sgs->group_weight is not set while gathering statistics in
update_sg_wakeup_stats(). This means that a group can be classified as
fully busy with 0 running tasks if utilization is high enough.
This path is mainly used for fork and exec.
Fixes: 57abff067a ("sched/fair: Rework find_idlest_group()")
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Mel Gorman <mgorman@techsingularity.net>
Link: https://lore.kernel.org/r/20200218144534.4564-1-vincent.guittot@linaro.org
Fix kernel-doc warning in kernel/sched/fair.c, caused by a recent
function parameter removal:
../kernel/sched/fair.c:3526: warning: Excess function parameter 'flags' description in 'attach_entity_load_avg'
Fixes: a4f9a0e51b ("sched/fair: Remove redundant call to cpufreq_update_util()")
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/cbe964e4-6879-fd08-41c9-ef1917414af4@infradead.org
The following XFS commit:
8ab39f11d9 ("xfs: prevent CIL push holdoff in log recovery")
changed the logic from using bound workqueues to using unbound
workqueues. Functionally this makes sense but it was observed at the
time that the dbench performance dropped quite a lot and CPU migrations
were increased.
The current pattern of the task migration is straight-forward. With XFS,
an IO issuer delegates work to xlog_cil_push_work ()on an unbound kworker.
This runs on a nearby CPU and on completion, dbench wakes up on its old CPU
as it is still idle and no migration occurs. dbench then queues the real
IO on the blk_mq_requeue_work() work item which runs on a bound kworker
which is forced to run on the same CPU as dbench. When IO completes,
the bound kworker wakes dbench but as the kworker is a bound but,
real task, the CPU is not considered idle and dbench gets migrated by
select_idle_sibling() to a new CPU. dbench may ping-pong between two CPUs
for a while but ultimately it starts a round-robin of all CPUs sharing
the same LLC. High-frequency migration on each IO completion has poor
performance overall. It has negative implications both in commication
costs and power management. mpstat confirmed that at low thread counts
that all CPUs sharing an LLC has low level of activity.
Note that even if the CIL patch was reverted, there still would
be migrations but the impact is less noticeable. It turns out that
individually the scheduler, XFS, blk-mq and workqueues all made sensible
decisions but in combination, the overall effect was sub-optimal.
This patch special cases the IO issue/completion pattern and allows
a bound kworker waker and a task wakee to stack on the same CPU if
there is a strong chance they are directly related. The expectation
is that the kworker is likely going back to sleep shortly. This is not
guaranteed as the IO could be queued asynchronously but there is a very
strong relationship between the task and kworker in this case that would
justify stacking on the same CPU instead of migrating. There should be
few concerns about kworker starvation given that the special casing is
only when the kworker is the waker.
DBench on XFS
MMTests config: io-dbench4-async modified to run on a fresh XFS filesystem
UMA machine with 8 cores sharing LLC
5.5.0-rc7 5.5.0-rc7
tipsched-20200124 kworkerstack
Amean 1 22.63 ( 0.00%) 20.54 * 9.23%*
Amean 2 25.56 ( 0.00%) 23.40 * 8.44%*
Amean 4 28.63 ( 0.00%) 27.85 * 2.70%*
Amean 8 37.66 ( 0.00%) 37.68 ( -0.05%)
Amean 64 469.47 ( 0.00%) 468.26 ( 0.26%)
Stddev 1 1.00 ( 0.00%) 0.72 ( 28.12%)
Stddev 2 1.62 ( 0.00%) 1.97 ( -21.54%)
Stddev 4 2.53 ( 0.00%) 3.58 ( -41.19%)
Stddev 8 5.30 ( 0.00%) 5.20 ( 1.92%)
Stddev 64 86.36 ( 0.00%) 94.53 ( -9.46%)
NUMA machine, 48 CPUs total, 24 CPUs share cache
5.5.0-rc7 5.5.0-rc7
tipsched-20200124 kworkerstack-v1r2
Amean 1 58.69 ( 0.00%) 30.21 * 48.53%*
Amean 2 60.90 ( 0.00%) 35.29 * 42.05%*
Amean 4 66.77 ( 0.00%) 46.55 * 30.28%*
Amean 8 81.41 ( 0.00%) 68.46 * 15.91%*
Amean 16 113.29 ( 0.00%) 107.79 * 4.85%*
Amean 32 199.10 ( 0.00%) 198.22 * 0.44%*
Amean 64 478.99 ( 0.00%) 477.06 * 0.40%*
Amean 128 1345.26 ( 0.00%) 1372.64 * -2.04%*
Stddev 1 2.64 ( 0.00%) 4.17 ( -58.08%)
Stddev 2 4.35 ( 0.00%) 5.38 ( -23.73%)
Stddev 4 6.77 ( 0.00%) 6.56 ( 3.00%)
Stddev 8 11.61 ( 0.00%) 10.91 ( 6.04%)
Stddev 16 18.63 ( 0.00%) 19.19 ( -3.01%)
Stddev 32 38.71 ( 0.00%) 38.30 ( 1.06%)
Stddev 64 100.28 ( 0.00%) 91.24 ( 9.02%)
Stddev 128 186.87 ( 0.00%) 160.34 ( 14.20%)
Dbench has been modified to report the time to complete a single "load
file". This is a more meaningful metric for dbench that a throughput
metric as the benchmark makes many different system calls that are not
throughput-related
Patch shows a 9.23% and 48.53% reduction in the time to process a load
file with the difference partially explained by the number of CPUs sharing
a LLC. In a separate run, task migrations were almost eliminated by the
patch for low client counts. In case people have issue with the metric
used for the benchmark, this is a comparison of the throughputs as
reported by dbench on the NUMA machine.
dbench4 Throughput (misleading but traditional)
5.5.0-rc7 5.5.0-rc7
tipsched-20200124 kworkerstack-v1r2
Hmean 1 321.41 ( 0.00%) 617.82 * 92.22%*
Hmean 2 622.87 ( 0.00%) 1066.80 * 71.27%*
Hmean 4 1134.56 ( 0.00%) 1623.74 * 43.12%*
Hmean 8 1869.96 ( 0.00%) 2212.67 * 18.33%*
Hmean 16 2673.11 ( 0.00%) 2806.13 * 4.98%*
Hmean 32 3032.74 ( 0.00%) 3039.54 ( 0.22%)
Hmean 64 2514.25 ( 0.00%) 2498.96 * -0.61%*
Hmean 128 1778.49 ( 0.00%) 1746.05 * -1.82%*
Note that this is somewhat specific to XFS and ext4 shows no performance
difference as it does not rely on kworkers in the same way. No major
problem was observed running other workloads on different machines although
not all tests have completed yet.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200128154006.GD3466@techsingularity.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The CPU load balancer balances between different domains to spread load
and strives to have equal balance everywhere. Communicating tasks can
migrate so they are topologically close to each other but these decisions
are independent. On a lightly loaded NUMA machine, two communicating tasks
pulled together at wakeup time can be pushed apart by the load balancer.
In isolation, the load balancer decision is fine but it ignores the tasks
data locality and the wakeup/LB paths continually conflict. NUMA balancing
is also a factor but it also simply conflicts with the load balancer.
This patch allows a fixed degree of imbalance of two tasks to exist
between NUMA domains regardless of utilisation levels. In many cases,
this prevents communicating tasks being pulled apart. It was evaluated
whether the imbalance should be scaled to the domain size. However, no
additional benefit was measured across a range of workloads and machines
and scaling adds the risk that lower domains have to be rebalanced. While
this could change again in the future, such a change should specify the
use case and benefit.
The most obvious impact is on netperf TCP_STREAM -- two simple
communicating tasks with some softirq offload depending on the
transmission rate.
2-socket Haswell machine 48 core, HT enabled
netperf-tcp -- mmtests config config-network-netperf-unbound
baseline lbnuma-v3
Hmean 64 568.73 ( 0.00%) 577.56 * 1.55%*
Hmean 128 1089.98 ( 0.00%) 1128.06 * 3.49%*
Hmean 256 2061.72 ( 0.00%) 2104.39 * 2.07%*
Hmean 1024 7254.27 ( 0.00%) 7557.52 * 4.18%*
Hmean 2048 11729.20 ( 0.00%) 13350.67 * 13.82%*
Hmean 3312 15309.08 ( 0.00%) 18058.95 * 17.96%*
Hmean 4096 17338.75 ( 0.00%) 20483.66 * 18.14%*
Hmean 8192 25047.12 ( 0.00%) 27806.84 * 11.02%*
Hmean 16384 27359.55 ( 0.00%) 33071.88 * 20.88%*
Stddev 64 2.16 ( 0.00%) 2.02 ( 6.53%)
Stddev 128 2.31 ( 0.00%) 2.19 ( 5.05%)
Stddev 256 11.88 ( 0.00%) 3.22 ( 72.88%)
Stddev 1024 23.68 ( 0.00%) 7.24 ( 69.43%)
Stddev 2048 79.46 ( 0.00%) 71.49 ( 10.03%)
Stddev 3312 26.71 ( 0.00%) 57.80 (-116.41%)
Stddev 4096 185.57 ( 0.00%) 96.15 ( 48.19%)
Stddev 8192 245.80 ( 0.00%) 100.73 ( 59.02%)
Stddev 16384 207.31 ( 0.00%) 141.65 ( 31.67%)
In this case, there was a sizable improvement to performance and
a general reduction in variance. However, this is not univeral.
For most machines, the impact was roughly a 3% performance gain.
Ops NUMA base-page range updates 19796.00 292.00
Ops NUMA PTE updates 19796.00 292.00
Ops NUMA PMD updates 0.00 0.00
Ops NUMA hint faults 16113.00 143.00
Ops NUMA hint local faults % 8407.00 142.00
Ops NUMA hint local percent 52.18 99.30
Ops NUMA pages migrated 4244.00 1.00
Without the patch, only 52.18% of sampled accesses are local. In an
earlier changelog, 100% of sampled accesses are local and indeed on
most machines, this was still the case. In this specific case, the
local sampled rates was 99.3% but note the "base-page range updates"
and "PTE updates". The activity with the patch is negligible as were
the number of faults. The small number of pages migrated were related to
shared libraries. A 2-socket Broadwell showed better results on average
but are not presented for brevity as the performance was similar except
it showed 100% of the sampled NUMA hints were local. The patch holds up
for a 4-socket Haswell, an AMD EPYC and AMD Epyc 2 machine.
For dbench, the impact depends on the filesystem used and the number of
clients. On XFS, there is little difference as the clients typically
communicate with workqueues which have a separate class of scheduler
problem at the moment. For ext4, performance is generally better,
particularly for small numbers of clients as NUMA balancing activity is
negligible with the patch applied.
A more interesting example is the Facebook schbench which uses a
number of messaging threads to communicate with worker threads. In this
configuration, one messaging thread is used per NUMA node and the number of
worker threads is varied. The 50, 75, 90, 95, 99, 99.5 and 99.9 percentiles
for response latency is then reported.
Lat 50.00th-qrtle-1 44.00 ( 0.00%) 37.00 ( 15.91%)
Lat 75.00th-qrtle-1 53.00 ( 0.00%) 41.00 ( 22.64%)
Lat 90.00th-qrtle-1 57.00 ( 0.00%) 42.00 ( 26.32%)
Lat 95.00th-qrtle-1 63.00 ( 0.00%) 43.00 ( 31.75%)
Lat 99.00th-qrtle-1 76.00 ( 0.00%) 51.00 ( 32.89%)
Lat 99.50th-qrtle-1 89.00 ( 0.00%) 52.00 ( 41.57%)
Lat 99.90th-qrtle-1 98.00 ( 0.00%) 55.00 ( 43.88%)
Lat 50.00th-qrtle-2 42.00 ( 0.00%) 42.00 ( 0.00%)
Lat 75.00th-qrtle-2 48.00 ( 0.00%) 47.00 ( 2.08%)
Lat 90.00th-qrtle-2 53.00 ( 0.00%) 52.00 ( 1.89%)
Lat 95.00th-qrtle-2 55.00 ( 0.00%) 53.00 ( 3.64%)
Lat 99.00th-qrtle-2 62.00 ( 0.00%) 60.00 ( 3.23%)
Lat 99.50th-qrtle-2 63.00 ( 0.00%) 63.00 ( 0.00%)
Lat 99.90th-qrtle-2 68.00 ( 0.00%) 66.00 ( 2.94%
For higher worker threads, the differences become negligible but it's
interesting to note the difference in wakeup latency at low utilisation
and mpstat confirms that activity was almost all on one node until
the number of worker threads increase.
Hackbench generally showed neutral results across a range of machines.
This is different to earlier versions of the patch which allowed imbalances
for higher degrees of utilisation. perf bench pipe showed negligible
differences in overall performance as the differences are very close to
the noise.
An earlier prototype of the patch showed major regressions for NAS C-class
when running with only half of the available CPUs -- 20-30% performance
hits were measured at the time. With this version of the patch, the impact
is negligible with small gains/losses within the noise measured. This is
because the number of threads far exceeds the small imbalance the aptch
cares about. Similarly, there were report of regressions for the autonuma
benchmark against earlier versions but again, normal load balancing now
applies for that workload.
In general, the patch simply seeks to avoid unnecessary cross-node
migrations in the basic case where imbalances are very small. For low
utilisation communicating workloads, this patch generally behaves better
with less NUMA balancing activity. For high utilisation, there is no
change in behaviour.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Acked-by: Phil Auld <pauld@redhat.com>
Tested-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200114101319.GO3466@techsingularity.net
sched_idle_cpu() isn't used for non SMP configuration and with a recent
change, we have started getting following warning:
kernel/sched/fair.c:5221:12: warning: ‘sched_idle_cpu’ defined but not used [-Wunused-function]
Fix that by defining sched_idle_cpu() only for SMP configurations.
Fixes: 323af6deaf ("sched/fair: Load balance aggressively for SCHED_IDLE CPUs")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Link: https://lore.kernel.org/r/f0554f590687478b33914a4aff9f0e6a62886d44.1579499907.git.viresh.kumar@linaro.org
With commit
bef69dd878 ("sched/cpufreq: Move the cfs_rq_util_change() call to cpufreq_update_util()")
update_load_avg() has become the central point for calling cpufreq
(not including the update of blocked load). This change helps to
simplify further the number of calls to cpufreq_update_util() and to
remove last redundant ones. With update_load_avg(), we are now sure
that cpufreq_update_util() will be called after every task attachment
to a cfs_rq and especially after propagating this event down to the
util_avg of the root cfs_rq, which is the level that is used by
cpufreq governors like schedutil to set the frequency of a CPU.
The SCHED_CPUFREQ_MIGRATION flag forces an early call to cpufreq when
the migration happens in a cgroup whereas util_avg of root cfs_rq is
not yet updated and this call is duplicated with the one that happens
immediately after when the migration event reaches the root cfs_rq.
The dedicated flag SCHED_CPUFREQ_MIGRATION is now useless and can be
removed. The interface of attach_entity_load_avg() can also be
simplified accordingly.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/1579083620-24943-1-git-send-email-vincent.guittot@linaro.org
commit bf475ce0a3 ("sched/fair: Add per-CPU min capacity to
sched_group_capacity") introduced per-cpu min_capacity.
commit e3d6d0cb66 ("sched/fair: Add sched_group per-CPU max capacity")
introduced per-cpu max_capacity.
In the SD_OVERLAP case, the local variable 'capacity' represents the sum
of CPU capacity of all CPUs in the first sched group (sg) of the sched
domain (sd).
It is erroneously used to calculate sg's min and max CPU capacity.
To fix this use capacity_of(cpu) instead of 'capacity'.
The code which achieves this via cpu_rq(cpu)->sd->groups->sgc->capacity
(for rq->sd != NULL) can be removed since it delivers the same value as
capacity_of(cpu) which is currently only used for the (!rq->sd) case
(see update_cpu_capacity()).
An sg of the lowest sd (rq->sd or sd->child == NULL) represents a single
CPU (and hence sg->sgc->capacity == capacity_of(cpu)).
Signed-off-by: Peng Liu <iwtbavbm@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20200104130828.GA7718@iZj6chx1xj0e0buvshuecpZ
Move the code of calculation for delta_sum/delta_avg to where
it is really needed to be done.
Signed-off-by: Peng Wang <rocking@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20200103114400.17668-1-rocking@linux.alibaba.com
The fair scheduler performs periodic load balance on every CPU to check
if it can pull some tasks from other busy CPUs. The duration of this
periodic load balance is set to sd->balance_interval for the idle CPUs
and is calculated by multiplying the sd->balance_interval with the
sd->busy_factor (set to 32 by default) for the busy CPUs. The
multiplication is done for busy CPUs to avoid doing load balance too
often and rather spend more time executing actual task. While that is
the right thing to do for the CPUs busy with SCHED_OTHER or SCHED_BATCH
tasks, it may not be the optimal thing for CPUs running only SCHED_IDLE
tasks.
With the recent enhancements in the fair scheduler around SCHED_IDLE
CPUs, we now prefer to enqueue a newly-woken task to a SCHED_IDLE
CPU instead of other busy or idle CPUs. The same reasoning should be
applied to the load balancer as well to make it migrate tasks more
aggressively to a SCHED_IDLE CPU, as that will reduce the scheduling
latency of the migrated (SCHED_OTHER) tasks.
This patch makes minimal changes to the fair scheduler to do the next
load balance soon after the last non SCHED_IDLE task is dequeued from a
runqueue, i.e. making the CPU SCHED_IDLE. Also the sd->busy_factor is
ignored while calculating the balance_interval for such CPUs. This is
done to avoid delaying the periodic load balance by few hundred
milliseconds for SCHED_IDLE CPUs.
This is tested on ARM64 Hikey620 platform (octa-core) with the help of
rt-app and it is verified, using kernel traces, that the newly
SCHED_IDLE CPU does load balancing shortly after it becomes SCHED_IDLE
and pulls tasks from other busy CPUs.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/e485827eb8fe7db0943d6f3f6e0f5a4a70272781.1578471925.git.viresh.kumar@linaro.org
Similarly to calculate_imbalance() and find_busiest_group(), using the
number of idle CPUs when there is only 1 CPU in the group is not efficient
because we can't make a difference between a CPU running 1 task and a CPU
running dozens of small tasks competing for the same CPU but not enough
to overload it. More generally speaking, we should use the number of
running tasks when there is the same number of idle CPUs in a group instead
of blindly select the 1st one.
When the groups have spare capacity and the same number of idle CPUs, we
compare the number of running tasks to select the busiest group.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/1576839893-26930-1-git-send-email-vincent.guittot@linaro.org
task_fits_capacity() has just been made uclamp-aware, and
find_energy_efficient_cpu() needs to go through the same treatment.
Things are somewhat different here however - using the task max clamp isn't
sufficient. Consider the following setup:
The target runqueue, rq:
rq.cpu_capacity_orig = 512
rq.cfs.avg.util_avg = 200
rq.uclamp.max = 768 // the max p.uclamp.max of all enqueued p's is 768
The waking task, p (not yet enqueued on rq):
p.util_est = 600
p.uclamp.max = 100
Now, consider the following code which doesn't use the rq clamps:
util = uclamp_task_util(p);
// Does the task fit in the spare CPU capacity?
cpu = cpu_of(rq);
fits_capacity(util, cpu_capacity(cpu) - cpu_util(cpu))
This would lead to:
util = 100;
fits_capacity(100, 512 - 200)
fits_capacity() would return true. However, enqueuing p on that CPU *will*
cause it to become overutilized since rq clamp values are max-aggregated,
so we'd remain with
rq.uclamp.max = 768
which comes from the other tasks already enqueued on rq. Thus, we could
select a high enough frequency to reach beyond 0.8 * 512 utilization
(== overutilized) after enqueuing p on rq. What find_energy_efficient_cpu()
needs here is uclamp_rq_util_with() which lets us peek at the future
utilization landscape, including rq-wide uclamp values.
Make find_energy_efficient_cpu() use uclamp_rq_util_with() for its
fits_capacity() check. This is in line with what compute_energy() ends up
using for estimating utilization.
Tested-By: Dietmar Eggemann <dietmar.eggemann@arm.com>
Suggested-by: Quentin Perret <qperret@google.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.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/20191211113851.24241-6-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
task_fits_capacity() drives CPU selection at wakeup time, and is also used
to detect misfit tasks. Right now it does so by comparing task_util_est()
with a CPU's capacity, but doesn't take into account uclamp restrictions.
There's a few interesting uses that can come out of doing this. For
instance, a low uclamp.max value could prevent certain tasks from being
flagged as misfit tasks, so they could merrily remain on low-capacity CPUs.
Similarly, a high uclamp.min value would steer tasks towards high capacity
CPUs at wakeup (and, should that fail, later steered via misfit balancing),
so such "boosted" tasks would favor CPUs of higher capacity.
Introduce uclamp_task_util() and make task_fits_capacity() use it.
Tested-By: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Quentin Perret <qperret@google.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.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/20191211113851.24241-5-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
There are instances where we keep searching for an idle CPU despite
already having a sched-idle CPU (in find_idlest_group_cpu(),
select_idle_smt() and select_idle_cpu() and then there are places where
we don't necessarily do that and return a sched-idle CPU as soon as we
find one (in select_idle_sibling()). This looks a bit inconsistent and
it may be worth having the same policy everywhere.
On the other hand, choosing a sched-idle CPU over a idle one shall be
beneficial from performance and power point of view as well, as we don't
need to get the CPU online from a deep idle state which wastes quite a
lot of time and energy and delays the scheduling of the newly woken up
task.
This patch tries to simplify code around sched-idle CPU selection and
make it consistent throughout.
Testing is done with the help of rt-app on hikey board (ARM64 octa-core,
2 clusters, 0-3 and 4-7). The cpufreq governor was set to performance to
avoid any side affects from CPU frequency. Following are the tests
performed:
Test 1: 1-cfs-task:
A single SCHED_NORMAL task is pinned to CPU5 which runs for 2333 us
out of 7777 us (so gives time for the cluster to go in deep idle
state).
Test 2: 1-cfs-1-idle-task:
A single SCHED_NORMAL task is pinned on CPU5 and single SCHED_IDLE
task is pinned on CPU6 (to make sure cluster 1 doesn't go in deep idle
state).
Test 3: 1-cfs-8-idle-task:
A single SCHED_NORMAL task is pinned on CPU5 and eight SCHED_IDLE
tasks are created which run forever (not pinned anywhere, so they run
on all CPUs). Checked with kernelshark that as soon as NORMAL task
sleeps, the SCHED_IDLE task starts running on CPU5.
And here are the results on mean latency (in us), using the "st" tool.
$ st 1-cfs-task/rt-app-cfs_thread-0.log
N min max sum mean stddev
642 90 592 197180 307.134 109.906
$ st 1-cfs-1-idle-task/rt-app-cfs_thread-0.log
N min max sum mean stddev
642 67 311 113850 177.336 41.4251
$ st 1-cfs-8-idle-task/rt-app-cfs_thread-0.log
N min max sum mean stddev
643 29 173 41364 64.3297 13.2344
The mean latency when we need to:
- wakeup from deep idle state is 307 us.
- wakeup from shallow idle state is 177 us.
- preempt a SCHED_IDLE task is 64 us.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-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/b90cbcce608cef4e02a7bbfe178335f76d201bab.1573728344.git.viresh.kumar@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
select_idle_cpu() will scan the LLC domain for idle CPUs,
it's always expensive. so the next commit :
1ad3aaf3fc ("sched/core: Implement new approach to scale select_idle_cpu()")
introduces a way to limit how many CPUs we scan.
But it consume some CPUs out of 'nr' that are not allowed
for the task and thus waste our attempts. The function
always return nr_cpumask_bits, and we can't find a CPU
which our task is allowed to run.
Cpumask may be too big, similar to select_idle_core(), use
per_cpu_ptr 'select_idle_mask' to prevent stack overflow.
Fixes: 1ad3aaf3fc ("sched/core: Implement new approach to scale select_idle_cpu()")
Signed-off-by: Cheng Jian <cj.chengjian@huawei.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20191213024530.28052-1-cj.chengjian@huawei.com
The runqueue of a fair task being remotely reniced is going to get a
resched IPI in order to reassess which task should be the current
running on the CPU. However that evaluation is useless if the fair task
is running alone, in which case we can spare that IPI, preventing
nohz_full CPUs from being disturbed.
Reported-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20191203160106.18806-2-frederic@kernel.org
The load balance can fail to find a suitable task during the periodic check
because the imbalance is smaller than half of the load of the waiting
tasks. This results in the increase of the number of failed load balance,
which can end up to start an active migration. This active migration is
useless because the current running task is not a better choice than the
waiting ones. In fact, the current task was probably not running but
waiting for the CPU during one of the previous attempts and it had already
not been selected.
When load balance fails too many times to migrate a task, we should relax
the contraint on the maximum load of the tasks that can be migrated
similarly to what is done with cache hotness.
Before the rework, load balance used to set the imbalance to the average
load_per_task in order to mitigate such situation. This increased the
likelihood of migrating a task but also of selecting a larger task than
needed while more appropriate ones were in the list.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/1575036287-6052-1-git-send-email-vincent.guittot@linaro.org
Because of CPU affinity, the local group can be skipped which breaks the
assumption that statistics are always collected for local group. With
uninitialized local_sgs, the comparison is meaningless and the behavior
unpredictable. This can even end up to use local pointer which is to
NULL in this case.
If the local group has been skipped because of CPU affinity, we return
the idlest group.
Fixes: 57abff067a ("sched/fair: Rework find_idlest_group()")
Reported-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Tested-by: John Stultz <john.stultz@linaro.org>
Cc: rostedt@goodmis.org
Cc: valentin.schneider@arm.com
Cc: mingo@redhat.com
Cc: mgorman@suse.de
Cc: juri.lelli@redhat.com
Cc: dietmar.eggemann@arm.com
Cc: bsegall@google.com
Cc: qais.yousef@arm.com
Link: https://lkml.kernel.org/r/1575483700-22153-1-git-send-email-vincent.guittot@linaro.org
update_cfs_rq_load_avg() calls cfs_rq_util_change() every time PELT decays,
which might be inefficient when the cpufreq driver has rate limitation.
When a task is attached on a CPU, we have this call path:
update_load_avg()
update_cfs_rq_load_avg()
cfs_rq_util_change -- > trig frequency update
attach_entity_load_avg()
cfs_rq_util_change -- > trig frequency update
The 1st frequency update will not take into account the utilization of the
newly attached task and the 2nd one might be discarded because of rate
limitation of the cpufreq driver.
update_cfs_rq_load_avg() is only called by update_blocked_averages()
and update_load_avg() so we can move the call to
cfs_rq_util_change/cpufreq_update_util() into these two functions.
It's also interesting to note that update_load_avg() already calls
cfs_rq_util_change() directly for the !SMP case.
This change will also ensure that cpufreq_update_util() is called even
when there is no more CFS rq in the leaf_cfs_rq_list to update, but only
IRQ, RT or DL PELT signals.
[ mingo: Minor updates. ]
Reported-by: Doug Smythies <dsmythies@telus.net>
Tested-by: Doug Smythies <dsmythies@telus.net>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: juri.lelli@redhat.com
Cc: linux-pm@vger.kernel.org
Cc: mgorman@suse.de
Cc: rostedt@goodmis.org
Cc: sargun@sargun.me
Cc: srinivas.pandruvada@linux.intel.com
Cc: tj@kernel.org
Cc: xiexiuqi@huawei.com
Cc: xiezhipeng1@huawei.com
Fixes: 039ae8bcf7 ("sched/fair: Fix O(nr_cgroups) in the load balancing path")
Link: https://lkml.kernel.org/r/1574083279-799-1-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Add comments to describe each state of goup_type and to add some details
about the load balance at NUMA level.
[ Valentin Schneider: Updates to the comments. ]
[ mingo: Other updates to the comments. ]
Reported-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Acked-by: Valentin Schneider <valentin.schneider@arm.com>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/1573570243-1903-1-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The task, for which the scheduler looks for the idlest group of CPUs, must
be discounted from all statistics in order to get a fair comparison
between groups. This includes utilization, load, nr_running and idle_cpus.
Such unfairness can be easily highlighted with the unixbench execl 1 task.
This test continuously call execve() and the scheduler looks for the idlest
group/CPU on which it should place the task. Because the task runs on the
local group/CPU, the latter seems already busy even if there is nothing
else running on it. As a result, the scheduler will always select another
group/CPU than the local one.
This recovers most of the performance regression on my system from the
recent load-balancer rewrite.
[ mingo: Minor cleanups. ]
Reported-by: kernel test robot <rong.a.chen@intel.com>
Tested-by: kernel test robot <rong.a.chen@intel.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: dietmar.eggemann@arm.com
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Fixes: 57abff067a ("sched/fair: Rework find_idlest_group()")
Link: https://lkml.kernel.org/r/1571762798-25900-1-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Ever since we moved the sched_class definitions into their own files,
the constant expression {fair,idle}_sched_class.pick_next_task() is
not in fact a compile time constant anymore and results in an indirect
call (barring LTO).
Fix that by exposing pick_next_task_{fair,idle}() directly, this gets
rid of the indirect call (and RETPOLINE) on the fast path.
Also remove the unlikely() from the idle case, it is in fact /the/ way
we select idle -- and that is a very common thing to do.
Performance for will-it-scale/sched_yield improves by 2% (as reported
by 0-day).
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>
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: juri.lelli@redhat.com
Cc: ktkhai@virtuozzo.com
Cc: mgorman@suse.de
Cc: qais.yousef@arm.com
Cc: qperret@google.com
Cc: rostedt@goodmis.org
Cc: valentin.schneider@arm.com
Cc: vincent.guittot@linaro.org
Link: https://lkml.kernel.org/r/20191108131909.603037345@infradead.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Commit 67692435c4 ("sched: Rework pick_next_task() slow-path")
inadvertly introduced a race because it changed a previously
unexplored dependency between dropping the rq->lock and
sched_class::put_prev_task().
The comments about dropping rq->lock, in for example
newidle_balance(), only mentions the task being current and ->on_cpu
being set. But when we look at the 'change' pattern (in for example
sched_setnuma()):
queued = task_on_rq_queued(p); /* p->on_rq == TASK_ON_RQ_QUEUED */
running = task_current(rq, p); /* rq->curr == p */
if (queued)
dequeue_task(...);
if (running)
put_prev_task(...);
/* change task properties */
if (queued)
enqueue_task(...);
if (running)
set_next_task(...);
It becomes obvious that if we do this after put_prev_task() has
already been called on @p, things go sideways. This is exactly what
the commit in question allows to happen when it does:
prev->sched_class->put_prev_task(rq, prev, rf);
if (!rq->nr_running)
newidle_balance(rq, rf);
The newidle_balance() call will drop rq->lock after we've called
put_prev_task() and that allows the above 'change' pattern to
interleave and mess up the state.
Furthermore, it turns out we lost the RT-pull when we put the last DL
task.
Fix both problems by extracting the balancing from put_prev_task() and
doing a multi-class balance() pass before put_prev_task().
Fixes: 67692435c4 ("sched: Rework pick_next_task() slow-path")
Reported-by: Quentin Perret <qperret@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Quentin Perret <qperret@google.com>
Tested-by: Valentin Schneider <valentin.schneider@arm.com>
The estimated utilization for a task:
util_est = max(util_avg, est.enqueue, est.ewma)
is defined based on:
- util_avg: the PELT defined utilization
- est.enqueued: the util_avg at the end of the last activation
- est.ewma: a exponential moving average on the est.enqueued samples
According to this definition, when a task suddenly changes its bandwidth
requirements from small to big, the EWMA will need to collect multiple
samples before converging up to track the new big utilization.
This slow convergence towards bigger utilization values is not
aligned to the default scheduler behavior, which is to optimize for
performance. Moreover, the est.ewma component fails to compensate for
temporarely utilization drops which spans just few est.enqueued samples.
To let util_est do a better job in the scenario depicted above, change
its definition by making util_est directly follow upward motion and
only decay the est.ewma on downward.
Signed-off-by: Patrick Bellasi <patrick.bellasi@matbug.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Douglas Raillard <douglas.raillard@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Quentin Perret <qperret@google.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20191023205630.14469-1-patrick.bellasi@matbug.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The slow wake up path computes per sched_group statisics to select the
idlest group, which is quite similar to what load_balance() is doing
for selecting busiest group. Rework find_idlest_group() to classify the
sched_group and select the idlest one following the same steps as
load_balance().
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-12-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Runnable load was originally introduced to take into account the case where
blocked load biases the wake up path which may end to select an overloaded
CPU with a large number of runnable tasks instead of an underutilized
CPU with a huge blocked load.
Tha wake up path now starts looking for idle CPUs before comparing
runnable load and it's worth aligning the wake up path with the
load_balance() logic.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-10-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Utilization is used to detect a misfit task but the load is then used to
select the task on the CPU which can lead to select a small task with
high weight instead of the task that triggered the misfit migration.
Check that task can't fit the CPU's capacity when selecting the misfit
task instead of using the load.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Acked-by: Valentin Schneider <valentin.schneider@arm.com>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Link: https://lkml.kernel.org/r/1571405198-27570-9-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
'runnable load' was originally introduced to take into account the case
where blocked load biases the load balance decision which was selecting
underutilized groups with huge blocked load whereas other groups were
overloaded.
The load is now only used when groups are overloaded. In this case,
it's worth being conservative and taking into account the sleeping
tasks that might wake up on the CPU.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-7-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
CFS load_balance() only takes care of CFS tasks whereas CPUs can be used by
other scheduling classes. Typically, a CFS task preempted by an RT or deadline
task will not get a chance to be pulled by another CPU because
load_balance() doesn't take into account tasks from other classes.
Add sum of nr_running in the statistics and use it to detect such
situations.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-6-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The load_balance() algorithm contains some heuristics which have become
meaningless since the rework of the scheduler's metrics like the
introduction of PELT.
Furthermore, load is an ill-suited metric for solving certain task
placement imbalance scenarios.
For instance, in the presence of idle CPUs, we should simply try to get at
least one task per CPU, whereas the current load-based algorithm can actually
leave idle CPUs alone simply because the load is somewhat balanced.
The current algorithm ends up creating virtual and meaningless values like
the avg_load_per_task or tweaks the state of a group to make it overloaded
whereas it's not, in order to try to migrate tasks.
load_balance() should better qualify the imbalance of the group and clearly
define what has to be moved to fix this imbalance.
The type of sched_group has been extended to better reflect the type of
imbalance. We now have:
group_has_spare
group_fully_busy
group_misfit_task
group_asym_packing
group_imbalanced
group_overloaded
Based on the type of sched_group, load_balance now sets what it wants to
move in order to fix the imbalance. It can be some load as before but also
some utilization, a number of task or a type of task:
migrate_task
migrate_util
migrate_load
migrate_misfit
This new load_balance() algorithm fixes several pending wrong tasks
placement:
- the 1 task per CPU case with asymmetric system
- the case of cfs task preempted by other class
- the case of tasks not evenly spread on groups with spare capacity
Also the load balance decisions have been consolidated in the 3 functions
below after removing the few bypasses and hacks of the current code:
- update_sd_pick_busiest() select the busiest sched_group.
- find_busiest_group() checks if there is an imbalance between local and
busiest group.
- calculate_imbalance() decides what have to be moved.
Finally, the now unused field total_running of struct sd_lb_stats has been
removed.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: riel@surriel.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-5-git-send-email-vincent.guittot@linaro.org
[ Small readability and spelling updates. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Rename sum_nr_running to sum_h_nr_running because it effectively tracks
cfs->h_nr_running so we can use sum_nr_running to track rq->nr_running
when needed.
There are no functional changes.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Acked-by: Rik van Riel <riel@surriel.com>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: srikar@linux.vnet.ibm.com
Link: https://lkml.kernel.org/r/1571405198-27570-3-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Clean up asym packing to follow the default load balance behavior:
- classify the group by creating a group_asym_packing field.
- calculate the imbalance in calculate_imbalance() instead of bypassing it.
We don't need to test twice same conditions anymore to detect asym packing
and we consolidate the calculation of imbalance in calculate_imbalance().
There is no functional changes.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Acked-by: Rik van Riel <riel@surriel.com>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: hdanton@sina.com
Cc: parth@linux.ibm.com
Cc: pauld@redhat.com
Cc: quentin.perret@arm.com
Cc: srikar@linux.vnet.ibm.com
Cc: valentin.schneider@arm.com
Link: https://lkml.kernel.org/r/1571405198-27570-2-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The quota/period ratio is used to ensure a child task group won't get
more bandwidth than the parent task group, and is calculated as:
normalized_cfs_quota() = [(quota_us << 20) / period_us]
If the quota/period ratio was changed during this scaling due to
precision loss, it will cause inconsistency between parent and child
task groups.
See below example:
A userspace container manager (kubelet) does three operations:
1) Create a parent cgroup, set quota to 1,000us and period to 10,000us.
2) Create a few children cgroups.
3) Set quota to 1,000us and period to 10,000us on a child cgroup.
These operations are expected to succeed. However, if the scaling of
147/128 happens before step 3, quota and period of the parent cgroup
will be changed:
new_quota: 1148437ns, 1148us
new_period: 11484375ns, 11484us
And when step 3 comes in, the ratio of the child cgroup will be
104857, which will be larger than the parent cgroup ratio (104821),
and will fail.
Scaling them by a factor of 2 will fix the problem.
Tested-by: Phil Auld <pauld@redhat.com>
Signed-off-by: Xuewei Zhang <xueweiz@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Phil Auld <pauld@redhat.com>
Cc: Anton Blanchard <anton@ozlabs.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Fixes: 2e8e192263 ("sched/fair: Limit sched_cfs_period_timer() loop to avoid hard lockup")
Link: https://lkml.kernel.org/r/20191004001243.140897-1-xueweiz@google.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The EAS wake-up path computes the system energy for several CPU
candidates: the CPU with maximum spare capacity in each performance
domain, and the prev_cpu. However, if prev_cpu also happens to be the
CPU with maximum spare capacity in its performance domain, the energy
calculation is still done twice, unnecessarily.
Add a condition to filter out this corner case before doing the energy
calculation.
Reported-by: Pavan Kondeti <pkondeti@codeaurora.org>
Signed-off-by: Quentin Perret <qperret@qperret.net>
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>
Cc: dietmar.eggemann@arm.com
Cc: juri.lelli@redhat.com
Cc: morten.rasmussen@arm.com
Cc: qais.yousef@arm.com
Cc: rjw@rjwysocki.net
Cc: tkjos@google.com
Cc: valentin.schneider@arm.com
Cc: vincent.guittot@linaro.org
Fixes: eb92692b25 ("sched/fair: Speed-up energy-aware wake-ups")
Link: https://lkml.kernel.org/r/20190920094115.GA11503@qperret.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
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>
cfs_rq_clock_task() was first introduced and used in:
f1b17280ef ("sched: Maintain runnable averages across throttled periods")
Over time its use has been graduately removed by the following commits:
d31b1a66cb ("sched/fair: Factorize PELT update")
2312729688 ("sched/fair: Update scale invariance of PELT")
Today, there is no single user left, so it can be safely removed.
Found via the -Wunused-function build warning.
Signed-off-by: Qian Cai <cai@lca.pw>
Cc: Ben Segall <bsegall@google.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/1568668775-2127-1-git-send-email-cai@lca.pw
[ Rewrote the changelog. ]
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>
