This extends SCHED_IDLE to cgroups.
Interface: cgroup/cpu.idle.
0: default behavior
1: SCHED_IDLE
Extending SCHED_IDLE to cgroups means that we incorporate the existing
aspects of SCHED_IDLE; a SCHED_IDLE cgroup will count all of its
descendant threads towards the idle_h_nr_running count of all of its
ancestor cgroups. Thus, sched_idle_rq() will work properly.
Additionally, SCHED_IDLE cgroups are configured with minimum weight.
There are two key differences between the per-task and per-cgroup
SCHED_IDLE interface:
- The cgroup interface allows tasks within a SCHED_IDLE hierarchy to
maintain their relative weights. The entity that is "idle" is the
cgroup, not the tasks themselves.
- Since the idle entity is the cgroup, our SCHED_IDLE wakeup preemption
decision is not made by comparing the current task with the woken
task, but rather by comparing their matching sched_entity.
A typical use-case for this is a user that creates an idle and a
non-idle subtree. The non-idle subtree will dominate competition vs
the idle subtree, but the idle subtree will still be high priority vs
other users on the system. The latter is accomplished via comparing
matching sched_entity in the waken preemption path (this could also be
improved by making the sched_idle_rq() decision dependent on the
perspective of a specific task).
For now, we maintain the existing SCHED_IDLE semantics. Future patches
may make improvements that extend how we treat SCHED_IDLE entities.
The per-task_group idle field is an integer that currently only holds
either a 0 or a 1. This is explicitly typed as an integer to allow for
further extensions to this API. For example, a negative value may
indicate a highly latency-sensitive cgroup that should be preferred
for preemption/placement/etc.
Signed-off-by: Josh Don <joshdon@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lore.kernel.org/r/20210730020019.1487127-2-joshdon@google.com
Eugene tripped over the case where rq_lock(), as called in a
for_each_possible_cpu() loop came apart because rq->core hadn't been
setup yet.
This is a somewhat unusual, but valid case.
Rework things such that rq->core is initialized to point at itself. IOW
initialize each CPU as a single threaded Core. CPU online will then join
the new CPU (thread) to an existing Core where needed.
For completeness sake, have CPU offline fully undo the state so as to
not presume the topology will match the next time it comes online.
Fixes: 9edeaea1bc ("sched: Core-wide rq->lock")
Reported-by: Eugene Syromiatnikov <esyr@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Josh Don <joshdon@google.com>
Tested-by: Eugene Syromiatnikov <esyr@redhat.com>
Link: https://lkml.kernel.org/r/YR473ZGeKqMs6kw+@hirez.programming.kicks-ass.net
It is possible for sched_getattr() to incorrectly report the state of
the reset_on_fork flag when called on a deadline task.
Indeed, if the flag was set on a deadline task using sched_setattr()
with flags (SCHED_FLAG_RESET_ON_FORK | SCHED_FLAG_KEEP_PARAMS), then
p->sched_reset_on_fork will be set, but __setscheduler() will bail out
early, which means that the dl_se->flags will not get updated by
__setscheduler_params()->__setparam_dl(). Consequently, if
sched_getattr() is then called on the task, __getparam_dl() will
override kattr.sched_flags with the now out-of-date copy in dl_se->flags
and report the stale value to userspace.
To fix this, make sure to only copy the flags that are relevant to
sched_deadline to and from the dl_se->flags field.
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20210727101103.2729607-2-qperret@google.com
When a task wakes up on an idle rq, uclamp_rq_util_with() would max
aggregate with rq value. But since there is no task enqueued yet, the
values are stale based on the last task that was running. When the new
task actually wakes up and enqueued, then the rq uclamp values should
reflect that of the newly woken up task effective uclamp values.
This is a problem particularly for uclamp_max because it default to
1024. If a task p with uclamp_max = 512 wakes up, then max aggregation
would ignore the capping that should apply when this task is enqueued,
which is wrong.
Fix that by ignoring max aggregation if the rq is idle since in that
case the effective uclamp value of the rq will be the ones of the task
that will wake up.
Fixes: 9d20ad7dfc ("sched/uclamp: Add uclamp_util_with()")
Signed-off-by: Xuewen Yan <xuewen.yan@unisoc.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
[qias: Changelog]
Reviewed-by: Qais Yousef <qais.yousef@arm.com>
Link: https://lore.kernel.org/r/20210630141204.8197-1-xuewen.yan94@gmail.com
Since commit '8a99b6833c88(sched: Move SCHED_DEBUG sysctl to debugfs)',
SCHED_DEBUG sysctls are moved to debugfs, so these extern sysctls in
include/linux/sched/sysctl.h are no longer needed for sysctl.c, even
some are no longer needed.
So move those extern sysctls that needed by kernel/sched/debug.c to
kernel/sched/sched.h, and remove others that are no longer needed.
Signed-off-by: Hailong Liu <liu.hailong6@zte.com.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210606115451.26745-1-liuhailongg6@163.com
The CFS bandwidth controller limits CPU requests of a task group to
quota during each period. However, parallel workloads might be bursty
so that they get throttled even when their average utilization is under
quota. And they are latency sensitive at the same time so that
throttling them is undesired.
We borrow time now against our future underrun, at the cost of increased
interference against the other system users. All nicely bounded.
Traditional (UP-EDF) bandwidth control is something like:
(U = \Sum u_i) <= 1
This guaranteeds both that every deadline is met and that the system is
stable. After all, if U were > 1, then for every second of walltime,
we'd have to run more than a second of program time, and obviously miss
our deadline, but the next deadline will be further out still, there is
never time to catch up, unbounded fail.
This work observes that a workload doesn't always executes the full
quota; this enables one to describe u_i as a statistical distribution.
For example, have u_i = {x,e}_i, where x is the p(95) and x+e p(100)
(the traditional WCET). This effectively allows u to be smaller,
increasing the efficiency (we can pack more tasks in the system), but at
the cost of missing deadlines when all the odds line up. However, it
does maintain stability, since every overrun must be paired with an
underrun as long as our x is above the average.
That is, suppose we have 2 tasks, both specify a p(95) value, then we
have a p(95)*p(95) = 90.25% chance both tasks are within their quota and
everything is good. At the same time we have a p(5)p(5) = 0.25% chance
both tasks will exceed their quota at the same time (guaranteed deadline
fail). Somewhere in between there's a threshold where one exceeds and
the other doesn't underrun enough to compensate; this depends on the
specific CDFs.
At the same time, we can say that the worst case deadline miss, will be
\Sum e_i; that is, there is a bounded tardiness (under the assumption
that x+e is indeed WCET).
The benefit of burst is seen when testing with schbench. Default value of
kernel.sched_cfs_bandwidth_slice_us(5ms) and CONFIG_HZ(1000) is used.
mkdir /sys/fs/cgroup/cpu/test
echo $$ > /sys/fs/cgroup/cpu/test/cgroup.procs
echo 100000 > /sys/fs/cgroup/cpu/test/cpu.cfs_quota_us
echo 100000 > /sys/fs/cgroup/cpu/test/cpu.cfs_burst_us
./schbench -m 1 -t 3 -r 20 -c 80000 -R 10
The average CPU usage is at 80%. I run this for 10 times, and got long tail
latency for 6 times and got throttled for 8 times.
Tail latencies are shown below, and it wasn't the worst case.
Latency percentiles (usec)
50.0000th: 19872
75.0000th: 21344
90.0000th: 22176
95.0000th: 22496
*99.0000th: 22752
99.5000th: 22752
99.9000th: 22752
min=0, max=22727
rps: 9.90 p95 (usec) 22496 p99 (usec) 22752 p95/cputime 28.12% p99/cputime 28.44%
The interferenece when using burst is valued by the possibilities for
missing the deadline and the average WCET. Test results showed that when
there many cgroups or CPU is under utilized, the interference is
limited. More details are shown in:
https://lore.kernel.org/lkml/5371BD36-55AE-4F71-B9D7-B86DC32E3D2B@linux.alibaba.com/
Co-developed-by: Shanpei Chen <shanpeic@linux.alibaba.com>
Signed-off-by: Shanpei Chen <shanpeic@linux.alibaba.com>
Co-developed-by: Tianchen Ding <dtcccc@linux.alibaba.com>
Signed-off-by: Tianchen Ding <dtcccc@linux.alibaba.com>
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>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20210621092800.23714-2-changhuaixin@linux.alibaba.com
This is a partial forward-port of Peter Ziljstra's work first posted
at:
https://lore.kernel.org/lkml/20180530142236.667774973@infradead.org/
Currently select_idle_cpu()'s proportional scheme uses the average idle
time *for when we are idle*, that is temporally challenged. When a CPU
is not at all idle, we'll happily continue using whatever value we did
see when the CPU goes idle. To fix this, introduce a separate average
idle and age it (the existing value still makes sense for things like
new-idle balancing, which happens when we do go idle).
The overall goal is to not spend more time scanning for idle CPUs than
we're idle for. Otherwise we're inhibiting work. This means that we need to
consider the cost over all the wake-ups between consecutive idle periods.
To track this, the scan cost is subtracted from the estimated average
idle time.
The impact of this patch is related to workloads that have domains that
are fully busy or overloaded. Without the patch, the scan depth may be
too high because a CPU is not reaching idle.
Due to the nature of the patch, this is a regression magnet. It
potentially wins when domains are almost fully busy or overloaded --
at that point searches are likely to fail but idle is not being aged
as CPUs are active so search depth is too large and useless. It will
potentially show regressions when there are idle CPUs and a deep search is
beneficial. This tbench result on a 2-socket broadwell machine partially
illustates the problem
5.13.0-rc2 5.13.0-rc2
vanilla sched-avgidle-v1r5
Hmean 1 445.02 ( 0.00%) 451.36 * 1.42%*
Hmean 2 830.69 ( 0.00%) 846.03 * 1.85%*
Hmean 4 1350.80 ( 0.00%) 1505.56 * 11.46%*
Hmean 8 2888.88 ( 0.00%) 2586.40 * -10.47%*
Hmean 16 5248.18 ( 0.00%) 5305.26 * 1.09%*
Hmean 32 8914.03 ( 0.00%) 9191.35 * 3.11%*
Hmean 64 10663.10 ( 0.00%) 10192.65 * -4.41%*
Hmean 128 18043.89 ( 0.00%) 18478.92 * 2.41%*
Hmean 256 16530.89 ( 0.00%) 17637.16 * 6.69%*
Hmean 320 16451.13 ( 0.00%) 17270.97 * 4.98%*
Note that 8 was a regression point where a deeper search would have helped
but it gains for high thread counts when searches are useless. Hackbench
is a more extreme example although not perfect as the tasks idle rapidly
hackbench-process-pipes
5.13.0-rc2 5.13.0-rc2
vanilla sched-avgidle-v1r5
Amean 1 0.3950 ( 0.00%) 0.3887 ( 1.60%)
Amean 4 0.9450 ( 0.00%) 0.9677 ( -2.40%)
Amean 7 1.4737 ( 0.00%) 1.4890 ( -1.04%)
Amean 12 2.3507 ( 0.00%) 2.3360 * 0.62%*
Amean 21 4.0807 ( 0.00%) 4.0993 * -0.46%*
Amean 30 5.6820 ( 0.00%) 5.7510 * -1.21%*
Amean 48 8.7913 ( 0.00%) 8.7383 ( 0.60%)
Amean 79 14.3880 ( 0.00%) 13.9343 * 3.15%*
Amean 110 21.2233 ( 0.00%) 19.4263 * 8.47%*
Amean 141 28.2930 ( 0.00%) 25.1003 * 11.28%*
Amean 172 34.7570 ( 0.00%) 30.7527 * 11.52%*
Amean 203 41.0083 ( 0.00%) 36.4267 * 11.17%*
Amean 234 47.7133 ( 0.00%) 42.0623 * 11.84%*
Amean 265 53.0353 ( 0.00%) 47.7720 * 9.92%*
Amean 296 60.0170 ( 0.00%) 53.4273 * 10.98%*
Stddev 1 0.0052 ( 0.00%) 0.0025 ( 51.57%)
Stddev 4 0.0357 ( 0.00%) 0.0370 ( -3.75%)
Stddev 7 0.0190 ( 0.00%) 0.0298 ( -56.64%)
Stddev 12 0.0064 ( 0.00%) 0.0095 ( -48.38%)
Stddev 21 0.0065 ( 0.00%) 0.0097 ( -49.28%)
Stddev 30 0.0185 ( 0.00%) 0.0295 ( -59.54%)
Stddev 48 0.0559 ( 0.00%) 0.0168 ( 69.92%)
Stddev 79 0.1559 ( 0.00%) 0.0278 ( 82.17%)
Stddev 110 1.1728 ( 0.00%) 0.0532 ( 95.47%)
Stddev 141 0.7867 ( 0.00%) 0.0968 ( 87.69%)
Stddev 172 1.0255 ( 0.00%) 0.0420 ( 95.91%)
Stddev 203 0.8106 ( 0.00%) 0.1384 ( 82.92%)
Stddev 234 1.1949 ( 0.00%) 0.1328 ( 88.89%)
Stddev 265 0.9231 ( 0.00%) 0.0820 ( 91.11%)
Stddev 296 1.0456 ( 0.00%) 0.1327 ( 87.31%)
Again, higher thread counts benefit and the standard deviation
shows that results are also a lot more stable when the idle
time is aged.
The patch potentially matters when a socket was multiple LLCs as the
maximum search depth is lower. However, some of the test results were
suspiciously good (e.g. specjbb2005 gaining 50% on a Zen1 machine) and
other results were not dramatically different to other mcahines.
Given the nature of the patch, Peter's full series is not being forward
ported as each part should stand on its own. Preferably they would be
merged at different times to reduce the risk of false bisections.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210615111611.GH30378@techsingularity.net
Make:
struct dl_rq::dl_nr_migratory
struct dl_rq::dl_nr_running
struct rt_rq::rt_nr_boosted
struct rt_rq::rt_nr_migratory
struct rt_rq::rt_nr_total
struct rq::nr_uninterruptible
32-bit.
If total number of tasks can't exceed 2**32 (and less due to futex pid
limits), then per-runqueue counters can't as well.
This patchset has been sponsored by REX Prefix Eradication Society.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20210422200228.1423391-4-adobriyan@gmail.com
In order to not have to use pid_struct, create a new, smaller,
structure to manage task cookies for core scheduling.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.919768100@infradead.org
- Don't migrate if there is a cookie mismatch
Load balance tries to move task from busiest CPU to the
destination CPU. When core scheduling is enabled, if the
task's cookie does not match with the destination CPU's
core cookie, this task may be skipped by this CPU. This
mitigates the forced idle time on the destination CPU.
- Select cookie matched idle CPU
In the fast path of task wakeup, select the first cookie matched
idle CPU instead of the first idle CPU.
- Find cookie matched idlest CPU
In the slow path of task wakeup, find the idlest CPU whose core
cookie matches with task's cookie
Signed-off-by: Aubrey Li <aubrey.li@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.860083871@infradead.org
When a sibling is forced-idle to match the core-cookie; search for
matching tasks to fill the core.
rcu_read_unlock() can incur an infrequent deadlock in
sched_core_balance(). Fix this by using the RCU-sched flavor instead.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.800048269@infradead.org
During force-idle, we end up doing cross-cpu comparison of vruntimes
during pick_next_task. If we simply compare (vruntime-min_vruntime)
across CPUs, and if the CPUs only have 1 task each, we will always
end up comparing 0 with 0 and pick just one of the tasks all the time.
This starves the task that was not picked. To fix this, take a snapshot
of the min_vruntime when entering force idle and use it for comparison.
This min_vruntime snapshot will only be used for cross-CPU vruntime
comparison, and nothing else.
A note about the min_vruntime snapshot and force idling:
During selection:
When we're not fi, we need to update snapshot.
when we're fi and we were not fi, we must update snapshot.
When we're fi and we were already fi, we must not update snapshot.
Which gives:
fib fi update
0 0 1
0 1 1
1 0 1
1 1 0
Where:
fi: force-idled now
fib: force-idled before
So the min_vruntime snapshot needs to be updated when: !(fib && fi).
Also, the cfs_prio_less() function needs to be aware of whether the
core is in force idle or not, since it will be use this information to
know whether to advance a cfs_rq's min_vruntime_fi in the hierarchy.
So pass this information along via pick_task() -> prio_less().
Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.738542617@infradead.org
If there is only one long running local task and the sibling is
forced idle, it might not get a chance to run until a schedule
event happens on any cpu in the core.
So we check for this condition during a tick to see if a sibling
is starved and then give it a chance to schedule.
Signed-off-by: Vineeth Pillai <viremana@linux.microsoft.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.617407840@infradead.org
Instead of only selecting a local task, select a task for all SMT
siblings for every reschedule on the core (irrespective which logical
CPU does the reschedule).
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.557559654@infradead.org
Introduce task_struct::core_cookie as an opaque identifier for core
scheduling. When enabled; core scheduling will only allow matching
task to be on the core; where idle matches everything.
When task_struct::core_cookie is set (and core scheduling is enabled)
these tasks are indexed in a second RB-tree, first on cookie value
then on scheduling function, such that matching task selection always
finds the most elegible match.
NOTE: *shudder* at the overhead...
NOTE: *sigh*, a 3rd copy of the scheduling function; the alternative
is per class tracking of cookies and that just duplicates a lot of
stuff for no raisin (the 2nd copy lives in the rt-mutex PI code).
[Joel: folded fixes]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.496975854@infradead.org
Because sched_class::pick_next_task() also implies
sched_class::set_next_task() (and possibly put_prev_task() and
newidle_balance) it is not state invariant. This makes it unsuitable
for remote task selection.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[Vineeth: folded fixes]
Signed-off-by: Vineeth Remanan Pillai <viremana@linux.microsoft.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.437092775@infradead.org
rq_lockp() includes a static_branch(), which is asm-goto, which is
asm volatile which defeats regular CSE. This means that:
if (!static_branch(&foo))
return simple;
if (static_branch(&foo) && cond)
return complex;
Doesn't fold and we get horrible code. Introduce __rq_lockp() without
the static_branch() on.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.316696988@infradead.org
Introduce the basic infrastructure to have a core wide rq->lock.
This relies on the rq->__lock order being in increasing CPU number
(inside a core). It is also constrained to SMT8 per lockdep (and
SMT256 per preempt_count).
Luckily SMT8 is the max supported SMT count for Linux (Mips, Sparc and
Power are known to have this).
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/YJUNfzSgptjX7tG6@hirez.programming.kicks-ass.net
When switching on core-sched, CPUs need to agree which lock to use for
their RQ.
The new rule will be that rq->core_enabled will be toggled while
holding all rq->__locks that belong to a core. This means we need to
double check the rq->core_enabled value after each lock acquire and
retry if it changed.
This also has implications for those sites that take multiple RQ
locks, they need to be careful that the second lock doesn't end up
being the first lock.
Verify the lock pointer after acquiring the first lock, because if
they're on the same core, holding any of the rq->__lock instances will
pin the core state.
While there, change the rq->__lock order to CPU number, instead of rq
address, this greatly simplifies the next patch.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/YJUNY0dmrJMD/BIm@hirez.programming.kicks-ass.net
In preparation of playing games with rq->lock, abstract the thing
using an accessor.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.136465446@infradead.org
In prepration for playing games with rq->lock, add some rq_lock
wrappers.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Don Hiatt <dhiatt@digitalocean.com>
Tested-by: Hongyu Ning <hongyu.ning@linux.intel.com>
Tested-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/20210422123308.075967879@infradead.org
CPU scheduler marks need_resched flag to signal a schedule() on a
particular CPU. But, schedule() may not happen immediately in cases
where the current task is executing in the kernel mode (no
preemption state) for extended periods of time.
This patch adds a warn_on if need_resched is pending for more than the
time specified in sysctl resched_latency_warn_ms. If it goes off, it is
likely that there is a missing cond_resched() somewhere. Monitoring is
done via the tick and the accuracy is hence limited to jiffy scale. This
also means that we won't trigger the warning if the tick is disabled.
This feature (LATENCY_WARN) is default disabled.
Signed-off-by: Paul Turner <pjt@google.com>
Signed-off-by: Josh Don <joshdon@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210416212936.390566-1-joshdon@google.com
CONFIG_SCHED_DEBUG is the build-time Kconfig knob, the boot param
sched_debug and the /debug/sched/debug_enabled knobs control the
sched_debug_enabled variable, but what they really do is make
SCHED_DEBUG more verbose, so rename the lot.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Move the #ifdef SCHED_DEBUG bits to kernel/sched/debug.c in order to
collect all the debugfs bits.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20210412102001.353833279@infradead.org
Stop polluting sysctl with undocumented knobs that really are debug
only, move them all to /debug/sched/ along with the existing
/debug/sched_* files that already exist.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Tested-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20210412102001.287610138@infradead.org
Use the new cpu_dying() state to simplify and fix the balance_push()
vs CPU hotplug rollback state.
Specifically, we currently rely on notifiers sched_cpu_dying() /
sched_cpu_activate() to terminate balance_push, however if the
cpu_down() fails when we're past sched_cpu_deactivate(), it should
terminate balance_push at that point and not wait until we hit
sched_cpu_activate().
Similarly, when cpu_up() fails and we're going back down, balance_push
should be active, where it currently is not.
So instead, make sure balance_push is enabled below SCHED_AP_ACTIVE
(when !cpu_active()), and gate it's utility with cpu_dying().
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/YHgAYef83VQhKdC2@hirez.programming.kicks-ass.net
Fix ~42 single-word typos in scheduler code comments.
We have accumulated a few fun ones over the years. :-)
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Segall <bsegall@google.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: linux-kernel@vger.kernel.org
A significant portion of __calc_delta() time is spent in the loop
shifting a u64 by 32 bits. Use `fls` instead of iterating.
This is ~7x faster on benchmarks.
The generic `fls` implementation (`generic_fls`) is still ~4x faster
than the loop.
Architectures that have a better implementation will make use of it. For
example, on x86 we get an additional factor 2 in speed without dedicated
implementation.
On GCC, the asm versions of `fls` are about the same speed as the
builtin. On Clang, the versions that use fls are more than twice as
slow as the builtin. This is because the way the `fls` function is
written, clang puts the value in memory:
https://godbolt.org/z/EfMbYe. This bug is filed at
https://bugs.llvm.org/show_bug.cgi?idI406.
```
name cpu/op
BM_Calc<__calc_delta_loop> 9.57ms Â=B112%
BM_Calc<__calc_delta_generic_fls> 2.36ms Â=B113%
BM_Calc<__calc_delta_asm_fls> 2.45ms Â=B113%
BM_Calc<__calc_delta_asm_fls_nomem> 1.66ms Â=B112%
BM_Calc<__calc_delta_asm_fls64> 2.46ms Â=B113%
BM_Calc<__calc_delta_asm_fls64_nomem> 1.34ms Â=B115%
BM_Calc<__calc_delta_builtin> 1.32ms Â=B111%
```
Signed-off-by: Clement Courbet <courbet@google.com>
Signed-off-by: Josh Don <joshdon@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20210303224653.2579656-1-joshdon@google.com
Syzbot reported a handful of occurrences where an sd->nr_balance_failed can
grow to much higher values than one would expect.
A successful load_balance() resets it to 0; a failed one increments
it. Once it gets to sd->cache_nice_tries + 3, this *should* trigger an
active balance, which will either set it to sd->cache_nice_tries+1 or reset
it to 0. However, in case the to-be-active-balanced task is not allowed to
run on env->dst_cpu, then the increment is done without any further
modification.
This could then be repeated ad nauseam, and would explain the absurdly high
values reported by syzbot (86, 149). VincentG noted there is value in
letting sd->cache_nice_tries grow, so the shift itself should be
fixed. That means preventing:
"""
If the value of the right operand is negative or is greater than or equal
to the width of the promoted left operand, the behavior is undefined.
"""
Thus we need to cap the shift exponent to
BITS_PER_TYPE(typeof(lefthand)) - 1.
I had a look around for other similar cases via coccinelle:
@expr@
position pos;
expression E1;
expression E2;
@@
(
E1 >> E2@pos
|
E1 >> E2@pos
)
@cst depends on expr@
position pos;
expression expr.E1;
constant cst;
@@
(
E1 >> cst@pos
|
E1 << cst@pos
)
@script:python depends on !cst@
pos << expr.pos;
exp << expr.E2;
@@
# Dirty hack to ignore constexpr
if exp.upper() != exp:
coccilib.report.print_report(pos[0], "Possible UB shift here")
The only other match in kernel/sched is rq_clock_thermal() which employs
sched_thermal_decay_shift, and that exponent is already capped to 10, so
that one is fine.
Fixes: 5a7f555904 ("sched/fair: Relax constraint on task's load during load balance")
Reported-by: syzbot+d7581744d5fd27c9fbe1@syzkaller.appspotmail.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>
Link: http://lore.kernel.org/r/000000000000ffac1205b9a2112f@google.com
Instead of waking up a random and already idle CPU, we can take advantage
of this_cpu being about to enter idle to run the ILB and update the
blocked load.
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>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20210224133007.28644-7-vincent.guittot@linaro.org
The HRTICK feature has traditionally been servicing configurations that
need precise preemptions point for NORMAL tasks. More recently, the
feature has been extended to also service DEADLINE tasks with stringent
runtime enforcement needs (e.g., runtime < 1ms with HZ=1000).
Enabling HRTICK sched feature currently enables the additional timer and
task tick for both classes, which might introduced undesired overhead
for no additional benefit if one needed it only for one of the cases.
Separate HRTICK sched feature in two (and leave the traditional case
name unmodified) so that it can be selectively enabled when needed.
With:
$ echo HRTICK > /sys/kernel/debug/sched_features
the NORMAL/fair hrtick gets enabled.
With:
$ echo HRTICK_DL > /sys/kernel/debug/sched_features
the DEADLINE hrtick gets enabled.
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Luis Claudio R. Goncalves <lgoncalv@redhat.com>
Signed-off-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20210208073554.14629-3-juri.lelli@redhat.com
Hung tasks and RCU stall cases were reported on systems which were not
100% busy. Investigation of such unexpected cases (no sign of potential
starvation caused by tasks hogging the system) pointed out that the
periodic sched tick timer wasn't serviced anymore after a certain point
and that caused all machinery that depends on it (timers, RCU, etc.) to
stop working as well. This issues was however only reproducible if
HRTICK was enabled.
Looking at core dumps it was found that the rbtree of the hrtimer base
used also for the hrtick was corrupted (i.e. next as seen from the base
root and actual leftmost obtained by traversing the tree are different).
Same base is also used for periodic tick hrtimer, which might get "lost"
if the rbtree gets corrupted.
Much alike what described in commit 1f71addd34 ("tick/sched: Do not
mess with an enqueued hrtimer") there is a race window between
hrtimer_set_expires() in hrtick_start and hrtimer_start_expires() in
__hrtick_restart() in which the former might be operating on an already
queued hrtick hrtimer, which might lead to corruption of the base.
Use hrtick_start() (which removes the timer before enqueuing it back) to
ensure hrtick hrtimer reprogramming is entirely guarded by the base
lock, so that no race conditions can occur.
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Luis Claudio R. Goncalves <lgoncalv@redhat.com>
Signed-off-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20210208073554.14629-2-juri.lelli@redhat.com
The only remaining use of MAX_USER_PRIO (and USER_PRIO) is the
SCALE_PRIO() definition in the PowerPC Cell architecture's Synergistic
Processor Unit (SPU) scheduler. TASK_USER_PRIO isn't used anymore.
Commit fe443ef2ac ("[POWERPC] spusched: Dynamic timeslicing for
SCHED_OTHER") copied SCALE_PRIO() from the task scheduler in v2.6.23.
Commit a4ec24b48d ("sched: tidy up SCHED_RR") removed it from the task
scheduler in v2.6.24.
Commit 3ee237dddc ("sched/prio: Add 3 macros of MAX_NICE, MIN_NICE and
NICE_WIDTH in prio.h") introduced NICE_WIDTH much later.
With:
MAX_USER_PRIO = USER_PRIO(MAX_PRIO)
= MAX_PRIO - MAX_RT_PRIO
MAX_PRIO = MAX_RT_PRIO + NICE_WIDTH
MAX_USER_PRIO = MAX_RT_PRIO + NICE_WIDTH - MAX_RT_PRIO
MAX_USER_PRIO = NICE_WIDTH
MAX_USER_PRIO can be replaced by NICE_WIDTH to be able to remove all the
{*_}USER_PRIO defines.
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@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/20210128131040.296856-3-dietmar.eggemann@arm.com
In preparation of using the balance_push state in ttwu() we need it to
provide a reliable and consistent state.
The immediate problem is that rq->balance_callback gets cleared every
schedule() and then re-set in the balance_push_callback() itself. This
is not a reliable signal, so add a variable that stays set during the
entire time.
Also move setting it before the synchronize_rcu() in
sched_cpu_deactivate(), such that we get guaranteed visibility to
ttwu(), which is a preempt-disable region.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Tested-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lkml.kernel.org/r/20210121103506.966069627@infradead.org
There is nothing schedutil specific in schedutil_cpu_util(), rename it
to effective_cpu_util(). Also create and expose another wrapper
sched_cpu_util() which can be used by other parts of the kernel, like
thermal core (that will be done in a later commit).
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/db011961fb3bb8bef1c0eda5cd64564637d3ef31.1607400596.git.viresh.kumar@linaro.org
The kernel test robot measured a -1.6% performance regression on
will-it-scale/sched_yield due to commit:
2558aacff8 ("sched/hotplug: Ensure only per-cpu kthreads run during hotplug")
Even though we were careful to replace a single load with another
single load from the same cacheline.
Restore finish_lock_switch() to the exact state before the offending
patch and solve the problem differently.
Fixes: 2558aacff8 ("sched/hotplug: Ensure only per-cpu kthreads run during hotplug")
Reported-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20201210161408.GX3021@hirez.programming.kicks-ass.net
Now that the scheduler can deal with migrate disable properly, there is no
real compelling reason to make it only available for RT.
There are quite some code pathes which needlessly disable preemption in
order to prevent migration and some constructs like kmap_atomic() enforce
it implicitly.
Making it available independent of RT allows to provide a preemptible
variant of kmap_atomic() and makes the code more consistent in general.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Grudgingly-Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20201118204007.269943012@linutronix.de
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
To remove the sd_flag parameter of select_task_rq(), we need another way of
encoding wakeup types. There already is a WF_FORK flag, add the missing two.
With that said, we still need an easy way to turn WF_foo into
SD_bar (e.g. WF_TTWU into SD_BALANCE_WAKE). As suggested by Peter, let's
make our lives easier and make them match exactly, and throw in some
compile-time checks for good measure.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20201102184514.2733-2-valentin.schneider@arm.com
In order to minimize the interference of migrate_disable() on lower
priority tasks, which can be deprived of runtime due to being stuck
below a higher priority task. Teach the RT/DL balancers to push away
these higher priority tasks when a lower priority task gets selected
to run on a freshly demoted CPU (pull).
This adds migration interference to the higher priority task, but
restores bandwidth to system that would otherwise be irrevocably lost.
Without this it would be possible to have all tasks on the system
stuck on a single CPU, each task preempted in a migrate_disable()
section with a single high priority task running.
This way we can still approximate running the M highest priority tasks
on the system.
Migrating the top task away is (ofcourse) still subject to
migrate_disable() too, which means the lower task is subject to an
interference equivalent to the worst case migrate_disable() section.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102347.499155098@infradead.org
On CPU unplug tasks which are in a migrate disabled region cannot be pushed
to a different CPU until they returned to migrateable state.
Account the number of tasks on a runqueue which are in a migrate disabled
section and make the hotplug wait mechanism respect that.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102347.067278757@infradead.org
Add the base migrate_disable() support (under protest).
While migrate_disable() is (currently) required for PREEMPT_RT, it is
also one of the biggest flaws in the system.
Notably this is just the base implementation, it is broken vs
sched_setaffinity() and hotplug, both solved in additional patches for
ease of review.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102346.818170844@infradead.org
Thread a u32 flags word through the *set_cpus_allowed*() callchain.
This will allow adding behavioural tweaks for future users.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102346.729082820@infradead.org
RT kernels need to ensure that all tasks which are not per CPU kthreads
have left the outgoing CPU to guarantee that no tasks are force migrated
within a migrate disabled section.
There is also some desire to (ab)use fine grained CPU hotplug control to
clear a CPU from active state to force migrate tasks which are not per CPU
kthreads away for power control purposes.
Add a mechanism which waits until all tasks which should leave the CPU
after the CPU active flag is cleared have moved to a different online CPU.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102346.377836842@infradead.org
In preparation for migrate_disable(), make sure only per-cpu kthreads
are allowed to run on !active CPUs.
This is ran (as one of the very first steps) from the cpu-hotplug
task which is a per-cpu kthread and completion of the hotplug
operation only requires such tasks.
This constraint enables the migrate_disable() implementation to wait
for completion of all migrate_disable regions on this CPU at hotplug
time without fear of any new ones starting.
This replaces the unlikely(rq->balance_callbacks) test at the tail of
context_switch with an unlikely(rq->balance_work), the fast path is
not affected.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102346.292709163@infradead.org
The intent of balance_callback() has always been to delay executing
balancing operations until the end of the current rq->lock section.
This is because balance operations must often drop rq->lock, and that
isn't safe in general.
However, as noted by Scott, there were a few holes in that scheme;
balance_callback() was called after rq->lock was dropped, which means
another CPU can interleave and touch the callback list.
Rework code to call the balance callbacks before dropping rq->lock
where possible, and otherwise splice the balance list onto a local
stack.
This guarantees that the balance list must be empty when we take
rq->lock. IOW, we'll only ever run our own balance callbacks.
Reported-by: Scott Wood <swood@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Link: https://lkml.kernel.org/r/20201023102346.203901269@infradead.org
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>
When change sched_rt_{runtime, period}_us, we validate that the new
settings should at least accommodate the currently allocated -dl
bandwidth:
sched_rt_handler()
--> sched_dl_bandwidth_validate()
{
new_bw = global_rt_runtime()/global_rt_period();
for_each_possible_cpu(cpu) {
dl_b = dl_bw_of(cpu);
if (new_bw < dl_b->total_bw) <-------
ret = -EBUSY;
}
}
But under CONFIG_SMP, dl_bw is per root domain , but not per CPU,
dl_b->total_bw is the allocated bandwidth of the whole root domain.
Instead, we should compare dl_b->total_bw against "cpus*new_bw",
where 'cpus' is the number of CPUs of the root domain.
Also, below annotation(in kernel/sched/sched.h) implied implementation
only appeared in SCHED_DEADLINE v2[1], then deadline scheduler kept
evolving till got merged(v9), but the annotation remains unchanged,
meaningless and misleading, update it.
* With respect to SMP, the bandwidth is given on a per-CPU basis,
* meaning that:
* - dl_bw (< 100%) is the bandwidth of the system (group) on each CPU;
* - dl_total_bw array contains, in the i-eth element, the currently
* allocated bandwidth on the i-eth CPU.
[1]: https://lore.kernel.org/lkml/1267385230.13676.101.camel@Palantir/
Fixes: 332ac17ef5 ("sched/deadline: Add bandwidth management for SCHED_DEADLINE tasks")
Signed-off-by: Peng Liu <iwtbavbm@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lkml.kernel.org/r/db6bbda316048cda7a1bbc9571defde193a8d67e.1602171061.git.iwtbavbm@gmail.com
Under CONFIG_SMP, dl_bw is per root domain, but not per CPU.
When checking or updating dl_bw, currently iterating every CPU is
overdoing, just need iterate each root domain once.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Peng Liu <iwtbavbm@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lkml.kernel.org/r/78d21ee792cc48ff79e8cd62a5f26208463684d6.1602171061.git.iwtbavbm@gmail.com
Commit:
765cc3a4b2 ("sched/core: Optimize sched_feat() for !CONFIG_SCHED_DEBUG builds")
made sched features static for !CONFIG_SCHED_DEBUG configurations, but
overlooked the CONFIG_SCHED_DEBUG=y and !CONFIG_JUMP_LABEL cases.
For the latter echoing changes to /sys/kernel/debug/sched_features has
the nasty effect of effectively changing what sched_features reports,
but without actually changing the scheduler behaviour (since different
translation units get different sysctl_sched_features).
Fix CONFIG_SCHED_DEBUG=y and !CONFIG_JUMP_LABEL configurations by properly
restructuring ifdefs.
Fixes: 765cc3a4b2 ("sched/core: Optimize sched_feat() for !CONFIG_SCHED_DEBUG builds")
Co-developed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Signed-off-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Patrick Bellasi <patrick.bellasi@matbug.net>
Reviewed-by: Valentin Schneider <valentin.schneider@arm.com>
Link: https://lore.kernel.org/r/20201013053114.160628-1-juri.lelli@redhat.com
In the following commit:
04f5c362ec6d: ("sched/fair: Replace zero-length array with flexible-array")
a zero-length array cpumask[0] has been replaced with cpumask[].
But there is still a cpumask[0] in 'struct sched_group_capacity'
which was missed.
The point of using [] instead of [0] is that with [] the compiler will
generate a build warning if it isn't the last member of a struct.
[ mingo: Rewrote the changelog. ]
Signed-off-by: zhuguangqing <zhuguangqing@xiaomi.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20201014140220.11384-1-zhuguangqing83@gmail.com
The count field is meant to tell if an update to nr_running
is an add or a subtract. Make it do so by adding the missing
minus sign.
Fixes: 9d246053a6 ("sched: Add a tracepoint to track rq->nr_running")
Signed-off-by: Phil Auld <pauld@redhat.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200805203138.1411-1-pauld@redhat.com
Rather that hide their purpose in some dark, damp corner of Documentation/,
add some documentation to the default implementations.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200731192016.7484-2-valentin.schneider@arm.com
In sched_update_tick_dependency() there's two calls that check
whether nohz_full is enabled: tick_nohz_full_cpu() does it
implicitly, while there's also an explicit call to tick_nohz_full_enabled().
Remove the duplicated, open coded check.
[ mingo: Amended the changelog. ]
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/1595935075-14223-1-git-send-email-linmiaohe@huawei.com
Dave hit the problem fixed by commit:
b6e13e8582 ("sched/core: Fix ttwu() race")
and failed to understand much of the code involved. Per his request a
few comments to (hopefully) clarify things.
Requested-by: Dave Chinner <david@fromorbit.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200702125211.GQ4800@hirez.programming.kicks-ass.net
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
There is a report that when uclamp is enabled, a netperf UDP test
regresses compared to a kernel compiled without uclamp.
https://lore.kernel.org/lkml/20200529100806.GA3070@suse.de/
While investigating the root cause, there were no sign that the uclamp
code is doing anything particularly expensive but could suffer from bad
cache behavior under certain circumstances that are yet to be
understood.
https://lore.kernel.org/lkml/20200616110824.dgkkbyapn3io6wik@e107158-lin/
To reduce the pressure on the fast path anyway, add a static key that is
by default will skip executing uclamp logic in the
enqueue/dequeue_task() fast path until it's needed.
As soon as the user start using util clamp by:
1. Changing uclamp value of a task with sched_setattr()
2. Modifying the default sysctl_sched_util_clamp_{min, max}
3. Modifying the default cpu.uclamp.{min, max} value in cgroup
We flip the static key now that the user has opted to use util clamp.
Effectively re-introducing uclamp logic in the enqueue/dequeue_task()
fast path. It stays on from that point forward until the next reboot.
This should help minimize the effect of util clamp on workloads that
don't need it but still allow distros to ship their kernels with uclamp
compiled in by default.
SCHED_WARN_ON() in uclamp_rq_dec_id() was removed since now we can end
up with unbalanced call to uclamp_rq_dec_id() if we flip the key while
a task is running in the rq. Since we know it is harmless we just
quietly return if we attempt a uclamp_rq_dec_id() when
rq->uclamp[].bucket[].tasks is 0.
In schedutil, we introduce a new uclamp_is_enabled() helper which takes
the static key into account to ensure RT boosting behavior is retained.
The following results demonstrates how this helps on 2 Sockets Xeon E5
2x10-Cores system.
nouclamp uclamp uclamp-static-key
Hmean send-64 162.43 ( 0.00%) 157.84 * -2.82%* 163.39 * 0.59%*
Hmean send-128 324.71 ( 0.00%) 314.78 * -3.06%* 326.18 * 0.45%*
Hmean send-256 641.55 ( 0.00%) 628.67 * -2.01%* 648.12 * 1.02%*
Hmean send-1024 2525.28 ( 0.00%) 2448.26 * -3.05%* 2543.73 * 0.73%*
Hmean send-2048 4836.14 ( 0.00%) 4712.08 * -2.57%* 4867.69 * 0.65%*
Hmean send-3312 7540.83 ( 0.00%) 7425.45 * -1.53%* 7621.06 * 1.06%*
Hmean send-4096 9124.53 ( 0.00%) 8948.82 * -1.93%* 9276.25 * 1.66%*
Hmean send-8192 15589.67 ( 0.00%) 15486.35 * -0.66%* 15819.98 * 1.48%*
Hmean send-16384 26386.47 ( 0.00%) 25752.25 * -2.40%* 26773.74 * 1.47%*
The perf diff between nouclamp and uclamp-static-key when uclamp is
disabled in the fast path:
8.73% -1.55% [kernel.kallsyms] [k] try_to_wake_up
0.07% +0.04% [kernel.kallsyms] [k] deactivate_task
0.13% -0.02% [kernel.kallsyms] [k] activate_task
The diff between nouclamp and uclamp-static-key when uclamp is enabled
in the fast path:
8.73% -0.72% [kernel.kallsyms] [k] try_to_wake_up
0.13% +0.39% [kernel.kallsyms] [k] activate_task
0.07% +0.38% [kernel.kallsyms] [k] deactivate_task
Fixes: 69842cba9a ("sched/uclamp: Add CPU's clamp buckets refcounting")
Reported-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Qais Yousef <qais.yousef@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Lukasz Luba <lukasz.luba@arm.com>
Link: https://lkml.kernel.org/r/20200630112123.12076-3-qais.yousef@arm.com
For some mysterious reason GCC-4.9 has a 64 byte section alignment for
structures, all other GCC versions (and Clang) tested (including 4.8
and 5.0) are fine with the 32 bytes alignment.
Getting this right is important for the new SCHED_DATA macro that
creates an explicitly ordered array of 'struct sched_class' in the
linker script and expect pointer arithmetic to work.
Fixes: c3a340f7e7 ("sched: Have sched_class_highest define by vmlinux.lds.h")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200630144905.GX4817@hirez.programming.kicks-ass.net
Use a better name for this poorly named flag, to avoid confusion...
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Link: https://lkml.kernel.org/r/20200622100825.785115830@infradead.org
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
Now that the sched_class descriptors are defined by the linker script, and
this needs to be aware of the existance of stop_sched_class when SMP is
enabled or not, as it is used as the "highest" priority when defined. Move
the declaration of sched_class_highest to the same location in the linker
script that inserts stop_sched_class, and this will also make it easier to
see what should be defined as the highest class, as this linker script
location defines the priorities as well.
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20191219214558.682913590@goodmis.org
The current SCHED_DEADLINE (DL) scheduler uses a global EDF scheduling
algorithm w/o considering CPU capacity or task utilization.
This works well on homogeneous systems where DL tasks are guaranteed
to have a bounded tardiness but presents issues on heterogeneous
systems.
A DL task can migrate to a CPU which does not have enough CPU capacity
to correctly serve the task (e.g. a task w/ 70ms runtime and 100ms
period on a CPU w/ 512 capacity).
Add the DL fitness function dl_task_fits_capacity() for DL admission
control on heterogeneous systems. A task fits onto a CPU if:
CPU original capacity / 1024 >= task runtime / task deadline
Use this function on heterogeneous systems to try to find a CPU which
meets this criterion during task wakeup, push and offline migration.
On homogeneous systems the original behavior of the DL admission
control should be retained.
Signed-off-by: Luca Abeni <luca.abeni@santannapisa.it>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lkml.kernel.org/r/20200520134243.19352-5-dietmar.eggemann@arm.com
The current SCHED_DEADLINE (DL) admission control ensures that
sum of reserved CPU bandwidth < x * M
where
x = /proc/sys/kernel/sched_rt_{runtime,period}_us
M = # CPUs in root domain.
DL admission control works well for homogeneous systems where the
capacity of all CPUs are equal (1024). I.e. bounded tardiness for DL
and non-starvation of non-DL tasks is guaranteed.
But on heterogeneous systems where capacity of CPUs are different it
could fail by over-allocating CPU time on smaller capacity CPUs.
On an Arm big.LITTLE/DynamIQ system DL tasks can easily starve other
tasks making it unusable.
Fix this by explicitly considering the CPU capacity in the DL admission
test by replacing M with the root domain CPU capacity sum.
Signed-off-by: Luca Abeni <luca.abeni@santannapisa.it>
Signed-off-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Juri Lelli <juri.lelli@redhat.com>
Link: https://lkml.kernel.org/r/20200520134243.19352-4-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
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 ttwu_queue_remote() got this wrong.
While on first reading ttwu_queue_remote() 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.
The actual race is vs the sched_ttwu_pending() call in the idle loop;
that can run the wakeup-list without consuming the CSD.
Instead of trying to chain the lists, merge them.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200526161908.129371594@infradead.org
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
Just like the ttwu_queue_remote() IPI, make use of _TIF_POLLING_NRFLAG
to avoid sending IPIs to idle CPUs.
[ mingo: Fix UP build bug. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lore.kernel.org/r/20200526161907.953304789@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
The previous commit:
c6e7bd7afaeb: ("sched/core: Optimize ttwu() spinning on p->on_cpu")
avoids spinning on p->on_rq when the task is descheduling, but only if the
wakee is on a CPU that does not share cache with the waker.
This patch offloads the activation of the wakee to the CPU that is about to
go idle if the task is the only one on the runqueue. This potentially allows
the waker task to continue making progress when the wakeup is not strictly
synchronous.
This is very obvious with netperf UDP_STREAM running on localhost. The
waker is sending packets as quickly as possible without waiting for any
reply. It frequently wakes the server for the processing of packets and
when netserver is using local memory, it quickly completes the processing
and goes back to idle. The waker often observes that netserver is on_rq
and spins excessively leading to a drop in throughput.
This is a comparison of 5.7-rc6 against "sched: Optimize ttwu() spinning
on p->on_cpu" and against this patch labeled vanilla, optttwu-v1r1 and
localwakelist-v1r2 respectively.
5.7.0-rc6 5.7.0-rc6 5.7.0-rc6
vanilla optttwu-v1r1 localwakelist-v1r2
Hmean send-64 251.49 ( 0.00%) 258.05 * 2.61%* 305.59 * 21.51%*
Hmean send-128 497.86 ( 0.00%) 519.89 * 4.43%* 600.25 * 20.57%*
Hmean send-256 944.90 ( 0.00%) 997.45 * 5.56%* 1140.19 * 20.67%*
Hmean send-1024 3779.03 ( 0.00%) 3859.18 * 2.12%* 4518.19 * 19.56%*
Hmean send-2048 7030.81 ( 0.00%) 7315.99 * 4.06%* 8683.01 * 23.50%*
Hmean send-3312 10847.44 ( 0.00%) 11149.43 * 2.78%* 12896.71 * 18.89%*
Hmean send-4096 13436.19 ( 0.00%) 13614.09 ( 1.32%) 15041.09 * 11.94%*
Hmean send-8192 22624.49 ( 0.00%) 23265.32 * 2.83%* 24534.96 * 8.44%*
Hmean send-16384 34441.87 ( 0.00%) 36457.15 * 5.85%* 35986.21 * 4.48%*
Note that this benefit is not universal to all wakeups, it only applies
to the case where the waker often spins on p->on_rq.
The impact can be seen from a "perf sched latency" report generated from
a single iteration of one packet size:
-----------------------------------------------------------------------------------------------------------------
Task | Runtime ms | Switches | Average delay ms | Maximum delay ms | Maximum delay at |
-----------------------------------------------------------------------------------------------------------------
vanilla
netperf:4337 | 21709.193 ms | 2932 | avg: 0.002 ms | max: 0.041 ms | max at: 112.154512 s
netserver:4338 | 14629.459 ms | 5146990 | avg: 0.001 ms | max: 1615.864 ms | max at: 140.134496 s
localwakelist-v1r2
netperf:4339 | 29789.717 ms | 2460 | avg: 0.002 ms | max: 0.059 ms | max at: 138.205389 s
netserver:4340 | 18858.767 ms | 7279005 | avg: 0.001 ms | max: 0.362 ms | max at: 135.709683 s
-----------------------------------------------------------------------------------------------------------------
Note that the average wakeup delay is quite small on both the vanilla
kernel and with the two patches applied. However, there are significant
outliers with the vanilla kernel with the maximum one measured as 1615
milliseconds with a vanilla kernel but never worse than 0.362 ms with
both patches applied and a much higher rate of context switching.
Similarly a separate profile of cycles showed that 2.83% of all cycles
were spent in try_to_wake_up() with almost half of the cycles spent
on spinning on p->on_rq. With the two patches, the percentage of cycles
spent in try_to_wake_up() drops to 1.13%
Signed-off-by: Mel Gorman <mgorman@techsingularity.net>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Jirka Hladky <jhladky@redhat.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: valentin.schneider@arm.com
Cc: Hillf Danton <hdanton@sina.com>
Cc: Rik van Riel <riel@surriel.com>
Link: https://lore.kernel.org/r/20200524202956.27665-3-mgorman@techsingularity.net
When users write some huge number into cpu.cfs_quota_us or
cpu.rt_runtime_us, overflow might happen during to_ratio() shifts of
schedulable checks.
to_ratio() could be altered to avoid unnecessary internal overflow, but
min_cfs_quota_period is less than 1 << BW_SHIFT, so a cutoff would still
be needed. Set a cap MAX_BW for cfs_quota_us and rt_runtime_us to
prevent overflow.
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>
Link: https://lkml.kernel.org/r/20200425105248.60093-1-changhuaixin@linux.alibaba.com
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
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
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
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
The following commit:
5e83eafbfd ("sched/fair: Remove the rq->cpu_load[] update code")
eliminated the last use case for rq->last_load_update_tick, so remove
the field as well.
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Vincent Donnefort <vincent.donnefort@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/1584710495-308969-1-git-send-email-vincent.donnefort@arm.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
- Support for locked CSD objects in smp_call_function_single_async()
which allows to simplify callsites in the scheduler core and MIPS
- Treewide consolidation of CPU hotplug functions which ensures the
consistency between the sysfs interface and kernel state. The low level
functions cpu_up/down() are now confined to the core code and not
longer accessible from random code.
-----BEGIN PGP SIGNATURE-----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=jwCg
-----END PGP SIGNATURE-----
Merge tag 'smp-core-2020-03-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
Pull core SMP updates from Thomas Gleixner:
"CPU (hotplug) updates:
- Support for locked CSD objects in smp_call_function_single_async()
which allows to simplify callsites in the scheduler core and MIPS
- Treewide consolidation of CPU hotplug functions which ensures the
consistency between the sysfs interface and kernel state. The low
level functions cpu_up/down() are now confined to the core code and
not longer accessible from random code"
* tag 'smp-core-2020-03-30' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (22 commits)
cpu/hotplug: Ignore pm_wakeup_pending() for disable_nonboot_cpus()
cpu/hotplug: Hide cpu_up/down()
cpu/hotplug: Move bringup of secondary CPUs out of smp_init()
torture: Replace cpu_up/down() with add/remove_cpu()
firmware: psci: Replace cpu_up/down() with add/remove_cpu()
xen/cpuhotplug: Replace cpu_up/down() with device_online/offline()
parisc: Replace cpu_up/down() with add/remove_cpu()
sparc: Replace cpu_up/down() with add/remove_cpu()
powerpc: Replace cpu_up/down() with add/remove_cpu()
x86/smp: Replace cpu_up/down() with add/remove_cpu()
arm64: hibernate: Use bringup_hibernate_cpu()
cpu/hotplug: Provide bringup_hibernate_cpu()
arm64: Use reboot_cpu instead of hardconding it to 0
arm64: Don't use disable_nonboot_cpus()
ARM: Use reboot_cpu instead of hardcoding it to 0
ARM: Don't use disable_nonboot_cpus()
ia64: Replace cpu_down() with smp_shutdown_nonboot_cpus()
cpu/hotplug: Create a new function to shutdown nonboot cpus
cpu/hotplug: Add new {add,remove}_cpu() functions
sched/core: Remove rq.hrtick_csd_pending
...
Pull scheduler updates from Ingo Molnar:
"The main changes in this cycle are:
- Various NUMA scheduling updates: harmonize the load-balancer and
NUMA placement logic to not work against each other. The intended
result is better locality, better utilization and fewer migrations.
- Introduce Thermal Pressure tracking and optimizations, to improve
task placement on thermally overloaded systems.
- Implement frequency invariant scheduler accounting on (some) x86
CPUs. This is done by observing and sampling the 'recent' CPU
frequency average at ~tick boundaries. The CPU provides this data
via the APERF/MPERF MSRs. This hopefully makes our capacity
estimates more precise and keeps tasks on the same CPU better even
if it might seem overloaded at a lower momentary frequency. (As
usual, turbo mode is a complication that we resolve by observing
the maximum frequency and renormalizing to it.)
- Add asymmetric CPU capacity wakeup scan to improve capacity
utilization on asymmetric topologies. (big.LITTLE systems)
- PSI fixes and optimizations.
- RT scheduling capacity awareness fixes & improvements.
- Optimize the CONFIG_RT_GROUP_SCHED constraints code.
- Misc fixes, cleanups and optimizations - see the changelog for
details"
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (62 commits)
threads: Update PID limit comment according to futex UAPI change
sched/fair: Fix condition of avg_load calculation
sched/rt: cpupri_find: Trigger a full search as fallback
kthread: Do not preempt current task if it is going to call schedule()
sched/fair: Improve spreading of utilization
sched: Avoid scale real weight down to zero
psi: Move PF_MEMSTALL out of task->flags
MAINTAINERS: Add maintenance information for psi
psi: Optimize switching tasks inside shared cgroups
psi: Fix cpu.pressure for cpu.max and competing cgroups
sched/core: Distribute tasks within affinity masks
sched/fair: Fix enqueue_task_fair warning
thermal/cpu-cooling, sched/core: Move the arch_set_thermal_pressure() API to generic scheduler code
sched/rt: Remove unnecessary push for unfit tasks
sched/rt: Allow pulling unfitting task
sched/rt: Optimize cpupri_find() on non-heterogenous systems
sched/rt: Re-instate old behavior in select_task_rq_rt()
sched/rt: cpupri_find: Implement fallback mechanism for !fit case
sched/fair: Fix reordering of enqueue/dequeue_task_fair()
sched/fair: Fix runnable_avg for throttled cfs
...
As a preparation to use simple wait queues for completions:
- Provide swake_up_all_locked() to support complete_all()
- Make __prepare_to_swait() public available
This is done to enable the usage of complete() within truly atomic contexts
on a PREEMPT_RT enabled kernel.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200321113242.228481202@linutronix.de
During our testing, we found a case that shares no longer
working correctly, the cgroup topology is like:
/sys/fs/cgroup/cpu/A (shares=102400)
/sys/fs/cgroup/cpu/A/B (shares=2)
/sys/fs/cgroup/cpu/A/B/C (shares=1024)
/sys/fs/cgroup/cpu/D (shares=1024)
/sys/fs/cgroup/cpu/D/E (shares=1024)
/sys/fs/cgroup/cpu/D/E/F (shares=1024)
The same benchmark is running in group C & F, no other tasks are
running, the benchmark is capable to consumed all the CPUs.
We suppose the group C will win more CPU resources since it could
enjoy all the shares of group A, but it's F who wins much more.
The reason is because we have group B with shares as 2, since
A->cfs_rq.load.weight == B->se.load.weight == B->shares/nr_cpus,
so A->cfs_rq.load.weight become very small.
And in calc_group_shares() we calculate shares as:
load = max(scale_load_down(cfs_rq->load.weight), cfs_rq->avg.load_avg);
shares = (tg_shares * load) / tg_weight;
Since the 'cfs_rq->load.weight' is too small, the load become 0
after scale down, although 'tg_shares' is 102400, shares of the se
which stand for group A on root cfs_rq become 2.
While the se of D on root cfs_rq is far more bigger than 2, so it
wins the battle.
Thus when scale_load_down() scale real weight down to 0, it's no
longer telling the real story, the caller will have the wrong
information and the calculation will be buggy.
This patch add check in scale_load_down(), so the real weight will
be >= MIN_SHARES after scale, after applied the group C wins as
expected.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Michael Wang <yun.wang@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/38e8e212-59a1-64b2-b247-b6d0b52d8dc1@linux.alibaba.com
Now smp_call_function_single_async() provides the protection that
we'll return with -EBUSY if the csd object is still pending, then we
don't need the rq.hrtick_csd_pending any more.
Signed-off-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20191216213125.9536-4-peterx@redhat.com
Since commit 06a76fe08d ("sched/deadline: Move DL related code
from sched/core.c to sched/deadline.c"), DL related code moved to
deadline.c.
Make the following two functions static since they're only used in
deadline.c:
dl_change_utilization()
init_dl_rq_bw_ratio()
Signed-off-by: Yu Chen <chen.yu@easystack.cn>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200228100329.16927-1-chen.yu@easystack.cn
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
Extrapolating on the existing framework to track rt/dl utilization using
pelt signals, add a similar mechanism to track thermal pressure. The
difference here from rt/dl utilization tracking is that, instead of
tracking time spent by a CPU running a RT/DL task through util_avg, the
average thermal pressure is tracked through load_avg. This is because
thermal pressure signal is weighted time "delta" capacity unlike util_avg
which is binary. "delta capacity" here means delta between the actual
capacity of a CPU and the decreased capacity a CPU due to a thermal event.
In order to track average thermal pressure, a new sched_avg variable
avg_thermal is introduced. Function update_thermal_load_avg can be called
to do the periodic bookkeeping (accumulate, decay and average) of the
thermal pressure.
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
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/20200222005213.3873-2-thara.gopinath@linaro.org
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 last remaining user of this macro has just been removed, get rid of it.
Suggested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
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/20200206191957.12325-4-valentin.schneider@arm.com
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>
Implement arch_scale_freq_capacity() for 'modern' x86. This function
is used by the scheduler to correctly account usage in the face of
DVFS.
The present patch addresses Intel processors specifically and has positive
performance and performance-per-watt implications for the schedutil cpufreq
governor, bringing it closer to, if not on-par with, the powersave governor
from the intel_pstate driver/framework.
Large performance gains are obtained when the machine is lightly loaded and
no regression are observed at saturation. The benchmarks with the largest
gains are kernel compilation, tbench (the networking version of dbench) and
shell-intensive workloads.
1. FREQUENCY INVARIANCE: MOTIVATION
* Without it, a task looks larger if the CPU runs slower
2. PECULIARITIES OF X86
* freq invariance accounting requires knowing the ratio freq_curr/freq_max
2.1 CURRENT FREQUENCY
* Use delta_APERF / delta_MPERF * freq_base (a.k.a "BusyMHz")
2.2 MAX FREQUENCY
* It varies with time (turbo). As an approximation, we set it to a
constant, i.e. 4-cores turbo frequency.
3. EFFECTS ON THE SCHEDUTIL FREQUENCY GOVERNOR
* The invariant schedutil's formula has no feedback loop and reacts faster
to utilization changes
4. KNOWN LIMITATIONS
* In some cases tasks can't reach max util despite how hard they try
5. PERFORMANCE TESTING
5.1 MACHINES
* Skylake, Broadwell, Haswell
5.2 SETUP
* baseline Linux v5.2 w/ non-invariant schedutil. Tested freq_max = 1-2-3-4-8-12
active cores turbo w/ invariant schedutil, and intel_pstate/powersave
5.3 BENCHMARK RESULTS
5.3.1 NEUTRAL BENCHMARKS
* NAS Parallel Benchmark (HPC), hackbench
5.3.2 NON-NEUTRAL BENCHMARKS
* tbench (10-30% better), kernbench (10-15% better),
shell-intensive-scripts (30-50% better)
* no regressions
5.3.3 SELECTION OF DETAILED RESULTS
5.3.4 POWER CONSUMPTION, PERFORMANCE-PER-WATT
* dbench (5% worse on one machine), kernbench (3% worse),
tbench (5-10% better), shell-intensive-scripts (10-40% better)
6. MICROARCH'ES ADDRESSED HERE
* Xeon Core before Scalable Performance processors line (Xeon Gold/Platinum
etc have different MSRs semantic for querying turbo levels)
7. REFERENCES
* MMTests performance testing framework, github.com/gormanm/mmtests
+-------------------------------------------------------------------------+
| 1. FREQUENCY INVARIANCE: MOTIVATION
+-------------------------------------------------------------------------+
For example; suppose a CPU has two frequencies: 500 and 1000 Mhz. When
running a task that would consume 1/3rd of a CPU at 1000 MHz, it would
appear to consume 2/3rd (or 66.6%) when running at 500 MHz, giving the
false impression this CPU is almost at capacity, even though it can go
faster [*]. In a nutshell, without frequency scale-invariance tasks look
larger just because the CPU is running slower.
[*] (footnote: this assumes a linear frequency/performance relation; which
everybody knows to be false, but given realities its the best approximation
we can make.)
+-------------------------------------------------------------------------+
| 2. PECULIARITIES OF X86
+-------------------------------------------------------------------------+
Accounting for frequency changes in PELT signals requires the computation of
the ratio freq_curr / freq_max. On x86 neither of those terms is readily
available.
2.1 CURRENT FREQUENCY
====================
Since modern x86 has hardware control over the actual frequency we run
at (because amongst other things, Turbo-Mode), we cannot simply use
the frequency as requested through cpufreq.
Instead we use the APERF/MPERF MSRs to compute the effective frequency
over the recent past. Also, because reading MSRs is expensive, don't
do so every time we need the value, but amortize the cost by doing it
every tick.
2.2 MAX FREQUENCY
=================
Obtaining freq_max is also non-trivial because at any time the hardware can
provide a frequency boost to a selected subset of cores if the package has
enough power to spare (eg: Turbo Boost). This means that the maximum frequency
available to a given core changes with time.
The approach taken in this change is to arbitrarily set freq_max to a constant
value at boot. The value chosen is the "4-cores (4C) turbo frequency" on most
microarchitectures, after evaluating the following candidates:
* 1-core (1C) turbo frequency (the fastest turbo state available)
* around base frequency (a.k.a. max P-state)
* something in between, such as 4C turbo
To interpret these options, consider that this is the denominator in
freq_curr/freq_max, and that ratio will be used to scale PELT signals such as
util_avg and load_avg. A large denominator will undershoot (util_avg looks a
bit smaller than it really is), viceversa with a smaller denominator PELT
signals will tend to overshoot. Given that PELT drives frequency selection
in the schedutil governor, we will have:
freq_max set to | effect on DVFS
--------------------+------------------
1C turbo | power efficiency (lower freq choices)
base freq | performance (higher util_avg, higher freq requests)
4C turbo | a bit of both
4C turbo proves to be a good compromise in a number of benchmarks (see below).
+-------------------------------------------------------------------------+
| 3. EFFECTS ON THE SCHEDUTIL FREQUENCY GOVERNOR
+-------------------------------------------------------------------------+
Once an architecture implements a frequency scale-invariant utilization (the
PELT signal util_avg), schedutil switches its frequency selection formula from
freq_next = 1.25 * freq_curr * util [non-invariant util signal]
to
freq_next = 1.25 * freq_max * util [invariant util signal]
where, in the second formula, freq_max is set to the 1C turbo frequency (max
turbo). The advantage of the second formula, whose usage we unlock with this
patch, is that freq_next doesn't depend on the current frequency in an
iterative fashion, but can jump to any frequency in a single update. This
absence of feedback in the formula makes it quicker to react to utilization
changes and more robust against pathological instabilities.
Compare it to the update formula of intel_pstate/powersave:
freq_next = 1.25 * freq_max * Busy%
where again freq_max is 1C turbo and Busy% is the percentage of time not spent
idling (calculated with delta_MPERF / delta_TSC); essentially the same as
invariant schedutil, and largely responsible for intel_pstate/powersave good
reputation. The non-invariant schedutil formula is derived from the invariant
one by approximating util_inv with util_raw * freq_curr / freq_max, but this
has limitations.
Testing shows improved performances due to better frequency selections when
the machine is lightly loaded, and essentially no change in behaviour at
saturation / overutilization.
+-------------------------------------------------------------------------+
| 4. KNOWN LIMITATIONS
+-------------------------------------------------------------------------+
It's been shown that it is possible to create pathological scenarios where a
CPU-bound task cannot reach max utilization, if the normalizing factor
freq_max is fixed to a constant value (see [Lelli-2018]).
If freq_max is set to 4C turbo as we do here, one needs to peg at least 5
cores in a package doing some busywork, and observe that none of those task
will ever reach max util (1024) because they're all running at less than the
4C turbo frequency.
While this concern still applies, we believe the performance benefit of
frequency scale-invariant PELT signals outweights the cost of this limitation.
[Lelli-2018]
https://lore.kernel.org/lkml/20180517150418.GF22493@localhost.localdomain/
+-------------------------------------------------------------------------+
| 5. PERFORMANCE TESTING
+-------------------------------------------------------------------------+
5.1 MACHINES
============
We tested the patch on three machines, with Skylake, Broadwell and Haswell
CPUs. The details are below, together with the available turbo ratios as
reported by the appropriate MSRs.
* 8x-SKYLAKE-UMA:
Single socket E3-1240 v5, Skylake 4 cores/8 threads
Max EFFiciency, BASE frequency and available turbo levels (MHz):
EFFIC 800 |********
BASE 3500 |***********************************
4C 3700 |*************************************
3C 3800 |**************************************
2C 3900 |***************************************
1C 3900 |***************************************
* 80x-BROADWELL-NUMA:
Two sockets E5-2698 v4, 2x Broadwell 20 cores/40 threads
Max EFFiciency, BASE frequency and available turbo levels (MHz):
EFFIC 1200 |************
BASE 2200 |**********************
8C 2900 |*****************************
7C 3000 |******************************
6C 3100 |*******************************
5C 3200 |********************************
4C 3300 |*********************************
3C 3400 |**********************************
2C 3600 |************************************
1C 3600 |************************************
* 48x-HASWELL-NUMA
Two sockets E5-2670 v3, 2x Haswell 12 cores/24 threads
Max EFFiciency, BASE frequency and available turbo levels (MHz):
EFFIC 1200 |************
BASE 2300 |***********************
12C 2600 |**************************
11C 2600 |**************************
10C 2600 |**************************
9C 2600 |**************************
8C 2600 |**************************
7C 2600 |**************************
6C 2600 |**************************
5C 2700 |***************************
4C 2800 |****************************
3C 2900 |*****************************
2C 3100 |*******************************
1C 3100 |*******************************
5.2 SETUP
=========
* The baseline is Linux v5.2 with schedutil (non-invariant) and the intel_pstate
driver in passive mode.
* The rationale for choosing the various freq_max values to test have been to
try all the 1-2-3-4C turbo levels (note that 1C and 2C turbo are identical
on all machines), plus one more value closer to base_freq but still in the
turbo range (8C turbo for both 80x-BROADWELL-NUMA and 48x-HASWELL-NUMA).
* In addition we've run all tests with intel_pstate/powersave for comparison.
* The filesystem is always XFS, the userspace is openSUSE Leap 15.1.
* 8x-SKYLAKE-UMA is capable of HWP (Hardware-Managed P-States), so the runs
with active intel_pstate on this machine use that.
This gives, in terms of combinations tested on each machine:
* 8x-SKYLAKE-UMA
* Baseline: Linux v5.2, non-invariant schedutil, intel_pstate passive
* intel_pstate active + powersave + HWP
* invariant schedutil, freq_max = 1C turbo
* invariant schedutil, freq_max = 3C turbo
* invariant schedutil, freq_max = 4C turbo
* both 80x-BROADWELL-NUMA and 48x-HASWELL-NUMA
* [same as 8x-SKYLAKE-UMA, but no HWP capable]
* invariant schedutil, freq_max = 8C turbo
(which on 48x-HASWELL-NUMA is the same as 12C turbo, or "all cores turbo")
5.3 BENCHMARK RESULTS
=====================
5.3.1 NEUTRAL BENCHMARKS
------------------------
Tests that didn't show any measurable difference in performance on any of the
test machines between non-invariant schedutil and our patch are:
* NAS Parallel Benchmarks (NPB) using either MPI or openMP for IPC, any
computational kernel
* flexible I/O (FIO)
* hackbench (using threads or processes, and using pipes or sockets)
5.3.2 NON-NEUTRAL BENCHMARKS
----------------------------
What follow are summary tables where each benchmark result is given a score.
* A tilde (~) means a neutral result, i.e. no difference from baseline.
* Scores are computed with the ratio result_new / result_baseline, so a tilde
means a score of 1.00.
* The results in the score ratio are the geometric means of results running
the benchmark with different parameters (eg: for kernbench: using 1, 2, 4,
... number of processes; for pgbench: varying the number of clients, and so
on).
* The first three tables show higher-is-better kind of tests (i.e. measured in
operations/second), the subsequent three show lower-is-better kind of tests
(i.e. the workload is fixed and we measure elapsed time, think kernbench).
* "gitsource" is a name we made up for the test consisting in running the
entire unit tests suite of the Git SCM and measuring how long it takes. We
take it as a typical example of shell-intensive serialized workload.
* In the "I_PSTATE" column we have the results for intel_pstate/powersave. Other
columns show invariant schedutil for different values of freq_max. 4C turbo
is circled as it's the value we've chosen for the final implementation.
80x-BROADWELL-NUMA (comparison ratio; higher is better)
+------+
I_PSTATE 1C 3C | 4C | 8C
pgbench-ro 1.14 ~ ~ | 1.11 | 1.14
pgbench-rw ~ ~ ~ | ~ | ~
netperf-udp 1.06 ~ 1.06 | 1.05 | 1.07
netperf-tcp ~ 1.03 ~ | 1.01 | 1.02
tbench4 1.57 1.18 1.22 | 1.30 | 1.56
+------+
8x-SKYLAKE-UMA (comparison ratio; higher is better)
+------+
I_PSTATE/HWP 1C 3C | 4C |
pgbench-ro ~ ~ ~ | ~ |
pgbench-rw ~ ~ ~ | ~ |
netperf-udp ~ ~ ~ | ~ |
netperf-tcp ~ ~ ~ | ~ |
tbench4 1.30 1.14 1.14 | 1.16 |
+------+
48x-HASWELL-NUMA (comparison ratio; higher is better)
+------+
I_PSTATE 1C 3C | 4C | 12C
pgbench-ro 1.15 ~ ~ | 1.06 | 1.16
pgbench-rw ~ ~ ~ | ~ | ~
netperf-udp 1.05 0.97 1.04 | 1.04 | 1.02
netperf-tcp 0.96 1.01 1.01 | 1.01 | 1.01
tbench4 1.50 1.05 1.13 | 1.13 | 1.25
+------+
In the table above we see that active intel_pstate is slightly better than our
4C-turbo patch (both in reference to the baseline non-invariant schedutil) on
read-only pgbench and much better on tbench. Both cases are notable in which
it shows that lowering our freq_max (to 8C-turbo and 12C-turbo on
80x-BROADWELL-NUMA and 48x-HASWELL-NUMA respectively) helps invariant
schedutil to get closer.
If we ignore active intel_pstate and focus on the comparison with baseline
alone, there are several instances of double-digit performance improvement.
80x-BROADWELL-NUMA (comparison ratio; lower is better)
+------+
I_PSTATE 1C 3C | 4C | 8C
dbench4 1.23 0.95 0.95 | 0.95 | 0.95
kernbench 0.93 0.83 0.83 | 0.83 | 0.82
gitsource 0.98 0.49 0.49 | 0.49 | 0.48
+------+
8x-SKYLAKE-UMA (comparison ratio; lower is better)
+------+
I_PSTATE/HWP 1C 3C | 4C |
dbench4 ~ ~ ~ | ~ |
kernbench ~ ~ ~ | ~ |
gitsource 0.92 0.55 0.55 | 0.55 |
+------+
48x-HASWELL-NUMA (comparison ratio; lower is better)
+------+
I_PSTATE 1C 3C | 4C | 8C
dbench4 ~ ~ ~ | ~ | ~
kernbench 0.94 0.90 0.89 | 0.90 | 0.90
gitsource 0.97 0.69 0.69 | 0.69 | 0.69
+------+
dbench is not very remarkable here, unless we notice how poorly active
intel_pstate is performing on 80x-BROADWELL-NUMA: 23% regression versus
non-invariant schedutil. We repeated that run getting consistent results. Out
of scope for the patch at hand, but deserving future investigation. Other than
that, we previously ran this campaign with Linux v5.0 and saw the patch doing
better on dbench a the time. We haven't checked closely and can only speculate
at this point.
On the NUMA boxes kernbench gets 10-15% improvements on average; we'll see in
the detailed tables that the gains concentrate on low process counts (lightly
loaded machines).
The test we call "gitsource" (running the git unit test suite, a long-running
single-threaded shell script) appears rather spectacular in this table (gains
of 30-50% depending on the machine). It is to be noted, however, that
gitsource has no adjustable parameters (such as the number of jobs in
kernbench, which we average over in order to get a single-number summary
score) and is exactly the kind of low-parallelism workload that benefits the
most from this patch. When looking at the detailed tables of kernbench or
tbench4, at low process or client counts one can see similar numbers.
5.3.3 SELECTION OF DETAILED RESULTS
-----------------------------------
Machine : 48x-HASWELL-NUMA
Benchmark : tbench4 (i.e. dbench4 over the network, actually loopback)
Varying parameter : number of clients
Unit : MB/sec (higher is better)
5.2.0 vanilla (BASELINE) 5.2.0 intel_pstate 5.2.0 1C-turbo
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Hmean 1 126.73 +- 0.31% ( ) 315.91 +- 0.66% ( 149.28%) 125.03 +- 0.76% ( -1.34%)
Hmean 2 258.04 +- 0.62% ( ) 614.16 +- 0.51% ( 138.01%) 269.58 +- 1.45% ( 4.47%)
Hmean 4 514.30 +- 0.67% ( ) 1146.58 +- 0.54% ( 122.94%) 533.84 +- 1.99% ( 3.80%)
Hmean 8 1111.38 +- 2.52% ( ) 2159.78 +- 0.38% ( 94.33%) 1359.92 +- 1.56% ( 22.36%)
Hmean 16 2286.47 +- 1.36% ( ) 3338.29 +- 0.21% ( 46.00%) 2720.20 +- 0.52% ( 18.97%)
Hmean 32 4704.84 +- 0.35% ( ) 4759.03 +- 0.43% ( 1.15%) 4774.48 +- 0.30% ( 1.48%)
Hmean 64 7578.04 +- 0.27% ( ) 7533.70 +- 0.43% ( -0.59%) 7462.17 +- 0.65% ( -1.53%)
Hmean 128 6998.52 +- 0.16% ( ) 6987.59 +- 0.12% ( -0.16%) 6909.17 +- 0.14% ( -1.28%)
Hmean 192 6901.35 +- 0.25% ( ) 6913.16 +- 0.10% ( 0.17%) 6855.47 +- 0.21% ( -0.66%)
5.2.0 3C-turbo 5.2.0 4C-turbo 5.2.0 12C-turbo
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Hmean 1 128.43 +- 0.28% ( 1.34%) 130.64 +- 3.81% ( 3.09%) 153.71 +- 5.89% ( 21.30%)
Hmean 2 311.70 +- 6.15% ( 20.79%) 281.66 +- 3.40% ( 9.15%) 305.08 +- 5.70% ( 18.23%)
Hmean 4 641.98 +- 2.32% ( 24.83%) 623.88 +- 5.28% ( 21.31%) 906.84 +- 4.65% ( 76.32%)
Hmean 8 1633.31 +- 1.56% ( 46.96%) 1714.16 +- 0.93% ( 54.24%) 2095.74 +- 0.47% ( 88.57%)
Hmean 16 3047.24 +- 0.42% ( 33.27%) 3155.02 +- 0.30% ( 37.99%) 3634.58 +- 0.15% ( 58.96%)
Hmean 32 4734.31 +- 0.60% ( 0.63%) 4804.38 +- 0.23% ( 2.12%) 4674.62 +- 0.27% ( -0.64%)
Hmean 64 7699.74 +- 0.35% ( 1.61%) 7499.72 +- 0.34% ( -1.03%) 7659.03 +- 0.25% ( 1.07%)
Hmean 128 6935.18 +- 0.15% ( -0.91%) 6942.54 +- 0.10% ( -0.80%) 7004.85 +- 0.12% ( 0.09%)
Hmean 192 6901.62 +- 0.12% ( 0.00%) 6856.93 +- 0.10% ( -0.64%) 6978.74 +- 0.10% ( 1.12%)
This is one of the cases where the patch still can't surpass active
intel_pstate, not even when freq_max is as low as 12C-turbo. Otherwise, gains are
visible up to 16 clients and the saturated scenario is the same as baseline.
The scores in the summary table from the previous sections are ratios of
geometric means of the results over different clients, as seen in this table.
Machine : 80x-BROADWELL-NUMA
Benchmark : kernbench (kernel compilation)
Varying parameter : number of jobs
Unit : seconds (lower is better)
5.2.0 vanilla (BASELINE) 5.2.0 intel_pstate 5.2.0 1C-turbo
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Amean 2 379.68 +- 0.06% ( ) 330.20 +- 0.43% ( 13.03%) 285.93 +- 0.07% ( 24.69%)
Amean 4 200.15 +- 0.24% ( ) 175.89 +- 0.22% ( 12.12%) 153.78 +- 0.25% ( 23.17%)
Amean 8 106.20 +- 0.31% ( ) 95.54 +- 0.23% ( 10.03%) 86.74 +- 0.10% ( 18.32%)
Amean 16 56.96 +- 1.31% ( ) 53.25 +- 1.22% ( 6.50%) 48.34 +- 1.73% ( 15.13%)
Amean 32 34.80 +- 2.46% ( ) 33.81 +- 0.77% ( 2.83%) 30.28 +- 1.59% ( 12.99%)
Amean 64 26.11 +- 1.63% ( ) 25.04 +- 1.07% ( 4.10%) 22.41 +- 2.37% ( 14.16%)
Amean 128 24.80 +- 1.36% ( ) 23.57 +- 1.23% ( 4.93%) 21.44 +- 1.37% ( 13.55%)
Amean 160 24.85 +- 0.56% ( ) 23.85 +- 1.17% ( 4.06%) 21.25 +- 1.12% ( 14.49%)
5.2.0 3C-turbo 5.2.0 4C-turbo 5.2.0 8C-turbo
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Amean 2 284.08 +- 0.13% ( 25.18%) 283.96 +- 0.51% ( 25.21%) 285.05 +- 0.21% ( 24.92%)
Amean 4 153.18 +- 0.22% ( 23.47%) 154.70 +- 1.64% ( 22.71%) 153.64 +- 0.30% ( 23.24%)
Amean 8 87.06 +- 0.28% ( 18.02%) 86.77 +- 0.46% ( 18.29%) 86.78 +- 0.22% ( 18.28%)
Amean 16 48.03 +- 0.93% ( 15.68%) 47.75 +- 1.99% ( 16.17%) 47.52 +- 1.61% ( 16.57%)
Amean 32 30.23 +- 1.20% ( 13.14%) 30.08 +- 1.67% ( 13.57%) 30.07 +- 1.67% ( 13.60%)
Amean 64 22.59 +- 2.02% ( 13.50%) 22.63 +- 0.81% ( 13.32%) 22.42 +- 0.76% ( 14.12%)
Amean 128 21.37 +- 0.67% ( 13.82%) 21.31 +- 1.15% ( 14.07%) 21.17 +- 1.93% ( 14.63%)
Amean 160 21.68 +- 0.57% ( 12.76%) 21.18 +- 1.74% ( 14.77%) 21.22 +- 1.00% ( 14.61%)
The patch outperform active intel_pstate (and baseline) by a considerable
margin; the summary table from the previous section says 4C turbo and active
intel_pstate are 0.83 and 0.93 against baseline respectively, so 4C turbo is
0.83/0.93=0.89 against intel_pstate (~10% better on average). There is no
noticeable difference with regard to the value of freq_max.
Machine : 8x-SKYLAKE-UMA
Benchmark : gitsource (time to run the git unit test suite)
Varying parameter : none
Unit : seconds (lower is better)
5.2.0 vanilla 5.2.0 intel_pstate/hwp 5.2.0 1C-turbo
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Amean 858.85 +- 1.16% ( ) 791.94 +- 0.21% ( 7.79%) 474.95 ( 44.70%)
5.2.0 3C-turbo 5.2.0 4C-turbo
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Amean 475.26 +- 0.20% ( 44.66%) 474.34 +- 0.13% ( 44.77%)
In this test, which is of interest as representing shell-intensive
(i.e. fork-intensive) serialized workloads, invariant schedutil outperforms
intel_pstate/powersave by a whopping 40% margin.
5.3.4 POWER CONSUMPTION, PERFORMANCE-PER-WATT
---------------------------------------------
The following table shows average power consumption in watt for each
benchmark. Data comes from turbostat (package average), which in turn is read
from the RAPL interface on CPUs. We know the patch affects CPU frequencies so
it's reasonable to ignore other power consumers (such as memory or I/O). Also,
we don't have a power meter available in the lab so RAPL is the best we have.
turbostat sampled average power every 10 seconds for the entire duration of
each benchmark. We took all those values and averaged them (i.e. with don't
have detail on a per-parameter granularity, only on whole benchmarks).
80x-BROADWELL-NUMA (power consumption, watts)
+--------+
BASELINE I_PSTATE 1C 3C | 4C | 8C
pgbench-ro 130.01 142.77 131.11 132.45 | 134.65 | 136.84
pgbench-rw 68.30 60.83 71.45 71.70 | 71.65 | 72.54
dbench4 90.25 59.06 101.43 99.89 | 101.10 | 102.94
netperf-udp 65.70 69.81 66.02 68.03 | 68.27 | 68.95
netperf-tcp 88.08 87.96 88.97 88.89 | 88.85 | 88.20
tbench4 142.32 176.73 153.02 163.91 | 165.58 | 176.07
kernbench 92.94 101.95 114.91 115.47 | 115.52 | 115.10
gitsource 40.92 41.87 75.14 75.20 | 75.40 | 75.70
+--------+
8x-SKYLAKE-UMA (power consumption, watts)
+--------+
BASELINE I_PSTATE/HWP 1C 3C | 4C |
pgbench-ro 46.49 46.68 46.56 46.59 | 46.52 |
pgbench-rw 29.34 31.38 30.98 31.00 | 31.00 |
dbench4 27.28 27.37 27.49 27.41 | 27.38 |
netperf-udp 22.33 22.41 22.36 22.35 | 22.36 |
netperf-tcp 27.29 27.29 27.30 27.31 | 27.33 |
tbench4 41.13 45.61 43.10 43.33 | 43.56 |
kernbench 42.56 42.63 43.01 43.01 | 43.01 |
gitsource 13.32 13.69 17.33 17.30 | 17.35 |
+--------+
48x-HASWELL-NUMA (power consumption, watts)
+--------+
BASELINE I_PSTATE 1C 3C | 4C | 12C
pgbench-ro 128.84 136.04 129.87 132.43 | 132.30 | 134.86
pgbench-rw 37.68 37.92 37.17 37.74 | 37.73 | 37.31
dbench4 28.56 28.73 28.60 28.73 | 28.70 | 28.79
netperf-udp 56.70 60.44 56.79 57.42 | 57.54 | 57.52
netperf-tcp 75.49 75.27 75.87 76.02 | 76.01 | 75.95
tbench4 115.44 139.51 119.53 123.07 | 123.97 | 130.22
kernbench 83.23 91.55 95.58 95.69 | 95.72 | 96.04
gitsource 36.79 36.99 39.99 40.34 | 40.35 | 40.23
+--------+
A lower power consumption isn't necessarily better, it depends on what is done
with that energy. Here are tables with the ratio of performance-per-watt on
each machine and benchmark. Higher is always better; a tilde (~) means a
neutral ratio (i.e. 1.00).
80x-BROADWELL-NUMA (performance-per-watt ratios; higher is better)
+------+
I_PSTATE 1C 3C | 4C | 8C
pgbench-ro 1.04 1.06 0.94 | 1.07 | 1.08
pgbench-rw 1.10 0.97 0.96 | 0.96 | 0.97
dbench4 1.24 0.94 0.95 | 0.94 | 0.92
netperf-udp ~ 1.02 1.02 | ~ | 1.02
netperf-tcp ~ 1.02 ~ | ~ | 1.02
tbench4 1.26 1.10 1.06 | 1.12 | 1.26
kernbench 0.98 0.97 0.97 | 0.97 | 0.98
gitsource ~ 1.11 1.11 | 1.11 | 1.13
+------+
8x-SKYLAKE-UMA (performance-per-watt ratios; higher is better)
+------+
I_PSTATE/HWP 1C 3C | 4C |
pgbench-ro ~ ~ ~ | ~ |
pgbench-rw 0.95 0.97 0.96 | 0.96 |
dbench4 ~ ~ ~ | ~ |
netperf-udp ~ ~ ~ | ~ |
netperf-tcp ~ ~ ~ | ~ |
tbench4 1.17 1.09 1.08 | 1.10 |
kernbench ~ ~ ~ | ~ |
gitsource 1.06 1.40 1.40 | 1.40 |
+------+
48x-HASWELL-NUMA (performance-per-watt ratios; higher is better)
+------+
I_PSTATE 1C 3C | 4C | 12C
pgbench-ro 1.09 ~ 1.09 | 1.03 | 1.11
pgbench-rw ~ 0.86 ~ | ~ | 0.86
dbench4 ~ 1.02 1.02 | 1.02 | ~
netperf-udp ~ 0.97 1.03 | 1.02 | ~
netperf-tcp 0.96 ~ ~ | ~ | ~
tbench4 1.24 ~ 1.06 | 1.05 | 1.11
kernbench 0.97 0.97 0.98 | 0.97 | 0.96
gitsource 1.03 1.33 1.32 | 1.32 | 1.33
+------+
These results are overall pleasing: in plenty of cases we observe
performance-per-watt improvements. The few regressions (read/write pgbench and
dbench on the Broadwell machine) are of small magnitude. kernbench loses a few
percentage points (it has a 10-15% performance improvement, but apparently the
increase in power consumption is larger than that). tbench4 and gitsource, which
benefit the most from the patch, keep a positive score in this table which is
a welcome surprise; that suggests that in those particular workloads the
non-invariant schedutil (and active intel_pstate, too) makes some rather
suboptimal frequency selections.
+-------------------------------------------------------------------------+
| 6. MICROARCH'ES ADDRESSED HERE
+-------------------------------------------------------------------------+
The patch addresses Xeon Core processors that use MSR_PLATFORM_INFO and
MSR_TURBO_RATIO_LIMIT to advertise their base frequency and turbo frequencies
respectively. This excludes the recent Xeon Scalable Performance processors
line (Xeon Gold, Platinum etc) whose MSRs have to be parsed differently.
Subsequent patches will address:
* Xeon Scalable Performance processors and Atom Goldmont/Goldmont Plus
* Xeon Phi (Knights Landing, Knights Mill)
* Atom Silvermont
+-------------------------------------------------------------------------+
| 7. REFERENCES
+-------------------------------------------------------------------------+
Tests have been run with the help of the MMTests performance testing
framework, see github.com/gormanm/mmtests. The configuration file names for
the benchmark used are:
db-pgbench-timed-ro-small-xfs
db-pgbench-timed-rw-small-xfs
io-dbench4-async-xfs
network-netperf-unbound
network-tbench
scheduler-unbound
workload-kerndevel-xfs
workload-shellscripts-xfs
hpc-nas-c-class-mpi-full-xfs
hpc-nas-c-class-omp-full
All those benchmarks are generally available on the web:
pgbench: https://www.postgresql.org/docs/10/pgbench.html
netperf: https://hewlettpackard.github.io/netperf/
dbench/tbench: https://dbench.samba.org/
gitsource: git unit test suite, github.com/git/git
NAS Parallel Benchmarks: https://www.nas.nasa.gov/publications/npb.html
hackbench: https://people.redhat.com/mingo/cfs-scheduler/tools/hackbench.c
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Giovanni Gherdovich <ggherdovich@suse.cz>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Acked-by: Doug Smythies <dsmythies@telus.net>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Link: https://lkml.kernel.org/r/20200122151617.531-2-ggherdovich@suse.cz
The current helper returns (CPU) rq utilization with uclamp restrictions
taken into account. A uclamp task utilization helper would be quite
helpful, but this requires some renaming.
Prepare the code for the introduction of a uclamp_task_util() by renaming
the existing uclamp_util_with() to uclamp_rq_util_with().
Tested-By: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Quentin Perret <qperret@google.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20191211113851.24241-4-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Vincent pointed out recently that the canonical type for utilization
values is 'unsigned long'. Internally uclamp uses 'unsigned int' values for
cache optimization, but this doesn't have to be exported to its users.
Make the uclamp helpers that deal with utilization use and return unsigned
long values.
Tested-By: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Quentin Perret <qperret@google.com>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20191211113851.24241-3-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The sole user of uclamp_util(), schedutil_cpu_util(), was made to use
uclamp_util_with() instead in commit:
af24bde8df ("sched/uclamp: Add uclamp support to energy_compute()")
From then on, uclamp_util() has remained unused. Being a simple wrapper
around uclamp_util_with(), we can get rid of it and win back a few lines.
Tested-By: Dietmar Eggemann <dietmar.eggemann@arm.com>
Suggested-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20191211113851.24241-2-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Some uclamp helpers had their return type changed from 'unsigned int' to
'enum uclamp_id' by commit
0413d7f33e ("sched/uclamp: Always use 'enum uclamp_id' for clamp_id values")
but it happens that some do return a value in the [0, SCHED_CAPACITY_SCALE]
range, which should really be unsigned int. The affected helpers are
uclamp_none(), uclamp_rq_max_value() and uclamp_eff_value(). Fix those up.
Note that this doesn't lead to any obj diff using a relatively recent
aarch64 compiler (8.3-2019.03). The current code of e.g. uclamp_eff_value()
properly returns an 11 bit value (bits_per(1024)) and doesn't seem to do
anything funny. I'm still marking this as fixing the above commit to be on
the safe side.
Signed-off-by: Valentin Schneider <valentin.schneider@arm.com>
Reviewed-by: Qais Yousef <qais.yousef@arm.com>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Dietmar.Eggemann@arm.com
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: patrick.bellasi@matbug.net
Cc: qperret@google.com
Cc: surenb@google.com
Cc: tj@kernel.org
Fixes: 0413d7f33e ("sched/uclamp: Always use 'enum uclamp_id' for clamp_id values")
Link: https://lkml.kernel.org/r/20191115103908.27610-1-valentin.schneider@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Josh Don