The CFS wakeup code will only ever go through EAS / its fast path on
"regular" wakeups (i.e. not on forks or execs). These are currently gated
by a check against 'sd_flag', which would be SD_BALANCE_WAKE at wakeup.
However, we now have a flag that explicitly tells us whether a wakeup is a
"regular" one, so hinge those conditions on that flag instead.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20201102184514.2733-4-valentin.schneider@arm.com
Only select_task_rq_fair() uses that parameter to do an actual domain
search, other classes only care about what kind of wakeup is happening
(fork, exec, or "regular") and thus just translate the flag into a wakeup
type.
WF_TTWU and WF_EXEC have just been added, use these along with WF_FORK to
encode the wakeup types we care about. For select_task_rq_fair(), we can
simply use the shiny new WF_flag : SD_flag mapping.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20201102184514.2733-3-valentin.schneider@arm.com
Since ab93a4bc95 ("sched/fair: Remove distribute_running fromCFS
bandwidth"), there is nothing to protect between
raw_spin_lock_irqsave/store() in do_sched_cfs_slack_timer().
Signed-off-by: Hui Su <sh_def@163.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Link: https://lkml.kernel.org/r/20201030144621.GA96974@rlk
During fast wakeup path, scheduler always check whether local or prev
cpus are good candidates for the task before looking for other cpus in
the domain. With commit b7a331615d ("sched/fair: Add asymmetric CPU
capacity wakeup scan") the heterogenous system gains a dedicated path
but doesn't try to reuse prev cpu whenever possible. If the previous
cpu is idle and belong to the LLC domain, we should check it 1st
before looking for another cpu because it stays one of the best
candidate and this also stabilizes task placement on the system.
This change aligns asymmetric path behavior with symmetric one and reduces
cases where the task migrates across all cpus of the sd_asym_cpucapacity
domains at wakeup.
This change does not impact normal EAS mode but only the overloaded case or
when EAS is not used.
- On hikey960 with performance governor (EAS disable)
./perf bench sched pipe -T -l 50000
mainline w/ patch
# migrations 999364 0
ops/sec 149313(+/-0.28%) 182587(+/- 0.40) +22%
- On hikey with performance governor
./perf bench sched pipe -T -l 50000
mainline w/ patch
# migrations 0 0
ops/sec 47721(+/-0.76%) 47899(+/- 0.56) +0.4%
According to test on hikey, the patch doesn't impact symmetric system
compared to current implementation (only tested on arm64)
Also read the uclamped value of task's utilization at most twice instead
instead each time we compare task's utilization with cpu's capacity.
Fixes: b7a331615d ("sched/fair: Add asymmetric CPU capacity wakeup scan")
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20201029161824.26389-1-vincent.guittot@linaro.org
schbench shows latency increase for 95 percentile above since:
commit 0b0695f2b3 ("sched/fair: Rework load_balance()")
Align the behavior of the load balancer with the wake up path, which tries
to select an idle CPU which belongs to the LLC for a waking task.
calculate_imbalance() will use nr_running instead of the spare
capacity when CPUs share resources (ie cache) at the domain level. This
will ensure a better spread of tasks on idle CPUs.
Running schbench on a hikey (8cores arm64) shows the problem:
tip/sched/core :
schbench -m 2 -t 4 -s 10000 -c 1000000 -r 10
Latency percentiles (usec)
50.0th: 33
75.0th: 45
90.0th: 51
95.0th: 4152
*99.0th: 14288
99.5th: 14288
99.9th: 14288
min=0, max=14276
tip/sched/core + patch :
schbench -m 2 -t 4 -s 10000 -c 1000000 -r 10
Latency percentiles (usec)
50.0th: 34
75.0th: 47
90.0th: 52
95.0th: 78
*99.0th: 94
99.5th: 94
99.9th: 94
min=0, max=94
Fixes: 0b0695f2b3 ("sched/fair: Rework load_balance()")
Reported-by: Chris Mason <clm@fb.com>
Suggested-by: Rik van Riel <riel@surriel.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Rik van Riel <riel@surriel.com>
Tested-by: Rik van Riel <riel@surriel.com>
Link: https://lkml.kernel.org/r/20201102102457.28808-1-vincent.guittot@linaro.org
As commit:
39f23ce07b ("sched/fair: Fix unthrottle_cfs_rq() for leaf_cfs_rq list")
does in unthrottle_cfs_rq(), throttle_cfs_rq() can also use the same
pattern as dequeue_task_fair().
No functional changes.
Signed-off-by: Peng Wang <rocking@linux.alibaba.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Phil Auld <pauld@redhat.com>
Cc: Ben Segall <bsegall@google.com>
Link: https://lore.kernel.org/r/f11dd2e3ab35cc538e2eb57bf0c99b6eaffce127.1604973978.git.rocking@linux.alibaba.com
In the case of a thread wakeup, wake_affine determines whether a core
will be chosen for the thread on the socket where the thread ran
previously or on the socket of the waker. This is done primarily by
comparing the load of the core where th thread ran previously (prev)
and the load of the waker (this).
commit 11f10e5420 ("sched/fair: Use load instead of runnable load
in wakeup path") changed the load computation from the runnable load
to the load average, where the latter includes the load of threads
that have already blocked on the core.
When a short-running daemon processes happens to run on prev, this
change raised the situation that prev could appear to have a greater
load than this, even when prev is actually idle. When prev and this
are on the same socket, the idle prev is detected later, in
select_idle_sibling. But if that does not hold, prev is completely
ignored, causing the waking thread to move to the socket of the waker.
In the case of N mostly active threads on N cores, this triggers other
migrations and hurts performance.
In contrast, before commit 11f10e5420, the load on an idle core
was 0, and in the case of a non-idle waker core, the effect of
wake_affine was to select prev as the target for searching for a core
for the waking thread.
To avoid unnecessary migrations, extend wake_affine_idle to check
whether the core where the thread previously ran is currently idle,
and if so simply return that core as the target.
[1] commit 11f10e5420 ("sched/fair: Use load instead of runnable
load in wakeup path")
This particularly has an impact when using the ondemand power manager,
where kworkers run every 0.004 seconds on all cores, increasing the
likelihood that an idle core will be considered to have a load.
The following numbers were obtained with the benchmarking tool
hyperfine (https://github.com/sharkdp/hyperfine) on the NAS parallel
benchmarks (https://www.nas.nasa.gov/publications/npb.html). The
tests were run on an 80-core Intel(R) Xeon(R) CPU E7-8870 v4 @
2.10GHz. Active (intel_pstate) and passive (intel_cpufreq) power
management were used. Times are in seconds. All experiments use all
160 hardware threads.
v5.9/intel-pstate v5.9+patch/intel-pstate
bt.C.c 24.725724+-0.962340 23.349608+-1.607214
lu.C.x 29.105952+-4.804203 25.249052+-5.561617
sp.C.x 31.220696+-1.831335 30.227760+-2.429792
ua.C.x 26.606118+-1.767384 25.778367+-1.263850
v5.9/ondemand v5.9+patch/ondemand
bt.C.c 25.330360+-1.028316 23.544036+-1.020189
lu.C.x 35.872659+-4.872090 23.719295+-3.883848
sp.C.x 32.141310+-2.289541 29.125363+-0.872300
ua.C.x 29.024597+-1.667049 25.728888+-1.539772
On the smaller data sets (A and B) and on the other NAS benchmarks
there is no impact on performance.
This also has a major impact on the splash2x.volrend benchmark of the
parsec benchmark suite that goes from 1m25 without this patch to 0m45,
in active (intel_pstate) mode.
Fixes: 11f10e5420 ("sched/fair: Use load instead of runnable load in wakeup path")
Signed-off-by: Julia Lawall <Julia.Lawall@inria.fr>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by Vincent Guittot <vincent.guittot@linaro.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Link: https://lkml.kernel.org/r/1603372550-14680-1-git-send-email-Julia.Lawall@inria.fr
Florian reported that all of kernel/sched/ is rebuild when
CONFIG_BLK_DEV_INITRD is changed, which, while not a bug is
unexpected. This is due to us including vmlinux.lds.h.
Jakub explained that the problem is that we put the alignment
requirement on the type instead of on a variable. Type alignment is a
minimum, the compiler is free to pick any larger alignment for a
specific instance of the type (eg. the variable).
So force the type alignment on all individual variable definitions and
remove the undesired dependency on vmlinux.lds.h.
Fixes: 85c2ce9104 ("sched, vmlinux.lds: Increase STRUCT_ALIGNMENT to 64 bytes for GCC-4.9")
Reported-by: Florian Fainelli <f.fainelli@gmail.com>
Suggested-by: Jakub Jelinek <jakub@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
It is possible for find_new_ilb() to select the current CPU, however,
this only happens from newidle balancing, in which case need_resched()
will be true, and consequently nohz_csd_func() will not trigger the
softirq.
Exclude the current CPU from becoming an ILB target.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
When the sched_schedstat changes from 0 to 1, some sched se maybe
already in the runqueue, the se->statistics.wait_start will be 0.
So it will let the (rq_of(cfs_rq)) - se->statistics.wait_start)
wrong. We need to avoid this scenario.
Signed-off-by: jun qian <qianjun.kernel@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Yafang Shao <laoar.shao@gmail.com>
Link: https://lkml.kernel.org/r/20201015064846.19809-1-qianjun.kernel@gmail.com
Use a more generic form for __section that requires quotes to avoid
complications with clang and gcc differences.
Remove the quote operator # from compiler_attributes.h __section macro.
Convert all unquoted __section(foo) uses to quoted __section("foo").
Also convert __attribute__((section("foo"))) uses to __section("foo")
even if the __attribute__ has multiple list entry forms.
Conversion done using the script at:
https://lore.kernel.org/lkml/75393e5ddc272dc7403de74d645e6c6e0f4e70eb.camel@perches.com/2-convert_section.pl
Signed-off-by: Joe Perches <joe@perches.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@gooogle.com>
Reviewed-by: Miguel Ojeda <ojeda@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A previous commit changed the notification mode from true/false to an
int, allowing notify-no, notify-yes, or signal-notify. This was
backwards compatible in the sense that any existing true/false user
would translate to either 0 (on notification sent) or 1, the latter
which mapped to TWA_RESUME. TWA_SIGNAL was assigned a value of 2.
Clean this up properly, and define a proper enum for the notification
mode. Now we have:
- TWA_NONE. This is 0, same as before the original change, meaning no
notification requested.
- TWA_RESUME. This is 1, same as before the original change, meaning
that we use TIF_NOTIFY_RESUME.
- TWA_SIGNAL. This uses TIF_SIGPENDING/JOBCTL_TASK_WORK for the
notification.
Clean up all the callers, switching their 0/1/false/true to using the
appropriate TWA_* mode for notifications.
Fixes: e91b481623 ("task_work: teach task_work_add() to do signal_wake_up()")
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
rq->cpu_capacity is a key element in several scheduler parts, such as EAS
task placement and load balancing. Tracking this value enables testing
and/or debugging by a toolkit.
Signed-off-by: Vincent Donnefort <vincent.donnefort@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/1598605249-72651-1-git-send-email-vincent.donnefort@arm.com
Currently, pick_next_entity(...) has the following structure
(simplified):
[...]
if (last_buddy_ok())
result = last_buddy;
if (next_buddy_ok())
result = next_buddy;
[...]
The intended behavior is to prefer next buddy over last buddy;
the current code somewhat obfuscates this, and also wastes
cycles checking the last buddy when eventually the next buddy is
picked up.
So this patch refactors two 'ifs' above into
[...]
if (next_buddy_ok())
result = next_buddy;
else if (last_buddy_ok())
result = last_buddy;
[...]
Signed-off-by: Peter Oskolkov <posk@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guitttot@linaro.org>
Link: https://lkml.kernel.org/r/20200930173532.1069092-1-posk@google.com
sched domains tend to trigger simultaneously the load balance loop but
the larger domains often need more time to collect statistics. This
slowness makes the larger domain trying to detach tasks from a rq whereas
tasks already migrated somewhere else at a sub-domain level. This is not
a real problem for idle LB because the period of smaller domains will
increase with its CPUs being busy and this will let time for higher ones
to pulled tasks. But this becomes a problem when all CPUs are already busy
because all domains stay synced when they trigger their LB.
A simple way to minimize simultaneous LB of all domains is to decrement the
the busy interval by 1 jiffies. Because of the busy_factor, the interval of
larger domain will not be a multiple of smaller ones anymore.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200921072424.14813-4-vincent.guittot@linaro.org
Some UCs like 9 always running tasks on 8 CPUs can't be balanced and the
load balancer currently migrates the waiting task between the CPUs in an
almost random manner. The success of a rq pulling a task depends of the
value of nr_balance_failed of its domains and its ability to be faster
than others to detach it. This behavior results in an unfair distribution
of the running time between tasks because some CPUs will run most of the
time, if not always, the same task whereas others will share their time
between several tasks.
Instead of using nr_balance_failed as a boolean to relax the condition
for detaching task, the LB will use nr_balanced_failed to relax the
threshold between the tasks'load and the imbalance. This mecanism
prevents the same rq or domain to always win the load balance fight.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200921072424.14813-2-vincent.guittot@linaro.org
In the file fair.c, sometims update_tg_load_avg(cfs_rq, 0) is used,
sometimes update_tg_load_avg(cfs_rq, false) is used.
update_tg_load_avg() has the parameter force, but in current code,
it never set 1 or true to it, so remove the force parameter.
Signed-off-by: Xianting Tian <tian.xianting@h3c.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200924014755.36253-1-tian.xianting@h3c.com
We've met problems that occasionally tasks with full cpumask
(e.g. by putting it into a cpuset or setting to full affinity)
were migrated to our isolated cpus in production environment.
After some analysis, we found that it is due to the current
select_idle_smt() not considering the sched_domain mask.
Steps to reproduce on my 31-CPU hyperthreads machine:
1. with boot parameter: "isolcpus=domain,2-31"
(thread lists: 0,16 and 1,17)
2. cgcreate -g cpu:test; cgexec -g cpu:test "test_threads"
3. some threads will be migrated to the isolated cpu16~17.
Fix it by checking the valid domain mask in select_idle_smt().
Fixes: 10e2f1acd0 ("sched/core: Rewrite and improve select_idle_siblings())
Reported-by: Wetp Zhang <wetp.zy@linux.alibaba.com>
Signed-off-by: Xunlei Pang <xlpang@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Jiang Biao <benbjiang@tencent.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/1600930127-76857-1-git-send-email-xlpang@linux.alibaba.com
Use runnable_avg to classify numa node state similarly to what is done for
normal load balancer. This helps to ensure that numa and normal balancers
use the same view of the state of the system.
Large arm64system: 2 nodes / 224 CPUs:
hackbench -l (256000/#grp) -g #grp
grp tip/sched/core +patchset improvement
1 14,008(+/- 4,99 %) 13,800(+/- 3.88 %) 1,48 %
4 4,340(+/- 5.35 %) 4.283(+/- 4.85 %) 1,33 %
16 3,357(+/- 0.55 %) 3.359(+/- 0.54 %) -0,06 %
32 3,050(+/- 0.94 %) 3.039(+/- 1,06 %) 0,38 %
64 2.968(+/- 1,85 %) 3.006(+/- 2.92 %) -1.27 %
128 3,290(+/-12.61 %) 3,108(+/- 5.97 %) 5.51 %
256 3.235(+/- 3.95 %) 3,188(+/- 2.83 %) 1.45 %
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Mel Gorman <mgorman@suse.de>
Link: https://lkml.kernel.org/r/20200921072959.16317-1-vincent.guittot@linaro.org
The code in reweight_entity() can be simplified.
For a sched entity on the rq, the entity accounting can be replaced by
cfs_rq instantaneous load updates currently called from within the
entity accounting.
Even though an entity on the rq can't represent a task in
reweight_entity() (a task is always dequeued before calling this
function) and so the numa task accounting and the rq->cfs_tasks list
management of the entity accounting are never called, the redundant
cfs_rq->nr_running decrement/increment will be avoided.
Signed-off-by: Jiang Biao <benbjiang@tencent.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20200811113209.34057-1-benbjiang@tencent.com
In find_energy_efficient_cpu() 'cpu_cap' could be less that 'util'.
It might be because of RT, DL (so higher sched class than CFS), irq or
thermal pressure signal, which reduce the capacity value.
In such situation the result of 'cpu_cap - util' might be negative but
stored in the unsigned long. Then it might be compared with other unsigned
long when uclamp_rq_util_with() reduced the 'util' such that is passes the
fits_capacity() check.
Prevent this situation and make the arithmetic more safe.
Fixes: 1d42509e47 ("sched/fair: Make EAS wakeup placement consider uclamp restrictions")
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20200810083004.26420-1-lukasz.luba@arm.com
SMT siblings share caches, so cache hotness should be irrelevant for
cross-sibling migration.
Signed-off-by: Josh Don <joshdon@google.com>
Proposed-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200804193413.510651-1-joshdon@google.com
- Make the Energy Model cover non-CPU devices (Lukasz Luba).
- Add Ice Lake server idle states table to the intel_idle driver
and eliminate a redundant static variable from it (Chen Yu,
Rafael Wysocki).
- Eliminate all W=1 build warnings from cpufreq (Lee Jones).
- Add support for Sapphire Rapids and for Power Limit 4 to the
Intel RAPL power capping driver (Sumeet Pawnikar, Zhang Rui).
- Fix function name in kerneldoc comments in the idle_inject power
capping driver (Yangtao Li).
- Fix locking issues with cpufreq governors and drop a redundant
"weak" function definition from cpufreq (Viresh Kumar).
- Rearrange cpufreq to register non-modular governors at the
core_initcall level and allow the default cpufreq governor to
be specified in the kernel command line (Quentin Perret).
- Extend, fix and clean up the intel_pstate driver (Srinivas
Pandruvada, Rafael Wysocki):
* Add a new sysfs attribute for disabling/enabling CPU
energy-efficiency optimizations in the processor.
* Make the driver avoid enabling HWP if EPP is not supported.
* Allow the driver to handle numeric EPP values in the sysfs
interface and fix the setting of EPP via sysfs in the active
mode.
* Eliminate a static checker warning and clean up a kerneldoc
comment.
- Clean up some variable declarations in the powernv cpufreq
driver (Wei Yongjun).
- Fix up the ->enter_s2idle callback definition to cover the case
when it points to the same function as ->idle correctly (Neal
Liu).
- Rearrange and clean up the PSCI cpuidle driver (Ulf Hansson).
- Make the PM core emit "changed" uevent when adding/removing the
"wakeup" sysfs attribute of devices (Abhishek Pandit-Subedi).
- Add a helper macro for declaring PM callbacks and use it in the
MMC jz4740 driver (Paul Cercueil).
- Fix white space in some places in the hibernate code and make the
system-wide PM code use "const char *" where appropriate (Xiang
Chen, Alexey Dobriyan).
- Add one more "unsafe" helper macro to the freezer to cover the NFS
use case (He Zhe).
- Change the language in the generic PM domains framework to use
parent/child terminology and clean up a typo and some comment
fromatting in that code (Kees Cook, Geert Uytterhoeven).
- Update the operating performance points OPP framework (Lukasz
Luba, Andrew-sh.Cheng, Valdis Kletnieks):
* Refactor dev_pm_opp_of_register_em() and update related drivers.
* Add a missing function export.
* Allow disabled OPPs in dev_pm_opp_get_freq().
- Update devfreq core and drivers (Chanwoo Choi, Lukasz Luba, Enric
Balletbo i Serra, Dmitry Osipenko, Kieran Bingham, Marc Zyngier):
* Add support for delayed timers to the devfreq core and make the
Samsung exynos5422-dmc driver use it.
* Unify sysfs interface to use "df-" as a prefix in instance names
consistently.
* Fix devfreq_summary debugfs node indentation.
* Add the rockchip,pmu phandle to the rk3399_dmc driver DT
bindings.
* List Dmitry Osipenko as the Tegra devfreq driver maintainer.
* Fix typos in the core devfreq code.
- Update the pm-graph utility to version 5.7 including a number of
fixes related to suspend-to-idle (Todd Brandt).
- Fix coccicheck errors and warnings in the cpupower utility (Shuah
Khan).
- Replace HTTP links with HTTPs ones in multiple places (Alexander
A. Klimov).
-----BEGIN PGP SIGNATURE-----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=8cUO
-----END PGP SIGNATURE-----
Merge tag 'pm-5.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull power management updates from Rafael Wysocki:
"The most significant change here is the extension of the Energy Model
to cover non-CPU devices (as well as CPUs) from Lukasz Luba.
There is also some new hardware support (Ice Lake server idle states
table for intel_idle, Sapphire Rapids and Power Limit 4 support in the
RAPL driver), some new functionality in the existing drivers (eg. a
new switch to disable/enable CPU energy-efficiency optimizations in
intel_pstate, delayed timers in devfreq), some assorted fixes (cpufreq
core, intel_pstate, intel_idle) and cleanups (eg. cpuidle-psci,
devfreq), including the elimination of W=1 build warnings from cpufreq
done by Lee Jones.
Specifics:
- Make the Energy Model cover non-CPU devices (Lukasz Luba).
- Add Ice Lake server idle states table to the intel_idle driver and
eliminate a redundant static variable from it (Chen Yu, Rafael
Wysocki).
- Eliminate all W=1 build warnings from cpufreq (Lee Jones).
- Add support for Sapphire Rapids and for Power Limit 4 to the Intel
RAPL power capping driver (Sumeet Pawnikar, Zhang Rui).
- Fix function name in kerneldoc comments in the idle_inject power
capping driver (Yangtao Li).
- Fix locking issues with cpufreq governors and drop a redundant
"weak" function definition from cpufreq (Viresh Kumar).
- Rearrange cpufreq to register non-modular governors at the
core_initcall level and allow the default cpufreq governor to be
specified in the kernel command line (Quentin Perret).
- Extend, fix and clean up the intel_pstate driver (Srinivas
Pandruvada, Rafael Wysocki):
* Add a new sysfs attribute for disabling/enabling CPU
energy-efficiency optimizations in the processor.
* Make the driver avoid enabling HWP if EPP is not supported.
* Allow the driver to handle numeric EPP values in the sysfs
interface and fix the setting of EPP via sysfs in the active
mode.
* Eliminate a static checker warning and clean up a kerneldoc
comment.
- Clean up some variable declarations in the powernv cpufreq driver
(Wei Yongjun).
- Fix up the ->enter_s2idle callback definition to cover the case
when it points to the same function as ->idle correctly (Neal Liu).
- Rearrange and clean up the PSCI cpuidle driver (Ulf Hansson).
- Make the PM core emit "changed" uevent when adding/removing the
"wakeup" sysfs attribute of devices (Abhishek Pandit-Subedi).
- Add a helper macro for declaring PM callbacks and use it in the MMC
jz4740 driver (Paul Cercueil).
- Fix white space in some places in the hibernate code and make the
system-wide PM code use "const char *" where appropriate (Xiang
Chen, Alexey Dobriyan).
- Add one more "unsafe" helper macro to the freezer to cover the NFS
use case (He Zhe).
- Change the language in the generic PM domains framework to use
parent/child terminology and clean up a typo and some comment
fromatting in that code (Kees Cook, Geert Uytterhoeven).
- Update the operating performance points OPP framework (Lukasz Luba,
Andrew-sh.Cheng, Valdis Kletnieks):
* Refactor dev_pm_opp_of_register_em() and update related drivers.
* Add a missing function export.
* Allow disabled OPPs in dev_pm_opp_get_freq().
- Update devfreq core and drivers (Chanwoo Choi, Lukasz Luba, Enric
Balletbo i Serra, Dmitry Osipenko, Kieran Bingham, Marc Zyngier):
* Add support for delayed timers to the devfreq core and make the
Samsung exynos5422-dmc driver use it.
* Unify sysfs interface to use "df-" as a prefix in instance
names consistently.
* Fix devfreq_summary debugfs node indentation.
* Add the rockchip,pmu phandle to the rk3399_dmc driver DT
bindings.
* List Dmitry Osipenko as the Tegra devfreq driver maintainer.
* Fix typos in the core devfreq code.
- Update the pm-graph utility to version 5.7 including a number of
fixes related to suspend-to-idle (Todd Brandt).
- Fix coccicheck errors and warnings in the cpupower utility (Shuah
Khan).
- Replace HTTP links with HTTPs ones in multiple places (Alexander A.
Klimov)"
* tag 'pm-5.9-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (71 commits)
cpuidle: ACPI: fix 'return' with no value build warning
cpufreq: intel_pstate: Fix EPP setting via sysfs in active mode
cpufreq: intel_pstate: Rearrange the storing of new EPP values
intel_idle: Customize IceLake server support
PM / devfreq: Fix the wrong end with semicolon
PM / devfreq: Fix indentaion of devfreq_summary debugfs node
PM / devfreq: Clean up the devfreq instance name in sysfs attr
memory: samsung: exynos5422-dmc: Add module param to control IRQ mode
memory: samsung: exynos5422-dmc: Adjust polling interval and uptreshold
memory: samsung: exynos5422-dmc: Use delayed timer as default
PM / devfreq: Add support delayed timer for polling mode
dt-bindings: devfreq: rk3399_dmc: Add rockchip,pmu phandle
PM / devfreq: tegra: Add Dmitry as a maintainer
PM / devfreq: event: Fix trivial spelling
PM / devfreq: rk3399_dmc: Fix kernel oops when rockchip,pmu is absent
cpuidle: change enter_s2idle() prototype
cpuidle: psci: Prevent domain idlestates until consumers are ready
cpuidle: psci: Convert PM domain to platform driver
cpuidle: psci: Fix error path via converting to a platform driver
cpuidle: psci: Fail cpuidle registration if set OSI mode failed
...
* pm-em:
OPP: refactor dev_pm_opp_of_register_em() and update related drivers
Documentation: power: update Energy Model description
PM / EM: change name of em_pd_energy to em_cpu_energy
PM / EM: remove em_register_perf_domain
PM / EM: add support for other devices than CPUs in Energy Model
PM / EM: update callback structure and add device pointer
PM / EM: introduce em_dev_register_perf_domain function
PM / EM: change naming convention from 'capacity' to 'performance'
* pm-core:
mmc: jz4740: Use pm_ptr() macro
PM: Make *_DEV_PM_OPS macros use __maybe_unused
PM: core: introduce pm_ptr() macro
In slow path, when selecting idlest group, if both groups have type
group_has_spare, only idle_cpus count gets compared.
As a result, if multiple tasks are created in a tight loop,
and go back to sleep immediately
(while waiting for all tasks to be created),
they may be scheduled on the same core, because CPU is back to idle
when the new fork happen.
For example:
sudo perf record -e sched:sched_wakeup_new -- \
sysbench threads --threads=4 run
...
total number of events: 61582
...
sudo perf script
sysbench 129378 [006] 74586.633466: sched:sched_wakeup_new:
sysbench:129380 [120] success=1 CPU:007
sysbench 129378 [006] 74586.634718: sched:sched_wakeup_new:
sysbench:129381 [120] success=1 CPU:007
sysbench 129378 [006] 74586.635957: sched:sched_wakeup_new:
sysbench:129382 [120] success=1 CPU:007
sysbench 129378 [006] 74586.637183: sched:sched_wakeup_new:
sysbench:129383 [120] success=1 CPU:007
This may have negative impact on performance for workloads with frequent
creation of multiple threads.
In this patch we are using group_util to select idlest group if both groups
have equal number of idle_cpus. Comparing the number of idle cpu is
not enough in this case, because the newly forked thread sleeps
immediately and before we select the cpu for the next one.
This is shown in the trace where the same CPU7 is selected for
all wakeup_new events.
That's why, looking at utilization when there is the same number of
CPU is a good way to see where the previous task was placed. Using
nr_running doesn't solve the problem because the newly forked task is not
running and the cpu would not have been idle in this case and an idle
CPU would have been selected instead.
With this patch newly created tasks would be better distributed.
With this patch:
sudo perf record -e sched:sched_wakeup_new -- \
sysbench threads --threads=4 run
...
total number of events: 74401
...
sudo perf script
sysbench 129455 [006] 75232.853257: sched:sched_wakeup_new:
sysbench:129457 [120] success=1 CPU:008
sysbench 129455 [006] 75232.854489: sched:sched_wakeup_new:
sysbench:129458 [120] success=1 CPU:009
sysbench 129455 [006] 75232.855732: sched:sched_wakeup_new:
sysbench:129459 [120] success=1 CPU:010
sysbench 129455 [006] 75232.856980: sched:sched_wakeup_new:
sysbench:129460 [120] success=1 CPU:011
We tested this patch with following benchmarks:
master: 'commit b3a9e3b962 ("Linux 5.8-rc1")'
100 iterations of: perf bench -f simple futex wake -s -t 128 -w 1
Lower result is better
| | BASELINE | +PATCH | DELTA (%) |
|---------|------------|----------|-------------|
| mean | 0.33 | 0.313 | +5.152 |
| std (%) | 10.433 | 7.563 | |
100 iterations of: sysbench threads --threads=8 run
Higher result is better
| | BASELINE | +PATCH | DELTA (%) |
|---------|------------|----------|-------------|
| mean | 5235.02 | 5863.73 | +12.01 |
| std (%) | 8.166 | 10.265 | |
100 iterations of: sysbench mutex --mutex-num=1 --threads=8 run
Lower result is better
| | BASELINE | +PATCH | DELTA (%) |
|---------|------------|----------|-------------|
| mean | 0.413 | 0.404 | +2.179 |
| std (%) | 3.791 | 1.816 | |
Signed-off-by: Peter Puhov <peter.puhov@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200714125941.4174-1-peter.puhov@linaro.org
task_h_load() can return 0 in some situations like running stress-ng
mmapfork, which forks thousands of threads, in a sched group on a 224 cores
system. The load balance doesn't handle this correctly because
env->imbalance never decreases and it will stop pulling tasks only after
reaching loop_max, which can be equal to the number of running tasks of
the cfs. Make sure that imbalance will be decreased by at least 1.
misfit task is the other feature that doesn't handle correctly such
situation although it's probably more difficult to face the problem
because of the smaller number of CPUs and running tasks on heterogenous
system.
We can't simply ensure that task_h_load() returns at least one because it
would imply to handle underflow in other places.
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>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: <stable@vger.kernel.org> # v4.4+
Link: https://lkml.kernel.org/r/20200710152426.16981-1-vincent.guittot@linaro.org
Add a bare tracepoint trace_sched_update_nr_running_tp which tracks
->nr_running CPU's rq. This is used to accurately trace this data and
provide a visualization of scheduler imbalances in, for example, the
form of a heat map. The tracepoint is accessed by loading an external
kernel module. An example module (forked from Qais' module and including
the pelt related tracepoints) can be found at:
https://github.com/auldp/tracepoints-helpers.git
A script to turn the trace-cmd report output into a heatmap plot can be
found at:
https://github.com/jirvoz/plot-nr-running
The tracepoints are added to add_nr_running() and sub_nr_running() which
are in kernel/sched/sched.h. In order to avoid CREATE_TRACE_POINTS in
the header a wrapper call is used and the trace/events/sched.h include
is moved before sched.h in kernel/sched/core.
Signed-off-by: Phil Auld <pauld@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200629192303.GC120228@lorien.usersys.redhat.com
Some performance regression on reaim benchmark have been raised with
commit 070f5e860e ("sched/fair: Take into account runnable_avg to classify group")
The problem comes from the init value of runnable_avg which is initialized
with max value. This can be a problem if the newly forked task is finally
a short task because the group of CPUs is wrongly set to overloaded and
tasks are pulled less agressively.
Set initial value of runnable_avg equals to util_avg to reflect that there
is no waiting time so far.
Fixes: 070f5e860e ("sched/fair: Take into account runnable_avg to classify group")
Reported-by: kernel test robot <rong.a.chen@intel.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200624154422.29166-1-vincent.guittot@linaro.org
While looking at enqueue_task_fair and dequeue_task_fair, it occurred
to me that dequeue_task_fair can also be optimized as Vincent described
in commit 7d148be69e ("sched/fair: Optimize enqueue_task_fair()").
When encountering throttled cfs_rq, dequeue_throttle label can ensure
se not to be NULL, and rq->nr_running remains unchanged, so we can also
skip the early balance check.
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/701eef9a40de93dcf5fe7063fd607bca5db38e05.1592287263.git.rocking@linux.alibaba.com
Now that the sched_class descriptors are defined in order via the linker
script vmlinux.lds.h, there's no reason to have a "next" pointer to the
previous priroity structure. The order of the sturctures can be aligned as
an array, and used to index and find the next sched_class descriptor.
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20191219214558.845353593@goodmis.org
In order to make a micro optimization in pick_next_task(), the order of the
sched class descriptor address must be in the same order as their priority
to each other. That is:
&idle_sched_class < &fair_sched_class < &rt_sched_class <
&dl_sched_class < &stop_sched_class
In order to guarantee this order of the sched class descriptors, add each
one into their own data section and force the order in the linker script.
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/157675913272.349305.8936736338884044103.stgit@localhost.localdomain
Energy Model framework now supports other devices than CPUs. Refactor some
of the functions in order to prevent wrong usage. The old function
em_pd_energy has to generic name. It must not be used without proper
cpumask pointer, which is possible only for CPU devices. Thus, rename it
and add proper description to warn of potential wrong usage for other
devices.
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Quentin Perret <qperret@google.com>
Signed-off-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Factorize in a single place the calculation of the divider to be used to
to compute *_avg from *_sum value
Suggested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200612154703.23555-1-vincent.guittot@linaro.org
With commit:
'b7031a02ec75 ("sched/fair: Add NOHZ_STATS_KICK")'
rebalance_domains of the local cfs_rq happens before others idle cpus have
updated nohz.next_balance and its value is overwritten.
Move the update of nohz.next_balance for other idles cpus before balancing
and updating the next_balance of local cfs_rq.
Also, the nohz.next_balance is now updated only if all idle cpus got a
chance to rebalance their domains and the idle balance has not been aborted
because of new activities on the CPU. In case of need_resched, the idle
load balance will be kick the next jiffie in order to address remaining
ilb.
Fixes: b7031a02ec ("sched/fair: Add NOHZ_STATS_KICK")
Reported-by: Peng Liu <iwtbavbm@gmail.com>
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>
Acked-by: Mel Gorman <mgorman@suse.de>
Link: https://lkml.kernel.org/r/20200609123748.18636-1-vincent.guittot@linaro.org
The util_est signals are key elements for EAS task placement and
frequency selection. Having tracepoints to track these signals enables
load-tracking and schedutil testing and/or debugging by a toolkit.
Signed-off-by: Vincent Donnefort <vincent.donnefort@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/1590597554-370150-1-git-send-email-vincent.donnefort@arm.com
Since commit 8ec59c0f5f ("sched/topology: Remove unused 'sd'
parameter from arch_scale_cpu_capacity()") it is no longer needed.
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20200603080304.16548-5-dietmar.eggemann@arm.com
Commit 6d1cafd8b5 ("sched: Resched proper CPU on yield_to()") moved
the code to resched the CPU from yield_to_task_fair() to yield_to()
making the preempt parameter in sched_class->yield_to_task()
unnecessary. Remove it. No other sched_class implements yield_to_task().
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200603080304.16548-3-dietmar.eggemann@arm.com
Pull networking updates from David Miller:
1) Allow setting bluetooth L2CAP modes via socket option, from Luiz
Augusto von Dentz.
2) Add GSO partial support to igc, from Sasha Neftin.
3) Several cleanups and improvements to r8169 from Heiner Kallweit.
4) Add IF_OPER_TESTING link state and use it when ethtool triggers a
device self-test. From Andrew Lunn.
5) Start moving away from custom driver versions, use the globally
defined kernel version instead, from Leon Romanovsky.
6) Support GRO vis gro_cells in DSA layer, from Alexander Lobakin.
7) Allow hard IRQ deferral during NAPI, from Eric Dumazet.
8) Add sriov and vf support to hinic, from Luo bin.
9) Support Media Redundancy Protocol (MRP) in the bridging code, from
Horatiu Vultur.
10) Support netmap in the nft_nat code, from Pablo Neira Ayuso.
11) Allow UDPv6 encapsulation of ESP in the ipsec code, from Sabrina
Dubroca. Also add ipv6 support for espintcp.
12) Lots of ReST conversions of the networking documentation, from Mauro
Carvalho Chehab.
13) Support configuration of ethtool rxnfc flows in bcmgenet driver,
from Doug Berger.
14) Allow to dump cgroup id and filter by it in inet_diag code, from
Dmitry Yakunin.
15) Add infrastructure to export netlink attribute policies to
userspace, from Johannes Berg.
16) Several optimizations to sch_fq scheduler, from Eric Dumazet.
17) Fallback to the default qdisc if qdisc init fails because otherwise
a packet scheduler init failure will make a device inoperative. From
Jesper Dangaard Brouer.
18) Several RISCV bpf jit optimizations, from Luke Nelson.
19) Correct the return type of the ->ndo_start_xmit() method in several
drivers, it's netdev_tx_t but many drivers were using
'int'. From Yunjian Wang.
20) Add an ethtool interface for PHY master/slave config, from Oleksij
Rempel.
21) Add BPF iterators, from Yonghang Song.
22) Add cable test infrastructure, including ethool interfaces, from
Andrew Lunn. Marvell PHY driver is the first to support this
facility.
23) Remove zero-length arrays all over, from Gustavo A. R. Silva.
24) Calculate and maintain an explicit frame size in XDP, from Jesper
Dangaard Brouer.
25) Add CAP_BPF, from Alexei Starovoitov.
26) Support terse dumps in the packet scheduler, from Vlad Buslov.
27) Support XDP_TX bulking in dpaa2 driver, from Ioana Ciornei.
28) Add devm_register_netdev(), from Bartosz Golaszewski.
29) Minimize qdisc resets, from Cong Wang.
30) Get rid of kernel_getsockopt and kernel_setsockopt in order to
eliminate set_fs/get_fs calls. From Christoph Hellwig.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (2517 commits)
selftests: net: ip_defrag: ignore EPERM
net_failover: fixed rollback in net_failover_open()
Revert "tipc: Fix potential tipc_aead refcnt leak in tipc_crypto_rcv"
Revert "tipc: Fix potential tipc_node refcnt leak in tipc_rcv"
vmxnet3: allow rx flow hash ops only when rss is enabled
hinic: add set_channels ethtool_ops support
selftests/bpf: Add a default $(CXX) value
tools/bpf: Don't use $(COMPILE.c)
bpf, selftests: Use bpf_probe_read_kernel
s390/bpf: Use bcr 0,%0 as tail call nop filler
s390/bpf: Maintain 8-byte stack alignment
selftests/bpf: Fix verifier test
selftests/bpf: Fix sample_cnt shared between two threads
bpf, selftests: Adapt cls_redirect to call csum_level helper
bpf: Add csum_level helper for fixing up csum levels
bpf: Fix up bpf_skb_adjust_room helper's skb csum setting
sfc: add missing annotation for efx_ef10_try_update_nic_stats_vf()
crypto/chtls: IPv6 support for inline TLS
Crypto/chcr: Fixes a coccinile check error
Crypto/chcr: Fixes compilations warnings
...
xdp_umem.c had overlapping changes between the 64-bit math fix
for the calculation of npgs and the removal of the zerocopy
memory type which got rid of the chunk_size_nohdr member.
The mlx5 Kconfig conflict is a case where we just take the
net-next copy of the Kconfig entry dependency as it takes on
the ESWITCH dependency by one level of indirection which is
what the 'net' conflicting change is trying to ensure.
Signed-off-by: David S. Miller <davem@davemloft.net>
In preparation of removing rq->wake_list, replace the
!list_empty(rq->wake_list) with rq->ttwu_pending. This is not fully
equivalent as this new variable is racy.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200526161908.070399698@infradead.org
The recent commit: 90b5363acd ("sched: Clean up scheduler_ipi()")
got smp_call_function_single_async() subtly wrong. Even though it will
return -EBUSY when trying to re-use a csd, that condition is not
atomic and still requires external serialization.
The change in kick_ilb() got this wrong.
While on first reading kick_ilb() has an atomic test-and-set that
appears to serialize the use, the matching 'release' is not in the
right place to actually guarantee this serialization.
Rework the nohz_idle_balance() trigger so that the release is in the
IPI callback and thus guarantees the required serialization for the
CSD.
Fixes: 90b5363acd ("sched: Clean up scheduler_ipi()")
Reported-by: Qian Cai <cai@lca.pw>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Cc: mgorman@techsingularity.net
Link: https://lore.kernel.org/r/20200526161907.778543557@infradead.org
Stefano reported a crash with using SQPOLL with io_uring:
BUG: kernel NULL pointer dereference, address: 00000000000003b0
CPU: 2 PID: 1307 Comm: io_uring-sq Not tainted 5.7.0-rc7 #11
RIP: 0010:task_numa_work+0x4f/0x2c0
Call Trace:
task_work_run+0x68/0xa0
io_sq_thread+0x252/0x3d0
kthread+0xf9/0x130
ret_from_fork+0x35/0x40
which is task_numa_work() oopsing on current->mm being NULL.
The task work is queued by task_tick_numa(), which checks if current->mm is
NULL at the time of the call. But this state isn't necessarily persistent,
if the kthread is using use_mm() to temporarily adopt the mm of a task.
Change the task_tick_numa() check to exclude kernel threads in general,
as it doesn't make sense to attempt ot balance for kthreads anyway.
Reported-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Link: https://lore.kernel.org/r/865de121-8190-5d30-ece5-3b097dc74431@kernel.dk
The MSCC bug fix in 'net' had to be slightly adjusted because the
register accesses are done slightly differently in net-next.
Signed-off-by: David S. Miller <davem@davemloft.net>
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.
This issue was found with the help of Coccinelle.
[1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
[2] https://github.com/KSPP/linux/issues/21
[3] commit 7649773293 ("cxgb3/l2t: Fix undefined behaviour")
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200507192141.GA16183@embeddedor
update_tg_cfs_*() propagate the impact of the attach/detach of an entity
down into the cfs_rq hierarchy and must keep the sync with the current pelt
window.
Even if we can't sync child cfs_rq and its group se, we can sync the group
se and its parent cfs_rq with current position in the PELT window. In fact,
we must keep them sync in order to stay also synced with others entities
and group entities that are already attached to the cfs_rq.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200506155301.14288-1-vincent.guittot@linaro.org
enqueue_task_fair jumps to enqueue_throttle label when cfs_rq_of(se) is
throttled which means that se can't be NULL in such case and we can move
the label after the if (!se) statement. Futhermore, the latter can be
removed because se is always NULL when reaching this point.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200513135502.4672-1-vincent.guittot@linaro.org
Although not exactly identical, unthrottle_cfs_rq() and enqueue_task_fair()
are quite close and follow the same sequence for enqueuing an entity in the
cfs hierarchy. Modify unthrottle_cfs_rq() to use the same pattern as
enqueue_task_fair(). This fixes a problem already faced with the latter and
add an optimization in the last for_each_sched_entity loop.
Fixes: fe61468b2c (sched/fair: Fix enqueue_task_fair warning)
Reported-by Tao Zhou <zohooouoto@zoho.com.cn>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Ben Segall <bsegall@google.com>
Link: https://lkml.kernel.org/r/20200513135528.4742-1-vincent.guittot@linaro.org
sched/fair: Fix enqueue_task_fair warning some more
The recent patch, fe61468b2c (sched/fair: Fix enqueue_task_fair warning)
did not fully resolve the issues with the rq->tmp_alone_branch !=
&rq->leaf_cfs_rq_list warning in enqueue_task_fair. There is a case where
the first for_each_sched_entity loop exits due to on_rq, having incompletely
updated the list. In this case the second for_each_sched_entity loop can
further modify se. The later code to fix up the list management fails to do
what is needed because se does not point to the sched_entity which broke out
of the first loop. The list is not fixed up because the throttled parent was
already added back to the list by a task enqueue in a parallel child hierarchy.
Address this by calling list_add_leaf_cfs_rq if there are throttled parents
while doing the second for_each_sched_entity loop.
Fixes: fe61468b2c ("sched/fair: Fix enqueue_task_fair warning")
Suggested-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Phil Auld <pauld@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20200512135222.GC2201@lorien.usersys.redhat.com
The scheduler IPI has grown weird and wonderful over the years, time
for spring cleaning.
Move all the non-trivial stuff out of it and into a regular smp function
call IPI. This then reduces the schedule_ipi() to most of it's former NOP
glory and ensures to keep the interrupt vector lean and mean.
Aside of that avoiding the full irq_enter() in the x86 IPI implementation
is incorrect as scheduler_ipi() can be instrumented. To work around that
scheduler_ipi() had an irq_enter/exit() hack when heavy work was
pending. This is gone now.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Alexandre Chartre <alexandre.chartre@oracle.com>
Link: https://lkml.kernel.org/r/20200505134058.361859938@linutronix.de
The code is executed with preemption(and interrupts) disabled,
so it's safe to use __this_cpu_write().
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200421144123.33580-1-songmuchun@bytedance.com
Function sched_init_granularity() is only called from __init
functions, so mark it __init as well.
Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Link: https://lkml.kernel.org/r/20200406074750.56533-1-songmuchun@bytedance.com
In order to prevent possible hardlockup of sched_cfs_period_timer()
loop, loop count is introduced to denote whether to scale quota and
period or not. However, scale is done between forwarding period timer
and refilling cfs bandwidth runtime, which means that period timer is
forwarded with old "period" while runtime is refilled with scaled
"quota".
Move do_sched_cfs_period_timer() before scaling to solve this.
Fixes: 2e8e192263 ("sched/fair: Limit sched_cfs_period_timer() loop to avoid hard lockup")
Signed-off-by: Huaixin Chang <changhuaixin@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Ben Segall <bsegall@google.com>
Reviewed-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200420024421.22442-3-changhuaixin@linux.alibaba.com
After Commit 6e2df0581f ("sched: Fix pick_next_task() vs 'change'
pattern race"), there is no need to expose newidle_balance() as it
is only used within fair.c file. Change this function back to static again.
No functional change.
Reported-by: kbuild test robot <lkp@intel.com>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/83cd3030b031ca5d646cd5e225be10e7a0fdd8f5.1587464698.git.yu.c.chen@intel.com
The SD_LOAD_BALANCE flag is set unconditionally for all domains in
sd_init(). By making the sched_domain->flags syctl interface read-only, we
have removed the last piece of code that could clear that flag - as such,
it will now be always present. Rather than to keep carrying it along, we
can work towards getting rid of it entirely.
cpusets don't need it because they can make CPUs be attached to the NULL
domain (e.g. cpuset with sched_load_balance=0), or to a partitioned
root_domain, i.e. a sched_domain hierarchy that doesn't span the entire
system (e.g. root cpuset with sched_load_balance=0 and sibling cpusets with
sched_load_balance=1).
isolcpus apply the same "trick": isolated CPUs are explicitly taken out of
the sched_domain rebuild (using housekeeping_cpumask()), so they get the
NULL domain treatment as well.
Remove the checks against SD_LOAD_BALANCE.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200415210512.805-4-valentin.schneider@arm.com
The last use of that parameter was removed by commit
57abff067a ("sched/fair: Rework find_idlest_group()")
Get rid of the parameter.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20200415210512.805-2-valentin.schneider@arm.com
We only consider group_balance_cpu() after there is no idle
cpu. So, just do comparison before return at these two cases.
Signed-off-by: Peng Wang <rocking@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/245c792f0e580b3ca342ad61257f4c066ee0f84f.1586594833.git.rocking@linux.alibaba.com
This is mostly a revert of commit:
baa9be4ffb ("sched/fair: Fix throttle_list starvation with low CFS quota")
The primary use of distribute_running was to determine whether to add
throttled entities to the head or the tail of the throttled list. Now
that we always add to the tail, we can remove this field.
The other use of distribute_running is in the slack_timer, so that we
don't start a distribution while one is already running. However, even
in the event that this race occurs, it is fine to have two distributions
running (especially now that distribute grabs the cfs_b->lock to
determine remaining quota before assigning).
Signed-off-by: Josh Don <joshdon@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Tested-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200410225208.109717-3-joshdon@google.com
There is a race window in which an entity begins throttling before quota
is added to the pool, but does not finish throttling until after we have
finished with distribute_cfs_runtime(). This entity is not observed by
distribute_cfs_runtime() because it was not on the throttled list at the
time that distribution was running. This race manifests as rare
period-length statlls for such entities.
Rather than heavy-weight the synchronization with the progress of
distribution, we can fix this by aborting throttling if bandwidth has
become available. Otherwise, we immediately add the entity to the
throttled list so that it can be observed by a subsequent distribution.
Additionally, we can remove the case of adding the throttled entity to
the head of the throttled list, and simply always add to the tail.
Thanks to 26a8b12747, distribute_cfs_runtime() no longer holds onto
its own pool of runtime. This means that if we do hit the !assign and
distribute_running case, we know that distribution is about to end.
Signed-off-by: Paul Turner <pjt@google.com>
Signed-off-by: Ben Segall <bsegall@google.com>
Signed-off-by: Josh Don <joshdon@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Link: https://lkml.kernel.org/r/20200410225208.109717-2-joshdon@google.com
Instead of having all the sysctl handlers deal with user pointers, which
is rather hairy in terms of the BPF interaction, copy the input to and
from userspace in common code. This also means that the strings are
always NUL-terminated by the common code, making the API a little bit
safer.
As most handler just pass through the data to one of the common handlers
a lot of the changes are mechnical.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Andrey Ignatov <rdna@fb.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
A negative imbalance value was observed after imbalance calculation,
this happens when the local sched group type is group_fully_busy,
and the average load of local group is greater than the selected
busiest group. Fix this problem by comparing the average load of the
local and busiest group before imbalance calculation formula.
Suggested-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Phil Auld <pauld@redhat.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Aubrey Li <aubrey.li@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/1585201349-70192-1-git-send-email-aubrey.li@intel.com
Currently, there is a potential race between distribute_cfs_runtime()
and assign_cfs_rq_runtime(). Race happens when cfs_b->runtime is read,
distributes without holding lock and finds out there is not enough
runtime to charge against after distribution. Because
assign_cfs_rq_runtime() might be called during distribution, and use
cfs_b->runtime at the same time.
Fibtest is the tool to test this race. Assume all gcfs_rq is throttled
and cfs period timer runs, slow threads might run and sleep, returning
unused cfs_rq runtime and keeping min_cfs_rq_runtime in their local
pool. If all this happens sufficiently quickly, cfs_b->runtime will drop
a lot. If runtime distributed is large too, over-use of runtime happens.
A runtime over-using by about 70 percent of quota is seen when we
test fibtest on a 96-core machine. We run fibtest with 1 fast thread and
95 slow threads in test group, configure 10ms quota for this group and
see the CPU usage of fibtest is 17.0%, which is far more than the
expected 10%.
On a smaller machine with 32 cores, we also run fibtest with 96
threads. CPU usage is more than 12%, which is also more than expected
10%. This shows that on similar workloads, this race do affect CPU
bandwidth control.
Solve this by holding lock inside distribute_cfs_runtime().
Fixes: c06f04c704 ("sched: Fix potential near-infinite distribute_cfs_runtime() loop")
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Huaixin Chang <changhuaixin@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/lkml/20200325092602.22471-1-changhuaixin@linux.alibaba.com/
sched/core.c uses update_avg() for rq->avg_idle and sched/fair.c uses an
open-coded version (with the exact same decay factor) for
rq->avg_scan_cost. On top of that, select_idle_cpu() expects to be able to
compare these two fields.
The only difference between the two is that rq->avg_scan_cost is computed
using a pure division rather than a shift. Turns out it actually matters,
first of all because the shifted value can be negative, and the standard
has this to say about it:
"""
The result of E1 >> E2 is E1 right-shifted E2 bit positions. [...] If E1
has a signed type and a negative value, the resulting value is
implementation-defined.
"""
Not only this, but (arithmetic) right shifting a negative value (using 2's
complement) is *not* equivalent to dividing it by the corresponding power
of 2. Let's look at a few examples:
-4 -> 0xF..FC
-4 >> 3 -> 0xF..FF == -1 != -4 / 8
-8 -> 0xF..F8
-8 >> 3 -> 0xF..FF == -1 == -8 / 8
-9 -> 0xF..F7
-9 >> 3 -> 0xF..FE == -2 != -9 / 8
Make update_avg() use a division, and export it to the private scheduler
header to reuse it where relevant. Note that this still lets compilers use
a shift here, but should prevent any unwanted surprise. The disassembly of
select_idle_cpu() remains unchanged on arm64, and ttwu_do_wakeup() gains 2
instructions; the diff sort of looks like this:
- sub x1, x1, x0
+ subs x1, x1, x0 // set condition codes
+ add x0, x1, #0x7
+ csel x0, x0, x1, mi // x0 = x1 < 0 ? x0 : x1
add x0, x3, x0, asr #3
which does the right thing (i.e. gives us the expected result while still
using an arithmetic shift)
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200330090127.16294-1-valentin.schneider@arm.com
In update_sg_wakeup_stats(), the comment says:
Computing avg_load makes sense only when group is fully
busy or overloaded.
But, the code below this comment does not check like this.
From reading the code about avg_load in other functions, I
confirm that avg_load should be calculated in fully busy or
overloaded case. The comment is correct and the checking
condition is wrong. So, change that condition.
Fixes: 57abff067a ("sched/fair: Rework find_idlest_group()")
Signed-off-by: Tao Zhou <ouwen210@hotmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Link: https://lkml.kernel.org/r/Message-ID:
During load_balancing, a group with spare capacity will try to pull some
utilizations from an overloaded group. In such case, the load balance
looks for the runqueue with the highest utilization. Nevertheless, it
should also ensure that there are some pending tasks to pull otherwise
the load balance will fail to pull a task and the spread of the load will
be delayed.
This situation is quite transient but it's possible to highlight the
effect with a short run of sysbench test so the time to spread task impacts
the global result significantly.
Below are the average results for 15 iterations on an arm64 octo core:
sysbench --test=cpu --num-threads=8 --max-requests=1000 run
tip/sched/core +patchset
total time: 172ms 158ms
per-request statistics:
avg: 1.337ms 1.244ms
max: 21.191ms 10.753ms
The average max doesn't fully reflect the wide spread of the value which
ranges from 1.350ms to more than 41ms for the tip/sched/core and from
1.350ms to 21ms with the patch.
Other factors like waiting for an idle load balance or cache hotness
can delay the spreading of the tasks which explains why we can still
have up to 21ms with the patch.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200312165429.990-1-vincent.guittot@linaro.org
When a cfs rq is throttled, the latter and its child are removed from the
leaf list but their nr_running is not changed which includes staying higher
than 1. When a task is enqueued in this throttled branch, the cfs rqs must
be added back in order to ensure correct ordering in the list but this can
only happens if nr_running == 1.
When cfs bandwidth is used, we call unconditionnaly list_add_leaf_cfs_rq()
when enqueuing an entity to make sure that the complete branch will be
added.
Similarly unthrottle_cfs_rq() can stop adding cfs in the list when a parent
is throttled. Iterate the remaining entity to ensure that the complete
branch will be added in the list.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
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>
Tested-by: Christian Borntraeger <borntraeger@de.ibm.com>
Tested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: stable@vger.kernel.org
Cc: stable@vger.kernel.org #v5.1+
Link: https://lkml.kernel.org/r/20200306135257.25044-1-vincent.guittot@linaro.org
Even when a cgroup is throttled, the group se of a child cgroup can still
be enqueued and its gse->on_rq stays true. When a task is enqueued on such
child, we still have to update the load_avg and increase
h_nr_running of the throttled cfs. Nevertheless, the 1st
for_each_sched_entity() loop is skipped because of gse->on_rq == true and the
2nd loop because the cfs is throttled whereas we have to update both
load_avg with the old h_nr_running and increase h_nr_running in such case.
The same sequence can happen during dequeue when se moves to parent before
breaking in the 1st loop.
Note that the update of load_avg will effectively happen only once in order
to sync up to the throttled time. Next call for updating load_avg will stop
early because the clock stays unchanged.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Fixes: 6d4d22468d ("sched/fair: Reorder enqueue/dequeue_task_fair path")
Link: https://lkml.kernel.org/r/20200306084208.12583-1-vincent.guittot@linaro.org
When a cfs_rq is throttled, its group entity is dequeued and its running
tasks are removed. We must update runnable_avg with the old h_nr_running
and update group_se->runnable_weight with the new h_nr_running at each
level of the hierarchy.
Reviewed-by: Ben Segall <bsegall@google.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Fixes: 9f68395333 ("sched/pelt: Add a new runnable average signal")
Link: https://lkml.kernel.org/r/20200227154115.8332-1-vincent.guittot@linaro.org
Qian Cai reported the following bug:
The linux-next commit ff7db0bf24 ("sched/numa: Prefer using an idle CPU as a
migration target instead of comparing tasks") introduced a boot warning,
[ 86.520534][ T1] WARNING: suspicious RCU usage
[ 86.520540][ T1] 5.6.0-rc3-next-20200227 #7 Not tainted
[ 86.520545][ T1] -----------------------------
[ 86.520551][ T1] kernel/sched/fair.c:5914 suspicious rcu_dereference_check() usage!
[ 86.520555][ T1]
[ 86.520555][ T1] other info that might help us debug this:
[ 86.520555][ T1]
[ 86.520561][ T1]
[ 86.520561][ T1] rcu_scheduler_active = 2, debug_locks = 1
[ 86.520567][ T1] 1 lock held by systemd/1:
[ 86.520571][ T1] #0: ffff8887f4b14848 (&mm->mmap_sem#2){++++}, at: do_page_fault+0x1d2/0x998
[ 86.520594][ T1]
[ 86.520594][ T1] stack backtrace:
[ 86.520602][ T1] CPU: 1 PID: 1 Comm: systemd Not tainted 5.6.0-rc3-next-20200227 #7
task_numa_migrate() checks for idle cores when updating NUMA-related statistics.
This relies on reading a RCU-protected structure in test_idle_cores() via this
call chain
task_numa_migrate
-> update_numa_stats
-> numa_idle_core
-> test_idle_cores
While the locking could be fine-grained, it is more appropriate to acquire
the RCU lock for the entire scan of the domain. This patch removes the
warning triggered at boot time.
Reported-by: Qian Cai <cai@lca.pw>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
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>
Fixes: ff7db0bf24 ("sched/numa: Prefer using an idle CPU as a migration target instead of comparing tasks")
Link: https://lkml.kernel.org/r/20200227191804.GJ3818@techsingularity.net
Building against the tip/sched/core as ff7db0bf24 ("sched/numa: Prefer
using an idle CPU as a migration target instead of comparing tasks") with
the arm64 defconfig (which doesn't have CONFIG_SCHED_SMT set) leads to:
kernel/sched/fair.c:1525:20: warning: 'test_idle_cores' declared 'static' but never defined [-Wunused-function]
static inline bool test_idle_cores(int cpu, bool def);
^~~~~~~~~~~~~~~
Rather than define it in its own CONFIG_SCHED_SMT #define island, bunch it
up with test_idle_cores().
Reported-by: Anshuman Khandual <anshuman.khandual@arm.com>
Reported-by: Naresh Kamboju <naresh.kamboju@linaro.org>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
[mgorman@techsingularity.net: Edit changelog, minor style change]
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
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>
Fixes: ff7db0bf24 ("sched/numa: Prefer using an idle CPU as a migration target instead of comparing tasks")
Link: https://lkml.kernel.org/r/20200303110258.1092-3-mgorman@techsingularity.net
Thermal pressure follows pelt signals which means the decay period for
thermal pressure is the default pelt decay period. Depending on SoC
characteristics and thermal activity, it might be beneficial to decay
thermal pressure slower, but still in-tune with the pelt signals. One way
to achieve this is to provide a command line parameter to set a decay
shift parameter to an integer between 0 and 10.
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200222005213.3873-10-thara.gopinath@linaro.org
cpu_capacity initially reflects the maximum possible capacity of a CPU.
Thermal pressure on a CPU means this maximum possible capacity is
unavailable due to thermal events. This patch subtracts the average
thermal pressure for a CPU from its maximum possible capacity so that
cpu_capacity reflects the remaining maximum capacity.
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200222005213.3873-8-thara.gopinath@linaro.org
Introduce support in scheduler periodic tick and other CFS bookkeeping
APIs to trigger the process of computing average thermal pressure for a
CPU. Also consider avg_thermal.load_avg in others_have_blocked which
allows for decay of pelt signals.
Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200222005213.3873-7-thara.gopinath@linaro.org
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
When domains are imbalanced or overloaded a search of all CPUs on the
target domain is searched and compared with task_numa_compare. In some
circumstances, a candidate is found that is an obvious win.
o A task can move to an idle CPU and an idle CPU is found
o A swap candidate is found that would move to its preferred domain
This patch terminates the search when either condition is met.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-14-mgorman@techsingularity.net
When swapping tasks for NUMA balancing, it is preferred that tasks move
to or remain on their preferred node. When considering an imbalance,
encourage tasks to move to their preferred node and discourage tasks from
moving away from their preferred node.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-13-mgorman@techsingularity.net
Multiple tasks can attempt to select and idle CPU but fail because
numa_migrate_on is already set and the migration fails. Instead of failing,
scan for an alternative idle CPU. select_idle_sibling is not used because
it requires IRQs to be disabled and it ignores numa_migrate_on allowing
multiple tasks to stack. This scan may still fail if there are idle
candidate CPUs due to races but if this occurs, it's best that a task
stay on an available CPU that move to a contended one.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-12-mgorman@techsingularity.net
task_numa_find_cpu() can scan a node multiple times. Minimally it scans to
gather statistics and later to find a suitable target. In some cases, the
second scan will simply pick an idle CPU if the load is not imbalanced.
This patch caches information on an idle core while gathering statistics
and uses it immediately if load is not imbalanced to avoid a second scan
of the node runqueues. Preference is given to an idle core rather than an
idle SMT sibling to avoid packing HT siblings due to linearly scanning the
node cpumask.
As a side-effect, even when the second scan is necessary, the importance
of using select_idle_sibling is much reduced because information on idle
CPUs is cached and can be reused.
Note that this patch actually makes is harder to move to an idle CPU
as multiple tasks can race for the same idle CPU due to a race checking
numa_migrate_on. This is addressed in the next patch.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-11-mgorman@techsingularity.net
Take into account the new runnable_avg signal to classify a group and to
mitigate the volatility of util_avg in face of intensive migration or
new task with random utilization.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: "Dietmar Eggemann <dietmar.eggemann@arm.com>"
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-10-mgorman@techsingularity.net
Now that runnable_load_avg has been removed, we can replace it by a new
signal that will highlight the runnable pressure on a cfs_rq. This signal
track the waiting time of tasks on rq and can help to better define the
state of rqs.
At now, only util_avg is used to define the state of a rq:
A rq with more that around 80% of utilization and more than 1 tasks is
considered as overloaded.
But the util_avg signal of a rq can become temporaly low after that a task
migrated onto another rq which can bias the classification of the rq.
When tasks compete for the same rq, their runnable average signal will be
higher than util_avg as it will include the waiting time and we can use
this signal to better classify cfs_rqs.
The new runnable_avg will track the runnable time of a task which simply
adds the waiting time to the running time. The runnable _avg of cfs_rq
will be the /Sum of se's runnable_avg and the runnable_avg of group entity
will follow the one of the rq similarly to util_avg.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: "Dietmar Eggemann <dietmar.eggemann@arm.com>"
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-9-mgorman@techsingularity.net
Now that runnable_load_avg is no more used, we can remove it to make
space for a new signal.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: "Dietmar Eggemann <dietmar.eggemann@arm.com>"
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-8-mgorman@techsingularity.net
The standard load balancer generally tries to keep the number of running
tasks or idle CPUs balanced between NUMA domains. The NUMA balancer allows
tasks to move if there is spare capacity but this causes a conflict and
utilisation between NUMA nodes gets badly skewed. This patch uses similar
logic between the NUMA balancer and load balancer when deciding if a task
migrating to its preferred node can use an idle CPU.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-7-mgorman@techsingularity.net
Similarly to what has been done for the normal load balancer, we can
replace runnable_load_avg by load_avg in numa load balancing and track the
other statistics like the utilization and the number of running tasks to
get to better view of the current state of a node.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: "Dietmar Eggemann <dietmar.eggemann@arm.com>"
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-6-mgorman@techsingularity.net
The walk through the cgroup hierarchy during the enqueue/dequeue of a task
is split in 2 distinct parts for throttled cfs_rq without any added value
but making code less readable.
Change the code ordering such that everything related to a cfs_rq
(throttled or not) will be done in the same loop.
In addition, the same steps ordering is used when updating a cfs_rq:
- update_load_avg
- update_cfs_group
- update *h_nr_running
This reordering enables the use of h_nr_running in PELT algorithm.
No functional and performance changes are expected and have been noticed
during tests.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: "Dietmar Eggemann <dietmar.eggemann@arm.com>"
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-5-mgorman@techsingularity.net
sched:sched_stick_numa is meant to fire when a task is unable to migrate
to the preferred node but from the trace, it's possibile to tell the
difference between "no CPU found", "migration to idle CPU failed" and
"tasks could not be swapped". Extend the tracepoint accordingly.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
[ Minor edits. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-4-mgorman@techsingularity.net
sched:sched_stick_numa is meant to fire when a task is unable to migrate
to the preferred node. The case where no candidate CPU could be found is
not traced which is an important gap. The tracepoint is not fired when
the task is not allowed to run on any CPU on the preferred node or the
task is already running on the target CPU but neither are interesting
corner cases.
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Valentin Schneider <valentin.schneider@arm.com>
Cc: Phil Auld <pauld@redhat.com>
Cc: Hillf Danton <hdanton@sina.com>
Link: https://lore.kernel.org/r/20200224095223.13361-3-mgorman@techsingularity.net
Capacity-awareness in the wake-up path previously involved disabling
wake_affine in certain scenarios. We have just made select_idle_sibling()
capacity-aware, so this isn't needed anymore.
Remove wake_cap() entirely.
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
[Changelog tweaks]
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[Changelog tweaks]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20200206191957.12325-5-valentin.schneider@arm.com
Issue
=====
On asymmetric CPU capacity topologies, we currently rely on wake_cap() to
drive select_task_rq_fair() towards either:
- its slow-path (find_idlest_cpu()) if either the previous or
current (waking) CPU has too little capacity for the waking task
- its fast-path (select_idle_sibling()) otherwise
Commit:
3273163c67 ("sched/fair: Let asymmetric CPU configurations balance at wake-up")
points out that this relies on the assumption that "[...]the CPU capacities
within an SD_SHARE_PKG_RESOURCES domain (sd_llc) are homogeneous".
This assumption no longer holds on newer generations of big.LITTLE
systems (DynamIQ), which can accommodate CPUs of different compute capacity
within a single LLC domain. To hopefully paint a better picture, a regular
big.LITTLE topology would look like this:
+---------+ +---------+
| L2 | | L2 |
+----+----+ +----+----+
|CPU0|CPU1| |CPU2|CPU3|
+----+----+ +----+----+
^^^ ^^^
LITTLEs bigs
which would result in the following scheduler topology:
DIE [ ] <- sd_asym_cpucapacity
MC [ ] [ ] <- sd_llc
0 1 2 3
Conversely, a DynamIQ topology could look like:
+-------------------+
| L3 |
+----+----+----+----+
| L2 | L2 | L2 | L2 |
+----+----+----+----+
|CPU0|CPU1|CPU2|CPU3|
+----+----+----+----+
^^^^^ ^^^^^
LITTLEs bigs
which would result in the following scheduler topology:
MC [ ] <- sd_llc, sd_asym_cpucapacity
0 1 2 3
What this means is that, on DynamIQ systems, we could pass the wake_cap()
test (IOW presume the waking task fits on the CPU capacities of some LLC
domain), thus go through select_idle_sibling().
This function operates on an LLC domain, which here spans both bigs and
LITTLEs, so it could very well pick a CPU of too small capacity for the
task, despite there being fitting idle CPUs - it very much depends on the
CPU iteration order, on which we have absolutely no guarantees
capacity-wise.
Implementation
==============
Introduce yet another select_idle_sibling() helper function that takes CPU
capacity into account. The policy is to pick the first idle CPU which is
big enough for the task (task_util * margin < cpu_capacity). If no
idle CPU is big enough, we pick the idle one with the highest capacity.
Unlike other select_idle_sibling() helpers, this one operates on the
sd_asym_cpucapacity sched_domain pointer, which is guaranteed to span all
known CPU capacities in the system. As such, this will work for both
"legacy" big.LITTLE (LITTLEs & bigs split at MC, joined at DIE) and for
newer DynamIQ systems (e.g. LITTLEs and bigs in the same MC domain).
Note that this limits the scope of select_idle_sibling() to
select_idle_capacity() for asymmetric CPU capacity systems - the LLC domain
will not be scanned, and no further heuristic will be applied.
Signed-off-by: Morten Rasmussen <morten.rasmussen@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Quentin Perret <qperret@google.com>
Link: https://lkml.kernel.org/r/20200206191957.12325-2-valentin.schneider@arm.com
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>
Currently we loop through all threads of a core to evaluate if the core is
idle or not. This is unnecessary. If a thread of a core is not idle, skip
evaluating other threads of a core. Also while clearing the cpumask, bits
of all CPUs of a core can be cleared in one-shot.
Collecting ticks on a Power 9 SMT 8 system around select_idle_core
while running schbench shows us
(units are in ticks, hence lesser is better)
Without patch
N Min Max Median Avg Stddev
x 130 151 1083 284 322.72308 144.41494
With patch
N Min Max Median Avg Stddev Improvement
x 164 88 610 201 225.79268 106.78943 30.03%
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
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>
Acked-by: Mel Gorman <mgorman@techsingularity.net>
Link: https://lkml.kernel.org/r/20191206172422.6578-1-srikar@linux.vnet.ibm.com
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
