context *ctx = spu->ctx;
set_bit(SPU_SCHED_NOTIFY_ACTIVE, &ctx->sched_flags);
- mb();
+ mb(); /* make sure any tasks woken up below */
+ /* can see the bit(s) set above */
wake_up_all(&ctx->stop_wq);
}
mutex_unlock(&spu_prio->active_mutex[node]);
_
I still wonder about that barrier. At the least it should be smp_mb().
But aren't our set_bit() semantics _alone_ sufficient to make this barrier
unneeded?
If it _is_ possible for the effects of a set_bit() to not be visible to a
woken-up thread then I suspect we'll have nasty little problems in quite a
few places. Maybe wake_up() should itself have a barrier to prevent such
things?
Doing that would be a documentation-only change, I suspect, given that the
current implementation of wake_up() starts out with a spin_lock_irqsave().
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@ozlabs.org
https://ozlabs.org/mailman/listinfo/linuxppc-dev
