From: Joao Martins <joao.m.martins@oracle.com>
To: "Rafael J. Wysocki" <rafael@kernel.org>,
Marcelo Tosatti <mtosatti@redhat.com>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>,
kvm-devel <kvm@vger.kernel.org>,
"Linux Kernel Mailing List" <linux-kernel@vger.kernel.org>,
"Paolo Bonzini" <pbonzini@redhat.com>,
"Radim Krčmář" <rkrcmar@redhat.com>,
"Sean Christopherson" <sean.j.christopherson@intel.com>,
"Vitaly Kuznetsov" <vkuznets@redhat.com>,
"Wanpeng Li" <wanpengli@tencent.com>,
"Jim Mattson" <jmattson@google.com>,
"Joerg Roedel" <joro@8bytes.org>,
"Daniel Lezcano" <daniel.lezcano@linaro.org>,
"Linux PM" <linux-pm@vger.kernel.org>,
"Boris Ostrovsky" <boris.ostrovsky@oracle.com>
Subject: Re: Is: Default governor regardless of cpuidle driver Was: [PATCH v2] cpuidle-haltpoll: vcpu hotplug support
Date: Tue, 3 Sep 2019 11:13:31 +0100 [thread overview]
Message-ID: <457e8ca1-beb3-ca39-b257-e7bc6bb35d4d@oracle.com> (raw)
In-Reply-To: <CAJZ5v0j8BEjdNyAn=ut9BxSH5Gphs_AivADPwXX=rJ1TF1+88A@mail.gmail.com>
On 9/2/19 10:55 PM, Rafael J. Wysocki wrote:
> On Thu, Aug 29, 2019 at 7:24 PM Marcelo Tosatti <mtosatti@redhat.com> wrote:
>>
>> On Thu, Aug 29, 2019 at 06:16:05PM +0100, Joao Martins wrote:
>>> On 8/29/19 4:10 PM, Joao Martins wrote:
>>>> When cpus != maxcpus cpuidle-haltpoll will fail to register all vcpus
>>>> past the online ones and thus fail to register the idle driver.
>>>> This is because cpuidle_add_sysfs() will return with -ENODEV as a
>>>> consequence from get_cpu_device() return no device for a non-existing
>>>> CPU.
>>>>
>>>> Instead switch to cpuidle_register_driver() and manually register each
>>>> of the present cpus through cpuhp_setup_state() callback and future
>>>> ones that get onlined. This mimmics similar logic that intel_idle does.
>>>>
>>>> Fixes: fa86ee90eb11 ("add cpuidle-haltpoll driver")
>>>> Signed-off-by: Joao Martins <joao.m.martins@oracle.com>
>>>> Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
>>>> ---
>>>
>>> While testing the above, I found out another issue on the haltpoll series.
>>> But I am not sure what is best suited to cpuidle framework, hence requesting
>>> some advise if below is a reasonable solution or something else is preferred.
>>>
>>> Essentially after haltpoll governor got introduced and regardless of the cpuidle
>>> driver the default governor is gonna be haltpoll for a guest (given haltpoll
>>> governor doesn't get registered for baremetal).
>>
>> Right.
>>
>>> Right now, for a KVM guest, the
>>> idle governors have these ratings:
>>>
>>> * ladder -> 10
>>> * teo -> 19
>>> * menu -> 20
>>> * haltpoll -> 21
>>> * ladder + nohz=off -> 25
>>
>> Yes. PowerPC KVM guests crash currently due to the use of the haltpoll
>> governor (have a patch in my queue to fix this, but your solution
>> embraces more cases).
>>
>>> When a guest is booted with MWAIT and intel_idle is probed and sucessfully
>>> registered, we will end up with a haltpoll governor being used as opposed to
>>> 'menu' (which used to be the default case). This would prevent IIUC that other
>>> C-states get used other than poll_state (state 0) and state 1.
>>>
>>> Given that haltpoll governor is largely only useful with a cpuidle-haltpoll
>>> it doesn't look reasonable to be the default? What about using haltpoll governor
>>> as default when haltpoll idle driver registers or modloads.
>>>
>>> My idea to achieve the above would be to decrease the rating to 9 (before the
>>> lowest rated governor) and retain old defaults before haltpoll. Then we would
>>> allow a cpuidle driver to define a preferred governor to switch on idle driver
>>> registration. Naturally all of would be ignored if overidden by
>>> cpuidle.governor=.
>>>
>>> The diff below the scissors line is an example of that.
>>>
>>> Thoughts?
>>
>> Works for me. Rafael?
>
> It works for me too, basically, except that I would rename
> cpuidle_default_governor in the patch to cpuidle_prev_governor.
>
Great! I'll send over a series with this then (splitted accordingly).
Also, In the course of hotplug/hotunplug testing, I found two small issues with
modload/modunload -- regardless of the hotplug patch. So I am gonna add that to
the series too.
Joao
next prev parent reply other threads:[~2019-09-03 10:15 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-29 15:10 [PATCH v2] cpuidle-haltpoll: vcpu hotplug support Joao Martins
2019-08-29 15:27 ` Marcelo Tosatti
2019-09-02 10:48 ` Joao Martins
2019-08-29 17:16 ` Is: Default governor regardless of cpuidle driver Was: " Joao Martins
2019-08-29 17:23 ` Marcelo Tosatti
2019-09-02 21:55 ` Rafael J. Wysocki
2019-09-03 10:13 ` Joao Martins [this message]
2019-08-29 17:42 ` Daniel Lezcano
2019-08-29 18:07 ` Joao Martins
2019-08-29 18:28 ` Daniel Lezcano
2019-08-29 19:11 ` Default governor regardless of cpuidle driver Joao Martins
2019-08-29 20:22 ` Daniel Lezcano
2019-08-29 21:12 ` Joao Martins
2019-08-29 21:51 ` Daniel Lezcano
2019-08-30 11:07 ` Joao Martins
2019-09-02 21:58 ` Rafael J. Wysocki
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=457e8ca1-beb3-ca39-b257-e7bc6bb35d4d@oracle.com \
--to=joao.m.martins@oracle.com \
--cc=boris.ostrovsky@oracle.com \
--cc=daniel.lezcano@linaro.org \
--cc=jmattson@google.com \
--cc=joro@8bytes.org \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=mtosatti@redhat.com \
--cc=pbonzini@redhat.com \
--cc=rafael@kernel.org \
--cc=rjw@rjwysocki.net \
--cc=rkrcmar@redhat.com \
--cc=sean.j.christopherson@intel.com \
--cc=vkuznets@redhat.com \
--cc=wanpengli@tencent.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).