From: Yuyang Du <yuyang.du@intel.com>
To: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
Rabin Vincent <rabin.vincent@axis.com>,
Mike Galbraith <umgwanakikbuti@gmail.com>,
"mingo@redhat.com" <mingo@redhat.com>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
Paul Turner <pjt@google.com>, Ben Segall <bsegall@google.com>
Subject: Re: [PATCH?] Livelock in pick_next_task_fair() / idle_balance()
Date: Fri, 3 Jul 2015 02:42:34 +0800 [thread overview]
Message-ID: <20150702184234.GC5197@intel.com> (raw)
In-Reply-To: <20150702114454.GB7598@e105550-lin.cambridge.arm.com>
On Thu, Jul 02, 2015 at 12:44:55PM +0100, Morten Rasmussen wrote:
> On Thu, Jul 02, 2015 at 12:53:59PM +0200, Peter Zijlstra wrote:
> > On Thu, Jul 02, 2015 at 07:25:11AM +0800, Yuyang Du wrote:
> > > And obviously, the idle balancing livelock SHOULD happen: one CPU pulls
> > > tasks from the other, makes the other idle, and this iterates...
> > >
> > > That being said, it is also obvious to prevent the livelock from happening:
> > > idle pulling until the source rq's nr_running is 1, becuase otherwise we
> > > just avoid idleness by making another idleness.
> >
> > Well, ideally the imbalance calculation would be so that it would avoid
> > this from happening in the first place. Its a 'balance' operation, not a
> > 'steal everything'.
> >
> > We want to take work -- as we have none -- but we want to ensure that
> > afterwards we have equal work, ie we're balanced.
>
> Agreed, I think this is the true problem. See my other reply.
Yes, this is agreed at all time. Like I said load_balance() (for idle balancing)
should compute the right imbalance and move a fair amount, to achieve we are
balanced. Whatever is wrong in how much computed and moved "right imbalance" is
should be fixed anyway.
But still, I think, even with the above, in idle balancing, pulling until the source
rq's nr_running == 1 is not just "a short term fix", but should be there permanently
acting like a last guard with no overhead, why not.
>
> >
> > So clearly that all is hosed. Now Morten was looking into simplifying
> > calculate_imbalance() recently.
>
> Yes. I'm held up doing other stuff at the moment, but I think
> calculate_imbalance() needs some attention and I'm planning on looking at
> that next.
Thanks,
Yuyang
next prev parent reply other threads:[~2015-07-03 2:34 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-06-30 14:30 [PATCH?] Livelock in pick_next_task_fair() / idle_balance() Rabin Vincent
2015-07-01 5:36 ` Mike Galbraith
2015-07-01 14:55 ` Rabin Vincent
2015-07-01 15:47 ` Mike Galbraith
2015-07-01 20:44 ` Peter Zijlstra
2015-07-01 23:25 ` Yuyang Du
2015-07-02 8:05 ` Mike Galbraith
2015-07-02 1:05 ` Yuyang Du
2015-07-02 10:25 ` Mike Galbraith
2015-07-02 11:40 ` Morten Rasmussen
2015-07-02 19:37 ` Yuyang Du
2015-07-03 9:34 ` Morten Rasmussen
2015-07-03 16:38 ` Peter Zijlstra
2015-07-05 22:31 ` Yuyang Du
2015-07-09 14:32 ` Morten Rasmussen
2015-07-09 23:24 ` Yuyang Du
2015-07-05 20:12 ` Yuyang Du
2015-07-06 17:36 ` Dietmar Eggemann
2015-07-07 11:17 ` Rabin Vincent
2015-07-13 17:43 ` Dietmar Eggemann
2015-07-09 13:53 ` Morten Rasmussen
2015-07-09 22:34 ` Yuyang Du
2015-07-02 10:53 ` Peter Zijlstra
2015-07-02 11:44 ` Morten Rasmussen
2015-07-02 18:42 ` Yuyang Du [this message]
2015-07-03 4:42 ` Mike Galbraith
2015-07-03 16:39 ` Peter Zijlstra
2015-07-05 22:11 ` Yuyang Du
2015-07-09 6:15 ` Stefan Ekenberg
2015-07-26 18:57 ` Yuyang Du
2015-08-03 17:05 ` [tip:sched/core] sched/fair: Avoid pulling all tasks in idle balancing tip-bot for Yuyang Du
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=20150702184234.GC5197@intel.com \
--to=yuyang.du@intel.com \
--cc=bsegall@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=morten.rasmussen@arm.com \
--cc=peterz@infradead.org \
--cc=pjt@google.com \
--cc=rabin.vincent@axis.com \
--cc=umgwanakikbuti@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.