linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Patrick Bellasi <patrick.bellasi@arm.com>
To: Quentin Perret <quentin.perret@arm.com>
Cc: peterz@infradead.org, rjw@rjwysocki.net,
	linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org,
	gregkh@linuxfoundation.org, mingo@redhat.com,
	dietmar.eggemann@arm.com, morten.rasmussen@arm.com,
	chris.redpath@arm.com, valentin.schneider@arm.com,
	vincent.guittot@linaro.org, thara.gopinath@linaro.org,
	viresh.kumar@linaro.org, tkjos@google.com,
	joel@joelfernandes.org, smuckle@google.com,
	adharmap@codeaurora.org, skannan@codeaurora.org,
	pkondeti@codeaurora.org, juri.lelli@redhat.com,
	edubezval@gmail.com, srinivas.pandruvada@linux.intel.com,
	currojerez@riseup.net, javi.merino@kernel.org
Subject: Re: [PATCH v6 03/14] PM: Introduce an Energy Model management framework
Date: Fri, 31 Aug 2018 10:04:42 +0100	[thread overview]
Message-ID: <20180831090442.GA25636@e110439-lin> (raw)
In-Reply-To: <20180829132811.iacfltcos6kfgp7e@queper01-lin>

On 29-Aug 14:28, Quentin Perret wrote:
> Hi Patrick,
> 
> On Wednesday 29 Aug 2018 at 11:04:35 (+0100), Patrick Bellasi wrote:
> > In the loop above we use smp_store_release() to propagate the pointer
> > setting in a PER_CPU(em_data), which ultimate goal is to protect
> > em_register_perf_domain() from multiple clients registering the same
> > power domain.
> > 
> > I think there are two possible optimizations there:
> > 
> > 1. use of a single memory barrier
> > 
> >    Since we are already em_pd_mutex protected, i.e. there cannot be a
> >    concurrent writers, we can use one single memory barrier after the
> >    loop, i.e.
> > 
> >         for_each_cpu(cpu, span)
> >                 WRITE_ONCE()
> >         smp_wmb()
> > 
> >    which should be just enough to ensure that all other CPUs will see
> >    the pointer set once we release the mutex
> 
> Right, I'm actually wondering if the memory barrier is needed at all ...
> The mutex lock()/unlock() should already ensure the ordering I want no ?
> 
> WRITE_ONCE() should prevent load/store tearing with concurrent em_cpu_get(),
> and the release/acquire semantics of mutex lock/unlock should be enough to
> serialize the memory accesses of concurrent em_register_perf_domain() calls
> properly ...
> 
> Hmm, let me read memory-barriers.txt again.

Yes, I think it should... but better double check.

> > 2. avoid using PER_CPU variables
> > 
> >    Apart from the initialization code, i.e. boot time, the em_data is
> >    expected to be read only, isn't it?
> 
> That's right. It's not only read only, it's also not read very often (in
> the use-cases I have in mind at least). The scheduler for example will
> call em_cpu_get() once when sched domains are built, and keep the
> reference instead of calling it again.
> 
> >    If that's the case, I think that using PER_CPU variables is not
> >    strictly required while it unnecessarily increases the cache pressure.
> > 
> >    In the worst case we can end up with one cache line for each CPU to
> >    host just an 8B pointer, instead of using that single cache line to host
> >    up to 8 pointers if we use just an array, i.e.
> > 
> >         struct em_perf_domain *em_data[NR_CPUS]
> >                 ____cacheline_aligned_in_smp __read_mostly;
> > 
> >    Consider also that: up to 8 pointers in a single cache line means
> >    also that single cache line can be enough to access the EM from all
> >    the CPUs of almost every modern mobile phone SoC.
> > 
> >    Note entirely sure if PER_CPU uses less overall memory in case you
> >    have much less CPUs then the compile time defined NR_CPUS.
> >    But still, if the above makes sense, you still have a 8x gain
> >    factor between number Write allocated .data..percp sections and
> >    the value of NR_CPUS. Meaning that in the worst case we allocate
> >    the same amount of memory using NR_CPUS=64 (the default on arm64)
> >    while running on an 8 CPUs system... but still we should get less
> >    cluster caches pressure at run-time with the array approach, 1
> >    cache line vs 4.
> 
> Right, using per_cpu() might cause to bring in cache things you don't
> really care about (other non-related per_cpu stuff), but that shouldn't
> waste memory I think. I mean, if my em_data var is the first in a cache
> line, the rest of the cache line will most likely be used by other
> per_cpu variables anyways ...
> 
> As you suggested, the alternative would be to have a simple array. I'm
> fine with this TBH. But I would probably allocate it dynamically using
> nr_cpu_ids instead of using a static NR_CPUS-wide thing though -- the
> registration of perf domains usually happens late enough in the boot
> process.
> 
> What do you think ?

Sound all reasonable to me.

> Thanks
> Quentin

Best Patrick

-- 
#include <best/regards.h>

Patrick Bellasi

  reply	other threads:[~2018-08-31  9:04 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-08-20  9:44 [PATCH v6 00/14] Energy Aware Scheduling Quentin Perret
2018-08-20  9:44 ` [PATCH v6 01/14] sched: Relocate arch_scale_cpu_capacity Quentin Perret
2018-08-20  9:44 ` [PATCH v6 02/14] sched/cpufreq: Factor out utilization to frequency mapping Quentin Perret
2018-09-10  9:29   ` Rafael J. Wysocki
2018-08-20  9:44 ` [PATCH v6 03/14] PM: Introduce an Energy Model management framework Quentin Perret
2018-08-29 10:04   ` Patrick Bellasi
2018-08-29 13:28     ` Quentin Perret
2018-08-31  9:04       ` Patrick Bellasi [this message]
2018-09-11  9:34       ` Andrea Parri
2018-09-11 12:32         ` Quentin Perret
2018-09-11 13:31           ` Andrea Parri
2018-09-10  9:44   ` Rafael J. Wysocki
2018-09-10 10:38     ` Quentin Perret
2018-09-10 10:40       ` Rafael J. Wysocki
2018-08-20  9:44 ` [PATCH v6 04/14] PM / EM: Expose the Energy Model in sysfs Quentin Perret
2018-09-06  6:56   ` Dietmar Eggemann
2018-09-06 14:09     ` Quentin Perret
2018-09-07  0:14       ` Dietmar Eggemann
2018-08-20  9:44 ` [PATCH v6 05/14] sched/topology: Reference the Energy Model of CPUs when available Quentin Perret
2018-08-29 16:22   ` Patrick Bellasi
2018-08-29 16:56     ` Quentin Perret
2018-08-30 10:00       ` Patrick Bellasi
2018-08-30 10:47         ` Quentin Perret
2018-08-30 12:50           ` Patrick Bellasi
2018-08-20  9:44 ` [PATCH v6 06/14] sched/topology: Lowest CPU asymmetry sched_domain level pointer Quentin Perret
2018-08-20  9:44 ` [PATCH v6 07/14] sched/topology: Introduce sched_energy_present static key Quentin Perret
2018-08-29 16:50   ` Patrick Bellasi
2018-08-29 17:20     ` Quentin Perret
2018-08-30  9:23       ` Patrick Bellasi
2018-08-30  9:57         ` Quentin Perret
2018-08-30 10:18           ` Patrick Bellasi
2018-09-06  6:06   ` Dietmar Eggemann
2018-09-06  9:29     ` Quentin Perret
2018-09-06 23:49       ` Dietmar Eggemann
2018-09-07  8:24         ` Quentin Perret
2018-08-20  9:44 ` [PATCH v6 08/14] sched/fair: Clean-up update_sg_lb_stats parameters Quentin Perret
2018-08-20  9:44 ` [PATCH v6 09/14] sched: Add over-utilization/tipping point indicator Quentin Perret
2018-08-20  9:44 ` [PATCH v6 10/14] sched/cpufreq: Refactor the utilization aggregation method Quentin Perret
2018-09-10  9:53   ` Rafael J. Wysocki
2018-09-10 10:07     ` Quentin Perret
2018-09-10 10:25       ` Rafael J. Wysocki
2018-08-20  9:44 ` [PATCH v6 11/14] sched/fair: Introduce an energy estimation helper function Quentin Perret
2018-08-20  9:44 ` [PATCH v6 12/14] sched/fair: Select an energy-efficient CPU on task wake-up Quentin Perret
2018-08-20  9:44 ` [PATCH v6 13/14] sched/topology: Make Energy Aware Scheduling depend on schedutil Quentin Perret
2018-09-04 10:59   ` Quentin Perret
2018-09-06  9:18     ` Rafael J. Wysocki
2018-09-06 14:38       ` Quentin Perret
2018-09-07  8:52         ` Rafael J. Wysocki
2018-09-07  8:56           ` Rafael J. Wysocki
2018-09-07  9:02             ` Quentin Perret
2018-09-07 15:29           ` Quentin Perret
2018-09-09 20:13             ` Rafael J. Wysocki
2018-09-10  8:24               ` Quentin Perret
2018-09-10  8:55                 ` Rafael J. Wysocki
2018-09-10  9:43                   ` Quentin Perret
2018-08-20  9:44 ` [PATCH v6 14/14] OPTIONAL: cpufreq: dt: Register an Energy Model Quentin Perret
2018-09-10  9:12 ` [PATCH v6 00/14] Energy Aware Scheduling 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=20180831090442.GA25636@e110439-lin \
    --to=patrick.bellasi@arm.com \
    --cc=adharmap@codeaurora.org \
    --cc=chris.redpath@arm.com \
    --cc=currojerez@riseup.net \
    --cc=dietmar.eggemann@arm.com \
    --cc=edubezval@gmail.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=javi.merino@kernel.org \
    --cc=joel@joelfernandes.org \
    --cc=juri.lelli@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pm@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=morten.rasmussen@arm.com \
    --cc=peterz@infradead.org \
    --cc=pkondeti@codeaurora.org \
    --cc=quentin.perret@arm.com \
    --cc=rjw@rjwysocki.net \
    --cc=skannan@codeaurora.org \
    --cc=smuckle@google.com \
    --cc=srinivas.pandruvada@linux.intel.com \
    --cc=thara.gopinath@linaro.org \
    --cc=tkjos@google.com \
    --cc=valentin.schneider@arm.com \
    --cc=vincent.guittot@linaro.org \
    --cc=viresh.kumar@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).