To support task performance boosting, the usage of a single knob has the
advantage to be a simple solution, both from the implementation and the
usability standpoint. However, on a real system it can be difficult to
identify a single value for the knob which fits the needs of multiple
different tasks. For example, some kernel threads and/or user-space
background services should be better managed the "standard" way while we
still want to be able to boost the performance of specific workloads.
In order to improve the flexibility of the task boosting mechanism this
patch is the first of a small series which extends the previous
implementation to introduce a "per task group" support.
This first patch introduces just the basic CGroups support, a new
"schedtune" CGroups controller is added which allows to configure
different boost value for different groups of tasks.
To keep the implementation simple but still effective for a boosting
strategy, the new controller:
1. allows only a two layer hierarchy
2. supports only a limited number of boost groups
A two layer hierarchy allows to place each task either:
a) in the root control group
thus being subject to a system-wide boosting value
b) in a child of the root group
thus being subject to the specific boost value defined by that
"boost group"
The limited number of "boost groups" supported is mainly motivated by
the observation that in a real system it could be useful to have only
few classes of tasks which deserve different treatment.
For example, background vs foreground or interactive vs low-priority.
As an additional benefit, a limited number of boost groups allows also
to have a simpler implementation especially for the code required to
compute the boost value for CPUs which have runnable tasks belonging to
different boost groups.
cc: Tejun Heo <tj@kernel.org>
cc: Li Zefan <lizefan@huawei.com>
cc: Johannes Weiner <hannes@cmpxchg.org>
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
The CPU usage signal is used by the scheduler as an estimation of the
overall bandwidth currently allocated on a CPU. When SchedDVFS is in
use, this signal affects the selection of the operating points (OPP)
required to accommodate all the workload allocated in a CPU.
A convenient way to boost the performance of tasks running on a CPU,
which is also little intrusive, is to boost the CPU usage signal each
time it is used to select an OPP.
This patch introduces a new function:
get_boosted_cpu_usage(cpu)
to return a boosted value for the usage of a specified CPU.
The margin added to the original usage is:
1. computed based on the "boosting strategy" in use
2. proportional to the system-wide boost value defined by provided
user-space interface
The boosted signal is used by SchedDVFS (transparently) each time it
requires to get an estimation of the capacity required for a CPU.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
The basic idea of the boost knob is to "artificially inflate" a signal
to make a task or logical CPU appears more demanding than it actually
is. Independently from the specific signal, a consistent and possibly
simple semantic for the concept of "signal boosting" must define:
1. how we translate the boost percentage into a "margin" value to be added
to the original signal to inflate
2. what is the meaning of a boost value from a user-space perspective
This patch provides the implementation of a possible boost semantic,
named "Signal Proportional Compensation" (SPC), where the boost
percentage (BP) is used to compute a margin (M) which is proportional to
the complement of the original signal (OS):
M = BP * (SCHED_LOAD_SCALE - OS)
The computed margin then added to the OS to obtain the Boosted Signal (BS)
BS = OS + M
The proposed boost semantic has these main features:
- each signal gets a boost which is proportional to its delta with respect
to the maximum available capacity in the system (i.e. SCHED_LOAD_SCALE)
- a 100% boosting has a clear understanding from a user-space perspective,
since it means simply to run (possibly) "all" tasks at the max OPP
- each boosting value means to improve the task performance by a quantity
which is proportional to the maximum achievable performance on that
system
Thus this semantics is somehow forcing a behaviour which is:
50% boosting means to run at half-way between the current and the
maximum performance which a task could achieve on that system
This patch provides the code to implement a fast integer division to
convert a boost percentage (BP) value into a margin (M).
NOTE: this code is suitable for all signals operating in range
[0..SCHED_LOAD_SCALE]
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
The current (CFS) scheduler implementation does not allow "to boost"
tasks performance by running them at a higher OPP compared to the
minimum required to meet their workload demands.
To support tasks performance boosting the scheduler should provide a
"knob" which allows to tune how much the system is going to be optimised
for energy efficiency vs performance.
This patch is the first of a series which provides a simple interface to
define a tuning knob. One system-wide "boost" tunable is exposed via:
/proc/sys/kernel/sched_cfs_boost
which can be configured in the range [0..100], to define a percentage
where:
- 0% boost requires to operate in "standard" mode by scheduling
tasks at the minimum capacities required by the workload demand
- 100% boost requires to push at maximum the task performances,
"regardless" of the incurred energy consumption
A boost value in between these two boundaries is used to bias the
power/performance trade-off, the higher the boost value the more the
scheduler is biased toward performance boosting instead of energy
efficiency.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
RT tasks don't provide any running constraints like deadline ones
except their running priority. The only current usable input to
estimate the capacity needed by RT tasks is the rt_avg metric. We use
it to estimate the CPU capacity needed for the RT scheduler class.
In order to monitor the evolution for RT task load, we must
peridiocally check it during the tick.
Then, we use the estimated capacity of the last activity to estimate
the next one which can not be that accurate but is a good starting
point without any impact on the wake up path of RT tasks.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
Instead of monitoring the exec time of deadline tasks to evaluate the
CPU capacity consumed by deadline scheduler class, we can directly
calculate it thanks to the sum of utilization of deadline tasks on the
CPU. We can remove deadline tasks from rt_avg metric and directly use
the average bandwidth of deadline scheduler in scale_rt_capacity.
Based in part on a similar patch from Luca Abeni <luca.abeni@unitn.it>.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
rt_avg is only used to scale the available CPU's capacity for CFS
tasks. As the update of this scaling is done during periodic load
balance, we only have to ensure that sched_avg_update has been called
before any periodic load balancing. This requirement is already
fulfilled by __update_cpu_load so the call in sched_rt_avg_update,
which is part of the hotpath, is useless.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
Since the true utilization of a long running task is not detectable
while it is running and might be bigger than the current cpu capacity,
create the maximum cpu capacity head room by requesting the maximum
cpu capacity once the cpu usage plus the capacity margin exceeds the
current capacity. This is also done to try to harm the performance of
a task the least.
Original fair-class only version authored by Juri Lelli
<juri.lelli@arm.com>.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
As we don't trigger freq changes from {en,de}queue_task_fair() during load
balancing, we need to do explicitly so on load balancing paths.
[smuckle@linaro.org: move update_capacity_of calls so rq lock is held]
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
Patch "sched/fair: add triggers for OPP change requests" introduced OPP
change triggers for enqueue_task_fair(), but the trigger was operating only
for wakeups. Fact is that it makes sense to consider wakeup_new also (i.e.,
fork()), as we don't know anything about a newly created task and thus we
most certainly want to jump to max OPP to not harm performance too much.
However, it is not currently possible (or at least it wasn't evident to me
how to do so :/) to tell new wakeups from other (non wakeup) operations.
This patch introduces an additional flag in sched.h that is only set at
fork() time and it is then consumed in enqueue_task_fair() for our purpose.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
Each time a task is {en,de}queued we might need to adapt the current
frequency to the new usage. Add triggers on {en,de}queue_task_fair() for
this purpose. Only trigger a freq request if we are effectively waking up
or going to sleep. Filter out load balancing related calls to reduce the
number of triggers.
[smuckle@linaro.org: resolve merge conflicts, define task_new,
use renamed static key sched_freq]
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
Scheduler-driven CPU frequency selection hopes to exploit both
per-task and global information in the scheduler to improve frequency
selection policy, achieving lower power consumption, improved
responsiveness/performance, and less reliance on heuristics and
tunables. For further discussion on the motivation of this integration
see [0].
This patch implements a shim layer between the Linux scheduler and the
cpufreq subsystem. The interface accepts capacity requests from the
CFS, RT and deadline sched classes. The requests from each sched class
are summed on each CPU with a margin applied to the CFS and RT
capacity requests to provide some headroom. Deadline requests are
expected to be precise enough given their nature to not require
headroom. The maximum total capacity request for a CPU in a frequency
domain drives the requested frequency for that domain.
Policy is determined by both the sched classes and this shim layer.
Note that this algorithm is event-driven. There is no polling loop to
check cpu idle time nor any other method which is unsynchronized with
the scheduler, aside from a throttling mechanism to ensure frequency
changes are not attempted faster than the hardware can accommodate them.
Thanks to Juri Lelli <juri.lelli@arm.com> for contributing design ideas,
code and test results, and to Ricky Liang <jcliang@chromium.org>
for initialization and static key inc/dec fixes.
[0] http://article.gmane.org/gmane.linux.kernel/1499836
[smuckle@linaro.org: various additions and fixes, revised commit text]
CC: Ricky Liang <jcliang@chromium.org>
Signed-off-by: Michael Turquette <mturquette@baylibre.com>
Signed-off-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Steve Muckle <smuckle@linaro.org>
With the new group_misfit_task load-balancing scenario additional policy
conditions are needed when load-balancing. Misfit task balancing only
makes sense between source group with lower capacity than the target
group. If capacities are the same, fallback to normal group_other
balancing. The aim is to balance tasks such that no task has its
throughput hindered by compute capacity if a cpu with more capacity is
available. Load-balancing is generally based on average load in the
sched_groups, but for misfitting tasks it is necessary to introduce
exceptions to migrate tasks against usual metrics and optimize
throughput.
This patch ensures the following load-balance for mixed capacity systems
(e.g. ARM big.LITTLE) for always-running tasks:
1. Place a task on each cpu starting in order from cpus with highest
capacity to lowest until all cpus are in use (i.e. one task on each
cpu).
2. Once all cpus are in use balance according to compute capacity such
that load per capacity is approximately the same regardless of the
compute capacity (i.e. big cpus get more tasks than little cpus).
Necessary changes are introduced in find_busiest_group(),
calculate_imbalance(), and find_busiest_queue(). This includes passing
the group_type on to find_busiest_queue() through struct lb_env, which
is currently only considers imbalance and not the imbalance situation
(group_type).
To avoid taking remote rq locks to examine source sched_groups for
misfit tasks, each cpu is responsible for tracking misfit tasks
themselves and update the rq->misfit_task flag. This means checking task
utilization when tasks are scheduled and on sched_tick.
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
To maximize throughput in systems with reduced capacity cpus (e.g.
high RT/IRQ load and/or ARM big.LITTLE) load-balancing has to consider
task and cpu utilization as well as per-cpu compute capacity when
load-balancing in addition to the current average load based
load-balancing policy. Tasks that are scheduled on a reduced capacity
cpu need to be identified and migrated to a higher capacity cpu if
possible.
To implement this additional policy an additional group_type
(load-balance scenario) is added: group_misfit_task. This represents
scenarios where a sched_group has tasks that are not suitable for its
per-cpu capacity. group_misfit_task is only considered if the system is
not overloaded in any other way (group_imbalanced or group_overloaded).
Identifying misfit tasks requires the rq lock to be held. To avoid
taking remote rq locks to examine source sched_groups for misfit tasks,
each cpu is responsible for tracking misfit tasks themselves and update
the rq->misfit_task flag. This means checking task utilization when
tasks are scheduled and on sched_tick.
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
struct sched_group_capacity currently represents the compute capacity
sum of all cpus in the sched_group. Unless it is divided by the
group_weight to get the average capacity per cpu it hides differences in
cpu capacity for mixed capacity systems (e.g. high RT/IRQ utilization or
ARM big.LITTLE). But even the average may not be sufficient if the group
covers cpus of different capacities. Instead, by extending struct
sched_group_capacity to indicate max per-cpu capacity in the group a
suitable group for a given task utilization can easily be found such
that cpus with reduced capacity can be avoided for tasks with high
utilization (not implemented by this patch).
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
EAS relies on idle balance to migrate a misfit task towards a cpu with
higher capacity.
When such a cpu becomes idle, idle balance should happen even if the rq
avg idle is smaller than the sched migration cost (default 500us).
The rq avg idle is updated during the wakeup of a task in case the rq has
a non-null idle_stamp. This value stays unchanged and valid until the next
task wakes up on this cpu after an idle period.
So rq avg idle could be smaller than sched migration cost preventing the
idle balance from happening. In this case we would be at the mercy of
wakeup, periodic or nohz-idle load balancing to put another task on this
cpu.
To break this dependency towards rq avg idle make EAS idle balance
independent from this rq avg idle has to be larger than sched migration
cost.
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Wakeup balancing uses cpu capacity awareness and needs to know the
system-wide maximum cpu capacity.
Patch "sched: Store system-wide maximum cpu capacity in root domain"
finds the system-wide maximum cpu capacity during scheduler domain
hierarchy setup. This is sufficient as long as maximum frequency
invariance is not enabled.
If it is enabled, the system-wide maximum cpu capacity can change
between scheduler domain hierarchy setups due to frequency capping.
The cpu capacity is changed in update_cpu_capacity() which is called in
load balance on the lowest scheduler domain hierarchy level. To be able
to know if a change in cpu capacity for a certain cpu also has an effect
on the system-wide maximum cpu capacity it is normally necessary to
iterate over all cpus. This would be way too costly. That's why this
patch follows a different approach.
The unsigned long max_cpu_capacity value in struct root_domain is
replaced with a struct max_cpu_capacity, containing value (the
max_cpu_capacity) and cpu (the cpu index of the cpu providing the
maximum cpu_capacity).
Changes to the system-wide maximum cpu capacity and the cpu index are
made if:
1 System-wide maximum cpu capacity < cpu capacity
2 System-wide maximum cpu capacity > cpu capacity and cpu index == cpu
There are no changes to the system-wide maximum cpu capacity in all
other cases.
Atomic read and write access to the pair (max_cpu_capacity.val,
max_cpu_capacity.cpu) is enforced by max_cpu_capacity.lock.
The access to max_cpu_capacity.val in task_fits_max() is still performed
without taking the max_cpu_capacity.lock.
The code to set max cpu capacity in build_sched_domains() has been
removed because the whole functionality is now provided by
update_cpu_capacity() instead.
This approach can introduce errors temporarily, e.g. in case the cpu
currently providing the max cpu capacity has its cpu capacity lowered
due to frequency capping and calls update_cpu_capacity() before any cpu
which might provide the max cpu now.
There is also an outstanding question:
Should the cpu capacity of a cpu going idle be set to a very small
value?
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
This patch implements support for extracting energy cost data from DT.
The data should conform to the DT bindings for energy cost data needed
by EAS (energy aware scheduling).
Signed-off-by: Robin Randhawa <robin.randhawa@arm.com>
With energy-aware scheduling enabled nohz_kick_needed() generates many
nohz idle-balance kicks which lead to nothing when multiple tasks get
packed on a single cpu to save energy. This causes unnecessary wake-ups
and hence wastes energy. Make these conditions depend on !energy_aware()
for now until the energy-aware nohz story gets sorted out.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
In case the system operates below the tipping point indicator,
introduced in ("sched: Add over-utilization/tipping point
indicator"), bail out in find_busiest_group after the dst and src
group statistics have been checked.
There is simply no need to move usage around because all involved
cpus still have spare cycles available.
For an energy-aware system below its tipping point, we rely on the
task placement of the wakeup path. This works well for short running
tasks.
The existence of long running tasks on one of the involved cpus lets
the system operate over its tipping point. To be able to move such
a task (whose load can't be used to average the load among the cpus)
from a src cpu with lower capacity than the dst_cpu, an additional
rule has to be implemented in need_active_balance.
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Let available compute capacity and estimated energy impact select
wake-up target cpu when energy-aware scheduling is enabled and the
system in not over-utilized (above the tipping point).
energy_aware_wake_cpu() attempts to find group of cpus with sufficient
compute capacity to accommodate the task and find a cpu with enough spare
capacity to handle the task within that group. Preference is given to
cpus with enough spare capacity at the current OPP. Finally, the energy
impact of the new target and the previous task cpu is compared to select
the wake-up target cpu.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
To estimate the energy consumption of a sched_group in
sched_group_energy() it is necessary to know which idle-state the group
is in when it is idle. For now, it is assumed that this is the current
idle-state (though it might be wrong). Based on the individual cpu
idle-states group_idle_state() finds the group idle-state.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
The idle-state of each cpu is currently pointed to by rq->idle_state but
there isn't any information in the struct cpuidle_state that can used to
look up the idle-state energy model data stored in struct
sched_group_energy. For this purpose is necessary to store the idle
state index as well. Ideally, the idle-state data should be unified.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Energy-aware scheduling is only meant to be active while the system is
_not_ over-utilized. That is, there are spare cycles available to shift
tasks around based on their actual utilization to get a more
energy-efficient task distribution without depriving any tasks. When
above the tipping point task placement is done the traditional way based
on load_avg, spreading the tasks across as many cpus as possible based
on priority scaled load to preserve smp_nice. Below the tipping point we
want to use util_avg instead. We need to define a criteria for when we
make the switch.
The util_avg for each cpu converges towards 100% (1024) regardless of
how many task additional task we may put on it. If we define
over-utilized as:
sum_{cpus}(rq.cfs.avg.util_avg) + margin > sum_{cpus}(rq.capacity)
some individual cpus may be over-utilized running multiple tasks even
when the above condition is false. That should be okay as long as we try
to spread the tasks out to avoid per-cpu over-utilization as much as
possible and if all tasks have the _same_ priority. If the latter isn't
true, we have to consider priority to preserve smp_nice.
For example, we could have n_cpus nice=-10 util_avg=55% tasks and
n_cpus/2 nice=0 util_avg=60% tasks. Balancing based on util_avg we are
likely to end up with nice=-10 tasks sharing cpus and nice=0 tasks
getting their own as we 1.5*n_cpus tasks in total and 55%+55% is less
over-utilized than 55%+60% for those cpus that have to be shared. The
system utilization is only 85% of the system capacity, but we are
breaking smp_nice.
To be sure not to break smp_nice, we have defined over-utilization
conservatively as when any cpu in the system is fully utilized at it's
highest frequency instead:
cpu_rq(any).cfs.avg.util_avg + margin > cpu_rq(any).capacity
IOW, as soon as one cpu is (nearly) 100% utilized, we switch to load_avg
to factor in priority to preserve smp_nice.
With this definition, we can skip periodic load-balance as no cpu has an
always-running task when the system is not over-utilized. All tasks will
be periodic and we can balance them at wake-up. This conservative
condition does however mean that some scenarios that could benefit from
energy-aware decisions even if one cpu is fully utilized would not get
those benefits.
For system where some cpus might have reduced capacity on some cpus
(RT-pressure and/or big.LITTLE), we want periodic load-balance checks as
soon a just a single cpu is fully utilized as it might one of those with
reduced capacity and in that case we want to migrate it.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Adds a generic energy-aware helper function, energy_diff(), that
calculates energy impact of adding, removing, and migrating utilization
in the system.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Extended sched_group_energy() to support energy prediction with usage
(tasks) added/removed from a specific cpu or migrated between a pair of
cpus. Useful for load-balancing decision making.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
For energy-aware load-balancing decisions it is necessary to know the
energy consumption estimates of groups of cpus. This patch introduces a
basic function, sched_group_energy(), which estimates the energy
consumption of the cpus in the group and any resources shared by the
members of the group.
NOTE: The function has five levels of identation and breaks the 80
character limit. Refactoring is necessary.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Add another member to the family of per-cpu sched_domain shortcut
pointers. This one, sd_ea, points to the highest level at which energy
model is provided. At this level and all levels below all sched_groups
have energy model data attached.
Partial energy model information is possible but restricted to providing
energy model data for lower level sched_domains (sd_ea and below) and
leaving load-balancing on levels above to non-energy-aware
load-balancing. For example, it is possible to apply energy-aware
scheduling within each socket on a multi-socket system and let normal
scheduling handle load-balancing between sockets.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Move cpu_util() to an earlier position in fair.c and change return
type to unsigned long as negative usage doesn't make much sense. All
other load and capacity related functions use unsigned long including
the caller of cpu_util().
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
capacity_orig_of() returns the max available compute capacity of a cpu.
For scale-invariant utilization tracking and energy-aware scheduling
decisions it is useful to know the compute capacity available at the
current OPP of a cpu.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
cpufreq is currently keeping it a secret which cpus are sharing
clock source. The scheduler needs to know about clock domains as well
to become more energy aware. The SD_SHARE_CAP_STATES domain flag
indicates whether cpus belonging to the sched_domain share capacity
states (P-states).
There is no connection with cpufreq (yet). The flag must be set by
the arch specific topology code.
cc: Russell King <linux@arm.linux.org.uk>
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
The sched_group_energy (sge) pointer of the first sched_group (sg) in
the sched_domain (sd) is initialized to point to the appropriate (in
terms of sd level and cpu) sge data defined in the arch and so to the
correct part of the Energy Model (EM).
Energy-aware scheduling allows that a system has only EM data up to a
certain sd level (so called highest energy aware balancing sd level).
A check in init_sched_energy() enforces that all sd's below this sd
level contain EM data.
The 'int cpu' parameter of sched_domain_energy_f requires that
check_sched_energy_data() makes sure that all cpus spanned by a sg
are provisioned with the same EM data.
This patch has also been tested with feature FORCE_SD_OVERLAP enabled.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
The struct sched_group_energy represents the per sched_group related
data which is needed for energy aware scheduling. It contains:
(1) number of elements of the idle state array
(2) pointer to the idle state array which comprises 'power consumption'
for each idle state
(3) number of elements of the capacity state array
(4) pointer to the capacity state array which comprises 'compute
capacity and power consumption' tuples for each capacity state
The struct sched_group obtains a pointer to a struct sched_group_energy.
The function pointer sched_domain_energy_f is introduced into struct
sched_domain_topology_level which will allow the arch to pass a particular
struct sched_group_energy from the topology shim layer into the scheduler
core.
The function pointer sched_domain_energy_f has an 'int cpu' parameter
since the folding of two adjacent sd levels via sd degenerate doesn't work
for all sd levels. I.e. it is not possible for example to use this feature
to provide per-cpu energy in sd level DIE on ARM's TC2 platform.
It was discussed that the folding of sd levels approach is preferable
over the cpu parameter approach, simply because the user (the arch
specifying the sd topology table) can introduce less errors. But since
it is not working, the 'int cpu' parameter is the only way out. It's
possible to use the folding of sd levels approach for
sched_domain_flags_f and the cpu parameter approach for the
sched_domain_energy_f at the same time though. With the use of the
'int cpu' parameter, an extra check function has to be provided to make
sure that all cpus spanned by a sched group are provisioned with the same
energy data.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
This patch introduces the ENERGY_AWARE sched feature, which is
implemented using jump labels when SCHED_DEBUG is defined. It is
statically set false when SCHED_DEBUG is not defined. Hence this doesn't
allow energy awareness to be enabled without SCHED_DEBUG. This
sched_feature knob will be replaced later with a more appropriate
control knob when things have matured a bit.
ENERGY_AWARE is based on per-entity load-tracking hence FAIR_GROUP_SCHED
must be enable. This dependency isn't checked at compile time yet.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Scenarios with the busiest group having just one task and the local
being idle on topologies with sched groups with different numbers of
cpus manage to dodge all load-balance bailout conditions resulting the
nr_balance_failed counter to be incremented. This eventually causes a
pointless active migration of the task. This patch prevents this by not
incrementing the counter when the busiest group only has one task.
ASYM_PACKING migrations and migrations due to reduced capacity should
still take place as these are explicitly captured by
need_active_balance().
A better solution would be to not attempt the load-balance in the first
place, but that requires significant changes to the order of bailout
conditions and statistics gathering.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
We do not want to miss out on the ability to pull a single remaining
task from a potential source cpu towards an idle destination cpu. Add an
extra criteria to need_active_balance() to kick off active load balance
if the source cpu is over-utilized and has lower capacity than the
destination cpu.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
find_idlest_group() selects the wake-up target group purely
based on group load which leads to suboptimal choices in low load
scenarios. An idle group with reduced capacity (due to RT tasks or
different cpu type) isn't necessarily a better target than a lightly
loaded group with higher capacity.
The patch adds spare capacity as an additional group selection
parameter. The target group is now selected based on the following
criteria:
1. Return the group with the cpu with most spare capacity and this
capacity is significant if such group exists. Significant spare capacity
is currently at least 20% to spare.
2. Return the group with the lowest load, unless it is the local group
in which case NULL is returned and the search is continued at the next
(lower) level.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Wakeup balancing is completely unaware of cpu capacity, cpu utilization
and task utilization. The task is preferably placed on a cpu which is
idle in the instant the wakeup happens. New tasks
(SD_BALANCE_{FORK,EXEC} are placed on an idle cpu in the idlest group if
such can be found, otherwise it goes on the least loaded one. Existing
tasks (SD_BALANCE_WAKE) are placed on the previous cpu or an idle cpu
sharing the same last level cache unless the wakee_flips heuristic in
wake_wide() decides to fallback to considering cpus outside SD_LLC.
Hence existing tasks are not guaranteed to get a chance to migrate to a
different group at wakeup in case the current one has reduced cpu
capacity (due RT/IRQ pressure or different uarch e.g. ARM big.LITTLE).
They may eventually get pulled by other cpus doing
periodic/idle/nohz_idle balance, but it may take quite a while before it
happens.
This patch adds capacity awareness to find_idlest_{group,queue} (used by
SD_BALANCE_{FORK,EXEC} and SD_BALANCE_WAKE under certain circumstances)
such that groups/cpus that can accommodate the waking task based on task
utilization are preferred. In addition, wakeup of existing tasks
(SD_BALANCE_WAKE) is sent through find_idlest_{group,queue} also if the
task doesn't fit the capacity of the previous cpu to allow it to escape
(override wake_affine) when necessary instead of relying on
periodic/idle/nohz_idle balance to eventually sort it out.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
To be able to compare the capacity of the target cpu with the highest
cpu capacity of the system in the wakeup path, store the system-wide
maximum cpu capacity in the root domain.
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
If a newly created task is selected to go to a different CPU in fork
balance when it wakes up the first time, its load averages should
not be removed from the source CPU since they are never added to
it before. The same is also applicable to a never used group entity.
Fix it in remove_entity_load_avg(): when entity's last_update_time
is 0, simply return. This should precisely identify the case in
question, because in other migrations, the last_update_time is set
to 0 after remove_entity_load_avg().
Reported-by: Steve Muckle <steve.muckle@linaro.org>
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
[peterz: cfs_rq_last_update_time]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <Juri.Lelli@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Patrick Bellasi <patrick.bellasi@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: http://lkml.kernel.org/r/20151216233427.GJ28098@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This deliberately changes the behavior of the per-cpuset
cpus file to not be effected by hotplug. When a cpu is offlined,
it will be removed from the cpuset/cpus file. When a cpu is onlined,
if the cpuset originally requested that that cpu was part of the cpuset,
that cpu will be restored to the cpuset. The cpus files still
have to be hierachical, but the ranges no longer have to be out of
the currently online cpus, just the physically present cpus.
Change-Id: I22cdf33e7d312117bcefba1aeb0125e1ada289a9
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>
x86_64:allmodconfig fails to build with the following error.
ERROR: "rcu_sync_lockdep_assert" [kernel/locking/locktorture.ko] undefined!
Introduced by commit 3228c5eb7a ("RFC: FROMLIST: locking/percpu-rwsem:
Optimize readers and reduce global impact"). The applied upstream version
exports the missing symbol, so let's do the same.
Change-Id: If4e516715c3415fe8c82090f287174857561550d
Fixes: 3228c5eb7a ("RFC: FROMLIST: locking/percpu-rwsem: Optimize ...")
Signed-off-by: Guenter Roeck <groeck@chromium.org>
cgroup_threadgroup_rwsem is acquired in read mode during process exit
and fork. It is also grabbed in write mode during
__cgroups_proc_write(). I've recently run into a scenario with lots
of memory pressure and OOM and I am beginning to see
systemd
__switch_to+0x1f8/0x350
__schedule+0x30c/0x990
schedule+0x48/0xc0
percpu_down_write+0x114/0x170
__cgroup_procs_write.isra.12+0xb8/0x3c0
cgroup_file_write+0x74/0x1a0
kernfs_fop_write+0x188/0x200
__vfs_write+0x6c/0xe0
vfs_write+0xc0/0x230
SyS_write+0x6c/0x110
system_call+0x38/0xb4
This thread is waiting on the reader of cgroup_threadgroup_rwsem to
exit. The reader itself is under memory pressure and has gone into
reclaim after fork. There are times the reader also ends up waiting on
oom_lock as well.
__switch_to+0x1f8/0x350
__schedule+0x30c/0x990
schedule+0x48/0xc0
jbd2_log_wait_commit+0xd4/0x180
ext4_evict_inode+0x88/0x5c0
evict+0xf8/0x2a0
dispose_list+0x50/0x80
prune_icache_sb+0x6c/0x90
super_cache_scan+0x190/0x210
shrink_slab.part.15+0x22c/0x4c0
shrink_zone+0x288/0x3c0
do_try_to_free_pages+0x1dc/0x590
try_to_free_pages+0xdc/0x260
__alloc_pages_nodemask+0x72c/0xc90
alloc_pages_current+0xb4/0x1a0
page_table_alloc+0xc0/0x170
__pte_alloc+0x58/0x1f0
copy_page_range+0x4ec/0x950
copy_process.isra.5+0x15a0/0x1870
_do_fork+0xa8/0x4b0
ppc_clone+0x8/0xc
In the meanwhile, all processes exiting/forking are blocked almost
stalling the system.
This patch moves the threadgroup_change_begin from before
cgroup_fork() to just before cgroup_canfork(). There is no nee to
worry about threadgroup changes till the task is actually added to the
threadgroup. This avoids having to call reclaim with
cgroup_threadgroup_rwsem held.
tj: Subject and description edits.
Signed-off-by: Balbir Singh <bsingharora@gmail.com>
Acked-by: Zefan Li <lizefan@huawei.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: stable@vger.kernel.org # v4.2+
Signed-off-by: Tejun Heo <tj@kernel.org>
[jstultz: Cherry-picked from:
git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup.git 568ac888215c7f]
Change-Id: Ie8ece84fb613cf6a7b08cea1468473a8df2b9661
Signed-off-by: John Stultz <john.stultz@linaro.org>
The current percpu-rwsem read side is entirely free of serializing insns
at the cost of having a synchronize_sched() in the write path.
The latency of the synchronize_sched() is too high for cgroups. The
commit 1ed1328792 talks about the write path being a fairly cold path
but this is not the case for Android which moves task to the foreground
cgroup and back around binder IPC calls from foreground processes to
background processes, so it is significantly hotter than human initiated
operations.
Switch cgroup_threadgroup_rwsem into the slow mode for now to avoid the
problem, hopefully it should not be that slow after another commit
80127a39681b ("locking/percpu-rwsem: Optimize readers and reduce global
impact").
We could just add rcu_sync_enter() into cgroup_init() but we do not want
another synchronize_sched() at boot time, so this patch adds the new helper
which doesn't block but currently can only be called before the first use.
Cc: Tejun Heo <tj@kernel.org>
Cc: Paul McKenney <paulmck@linux.vnet.ibm.com>
Reported-by: John Stultz <john.stultz@linaro.org>
Reported-by: Dmitry Shmidt <dimitrysh@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
[jstultz: backported to 4.4]
Change-Id: I34aa9c394d3052779b56976693e96d861bd255f2
Mailing-list-URL: https://lkml.org/lkml/2016/8/11/557
Signed-off-by: John Stultz <john.stultz@linaro.org>
Currently the percpu-rwsem switches to (global) atomic ops while a
writer is waiting; which could be quite a while and slows down
releasing the readers.
This patch cures this problem by ordering the reader-state vs
reader-count (see the comments in __percpu_down_read() and
percpu_down_write()). This changes a global atomic op into a full
memory barrier, which doesn't have the global cacheline contention.
This also enables using the percpu-rwsem with rcu_sync disabled in order
to bias the implementation differently, reducing the writer latency by
adding some cost to readers.
Mailing-list-URL: https://lkml.org/lkml/2016/8/9/181
Cc: Paul McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[jstultz: Backported to 4.4]
Change-Id: I8ea04b4dca2ec36f1c2469eccafde1423490572f
Signed-off-by: John Stultz <john.stultz@linaro.org>
I'm looking at trying to possibly merge the 32-bit and 64-bit versions
of the x86 uaccess.h implementation, but first this needs to be cleaned
up.
For example, the 32-bit version of "__copy_from_user_inatomic()" is
mostly the special cases for the constant size, and it's actually almost
never relevant. Most users aren't actually using a constant size
anyway, and the few cases that do small constant copies are better off
just using __get_user() instead.
So get rid of the unnecessary complexity.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit bd28b14591b98f696bc9f94c5ba2e598ca487dfd)
Signed-off-by: Alex Shi <alex.shi@linaro.org>
