The code in find_idlest_cpu() looks for the CPU with the smallest load.
However, if multiple CPUs are idle, the first idle CPU is selected
irrespective of the depth of its idle state.
Among the idle CPUs we should pick the one with with the shallowest idle
state, or the latest to have gone idle if all idle CPUs are in the same
state. The later applies even when cpuidle is configured out.
This patch doesn't cover the following issues:
- The idle exit latency of a CPU might be larger than the time needed
to migrate the waking task to an already running CPU with sufficient
capacity, and therefore performance would benefit from task packing
in such case (in most cases task packing is about power saving).
- Some idle states have a non negligible and non abortable entry latency
which needs to run to completion before the exit latency can start.
A concurrent patch series is making this info available to the cpuidle
core. Once available, the entry latency with the idle timestamp could
determine when the exit latency may be effective.
Those issues will be handled in due course. In the mean time, what
is implemented here should improve things already compared to the current
state of affairs.
Based on an initial patch from Daniel Lezcano.
Signed-off-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-pm@vger.kernel.org
Cc: linaro-kernel@lists.linaro.org
Link: http://lkml.kernel.org/n/tip-@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
When the cpu enters idle, it stores the cpuidle state pointer in its
struct rq instance which in turn could be used to make a better decision
when balancing tasks.
As soon as the cpu exits its idle state, the struct rq reference is
cleared.
There are a couple of situations where the idle state pointer could be changed
while it is being consulted:
1. For x86/acpi with dynamic c-states, when a laptop switches from battery
to AC that could result on removing the deeper idle state. The acpi driver
triggers:
'acpi_processor_cst_has_changed'
'cpuidle_pause_and_lock'
'cpuidle_uninstall_idle_handler'
'kick_all_cpus_sync'.
All cpus will exit their idle state and the pointed object will be set to
NULL.
2. The cpuidle driver is unloaded. Logically that could happen but not
in practice because the drivers are always compiled in and 95% of them are
not coded to unregister themselves. In any case, the unloading code must
call 'cpuidle_unregister_device', that calls 'cpuidle_pause_and_lock'
leading to 'kick_all_cpus_sync' as mentioned above.
A race can happen if we use the pointer and then one of these two scenarios
occurs at the same moment.
In order to be safe, the idle state pointer stored in the rq must be
used inside a rcu_read_lock section where we are protected with the
'rcu_barrier' in the 'cpuidle_uninstall_idle_handler' function. The
idle_get_state() and idle_put_state() accessors should be used to that
effect.
Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: "Rafael J. Wysocki" <rjw@rjwysocki.net>
Cc: linux-pm@vger.kernel.org
Cc: linaro-kernel@lists.linaro.org
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/n/tip-@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In wake_affine() I have tried to understand the meaning of the condition:
(this_load <= load &&
this_load + target_load(prev_cpu, idx) <= tl_per_task)
but I failed to find a use case that can take advantage of it and I haven't
found clear description in the previous commit's log.
Futhermore, the comment of the condition refers to the task_hot function that
was used before being replaced by the current condition:
/*
* This domain has SD_WAKE_AFFINE and
* p is cache cold in this domain, and
* there is no bad imbalance.
*/
If we look more deeply the below condition:
this_load + target_load(prev_cpu, idx) <= tl_per_task
When sync is clear, we have:
tl_per_task = runnable_load_avg / nr_running
this_load = max(runnable_load_avg, cpuload[idx])
target_load = max(runnable_load_avg', cpuload'[idx])
It implies that runnable_load_avg == 0 and nr_running <= 1 in order to match the
condition. This implies that runnable_load_avg == 0 too because of the
condition: this_load <= load.
but if this _load is null, 'balanced' is already set and the test is redundant.
If sync is set, it's not as straight forward as above (especially if cgroup
are involved) but the policy should be similar as we have removed a task that's
going to sleep in order to get a more accurate load and this_load values.
The current conclusion is that these additional condition don't give any benefit
so we can remove them.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: preeti@linux.vnet.ibm.com
Cc: riel@redhat.com
Cc: Morten.Rasmussen@arm.com
Cc: efault@gmx.de
Cc: nicolas.pitre@linaro.org
Cc: daniel.lezcano@linaro.org
Cc: dietmar.eggemann@arm.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1409051215-16788-3-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The imbalance flag can stay set whereas there is no imbalance.
Let assume that we have 3 tasks that run on a dual cores /dual cluster system.
We will have some idle load balance which are triggered during tick.
Unfortunately, the tick is also used to queue background work so we can reach
the situation where short work has been queued on a CPU which already runs a
task. The load balance will detect this imbalance (2 tasks on 1 CPU and an idle
CPU) and will try to pull the waiting task on the idle CPU. The waiting task is
a worker thread that is pinned on a CPU so an imbalance due to pinned task is
detected and the imbalance flag is set.
Then, we will not be able to clear the flag because we have at most 1 task on
each CPU but the imbalance flag will trig to useless active load balance
between the idle CPU and the busy CPU.
We need to reset of the imbalance flag as soon as we have reached a balanced
state. If all tasks are pinned, we don't consider that as a balanced state and
let the imbalance flag set.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Preeti U Murthy <preeti@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: riel@redhat.com
Cc: Morten.Rasmussen@arm.com
Cc: efault@gmx.de
Cc: nicolas.pitre@linaro.org
Cc: daniel.lezcano@linaro.org
Cc: dietmar.eggemann@arm.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1409051215-16788-2-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The code in task_numa_compare() will only examine at most one idle CPU per node,
because they all have the same score. However, some idle CPUs are better
candidates than others, due to busy or idle SMT siblings, etc...
The scheduler has logic to find the best CPU within an LLC to place a
task. The NUMA code should probably use it.
This seems to reduce the standard deviation for single instance SPECjbb2005
with a low warehouse count on my 4 node test system.
Signed-off-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: mgorman@suse.de
Cc: Mike Galbraith <umgwanakikbuti@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20140904163530.189d410a@cuia.bos.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Current code can fail to migrate a waking task (silently) when TTWU_QUEUE is
enabled.
When a task is waking, it is pending on the wake_list of the rq, but it is not
queued (task->on_rq == 0). In this case, set_cpus_allowed_ptr() and
__migrate_task() will not migrate it because its invisible to them.
This behavior is incorrect, because the task has been already woken, it will be
running on the wrong CPU without correct placement until the next wake-up or
update for cpus_allowed.
To fix this problem, we need to finish the wakeup (so they appear on
the runqueue) before we migrate them.
Reported-by: Sasha Levin <sasha.levin@oracle.com>
Reported-by: Jason J. Herne <jjherne@linux.vnet.ibm.com>
Tested-by: Jason J. Herne <jjherne@linux.vnet.ibm.com>
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/538ED7EB.5050303@cn.fujitsu.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull ACPI and power management fixes from Rafael Wysocki:
"These are regression fixes (ACPI sysfs, ACPI video, suspend test),
ACPI cpuidle deadlock fix, missing runtime validation of ACPI _DSD
output, a fix and a new CPU ID for the RAPL driver, new blacklist
entry for the ACPI EC driver and a couple of trivial cleanups
(intel_pstate and generic PM domains).
Specifics:
- Fix for recently broken test_suspend= command line argument (Rafael
Wysocki).
- Fixes for regressions related to the ACPI video driver caused by
switching the default to native backlight handling in 3.16 from
Hans de Goede.
- Fix for a sysfs attribute of ACPI device objects that returns stale
values sometimes due to the fact that they are cached instead of
executing the appropriate method (_SUN) every time (broken in
3.14). From Yasuaki Ishimatsu.
- Fix for a deadlock between cpuidle_lock and cpu_hotplug.lock in the
ACPI processor driver from Jiri Kosina.
- Runtime output validation for the ACPI _DSD device configuration
object missing from the support for it that has been introduced
recently. From Mika Westerberg.
- Fix for an unuseful and misleading RAPL (Running Average Power
Limit) domain detection message in the RAPL driver from Jacob Pan.
- New Intel Haswell CPU ID for the RAPL driver from Jason Baron.
- New Clevo W350etq blacklist entry for the ACPI EC driver from Lan
Tianyu.
- Cleanup for the intel_pstate driver and the core generic PM domains
code from Gabriele Mazzotta and Geert Uytterhoeven"
* tag 'pm+acpi-3.17-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
ACPI / cpuidle: fix deadlock between cpuidle_lock and cpu_hotplug.lock
ACPI / scan: not cache _SUN value in struct acpi_device_pnp
cpufreq: intel_pstate: Remove unneeded variable
powercap / RAPL: change domain detection message
powercap / RAPL: add support for CPU model 0x3f
PM / domains: Make generic_pm_domain.name const
PM / sleep: Fix test_suspend= command line option
ACPI / EC: Add msi quirk for Clevo W350etq
ACPI / video: Disable native_backlight on HP ENVY 15 Notebook PC
ACPI / video: Add a disable_native_backlight quirk
ACPI / video: Fix use_native_backlight selection logic
ACPICA: ACPI 5.1: Add support for runtime validation of _DSD package.
Pull RCU fix from Ingo Molnar:
"A boot hang fix for the offloaded callback RCU model (RCU_NOCB_CPU=y
&& (TREE_CPU=y || TREE_PREEMPT_RC)) in certain bootup scenarios"
* 'core-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
rcu: Make nocb leader kthreads process pending callbacks after spawning
Pull timer fixes from Thomas Gleixner:
"Three fixlets from the timer departement:
- Update the timekeeper before updating vsyscall and pvclock. This
fixes the kvm-clock regression reported by Chris and Paolo.
- Use the proper irq work interface from NMI. This fixes the
regression reported by Catalin and Dave.
- Clarify the compat_nanosleep error handling mechanism to avoid
future confusion"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
timekeeping: Update timekeeper before updating vsyscall and pvclock
compat: nanosleep: Clarify error handling
nohz: Restore NMI safe local irq work for local nohz kick
An overrun could happen in function start_hrtick_dl()
when a task with SCHED_DEADLINE runs in the microseconds
range.
For example, if a task with SCHED_DEADLINE has the following parameters:
Task runtime deadline period
P1 200us 500us 500us
The deadline and period from task P1 are less than 1ms.
In order to achieve microsecond precision, we need to enable HRTICK feature
by the next command:
PC#echo "HRTICK" > /sys/kernel/debug/sched_features
PC#trace-cmd record -e sched_switch &
PC#./schedtool -E -t 200000:500000:500000 -e ./test
The binary test is in an endless while(1) loop here.
Some pieces of trace.dat are as follows:
<idle>-0 157.603157: sched_switch: :R ==> 2481:4294967295: test
test-2481 157.603203: sched_switch: 2481:R ==> 0:120: swapper/2
<idle>-0 157.605657: sched_switch: :R ==> 2481:4294967295: test
test-2481 157.608183: sched_switch: 2481:R ==> 2483:120: trace-cmd
trace-cmd-2483 157.609656: sched_switch:2483:R==>2481:4294967295: test
We can get the runtime of P1 from the information above:
runtime = 157.608183 - 157.605657
runtime = 0.002526(2.526ms)
The correct runtime should be less than or equal to 200us at some point.
The problem is caused by a conditional judgment "delta > 10000"
in function start_hrtick_dl().
Because no hrtimer start up to control the rest of runtime
when the reset of runtime is less than 10us.
So the process will continue to run until tick-period is coming.
Move the code with the limit of the least time slice
from hrtick_start_fair() to hrtick_start() because the
EDF schedule class also needs this function in start_hrtick_dl().
To fix this problem, we call hrtimer_start() unconditionally in
start_hrtick_dl(), and make sure the scheduling slice won't be smaller
than 10us in hrtimer_start().
Signed-off-by: Xiaofeng Yan <xiaofeng.yan@huawei.com>
Reviewed-by: Li Zefan <lizefan@huawei.com>
Acked-by: Juri Lelli <juri.lelli@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1409022941-5880-1-git-send-email-xiaofeng.yan@huawei.com
[ Massaged the changelog and the code. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The update_walltime() code works on the shadow timekeeper to make the
seqcount protected region as short as possible. But that update to the
shadow timekeeper does not update all timekeeper fields because it's
sufficient to do that once before it becomes life. One of these fields
is tkr.base_mono. That stays stale in the shadow timekeeper unless an
operation happens which copies the real timekeeper to the shadow.
The update function is called after the update calls to vsyscall and
pvclock. While not correct, it did not cause any problems because none
of the invoked update functions used base_mono.
commit cbcf2dd3b3 (x86: kvm: Make kvm_get_time_and_clockread()
nanoseconds based) changed that in the kvm pvclock update function, so
the stale mono_base value got used and caused kvm-clock to malfunction.
Put the update where it belongs and fix the issue.
Reported-by: Chris J Arges <chris.j.arges@canonical.com>
Reported-by: Paolo Bonzini <pbonzini@redhat.com>
Cc: Gleb Natapov <gleb@kernel.org>
Cc: John Stultz <john.stultz@linaro.org>
Link: http://lkml.kernel.org/r/alpine.DEB.2.10.1409050000570.3333@nanos
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The error handling in compat_sys_nanosleep() is correct, but
completely non obvious. Document it and restrict it to the
-ERESTART_RESTARTBLOCK return value for clarity.
Reported-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
The use of "rcu_assign_pointer()" is NULLing out the pointer.
According to RCU_INIT_POINTER()'s block comment:
"1. This use of RCU_INIT_POINTER() is NULLing out the pointer"
it is better to use it instead of rcu_assign_pointer() because it has a
smaller overhead.
The following Coccinelle semantic patch was used:
@@
@@
- rcu_assign_pointer
+ RCU_INIT_POINTER
(..., NULL)
Signed-off-by: Andreea-Cristina Bernat <bernat.ada@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: paulmck@linux.vnet.ibm.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20140822145043.GA580@ada
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The local nohz kick is currently used by perf which needs it to be
NMI-safe. Recent commit though (7d1311b93e)
changed its implementation to fire the local kick using the remote kick
API. It was convenient to make the code more generic but the remote kick
isn't NMI-safe.
As a result:
WARNING: CPU: 3 PID: 18062 at kernel/irq_work.c:72 irq_work_queue_on+0x11e/0x140()
CPU: 3 PID: 18062 Comm: trinity-subchil Not tainted 3.16.0+ #34
0000000000000009 00000000903774d1 ffff880244e06c00 ffffffff9a7f1e37
0000000000000000 ffff880244e06c38 ffffffff9a0791dd ffff880244fce180
0000000000000003 ffff880244e06d58 ffff880244e06ef8 0000000000000000
Call Trace:
<NMI> [<ffffffff9a7f1e37>] dump_stack+0x4e/0x7a
[<ffffffff9a0791dd>] warn_slowpath_common+0x7d/0xa0
[<ffffffff9a07930a>] warn_slowpath_null+0x1a/0x20
[<ffffffff9a17ca1e>] irq_work_queue_on+0x11e/0x140
[<ffffffff9a10a2c7>] tick_nohz_full_kick_cpu+0x57/0x90
[<ffffffff9a186cd5>] __perf_event_overflow+0x275/0x350
[<ffffffff9a184f80>] ? perf_event_task_disable+0xa0/0xa0
[<ffffffff9a01a4cf>] ? x86_perf_event_set_period+0xbf/0x150
[<ffffffff9a187934>] perf_event_overflow+0x14/0x20
[<ffffffff9a020386>] intel_pmu_handle_irq+0x206/0x410
[<ffffffff9a0b54d3>] ? arch_vtime_task_switch+0x63/0x130
[<ffffffff9a01937b>] perf_event_nmi_handler+0x2b/0x50
[<ffffffff9a007b72>] nmi_handle+0xd2/0x390
[<ffffffff9a007aa5>] ? nmi_handle+0x5/0x390
[<ffffffff9a0d131b>] ? lock_release+0xab/0x330
[<ffffffff9a008062>] default_do_nmi+0x72/0x1c0
[<ffffffff9a0c925f>] ? cpuacct_account_field+0xcf/0x200
[<ffffffff9a008268>] do_nmi+0xb8/0x100
Lets fix this by restoring the use of local irq work for the nohz local
kick.
Reported-by: Catalin Iacob <iacobcatalin@gmail.com>
Reported-and-tested-by: Dave Jones <davej@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Pull an RCU fix from Paul E. McKenney:
"This series contains a single commit fixing an initialization bug
reported by Amit Shah and fixed by Pranith Kumar (and tested by Amit).
This bug results in a boot-time hang in callback-offloaded configurations
where callbacks were posted before the offloading ('rcuo') kthreads
were created."
Signed-off-by: Ingo Molnar <mingo@kernel.org>
After commit d431cbc53c (PM / sleep: Simplify sleep states sysfs
interface code) the pm_states[] array is not populated initially,
which causes setup_test_suspend() to always fail and the suspend
testing during boot doesn't work any more.
Fix the problem by using pm_labels[] instead of pm_states[] in
setup_test_suspend() and storing a pointer to the label of the
sleep state to test rather than the number representing it,
because the connection between the state numbers and labels is
only established by suspend_set_ops().
Fixes: d431cbc53c (PM / sleep: Simplify sleep states sysfs interface code)
Reported-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Pull irq handling fixlet from Thomas Gleixner:
"Just an export for an interrupt flow handler which is now used in gpio
modules"
* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irq: Export handle_fasteoi_irq
