From: Quan Xu <quan.xu0@gmail.com>
To: Peter Zijlstra <peterz@infradead.org>,
Yang Zhang <yang.zhang.wz@gmail.com>
Cc: quan.xu0@gmail.com, linux-kernel@vger.kernel.org,
kvm@vger.kernel.org, wanpeng.li@hotmail.com, mst@redhat.com,
pbonzini@redhat.com, tglx@linutronix.de, rkrcmar@redhat.com,
dmatlack@google.com, agraf@suse.de, linux-doc@vger.kernel.org,
Ingo Molnar <mingo@redhat.com>
Subject: Re: [RFC PATCH v2 7/7] sched/idle: update poll time when wakeup from idle
Date: Fri, 29 Sep 2017 18:29:21 +0800 [thread overview]
Message-ID: <db8e812b-8f86-b996-1477-d8baaf7f562a@gmail.com> (raw)
In-Reply-To: <20170829124655.ann5cipoubl5lppn@hirez.programming.kicks-ass.net>
On 2017/8/29 20:46, Peter Zijlstra wrote:
> On Tue, Aug 29, 2017 at 11:46:41AM +0000, Yang Zhang wrote:
>> In ttwu_do_wakeup, it will update avg_idle when wakeup from idle. Here
>> we just reuse this logic to update the poll time. It may be a little
>> late to update the poll in ttwu_do_wakeup, but the test result shows no
>> obvious performance gap compare with updating poll in irq handler.
>>
>> one problem is that idle_stamp only used when using CFS scheduler. But
>> it is ok since it is the default policy for scheduler and only consider
>> it should enough.
>>
>> Signed-off-by: Yang Zhang <yang.zhang.wz@gmail.com>
>> Signed-off-by: Quan Xu <quan.xu0@gmail.com>
> Same broken SoB chain, and not a useful word on why you need to adjust
> this crap to begin with. What you want that poll duration to be related
> to is the cost of a VMEXIT/VMENTER cycle, not however long we happened
> to be idle.
>
> So no.
Peter,
I think you are right..
IIUC, the time we happened to be idle may contain a chain of
VMEXIT/VMENTER cycles,
which would be mainly (except the last VMEXIT/VMENTER cycles) for just
idle loops. right?
as you mentioned, poll duration to be related to is the cost of __a__
VMEXIT/VMENTER cycle.
howerver it is very difficult to measure a VMEXIT/VMENTER cycle
accurately from
kvm guest, we could find out an approximate one -- dropping the idle
loops from the
time we happened to be idle.. make sense?
Quan
next prev parent reply other threads:[~2017-09-29 10:29 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-29 11:46 [RFC PATCH v2 0/7] x86/idle: add halt poll support Yang Zhang
2017-08-29 11:46 ` [RFC PATCH v2 1/7] x86/paravirt: Add pv_idle_ops to paravirt ops Yang Zhang
2017-08-29 11:46 ` Yang Zhang
2017-08-29 13:55 ` Konrad Rzeszutek Wilk
2017-08-29 13:55 ` Konrad Rzeszutek Wilk
2017-08-30 7:33 ` Juergen Gross
2017-08-30 7:33 ` Juergen Gross
2017-09-01 6:50 ` Yang Zhang
2017-09-01 6:50 ` Yang Zhang
2017-08-29 11:46 ` [RFC PATCH v2 2/7] KVM guest: register kvm_idle_poll for pv_idle_ops Yang Zhang
2017-08-29 11:46 ` [RFC PATCH v2 3/7] sched/idle: Add poll before enter real idle path Yang Zhang
2017-08-29 12:45 ` Peter Zijlstra
2017-09-01 5:57 ` Quan Xu
2017-09-14 8:41 ` Quan Xu
2017-09-14 9:18 ` Borislav Petkov
2017-08-29 14:39 ` Borislav Petkov
2017-09-01 6:49 ` Quan Xu
2017-09-29 10:39 ` Quan Xu
2017-08-29 11:46 ` [RFC PATCH v2 4/7] x86/paravirt: Add update in x86/paravirt pv_idle_ops Yang Zhang
2017-08-29 11:46 ` Yang Zhang
2017-08-29 11:46 ` [RFC PATCH v2 5/7] Documentation: Add three sysctls for smart idle poll Yang Zhang
2017-08-29 11:46 ` Yang Zhang
2017-08-29 11:46 ` Yang Zhang
2017-08-29 17:20 ` Luis R. Rodriguez
2017-08-29 17:20 ` Luis R. Rodriguez
2017-08-29 17:20 ` Luis R. Rodriguez
2017-08-29 11:46 ` [RFC PATCH v2 6/7] KVM guest: introduce smart idle poll algorithm Yang Zhang
2017-08-29 11:46 ` [RFC PATCH v2 7/7] sched/idle: update poll time when wakeup from idle Yang Zhang
2017-08-29 12:46 ` Peter Zijlstra
2017-09-01 7:30 ` Yang Zhang
2017-09-29 10:29 ` Quan Xu [this message]
2017-08-29 11:58 ` [RFC PATCH v2 0/7] x86/idle: add halt poll support Alexander Graf
2017-09-01 6:21 ` Yang Zhang
2017-08-29 13:03 ` Andi Kleen
2017-08-29 14:02 ` Wanpeng Li
2017-08-29 14:27 ` Konrad Rzeszutek Wilk
2017-08-29 14:36 ` Michael S. Tsirkin
2017-09-01 6:32 ` Yang Zhang
2017-09-01 6:52 ` Wanpeng Li
2017-09-01 6:44 ` Yang Zhang
2017-09-01 6:58 ` Wanpeng Li
2017-09-01 7:53 ` Yang Zhang
2017-08-29 14:56 ` Michael S. Tsirkin
2017-09-13 11:56 ` Yang Zhang
2017-09-14 8:36 ` Quan Xu
2017-09-14 9:19 ` Wanpeng Li
2017-09-14 9:40 ` Quan Xu
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=db8e812b-8f86-b996-1477-d8baaf7f562a@gmail.com \
--to=quan.xu0@gmail.com \
--cc=agraf@suse.de \
--cc=dmatlack@google.com \
--cc=kvm@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peterz@infradead.org \
--cc=rkrcmar@redhat.com \
--cc=tglx@linutronix.de \
--cc=wanpeng.li@hotmail.com \
--cc=yang.zhang.wz@gmail.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.