Keir Fraser writes:
> Moving the line 'pt->scheduled += pt->period' inside the if statement looks
> pretty dubious. Does your fix still work if we don't move that line? For the
> default timer mode I'm not sure it should make any difference either way,
> but it would for timer mode 0.
Oops, my brain was dead. You are right.
New one is attached. Of course it works.
Thanks,
Kouya
>
> -- Keir
>
> On 18/09/2009 06:24, "Kouya Shimura" <kouya@xxxxxxxxxxxxxx> wrote:
>
> > With c/s 20218, timer ticks might be missed when IRQs of a timer are queued.
> > Next scheduled time is accumulated wrongly.
> >
> > Thanks to Christoph for the report.
> >
> > Thanks,
> > Kouya
> >
> > Signed-off-by: Kouya Shimura <kouya@xxxxxxxxxxxxxx>
> > Reported-by: Christoph Egger <Christoph.Egger@xxxxxxx>
> >
>
Signed-off-by: Kouya Shimura <kouya@xxxxxxxxxxxxxx>
Reported-by: Christoph Egger <Christoph.Egger@xxxxxxx>
diff -r 72d130772f36 xen/arch/x86/hvm/vpt.c
--- a/xen/arch/x86/hvm/vpt.c Wed Sep 16 09:30:41 2009 +0100
+++ b/xen/arch/x86/hvm/vpt.c Fri Sep 18 17:13:25 2009 +0900
@@ -309,7 +309,6 @@ void pt_intr_post(struct vcpu *v, struct
else
{
pt->scheduled += pt->period;
- pt_process_missed_ticks(pt);
if ( mode_is(v->domain, one_missed_tick_pending) ||
mode_is(v->domain, no_missed_ticks_pending) )
@@ -324,7 +323,11 @@ void pt_intr_post(struct vcpu *v, struct
}
if ( pt->pending_intr_nr == 0 )
+ {
+ pt_process_missed_ticks(pt);
+ pt->do_not_freeze = 0;
set_timer(&pt->timer, pt->scheduled);
+ }
}
if ( mode_is(v->domain, delay_for_missed_ticks) &&
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home