[vfio-users] cpu usage in guest != cpu usage in host, even with exclusive pinning

Quentin Deldycke quentindeldycke at gmail.com
Tue Apr 5 11:31:43 UTC 2016


I tried a few with unigine valley and cinebench benchmark.

Totally no difference on scores, no dpc latency difference also.

But it is effective for turbostat when pc is in idle:

at default value turbostat reports ~35w (cores of vm all time around 1.5ghz)
at 0 turbostat report 20w (cores of vm around 300-400 mhz (6% busy)

Did not try "real gaming" for the moment. Did this remotely...

--
Deldycke Quentin


On 5 April 2016 at 10:53, Jonathan Scruggs <j.scruggs at gmail.com> wrote:

> I don't know what halt_poll_ns does, but wouldn't setting it to 0 disable
> it and is disabling it bad?
>
> So, what does halt_poll_ns do anyways?
>
> On 4 April 2016 at 19:17, Okky Hendriansyah <okky.htf at gmail.com> wrote:
>
>> On Mon, Apr 4, 2016 at 8:20 AM, Jayme Howard <g.prime at gmail.com> wrote:
>>
>>> Okay wow, Alex.  That made a huge difference immediately.  That dropped
>>> me from 90% to 33% on the main thread, and all the children are below 10%.
>>>
>>
>>> On Sun, Apr 3, 2016 at 8:07 PM, Alex Williamson <
>>> alex.l.williamson at gmail.com> wrote:
>>>
>>>> If this is the issue I think it is, the unraid folks bisected this and
>>>> found the culprit as:
>>>>
>>>> aca6ff2 KVM: dynamic halt-polling
>>>>
>>>> (
>>>> https://github.com/torvalds/linux/commit/aca6ff29c4063a8d467cdee241e6b3bf7dc4a171
>>>> )
>>>>
>>>> It seems to be a poor interaction of the halt polling interval vs the
>>>> timer ticks on win10.  You can pick a different polling interval with the
>>>> kvm module option halt_poll_ns.  The default is 500000.  It seemed that
>>>> setting this to 400000 or lower resolves the issue.  You can do this via a
>>>> modprobe entry, 'options kvm halt_poll_ns=400000' or on the kernel command
>>>> line with kvm.halt_poll_ns=400000.  You can also change it dynamically via
>>>> 'echo 400000 > /sys/module/kvm/parameters/halt_poll_ns'
>>>>
>>>
>> I can also confirm that tuning the halt_poll_ns kvm module parameter on
>> kernel 4.4.5 drops the host CPU usage! Haven't done any game benchmark but
>> from listening to music the CPU load is as low as it is on 4.1.20. Although
>> reducing the value to what Alex (400000) had suggested is probably
>> sufficient, I'm currently trying out the value of 0 for it just like what
>> Unraid 6.20 beta 20 changelog had it set on default [1]. Thanks for the
>> info, Alex.
>>
>> [1]
>> http://dnld.lime-technology.com/beta/unRAIDServer-6.2.0-beta20-x86_64.txt
>>
>>
>> Best regards,
>> Okky Hendriansyah
>>
>> _______________________________________________
>> vfio-users mailing list
>> vfio-users at redhat.com
>> https://www.redhat.com/mailman/listinfo/vfio-users
>>
>>
>
> _______________________________________________
> vfio-users mailing list
> vfio-users at redhat.com
> https://www.redhat.com/mailman/listinfo/vfio-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20160405/9a3ad295/attachment.htm>


More information about the vfio-users mailing list