LKML Archive on lore.kernel.org
 help / color / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: Patrick Bellasi <patrick.bellasi@arm.com>
Cc: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org,
	Ingo Molnar <mingo@redhat.com>, Tejun Heo <tj@kernel.org>,
	"Rafael J . Wysocki" <rafael.j.wysocki@intel.com>,
	Viresh Kumar <viresh.kumar@linaro.org>,
	Vincent Guittot <vincent.guittot@linaro.org>,
	Paul Turner <pjt@google.com>,
	Quentin Perret <quentin.perret@arm.com>,
	Dietmar Eggemann <dietmar.eggemann@arm.com>,
	Morten Rasmussen <morten.rasmussen@arm.com>,
	Juri Lelli <juri.lelli@redhat.com>, Todd Kjos <tkjos@google.com>,
	Joel Fernandes <joelaf@google.com>,
	Steve Muckle <smuckle@google.com>,
	Suren Baghdasaryan <surenb@google.com>
Subject: Re: [PATCH v4 02/16] sched/core: uclamp: map TASK's clamp values into CPU's clamp groups
Date: Wed, 12 Sep 2018 18:12:18 +0200
Message-ID: <20180912161218.GW24082@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20180912155619.GG1413@e110439-lin>

On Wed, Sep 12, 2018 at 04:56:19PM +0100, Patrick Bellasi wrote:
> On 12-Sep 15:49, Peter Zijlstra wrote:
> > On Tue, Aug 28, 2018 at 02:53:10PM +0100, Patrick Bellasi wrote:

> > > +/**
> > > + * uclamp_map: reference counts a utilization "clamp value"
> > > + * @value:    the utilization "clamp value" required
> > > + * @se_count: the number of scheduling entities requiring the "clamp value"
> > > + * @se_lock:  serialize reference count updates by protecting se_count
> > 
> > Why do you have a spinlock to serialize a single value? Don't we have
> > atomics for that?
> 
> There are some code paths where it's used to protect clamp groups
> mapping and initialization, e.g.
> 
>       uclamp_group_get()
>           spin_lock()
>               // initialize clamp group (if required) and then...
>               se_count += 1
>           spin_unlock()
> 
> Almost all these paths are triggered from user-space and protected
> by a global uclamp_mutex, but fork/exit paths.
> 
> To serialize these paths I'm using the spinlock above, does it make
> sense ? Can we use the global uclamp_mutex on forks/exit too ?

OK, then your comment is misleading; it serializes both fields.

> One additional observations is that, if in the future we want to add a
> kernel space API, (e.g. driver asking for a new clamp value), maybe we
> will need to have a serialized non-sleeping uclamp_group_get() API ?

No idea; but if you want to go all fancy you can replace he whole
uclamp_map thing with something like:

struct uclamp_map {
	union {
		struct {
			unsigned long v : 10;
			unsigned long c : BITS_PER_LONG - 10;
		};
		atomic_long_t s;
	};
};

And use uclamp_map::c == 0 as unused (as per normal refcount
semantics) and atomic_long_cmpxchg() the whole thing using
uclamp_map::s.

> > > + *                          uclamp_maps is a matrix of
> > > + *          +------- UCLAMP_CNT by CONFIG_UCLAMP_GROUPS_COUNT+1 entries
> > > + *          |                                |
> > > + *          |                /---------------+---------------\
> > > + *          |               +------------+       +------------+
> > > + *          |  / UCLAMP_MIN | value      |       | value      |
> > > + *          |  |            | se_count   |...... | se_count   |
> > > + *          |  |            +------------+       +------------+
> > > + *          +--+            +------------+       +------------+
> > > + *             |            | value      |       | value      |
> > > + *             \ UCLAMP_MAX | se_count   |...... | se_count   |
> > > + *                          +-----^------+       +----^-------+
> > > + *                                |                   |
> > > + *                      uc_map =  +                   |
> > > + *                     &uclamp_maps[clamp_id][0]      +
> > > + *                                                clamp_value =
> > > + *                                       uc_map[group_id].value
> > > + */
> > > +static struct uclamp_map uclamp_maps[UCLAMP_CNT]
> > > +				    [CONFIG_UCLAMP_GROUPS_COUNT + 1]
> > > +				    ____cacheline_aligned_in_smp;
> > > +
> > 
> > I'm still completely confused by all this.
> > 
> > sizeof(uclamp_map) = 12
> > 
> > that array is 2*6=12 of those, so the whole thing is 144 bytes. which is
> > more than 2 (64 byte) cachelines.
> 
> This data structure is *not* used in the hot-path, that's why I did not
> care about fitting it exactly into few cache lines.
> 
> It's used to map a user-space "clamp value" into a kernel-space "clamp
> group" when user-space:
>  - changes a task specific clamp value
>  - changes a cgroup clamp value
>  - a task forks/exits
> 
> I assume we can consider all those as "slow" code paths, is that correct ?

Yep.

> > What's the purpose of that cacheline align statement?
> 
> In uclamp_maps, we still need to scan the array when a clamp value is
> changed from user-space, i.e. the cases reported above. Thus, that
> alignment is just to ensure that we minimize the number of cache lines
> used. Does that make sense ?
> 
> Maybe that alignment implicitly generated by the compiler ?

It is not, but if it really is a slow path, we shouldn't care about
alignment.

> > Note that without that apparently superfluous lock, it would be 8*12 =
> > 96 bytes, which is 1.5 lines and would indeed suggest you default to
> > GROUP_COUNT=7 by default to fill 2 lines.
> 
> Yes, will check better if we can count on just the uclamp_mutex

Well, if we don't care about performance (slow path) then keeping he
lock is fine, just the comment and alignment are misleading.

> > Why are the min and max things torn up like that?  I'm fairly sure I
> > asked some of that last time; but the above comments only try to explain
> > what, not why.
> 
> We use that organization to speedup scanning for clamp values of the
> same clamp_id. That's more important in the hot-path than above, where
> we need to scan struct uclamp_cpu when a new aggregated clamp value
> has to be computed. This is done in:
> 
>    [PATCH v4 03/16] sched/core: uclamp: add CPU's clamp groups accounting
>    https://lore.kernel.org/lkml/20180828135324.21976-4-patrick.bellasi@arm.com/
> 
> Specifically:
> 
>    dequeue_task()
>      uclamp_cpu_put()
>        uclamp_cpu_put_id(clamp_id)
>          uclamp_cpu_update(clamp_id)
>             // Here we have an array scan by clamp_id
> 
> With the given data layout I reported above, when we update the
> min_clamp value (boost) we have all the data required in a single
> cache line.
> 
> If that makes sense, I can certainly improve the comment above to
> justify its layout.

OK, let me read on.

  reply index

Thread overview: 80+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-08-28 13:53 [PATCH v4 00/16] Add utilization clamping support Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 01/16] sched/core: uclamp: extend sched_setattr to support utilization clamping Patrick Bellasi
2018-09-05 11:01   ` Juri Lelli
2018-08-28 13:53 ` [PATCH v4 02/16] sched/core: uclamp: map TASK's clamp values into CPU's clamp groups Patrick Bellasi
2018-09-05 10:45   ` Juri Lelli
2018-09-06 13:48     ` Patrick Bellasi
2018-09-06 14:13       ` Juri Lelli
2018-09-06  8:17   ` Juri Lelli
2018-09-06 14:00     ` Patrick Bellasi
2018-09-08 23:47   ` Suren Baghdasaryan
2018-09-12 10:32     ` Patrick Bellasi
2018-09-12 13:49   ` Peter Zijlstra
2018-09-12 15:56     ` Patrick Bellasi
2018-09-12 16:12       ` Peter Zijlstra [this message]
2018-09-12 17:35         ` Patrick Bellasi
2018-09-12 17:42           ` Peter Zijlstra
2018-09-12 17:52             ` Patrick Bellasi
2018-09-13 19:14               ` Peter Zijlstra
2018-09-14  8:51                 ` Patrick Bellasi
2018-09-12 16:24   ` Peter Zijlstra
2018-09-12 17:42     ` Patrick Bellasi
2018-09-13 19:20       ` Peter Zijlstra
2018-09-14  8:47         ` Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 03/16] sched/core: uclamp: add CPU's clamp groups accounting Patrick Bellasi
2018-09-12 17:34   ` Peter Zijlstra
2018-09-12 17:44     ` Patrick Bellasi
2018-09-13 19:12   ` Peter Zijlstra
2018-09-14  9:07     ` Patrick Bellasi
2018-09-14 11:52       ` Peter Zijlstra
2018-09-14 13:41         ` Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 04/16] sched/core: uclamp: update CPU's refcount on clamp changes Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 05/16] sched/core: uclamp: enforce last task UCLAMP_MAX Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 06/16] sched/cpufreq: uclamp: add utilization clamping for FAIR tasks Patrick Bellasi
2018-09-14  9:32   ` Peter Zijlstra
2018-09-14 13:19     ` Patrick Bellasi
2018-09-14 13:36       ` Peter Zijlstra
2018-09-14 13:57         ` Patrick Bellasi
2018-09-27 10:23           ` Quentin Perret
2018-08-28 13:53 ` [PATCH v4 07/16] sched/core: uclamp: extend cpu's cgroup controller Patrick Bellasi
2018-08-28 18:29   ` Randy Dunlap
2018-08-29  8:53     ` Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 08/16] sched/core: uclamp: propagate parent clamps Patrick Bellasi
2018-09-09  3:02   ` Suren Baghdasaryan
2018-09-12 12:51     ` Patrick Bellasi
2018-09-12 15:56       ` Suren Baghdasaryan
2018-09-11 15:18   ` Tejun Heo
2018-09-11 16:26     ` Patrick Bellasi
2018-09-11 16:28       ` Tejun Heo
2018-08-28 13:53 ` [PATCH v4 09/16] sched/core: uclamp: map TG's clamp values into CPU's clamp groups Patrick Bellasi
2018-09-09 18:52   ` Suren Baghdasaryan
2018-09-12 14:19     ` Patrick Bellasi
2018-09-12 15:53       ` Suren Baghdasaryan
2018-08-28 13:53 ` [PATCH v4 10/16] sched/core: uclamp: use TG's clamps to restrict Task's clamps Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 11/16] sched/core: uclamp: add system default clamps Patrick Bellasi
2018-09-10 16:20   ` Suren Baghdasaryan
2018-09-11 16:46     ` Patrick Bellasi
2018-09-11 19:25       ` Suren Baghdasaryan
2018-08-28 13:53 ` [PATCH v4 12/16] sched/core: uclamp: update CPU's refcount on TG's clamp changes Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 13/16] sched/core: uclamp: use percentage clamp values Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 14/16] sched/core: uclamp: request CAP_SYS_ADMIN by default Patrick Bellasi
2018-09-04 13:47   ` Juri Lelli
2018-09-06 14:40     ` Patrick Bellasi
2018-09-06 14:59       ` Juri Lelli
2018-09-06 17:21         ` Patrick Bellasi
2018-09-14 11:10       ` Peter Zijlstra
2018-09-14 14:07         ` Patrick Bellasi
2018-09-14 14:28           ` Peter Zijlstra
2018-09-17 12:27             ` Patrick Bellasi
2018-09-21  9:13               ` Peter Zijlstra
2018-09-24 15:14                 ` Patrick Bellasi
2018-09-24 15:56                   ` Peter Zijlstra
2018-09-24 17:23                     ` Patrick Bellasi
2018-09-24 16:26                   ` Peter Zijlstra
2018-09-24 17:19                     ` Patrick Bellasi
2018-09-25 15:49                   ` Peter Zijlstra
2018-09-26 10:43                     ` Patrick Bellasi
2018-09-27 10:00                     ` Quentin Perret
2018-09-26 17:51                 ` Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 15/16] sched/core: uclamp: add clamp group discretization support Patrick Bellasi
2018-08-28 13:53 ` [PATCH v4 16/16] sched/cpufreq: uclamp: add utilization clamping for RT tasks Patrick Bellasi

Reply instructions:

You may reply publically 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=20180912161218.GW24082@hirez.programming.kicks-ass.net \
    --to=peterz@infradead.org \
    --cc=dietmar.eggemann@arm.com \
    --cc=joelaf@google.com \
    --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=patrick.bellasi@arm.com \
    --cc=pjt@google.com \
    --cc=quentin.perret@arm.com \
    --cc=rafael.j.wysocki@intel.com \
    --cc=smuckle@google.com \
    --cc=surenb@google.com \
    --cc=tj@kernel.org \
    --cc=tkjos@google.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

LKML Archive on lore.kernel.org

Archives are clonable:
	git clone --mirror https://lore.kernel.org/lkml/0 lkml/git/0.git
	git clone --mirror https://lore.kernel.org/lkml/1 lkml/git/1.git
	git clone --mirror https://lore.kernel.org/lkml/2 lkml/git/2.git
	git clone --mirror https://lore.kernel.org/lkml/3 lkml/git/3.git
	git clone --mirror https://lore.kernel.org/lkml/4 lkml/git/4.git
	git clone --mirror https://lore.kernel.org/lkml/5 lkml/git/5.git
	git clone --mirror https://lore.kernel.org/lkml/6 lkml/git/6.git
	git clone --mirror https://lore.kernel.org/lkml/7 lkml/git/7.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 lkml lkml/ https://lore.kernel.org/lkml \
		linux-kernel@vger.kernel.org
	public-inbox-index lkml

Example config snippet for mirrors

Newsgroup available over NNTP:
	nntp://nntp.lore.kernel.org/org.kernel.vger.linux-kernel


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git