Add to find_energy_efficient_cpu() a latency sensitive case which mimics
what was done for prefer-idle in android-4.19 and before (see [1] for
reference).
This isn't strictly equivalent to the legacy algorithm but comes real
close, and isn't very invasive. Overall, the idea is to select the
biggest idle CPU we can find for latency-sensitive boosted tasks, and
the smallest CPU where the can fit for latency-sensitive non-boosted
tasks.
The main differences with the legacy behaviour are the following:
1. the policy for 'prefer idle' when there isn't a single idle CPU in
the system is simpler now. We just pick the CPU with the highest
spare capacity;
2. the cstate awareness is implemented by minimizing the exit latency
rather than the idle state index. This is how it is done in the slow
path (find_idlest_group_cpu()), it doesn't require us to keep hooks
into CPUIdle, and should actually be better because what we want is
a CPU that can wake up quickly;
3. non-latency-sensitive tasks just use the standard mainline
energy-aware wake-up path, which decides the placement using the
Energy Model;
4. the 'boosted' and 'latency_sensitive' attributes of a task come from
util_clamp (which now replaces schedtune).
[1] c27c56105d
Change-Id: Ia58516906e9cb5abe08385a8cd088097043d8703
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
If we can classify the group as overloaded, that overrides
any classification as misfit but we may still have misfit
tasks present. Check the rq we're looking at to see if
this is the case.
Change-Id: Ida8eb66aa625e34de3fe2ee1b0dd8a78926273d8
Signed-off-by: Chris Redpath <chris.redpath@arm.com>
[Removed stray reference to rq_has_misfit]
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
When load balancing in a system with misfit tasks present, if we always
pull a misfit task to the local group this can lead to pulling a running
task from a smaller capacity CPUs to a bigger CPU which is busy. In this
situation, the pulled task is likely not to get a chance to run before
an idle balance on another small CPU pulls it back. This penalises the
pulled task as it is stopped for a short amount of time and then likely
relocated to a different CPU (since the original CPU just did a NEWLY_IDLE
balance and reset the periodic interval).
If we only do this unconditionally for NEWLY_IDLE balance, we can be
sure that any tasks and load which are present on the local group are
related to short-running tasks which we are happy to displace for a
longer running task in a system with misfit tasks present.
However, other balance types should only pull a task if we think
that the local group is underutilized - checking the number of tasks
gives us a conservative estimate here since if they were short tasks
we would have been doing NEWLY_IDLE balances instead.
Change-Id: I710add1ab1139482620b6addc8370ad194791beb
Signed-off-by: Chris Redpath <chris.redpath@arm.com>
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
In some systems the capacity and group weights line up to defeat all the
small imbalance correction conditions in fix_small_imbalance, which can
cause bad task placement. Add a new condition if the existing code can't
see anything to fix:
If we have asymmetric capacity, and there are more tasks than CPUs in
the busiest group *and* there are less tasks than CPUs in the local group
then we try to pull something. There could be transient small tasks which
prevent this from working, but on the whole it is beneficial for those
systems with inconvenient capacity/cluster size relationships.
Change-Id: Icf81cde215c082a61f816534b7990ccb70aee409
Signed-off-by: Chris Redpath <chris.redpath@arm.com>
Signed-off-by: Quentin Perret <quentin.perret@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.
Change-Id: I28f69c72febe0211decbe77b7bc3e48839d3d7b3
cc: Ingo Molnar <mingo@redhat.com>
cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Signed-off-by: Quentin Perret <quentin.perret@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.
Change-Id: I5063befab088fbf49e5d5e484ce0c6ee6165283a
Signed-off-by: Ionela Voinescu <ionela.voinescu@arm.com>*
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
(- Fixed cherry-pick issues, and conflict with a0fe2cf086 "sched/fair:
Tune down misfit NOHZ kicks" which makes use of max_cpu_capacity
- Squashed "sched/fair: remove printk while schedule is in progress"
fix from Caesar Wang <wxt@rock-chips.com>)
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
To be able to scale the cpu capacity by this factor introduce a call to
the new arch scaling function arch_scale_max_freq_capacity() in
update_cpu_capacity() and provide a default implementation which returns
SCHED_CAPACITY_SCALE.
Another subsystem (e.g. cpufreq) or architectural or platform specific
code can overwrite this default implementation, exactly as for frequency
and cpu invariance. It has to be enabled by the arch by defining
arch_scale_max_freq_capacity to the actual implementation.
Change-Id: I770a8b1f4f7340e9e314f71c64a765bf880f4b4d
Signed-off-by: Ionela Voinescu <ionela.voinescu@arm.com>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
( Fixed conflict with scaling against the PELT-based scale_rt_capacity )
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
Since we don't do energy-aware wakeups when we are overutilized, always
honoring sync wakeups in this state does not prevent wake-wide mechanics
overruling the flag as normal.
This patch is based upon previous work to build EAS for android products.
sync-hint code taken from commit 4a5e890ec60d
"sched/fair: add tunable to force selection at cpu granularity" written
by Juri Lelli <juri.lelli@arm.com>
Change-Id: I4b3d79141fc8e53dc51cd63ac11096c2e3cb10f5
Signed-off-by: Chris Redpath <chris.redpath@arm.com>
(cherry-picked from commit f1ec666a62dec1083ed52fe1ddef093b84373aaf)
[ Moved the feature to find_energy_efficient_cpu() and removed the
sysctl knob ]
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
Utilization clamping can be used to boost the utilization of small tasks
or cap that of big tasks. Thus, one of its possible usages is to bias
tasks placement to "promote" small tasks on higher capacity (less energy
efficient) CPUs or "constraint" big tasks on small capacity (more energy
efficient) CPUs.
When the Energy Aware Scheduler (EAS) looks for the most energy
efficient CPU to run a task on, it currently considers only the
effective utiliation estimated for a task.
Fix this by adding an additional check to skip CPUs which capacity is
smaller then the task clamped utilization.
Change-Id: I43fa6fa27e27c1eb5272c6a45d1a6a5b0faae1aa
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
Commit:
de53fd7aed ("sched/fair: Fix low cpu usage with high throttling by removing expiration of cpu-local slices")
introduced a few compilation warnings:
kernel/sched/fair.c: In function '__refill_cfs_bandwidth_runtime':
kernel/sched/fair.c:4365:6: warning: variable 'now' set but not used [-Wunused-but-set-variable]
kernel/sched/fair.c: In function 'start_cfs_bandwidth':
kernel/sched/fair.c:4992:6: warning: variable 'overrun' set but not used [-Wunused-but-set-variable]
Also, __refill_cfs_bandwidth_runtime() does no longer update the
expiration time, so fix the comments accordingly.
Signed-off-by: Qian Cai <cai@lca.pw>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ben Segall <bsegall@google.com>
Reviewed-by: Dave Chiluk <chiluk+linux@indeed.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: pauld@redhat.com
Fixes: de53fd7aed ("sched/fair: Fix low cpu usage with high throttling by removing expiration of cpu-local slices")
Link: https://lkml.kernel.org/r/1566326455-8038-1-git-send-email-cai@lca.pw
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Remove work arounds that were written before there was a grace period
after tasks left the runqueue in finish_task_switch().
In particular now that there tasks exiting the runqueue exprience
a RCU grace period none of the work performed by task_rcu_dereference()
excpet the rcu_dereference() is necessary so replace task_rcu_dereference()
with rcu_dereference().
Remove the code in rcuwait_wait_event() that checks to ensure the current
task has not exited. It is no longer necessary as it is guaranteed
that any running task will experience a RCU grace period after it
leaves the run queueue.
Remove the comment in rcuwait_wake_up() as it is no longer relevant.
Ref: 8f95c90ceb ("sched/wait, RCU: Introduce rcuwait machinery")
Ref: 150593bf86 ("sched/api: Introduce task_rcu_dereference() and try_get_task_struct()")
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Chris Metcalf <cmetcalf@ezchip.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Kirill Tkhai <tkhai@yandex.ru>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Paul E. McKenney <paulmck@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russell King - ARM Linux admin <linux@armlinux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/87lfurdpk9.fsf_-_@x220.int.ebiederm.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull scheduler updates from Ingo Molnar:
- MAINTAINERS: Add Mark Rutland as perf submaintainer, Juri Lelli and
Vincent Guittot as scheduler submaintainers. Add Dietmar Eggemann,
Steven Rostedt, Ben Segall and Mel Gorman as scheduler reviewers.
As perf and the scheduler is getting bigger and more complex,
document the status quo of current responsibilities and interests,
and spread the review pain^H^H^H^H fun via an increase in the Cc:
linecount generated by scripts/get_maintainer.pl. :-)
- Add another series of patches that brings the -rt (PREEMPT_RT) tree
closer to mainline: split the monolithic CONFIG_PREEMPT dependencies
into a new CONFIG_PREEMPTION category that will allow the eventual
introduction of CONFIG_PREEMPT_RT. Still a few more hundred patches
to go though.
- Extend the CPU cgroup controller with uclamp.min and uclamp.max to
allow the finer shaping of CPU bandwidth usage.
- Micro-optimize energy-aware wake-ups from O(CPUS^2) to O(CPUS).
- Improve the behavior of high CPU count, high thread count
applications running under cpu.cfs_quota_us constraints.
- Improve balancing with SCHED_IDLE (SCHED_BATCH) tasks present.
- Improve CPU isolation housekeeping CPU allocation NUMA locality.
- Fix deadline scheduler bandwidth calculations and logic when cpusets
rebuilds the topology, or when it gets deadline-throttled while it's
being offlined.
- Convert the cpuset_mutex to percpu_rwsem, to allow it to be used from
setscheduler() system calls without creating global serialization.
Add new synchronization between cpuset topology-changing events and
the deadline acceptance tests in setscheduler(), which were broken
before.
- Rework the active_mm state machine to be less confusing and more
optimal.
- Rework (simplify) the pick_next_task() slowpath.
- Improve load-balancing on AMD EPYC systems.
- ... and misc cleanups, smaller fixes and improvements - please see
the Git log for more details.
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (53 commits)
sched/psi: Correct overly pessimistic size calculation
sched/fair: Speed-up energy-aware wake-ups
sched/uclamp: Always use 'enum uclamp_id' for clamp_id values
sched/uclamp: Update CPU's refcount on TG's clamp changes
sched/uclamp: Use TG's clamps to restrict TASK's clamps
sched/uclamp: Propagate system defaults to the root group
sched/uclamp: Propagate parent clamps
sched/uclamp: Extend CPU's cgroup controller
sched/topology: Improve load balancing on AMD EPYC systems
arch, ia64: Make NUMA select SMP
sched, perf: MAINTAINERS update, add submaintainers and reviewers
sched/fair: Use rq_lock/unlock in online_fair_sched_group
cpufreq: schedutil: fix equation in comment
sched: Rework pick_next_task() slow-path
sched: Allow put_prev_task() to drop rq->lock
sched/fair: Expose newidle_balance()
sched: Add task_struct pointer to sched_class::set_curr_task
sched: Rework CPU hotplug task selection
sched/{rt,deadline}: Fix set_next_task vs pick_next_task
sched: Fix kerneldoc comment for ia64_set_curr_task
...
EAS computes the energy impact of migrating a waking task when deciding
on which CPU it should run. However, the current approach is known to
have a high algorithmic complexity, which can result in prohibitively
high wake-up latencies on systems with complex energy models, such as
systems with per-CPU DVFS. On such systems, the algorithm complexity is
in O(n^2) (ignoring the cost of searching for performance states in the
EM) with 'n' the number of CPUs.
To address this, re-factor the EAS wake-up path to compute the energy
'delta' (with and without the task) on a per-performance domain basis,
rather than system-wide, which brings the complexity down to O(n).
No functional changes intended.
Test results
~~~~~~~~~~~~
* Setup: Tested on a Google Pixel 3, with a Snapdragon 845 (4+4 CPUs,
A55/A75). Base kernel is 5.3-rc5 + Pixel3 specific patches. Android
userspace, no graphics.
* Test case: Run a periodic rt-app task, with 16ms period, ramping down
from 70% to 10%, in 5% steps of 500 ms each (json avail. at [1]).
Frequencies of all CPUs are pinned to max (using scaling_min_freq
CPUFreq sysfs entries) to reduce variability. The time to run
select_task_rq_fair() is measured using the function profiler
(/sys/kernel/debug/tracing/trace_stat/function*). See the test script
for more details [2].
Test 1:
I hacked the DT to 'fake' per-CPU DVFS. That is, we end up with one
CPUFreq policy per CPU (8 policies in total). Since all frequencies are
pinned to max for the test, this should have no impact on the actual
frequency selection, but it does in the EAS calculation.
+---------------------------+----------------------------------+
| Without patch | With patch |
+-----+-----+----------+----------+-----+-----------------+----------+
| CPU | Hit | Avg (us) | s^2 (us) | Hit | Avg (us) | s^2 (us) |
|-----+-----+----------+----------+-----+-----------------+----------+
| 0 | 274 | 38.303 | 1750.239 | 401 | 14.126 (-63.1%) | 146.625 |
| 1 | 197 | 49.529 | 1695.852 | 314 | 16.135 (-67.4%) | 167.525 |
| 2 | 142 | 34.296 | 1758.665 | 302 | 14.133 (-58.8%) | 130.071 |
| 3 | 172 | 31.734 | 1490.975 | 641 | 14.637 (-53.9%) | 139.189 |
| 4 | 316 | 7.834 | 178.217 | 425 | 5.413 (-30.9%) | 20.803 |
| 5 | 447 | 8.424 | 144.638 | 556 | 5.929 (-29.6%) | 27.301 |
| 6 | 581 | 14.886 | 346.793 | 456 | 5.711 (-61.6%) | 23.124 |
| 7 | 456 | 10.005 | 211.187 | 997 | 4.708 (-52.9%) | 21.144 |
+-----+-----+----------+----------+-----+-----------------+----------+
* Hit, Avg and s^2 are as reported by the function profiler
Test 2:
I also ran the same test with a normal DT, with 2 CPUFreq policies, to
see if this causes regressions in the most common case.
+---------------------------+----------------------------------+
| Without patch | With patch |
+-----+-----+----------+----------+-----+-----------------+----------+
| CPU | Hit | Avg (us) | s^2 (us) | Hit | Avg (us) | s^2 (us) |
|-----+-----+----------+----------+-----+-----------------+----------+
| 0 | 345 | 22.184 | 215.321 | 580 | 18.635 (-16.0%) | 146.892 |
| 1 | 358 | 18.597 | 200.596 | 438 | 12.934 (-30.5%) | 104.604 |
| 2 | 359 | 25.566 | 200.217 | 397 | 10.826 (-57.7%) | 74.021 |
| 3 | 362 | 16.881 | 200.291 | 718 | 11.455 (-32.1%) | 102.280 |
| 4 | 457 | 3.822 | 9.895 | 757 | 4.616 (+20.8%) | 13.369 |
| 5 | 344 | 4.301 | 7.121 | 594 | 5.320 (+23.7%) | 18.798 |
| 6 | 472 | 4.326 | 7.849 | 464 | 5.648 (+30.6%) | 22.022 |
| 7 | 331 | 4.630 | 13.937 | 408 | 5.299 (+14.4%) | 18.273 |
+-----+-----+----------+----------+-----+-----------------+----------+
* Hit, Avg and s^2 are as reported by the function profiler
In addition to these two tests, I also ran 50 iterations of the Lisa
EAS functional test suite [3] with this patch applied on Arm Juno r0,
Arm Juno r2, Arm TC2 and Hikey960, and could not see any regressions
(all EAS functional tests are passing).
[1] https://paste.debian.net/1100055/
[2] https://paste.debian.net/1100057/
[3] https://github.com/ARM-software/lisa/blob/master/lisa/tests/scheduler/eas_behaviour.py
Signed-off-by: Quentin Perret <quentin.perret@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: dietmar.eggemann@arm.com
Cc: juri.lelli@redhat.com
Cc: morten.rasmussen@arm.com
Cc: qais.yousef@arm.com
Cc: qperret@qperret.net
Cc: rjw@rjwysocki.net
Cc: tkjos@google.com
Cc: valentin.schneider@arm.com
Cc: vincent.guittot@linaro.org
Link: https://lkml.kernel.org/r/20190912094404.13802-1-qperret@qperret.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
do_sched_cfs_period_timer() will refill cfs_b runtime and call
distribute_cfs_runtime to unthrottle cfs_rq, sometimes cfs_b->runtime
will allocate all quota to one cfs_rq incorrectly, then other cfs_rqs
attached to this cfs_b can't get runtime and will be throttled.
We find that one throttled cfs_rq has non-negative
cfs_rq->runtime_remaining and cause an unexpetced cast from s64 to u64
in snippet:
distribute_cfs_runtime() {
runtime = -cfs_rq->runtime_remaining + 1;
}
The runtime here will change to a large number and consume all
cfs_b->runtime in this cfs_b period.
According to Ben Segall, the throttled cfs_rq can have
account_cfs_rq_runtime called on it because it is throttled before
idle_balance, and the idle_balance calls update_rq_clock to add time
that is accounted to the task.
This commit prevents cfs_rq to be assgined new runtime if it has been
throttled until that distribute_cfs_runtime is called.
Signed-off-by: Liangyan <liangyan.peng@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: shanpeic@linux.alibaba.com
Cc: stable@vger.kernel.org
Cc: xlpang@linux.alibaba.com
Fixes: d3d9dc3302 ("sched: Throttle entities exceeding their allowed bandwidth")
Link: https://lkml.kernel.org/r/20190826121633.6538-1-liangyan.peng@linux.alibaba.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Enabling WARN_DOUBLE_CLOCK in /sys/kernel/debug/sched_features causes
warning to fire in update_rq_clock. This seems to be caused by onlining
a new fair sched group not using the rq lock wrappers.
[] rq->clock_update_flags & RQCF_UPDATED
[] WARNING: CPU: 5 PID: 54385 at kernel/sched/core.c:210 update_rq_clock+0xec/0x150
[] Call Trace:
[] online_fair_sched_group+0x53/0x100
[] cpu_cgroup_css_online+0x16/0x20
[] online_css+0x1c/0x60
[] cgroup_apply_control_enable+0x231/0x3b0
[] cgroup_mkdir+0x41b/0x530
[] kernfs_iop_mkdir+0x61/0xa0
[] vfs_mkdir+0x108/0x1a0
[] do_mkdirat+0x77/0xe0
[] do_syscall_64+0x55/0x1d0
[] entry_SYSCALL_64_after_hwframe+0x44/0xa9
Using the wrappers in online_fair_sched_group instead of the raw locking
removes this warning.
[ tglx: Use rq_*lock_irq() ]
Signed-off-by: Phil Auld <pauld@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20190801133749.11033-1-pauld@redhat.com
It has been observed, that highly-threaded, non-cpu-bound applications
running under cpu.cfs_quota_us constraints can hit a high percentage of
periods throttled while simultaneously not consuming the allocated
amount of quota. This use case is typical of user-interactive non-cpu
bound applications, such as those running in kubernetes or mesos when
run on multiple cpu cores.
This has been root caused to cpu-local run queue being allocated per cpu
bandwidth slices, and then not fully using that slice within the period.
At which point the slice and quota expires. This expiration of unused
slice results in applications not being able to utilize the quota for
which they are allocated.
The non-expiration of per-cpu slices was recently fixed by
'commit 512ac999d2 ("sched/fair: Fix bandwidth timer clock drift
condition")'. Prior to that it appears that this had been broken since
at least 'commit 51f2176d74 ("sched/fair: Fix unlocked reads of some
cfs_b->quota/period")' which was introduced in v3.16-rc1 in 2014. That
added the following conditional which resulted in slices never being
expired.
if (cfs_rq->runtime_expires != cfs_b->runtime_expires) {
/* extend local deadline, drift is bounded above by 2 ticks */
cfs_rq->runtime_expires += TICK_NSEC;
Because this was broken for nearly 5 years, and has recently been fixed
and is now being noticed by many users running kubernetes
(https://github.com/kubernetes/kubernetes/issues/67577) it is my opinion
that the mechanisms around expiring runtime should be removed
altogether.
This allows quota already allocated to per-cpu run-queues to live longer
than the period boundary. This allows threads on runqueues that do not
use much CPU to continue to use their remaining slice over a longer
period of time than cpu.cfs_period_us. However, this helps prevent the
above condition of hitting throttling while also not fully utilizing
your cpu quota.
This theoretically allows a machine to use slightly more than its
allotted quota in some periods. This overflow would be bounded by the
remaining quota left on each per-cpu runqueueu. This is typically no
more than min_cfs_rq_runtime=1ms per cpu. For CPU bound tasks this will
change nothing, as they should theoretically fully utilize all of their
quota in each period. For user-interactive tasks as described above this
provides a much better user/application experience as their cpu
utilization will more closely match the amount they requested when they
hit throttling. This means that cpu limits no longer strictly apply per
period for non-cpu bound applications, but that they are still accurate
over longer timeframes.
This greatly improves performance of high-thread-count, non-cpu bound
applications with low cfs_quota_us allocation on high-core-count
machines. In the case of an artificial testcase (10ms/100ms of quota on
80 CPU machine), this commit resulted in almost 30x performance
improvement, while still maintaining correct cpu quota restrictions.
That testcase is available at https://github.com/indeedeng/fibtest.
Fixes: 512ac999d2 ("sched/fair: Fix bandwidth timer clock drift condition")
Signed-off-by: Dave Chiluk <chiluk+linux@indeed.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: John Hammond <jhammond@indeed.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Kyle Anderson <kwa@yelp.com>
Cc: Gabriel Munos <gmunoz@netflix.com>
Cc: Peter Oskolkov <posk@posk.io>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Brendan Gregg <bgregg@netflix.com>
Link: https://lkml.kernel.org/r/1563900266-19734-2-git-send-email-chiluk+linux@indeed.com
The load_balance() has a dedicated mecanism to detect when an imbalance
is due to CPU affinity and must be handled at parent level. In this case,
the imbalance field of the parent's sched_group is set.
The description of sg_imbalanced() gives a typical example of two groups
of 4 CPUs each and 4 tasks each with a cpumask covering 1 CPU of the first
group and 3 CPUs of the second group. Something like:
{ 0 1 2 3 } { 4 5 6 7 }
* * * *
But the load_balance fails to fix this UC on my octo cores system
made of 2 clusters of quad cores.
Whereas the load_balance is able to detect that the imbalanced is due to
CPU affinity, it fails to fix it because the imbalance field is cleared
before letting parent level a chance to run. In fact, when the imbalance is
detected, the load_balance reruns without the CPU with pinned tasks. But
there is no other running tasks in the situation described above and
everything looks balanced this time so the imbalance field is immediately
cleared.
The imbalance field should not be cleared if there is no other task to move
when the imbalance is detected.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/1561996022-28829-1-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
When going through execve(), zero out the NUMA fault statistics instead of
freeing them.
During execve, the task is reachable through procfs and the scheduler. A
concurrent /proc/*/sched reader can read data from a freed ->numa_faults
allocation (confirmed by KASAN) and write it back to userspace.
I believe that it would also be possible for a use-after-free read to occur
through a race between a NUMA fault and execve(): task_numa_fault() can
lead to task_numa_compare(), which invokes task_weight() on the currently
running task of a different CPU.
Another way to fix this would be to make ->numa_faults RCU-managed or add
extra locking, but it seems easier to wipe the NUMA fault statistics on
execve.
Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Petr Mladek <pmladek@suse.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Will Deacon <will@kernel.org>
Fixes: 82727018b0 ("sched/numa: Call task_numa_free() from do_execve()")
Link: https://lkml.kernel.org/r/20190716152047.14424-1-jannh@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull scheduler updates from Ingo Molnar:
- Remove the unused per rq load array and all its infrastructure, by
Dietmar Eggemann.
- Add utilization clamping support by Patrick Bellasi. This is a
refinement of the energy aware scheduling framework with support for
boosting of interactive and capping of background workloads: to make
sure critical GUI threads get maximum frequency ASAP, and to make
sure background processing doesn't unnecessarily move to cpufreq
governor to higher frequencies and less energy efficient CPU modes.
- Add the bare minimum of tracepoints required for LISA EAS regression
testing, by Qais Yousef - which allows automated testing of various
power management features, including energy aware scheduling.
- Restructure the former tsk_nr_cpus_allowed() facility that the -rt
kernel used to modify the scheduler's CPU affinity logic such as
migrate_disable() - introduce the task->cpus_ptr value instead of
taking the address of &task->cpus_allowed directly - by Sebastian
Andrzej Siewior.
- Misc optimizations, fixes, cleanups and small enhancements - see the
Git log for details.
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (33 commits)
sched/uclamp: Add uclamp support to energy_compute()
sched/uclamp: Add uclamp_util_with()
sched/cpufreq, sched/uclamp: Add clamps for FAIR and RT tasks
sched/uclamp: Set default clamps for RT tasks
sched/uclamp: Reset uclamp values on RESET_ON_FORK
sched/uclamp: Extend sched_setattr() to support utilization clamping
sched/core: Allow sched_setattr() to use the current policy
sched/uclamp: Add system default clamps
sched/uclamp: Enforce last task's UCLAMP_MAX
sched/uclamp: Add bucket local max tracking
sched/uclamp: Add CPU's clamp buckets refcounting
sched/fair: Rename weighted_cpuload() to cpu_runnable_load()
sched/debug: Export the newly added tracepoints
sched/debug: Add sched_overutilized tracepoint
sched/debug: Add new tracepoint to track PELT at se level
sched/debug: Add new tracepoints to track PELT at rq level
sched/debug: Add a new sched_trace_*() helper functions
sched/autogroup: Make autogroup_path() always available
sched/wait: Deduplicate code with do-while
sched/topology: Remove unused 'sd' parameter from arch_scale_cpu_capacity()
...
The Energy Aware Scheduler (EAS) estimates the energy impact of waking
up a task on a given CPU. This estimation is based on:
a) an (active) power consumption defined for each CPU frequency
b) an estimation of which frequency will be used on each CPU
c) an estimation of the busy time (utilization) of each CPU
Utilization clamping can affect both b) and c).
A CPU is expected to run:
- on an higher than required frequency, but for a shorter time, in case
its estimated utilization will be smaller than the minimum utilization
enforced by uclamp
- on a smaller than required frequency, but for a longer time, in case
its estimated utilization is bigger than the maximum utilization
enforced by uclamp
While compute_energy() already accounts clamping effects on busy time,
the clamping effects on frequency selection are currently ignored.
Fix it by considering how CPU clamp values will be affected by a
task waking up and being RUNNABLE on that CPU.
Do that by refactoring schedutil_freq_util() to take an additional
task_struct* which allows EAS to evaluate the impact on clamp values of
a task being eventually queued in a CPU. Clamp values are applied to the
RT+CFS utilization only when a FREQUENCY_UTIL is required by
compute_energy().
Do note that switching from ENERGY_UTIL to FREQUENCY_UTIL in the
computation of the cpu_util signal implies that we are more likely to
estimate the highest OPP when a RT task is running in another CPU of
the same performance domain. This can have an impact on energy
estimation but:
- it's not easy to say which approach is better, since it depends on
the use case
- the original approach could still be obtained by setting a smaller
task-specific util_min whenever required
Since we are at that:
- rename schedutil_freq_util() into schedutil_cpu_util(),
since it's not only used for frequency selection.
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alessio Balsini <balsini@android.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Perret <quentin.perret@arm.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Steve Muckle <smuckle@google.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Todd Kjos <tkjos@google.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://lkml.kernel.org/r/20190621084217.8167-12-patrick.bellasi@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Each time a frequency update is required via schedutil, a frequency is
selected to (possibly) satisfy the utilization reported by each
scheduling class and irqs. However, when utilization clamping is in use,
the frequency selection should consider userspace utilization clamping
hints. This will allow, for example, to:
- boost tasks which are directly affecting the user experience
by running them at least at a minimum "requested" frequency
- cap low priority tasks not directly affecting the user experience
by running them only up to a maximum "allowed" frequency
These constraints are meant to support a per-task based tuning of the
frequency selection thus supporting a fine grained definition of
performance boosting vs energy saving strategies in kernel space.
Add support to clamp the utilization of RUNNABLE FAIR and RT tasks
within the boundaries defined by their aggregated utilization clamp
constraints.
Do that by considering the max(min_util, max_util) to give boosted tasks
the performance they need even when they happen to be co-scheduled with
other capped tasks.
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alessio Balsini <balsini@android.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Perret <quentin.perret@arm.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Steve Muckle <smuckle@google.com>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Todd Kjos <tkjos@google.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Link: https://lkml.kernel.org/r/20190621084217.8167-10-patrick.bellasi@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
When a cfs_rq sleeps and returns its quota, we delay for 5ms before
waking any throttled cfs_rqs to coalesce with other cfs_rqs going to
sleep, as this has to be done outside of the rq lock we hold.
The current code waits for 5ms without any sleeps, instead of waiting
for 5ms from the first sleep, which can delay the unthrottle more than
we want. Switch this around so that we can't push this forward forever.
This requires an extra flag rather than using hrtimer_active, since we
need to start a new timer if the current one is in the process of
finishing.
Signed-off-by: Ben Segall <bsegall@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Xunlei Pang <xlpang@linux.alibaba.com>
Acked-by: Phil Auld <pauld@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/xm26a7euy6iq.fsf_-_@bsegall-linux.svl.corp.google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
cfs_rq_has_blocked() and others_have_blocked() are only used within
update_blocked_averages(). The !CONFIG_FAIR_GROUP_SCHED version of the
latter calls them within a #define CONFIG_NO_HZ_COMMON block, whereas
the CONFIG_FAIR_GROUP_SCHED one calls them unconditionnally.
As reported by Qian, the above leads to this warning in
!CONFIG_NO_HZ_COMMON configs:
kernel/sched/fair.c: In function 'update_blocked_averages':
kernel/sched/fair.c:7750:7: warning: variable 'done' set but not used [-Wunused-but-set-variable]
It wouldn't be wrong to keep cfs_rq_has_blocked() and
others_have_blocked() as they are, but since their only current use is
to figure out when we can stop calling update_blocked_averages() on
fully decayed NOHZ idle CPUs, we can give them a new definition for
!CONFIG_NO_HZ_COMMON.
Change the definition of cfs_rq_has_blocked() and
others_have_blocked() for !CONFIG_NO_HZ_COMMON so that the
NOHZ-specific blocks of update_blocked_averages() become no-ops and
the 'done' variable gets optimised out.
While at it, remove the CONFIG_NO_HZ_COMMON block from the
!CONFIG_FAIR_GROUP_SCHED definition of update_blocked_averages() by
using the newly-introduced update_blocked_load_status() helper.
No change in functionality intended.
[ Additions by Peter Zijlstra. ]
Reported-by: Qian Cai <cai@lca.pw>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20190603115424.7951-1-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In commit:
4b53a3412d ("sched/core: Remove the tsk_nr_cpus_allowed() wrapper")
the tsk_nr_cpus_allowed() wrapper was removed. There was not
much difference in !RT but in RT we used this to implement
migrate_disable(). Within a migrate_disable() section the CPU mask is
restricted to single CPU while the "normal" CPU mask remains untouched.
As an alternative implementation Ingo suggested to use:
struct task_struct {
const cpumask_t *cpus_ptr;
cpumask_t cpus_mask;
};
with
t->cpus_ptr = &t->cpus_mask;
In -RT we then can switch the cpus_ptr to:
t->cpus_ptr = &cpumask_of(task_cpu(p));
in a migration disabled region. The rules are simple:
- Code that 'uses' ->cpus_allowed would use the pointer.
- Code that 'modifies' ->cpus_allowed would use the direct mask.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20190423142636.14347-1-bigeasy@linutronix.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
