On Fri, 2017-06-16 at 10:41 -0700, Stefano Stabellini wrote:
> On Fri, 16 Jun 2017, Dario Faggioli wrote:
> >
> > diff --git a/xen/arch/arm/domain.c b/xen/arch/arm/domain.c
> > index 76310ed..86cd612 100644
> > --- a/xen/arch/arm/domain.c
> > +++ b/xen/arch/arm/domain.c
> > @@ -41,20 +41,28 @@ DEF
On Fri, 16 Jun 2017, Dario Faggioli wrote:
> On Fri, 2017-06-16 at 02:54 -0600, Jan Beulich wrote:
> > > > > On 14.06.17 at 18:53, wrote:
> > >
> > > --- a/xen/arch/x86/domain.c
> > > +++ b/xen/arch/x86/domain.c
> > > @@ -112,12 +112,18 @@ static void play_dead(void)
> > >
> > > static void id
>>> On 16.06.17 at 15:34, wrote:
> Assuming answers are 'no' and 'yes', I think I'd leave
> cpu_is_haltable() as it is
Agreed, you analysis was quite helpful.
Jan
___
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel
On Fri, 2017-06-16 at 05:47 -0600, Jan Beulich wrote:
> > On Fri, 2017-06-16 at 02:54 -0600, Jan Beulich wrote:
> > > > > > On 14.06.17 at 18:53, wrote:
> > > > unsigned long *work_to_do = &per_cpu(tasklet_work_to_do,
> > > > cpu);
> > > > struct list_head *list = &per_cpu(tasklet_list,
>>> On 16.06.17 at 12:44, wrote:
> On Fri, 2017-06-16 at 02:54 -0600, Jan Beulich wrote:
>> > > > On 14.06.17 at 18:53, wrote:
>> >
>> > --- a/xen/arch/x86/domain.c
>> > +++ b/xen/arch/x86/domain.c
>> > @@ -112,12 +112,18 @@ static void play_dead(void)
>> >
>> > static void idle_loop(void)
>>
On Fri, 2017-06-16 at 02:54 -0600, Jan Beulich wrote:
> > > > On 14.06.17 at 18:53, wrote:
> >
> > --- a/xen/arch/x86/domain.c
> > +++ b/xen/arch/x86/domain.c
> > @@ -112,12 +112,18 @@ static void play_dead(void)
> >
> > static void idle_loop(void)
> > {
> > +unsigned int cpu = smp_proces
>>> On 14.06.17 at 18:53, wrote:
> --- a/xen/arch/x86/domain.c
> +++ b/xen/arch/x86/domain.c
> @@ -112,12 +112,18 @@ static void play_dead(void)
>
> static void idle_loop(void)
> {
> +unsigned int cpu = smp_processor_id();
> +
> for ( ; ; )
> {
> -if ( cpu_is_offline(smp_
In fact, there exists two kind of tasklets: vCPU and
softirq context tasklets. When we want to do vCPU
context tasklet work, we force the idle vCPU (of a
particular pCPU) into execution, and run it from there.
This means there are two possible reasons for choosing
to run the idle vCPU:
1) we want
Hi,
following up on this:
https://lists.xen.org/archives/html/xen-devel/2017-06/msg01260.html
I did make a patch that moves do_tasklet() up a bit, within idle_loop().
While there, I did a bit more than that... let's see what you guys think. :-D
I've verified that this builds on ARM too, but h
