From: Morten Rasmussen <morten.rasmussen@arm.com>
To: Vincent Guittot <vincent.guittot@linaro.org>
Cc: "peterz@infradead.org" <peterz@infradead.org>,
"mingo@kernel.org" <mingo@kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"preeti@linux.vnet.ibm.com" <preeti@linux.vnet.ibm.com>,
"kamalesh@linux.vnet.ibm.com" <kamalesh@linux.vnet.ibm.com>,
"linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>,
"riel@redhat.com" <riel@redhat.com>,
"efault@gmx.de" <efault@gmx.de>,
"nicolas.pitre@linaro.org" <nicolas.pitre@linaro.org>,
"linaro-kernel@lists.linaro.org" <linaro-kernel@lists.linaro.org>
Subject: Re: [PATCH v9 05/10] sched: make scale_rt invariant with frequency
Date: Wed, 26 Nov 2014 11:57:59 +0000 [thread overview]
Message-ID: <20141126115759.GB3430@e105550-lin.cambridge.arm.com> (raw)
In-Reply-To: <CAKfTPtAG1qtqSgG3Ek7CeKcvm2=KXbMdb7TgrVcEFivwGeDOGg@mail.gmail.com>
On Tue, Nov 25, 2014 at 01:48:02PM +0000, Vincent Guittot wrote:
> On 24 November 2014 at 18:05, Morten Rasmussen <morten.rasmussen@arm.com> wrote:
> > On Mon, Nov 24, 2014 at 02:24:00PM +0000, Vincent Guittot wrote:
> >> On 21 November 2014 at 13:35, Morten Rasmussen <morten.rasmussen@arm.com> wrote:
> >> > On Mon, Nov 03, 2014 at 04:54:42PM +0000, Vincent Guittot wrote:
> >>
> >> [snip]
> >>
> >> >> The average running time of RT tasks is used to estimate the remaining compute
> >> >> @@ -5801,19 +5801,12 @@ static unsigned long scale_rt_capacity(int cpu)
> >> >>
> >> >> total = sched_avg_period() + delta;
> >> >>
> >> >> - if (unlikely(total < avg)) {
> >> >> - /* Ensures that capacity won't end up being negative */
> >> >> - available = 0;
> >> >> - } else {
> >> >> - available = total - avg;
> >> >> - }
> >> >> + used = div_u64(avg, total);
> >> >
> >> > I haven't looked through all the details of the rt avg tracking, but if
> >> > 'used' is in the range [0..SCHED_CAPACITY_SCALE], I believe it should
> >> > work. Is it guaranteed that total > 0 so we don't get division by zero?
> >>
> >> static inline u64 sched_avg_period(void)
> >> {
> >> return (u64)sysctl_sched_time_avg * NSEC_PER_MSEC / 2;
> >> }
> >>
> >
> > I see.
> >
> >> >
> >> > It does get a slightly more complicated if we want to figure out the
> >> > available capacity at the current frequency (current < max) later. Say,
> >> > rt eats 25% of the compute capacity, but the current frequency is only
> >> > 50%. In that case get:
> >> >
> >> > curr_avail_capacity = (arch_scale_cpu_capacity() *
> >> > (arch_scale_freq_capacity() - (SCHED_SCALE_CAPACITY - scale_rt_capacity())))
> >> > >> SCHED_CAPACITY_SHIFT
> >>
> >> You don't have to be so complicated but simply need to do:
> >> curr_avail_capacity for CFS = (capacity_of(CPU) *
> >> arch_scale_freq_capacity()) >> SCHED_CAPACITY_SHIFT
> >>
> >> capacity_of(CPU) = 600 is the max available capacity for CFS tasks
> >> once we have removed the 25% of capacity that is used by RT tasks
> >> arch_scale_freq_capacity = 512 because we currently run at 50% of max freq
> >>
> >> so curr_avail_capacity for CFS = 300
> >
> > I don't think that is correct. It is at least not what I had in mind.
> >
> > capacity_orig_of(cpu) = 800, we run at 50% frequency which means:
> >
> > curr_capacity = capacity_orig_of(cpu) * arch_scale_freq_capacity()
> > >> SCHED_CAPACITY_SHIFT
> > = 400
> >
> > So the total capacity at the current frequency (50%) is 400, without
> > considering RT. scale_rt_capacity() is frequency invariant, so it takes
> > away capacity_orig_of(cpu) - capacity_of(cpu) = 200 worth of capacity
> > for RT. We need to subtract that from the current capacity to get the
> > available capacity at the current frequency.
> >
> > curr_available_capacity = curr_capacity - (capacity_orig_of(cpu) -
> > capacity_of(cpu)) = 200
>
> you're right, this one looks good to me too
Okay, thanks for confirming. It doesn't affect this patch set anyway, I just
wanted to be sure that I got all the scaling factors right :)
next prev parent reply other threads:[~2014-11-26 11:58 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-11-03 16:54 [PATCH v9 00/10] sched: consolidation of CPU capacity and usage Vincent Guittot
2014-11-03 16:54 ` [PATCH v9 01/10] sched: add utilization_avg_contrib Vincent Guittot
2014-11-21 12:34 ` Morten Rasmussen
2014-11-24 14:04 ` Vincent Guittot
2014-11-24 17:34 ` Morten Rasmussen
2014-11-03 16:54 ` [PATCH v9 02/10] sched: Track group sched_entity usage contributions Vincent Guittot
2014-11-21 12:35 ` Morten Rasmussen
2014-11-24 14:04 ` Vincent Guittot
2014-11-24 15:39 ` Morten Rasmussen
2014-11-03 16:54 ` [PATCH v9 03/10] sched: remove frequency scaling from cpu_capacity Vincent Guittot
2014-11-21 12:35 ` Morten Rasmussen
2014-11-03 16:54 ` [PATCH v9 04/10] sched: Make sched entity usage tracking scale-invariant Vincent Guittot
2014-11-21 12:35 ` Morten Rasmussen
2014-11-26 16:05 ` Dietmar Eggemann
2014-11-03 16:54 ` [PATCH v9 05/10] sched: make scale_rt invariant with frequency Vincent Guittot
2014-11-21 12:35 ` Morten Rasmussen
2014-11-24 14:24 ` Vincent Guittot
2014-11-24 17:05 ` Morten Rasmussen
2014-11-25 13:48 ` Vincent Guittot
2014-11-26 11:57 ` Morten Rasmussen [this message]
2014-11-25 2:24 ` Wanpeng Li
2014-11-25 13:52 ` Vincent Guittot
2014-11-26 5:18 ` Wanpeng Li
2014-11-26 8:27 ` Vincent Guittot
2014-11-03 16:54 ` [PATCH v9 06/10] sched: add per rq cpu_capacity_orig Vincent Guittot
2014-11-03 16:54 ` [PATCH v9 07/10] sched: get CPU's usage statistic Vincent Guittot
2014-11-21 12:36 ` Morten Rasmussen
2014-11-03 16:54 ` [PATCH v9 08/10] sched: replace capacity_factor by usage Vincent Guittot
2014-11-19 15:15 ` pang.xunlei
2014-11-19 17:30 ` Vincent Guittot
2014-11-21 12:37 ` Morten Rasmussen
2014-11-24 14:41 ` Vincent Guittot
2014-11-24 17:16 ` Morten Rasmussen
2014-11-03 16:54 ` [PATCH v9 09/10] sched: add SD_PREFER_SIBLING for SMT level Vincent Guittot
2014-11-03 16:54 ` [PATCH v9 10/10] sched: move cfs task on a CPU with higher capacity Vincent Guittot
2014-11-21 12:37 ` Morten Rasmussen
2014-11-24 14:45 ` Vincent Guittot
2014-11-24 17:30 ` Morten Rasmussen
2014-11-21 12:34 ` [PATCH v9 00/10] sched: consolidation of CPU capacity and usage Morten Rasmussen
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=20141126115759.GB3430@e105550-lin.cambridge.arm.com \
--to=morten.rasmussen@arm.com \
--cc=efault@gmx.de \
--cc=kamalesh@linux.vnet.ibm.com \
--cc=linaro-kernel@lists.linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=nicolas.pitre@linaro.org \
--cc=peterz@infradead.org \
--cc=preeti@linux.vnet.ibm.com \
--cc=riel@redhat.com \
--cc=vincent.guittot@linaro.org \
/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).