Dario,
>> Without the processor assignment
>> here the vcpu might go on being assigned to a processor it no longer
>> is allowed to run on.
>>
> Ok.
>
>> In that case, function runq_candidate may only
>> get called for the vcpu's old processor, and runq_candidate will no
>> longer let a vcpu run o
Hey, I came across this patch again for other reasons, and I realized
I've a few more (minor) comments:
On Sun, 2015-02-08 at 17:45 -1000, Justin T. Weaver wrote:
> From: "Justin T. Weaver"
> diff --git a/xen/common/sched_credit2.c b/xen/common/sched_credit2.c
> index cf53770..de8fb5a 100644
> -
On Sun, 2015-03-08 at 21:11 -1000, Justin Weaver wrote:
> Thanks to all for the comments! I've implemented most of the changes
> recommended here in the v2 review. I should have a new patch set ready
> this week (with an updated soft affinity patch as well).
>
Great! :-)
> > Oh, and this is what
On Mon, 2015-03-09 at 11:45 +, George Dunlap wrote:
> On 03/09/2015 07:11 AM, Justin Weaver wrote:
> > I don't think there's any way the mask can be empty after the
> > cpumask_and with the cpu_hard_affinity and the VCPU2ONLINE. In
> > schedule.c:vcpu_set_hard_affinity, if the intersection of
On 03/09/2015 07:11 AM, Justin Weaver wrote:
> +static int get_safe_pcpu(struct csched2_vcpu *svc)
> +{
>
I also don't like the name... __choose_cpu() maybe ?
>>>
>>> I'm not 100% happy with the name, but I think "get_safe_pcpu" is more
>>> descriptive than just "__choose_cpu".
>>>
Thanks to all for the comments! I've implemented most of the changes
recommended here in the v2 review. I should have a new patch set ready
this week (with an updated soft affinity patch as well). I'll just
address the comments where you asked for my feedback...
>> rqd_avgload = rqd->
On Fri, 2015-03-06 at 15:18 +, George Dunlap wrote:
> On 03/03/2015 03:15 AM, Dario Faggioli wrote:
> > On Sun, 2015-02-08 at 17:45 -1000, Justin T. Weaver wrote:
> >> /*
> >> + * Use this to avoid having too many cpumask_t structs on the stack
> >> + */
> >> +static cpumask_t **cpumask = NUL
On 03/03/2015 03:15 AM, Dario Faggioli wrote:
> On Sun, 2015-02-08 at 17:45 -1000, Justin T. Weaver wrote:
>> From: "Justin T. Weaver"
>>
>> by making sure that vcpus only run on the pcpu(s) they are allowed to
>> run on based on their hard affinity cpu masks.
>>
> Ok, here I am reviewing this, at
>>> On 04.03.15 at 14:08, wrote:
> On Wed, 2015-03-04 at 12:50 +, Jan Beulich wrote:
>> >>> On 04.03.15 at 12:03, wrote:
>> > I see the point you're making, and I can live with _choose_cpu(), but
>> > the result would look a bit inconsistent, IMO.
>>
>> I'm tempted to ask for a cleanup patch
On Wed, 2015-03-04 at 12:50 +, Jan Beulich wrote:
> >>> On 04.03.15 at 12:03, wrote:
> > I see the point you're making, and I can live with _choose_cpu(), but
> > the result would look a bit inconsistent, IMO.
>
> I'm tempted to ask for a cleanup patch then.
>
Yep, and I'd be fine with that.
>>> On 04.03.15 at 12:03, wrote:
> I see the point you're making, and I can live with _choose_cpu(), but
> the result would look a bit inconsistent, IMO.
I'm tempted to ask for a cleanup patch then.
Jan
___
Xen-devel mailing list
Xen-devel@lists.xen.
On Tue, 2015-03-03 at 09:12 +, Jan Beulich wrote:
> >>> On 03.03.15 at 04:15, wrote:
> > On Sun, 2015-02-08 at 17:45 -1000, Justin T. Weaver wrote:
> >> +#define csched2_cpumask cpumask[smp_processor_id()]
> >> +
> > I like the idea, but put the right side between parentheses.
>
> Parenthese
>>> On 03.03.15 at 04:15, wrote:
> On Sun, 2015-02-08 at 17:45 -1000, Justin T. Weaver wrote:
>> --- a/xen/common/sched_credit2.c
>> +++ b/xen/common/sched_credit2.c
>> @@ -194,6 +194,12 @@ int opt_overload_balance_tolerance=-3;
>> integer_param("credit2_balance_over", opt_overload_balance_tolera
On Sun, 2015-02-08 at 17:45 -1000, Justin T. Weaver wrote:
> From: "Justin T. Weaver"
>
> by making sure that vcpus only run on the pcpu(s) they are allowed to
> run on based on their hard affinity cpu masks.
>
Ok, here I am reviewing this, at last... sorry for the delay! :-(
> Signed-off-by: J
From: "Justin T. Weaver"
by making sure that vcpus only run on the pcpu(s) they are allowed to
run on based on their hard affinity cpu masks.
Signed-off-by: Justin T. Weaver
---
Changes in v2:
* Added dynamically allocated cpu masks to avoid putting them on the stack;
replaced temp masks fr
15 matches
Mail list logo
