Date:   Thu, 17 Jan 2019 15:34:06 +0000
From:   Valentin Schneider <valentin.schneider@....com>
To:     linux-kernel@...r.kernel.org
Cc:     mingo@...nel.org, peterz@...radead.org, vincent.guittot@...aro.org,
        morten.rasmussen@....com, Dietmar.Eggemann@....com
Subject: [PATCH 0/5] sched/fair: NOHZ cleanups and misfit improvement

In

  commit 5fbdfae5221a ("sched/fair: Kick nohz balance if rq->misfit_task_load")

was added a trigger for nohz kicks, which is required to offload misfit tasks
from LITTLE to big CPUs. However, those kicks could be issued a lot more
frequently than what is strictly needed.

This patch-set brings some unrelated nohz code cleanups, and tunes down
unneeded nohz kicks.

- Patches [1-3] do some cleaning up of the current nohz code
- Patches [4-5] tweak the nohz kick conditions for asymmetric systems
  
* Testing
** kick_ilb() hits
  This causes a large reduction in calls to kick_ilb() (and thus subsequent
  rescheduling interrupts & useless nohz balance calls) in most scenarios.

  The "best case" one is running NR_BIG_CPUS big tasks, which I tested with
  4 50% periodic tasks running for 5 seconds on my HiKey960 (4x4 big.LITTLE):

  | CPU | hits (baseline) | hits (patchset) |
  |-----+-----------------+-----------------|
  |   0 |              31 |              41 |
  |   1 |              21 |               3 |
  |   2 |              35 |               2 |
  |   3 |               9 |               4 |
  |-----+-----------------+-----------------|
  |   4 |             170 |               4 |
  |   5 |             573 |               4 |
  |   6 |             544 |               4 |
  |   7 |             579 |               4 |


  Something a bit less idealistic with NR_CPUS-1 big tasks still shows some
  improvements (7 100% tasks running for 5 seconds on my HiKey960):

  | CPU | hits (baseline) | hits (patchset) |
  |-----+-----------------+-----------------|
  |   0 |              14 |             122 |
  |   1 |              47 |             162 |
  |   2 |              11 |             156 |
  |   3 |               9 |               3 |
  |-----+-----------------+-----------------|
  |   4 |              53 |               6 |
  |   5 |             276 |              13 |
  |   6 |             312 |               7 |
  |   7 |             250 |              11 |

  I was surprised to see such an increase in calls to kick_ilb() from LITTLE
  CPUs ([0-3]), but after a bit of investigation it turns out that the big
  CPUs would always run nohz_balancer_kick() a jiffy before the LITTLEs, so
  the LITTLEs would always bail out because nohz.next_balance had just been
  updated before they called nohz_balancer_kick(). IOW,

      time_before(now, nohz.next_balance)

  would always be true on CPUs [0-3] during my workload. Quieting the kicks
  issued by the big CPUs allowed the LITTLEs to execute nohz_balancer_kick()
  past that condition, explaining the higher number of kicks issued from LITTLE
  CPUs.
  
** misfit behaviour
  For good measure I also ran the usual misfit tests [1] which showed no
  particular change.

* Notes

  Since patch 05 introduces yet another label in kick_nohz_balancer(),
  I benchmarked splitting its "slowpath"  (all of the rcu derefs) into
  its own function and didn't find any real overhead. If folks don't
  hate the idea I can add it to a v2 or post it separately.

[1]: https://github.com/ARM-software/lisa/blob/next/lisa/tests/kernel/scheduler/misfit.py

Valentin Schneider (5):
  sched/fair: Use for_each_cpu_and for asym-packing nohz kicks
  sched/fair: Explain LLC nohz kick condition
  sched/fair: Prune nohz_balancer_kick() comment block
  sched/fair: Tune down misfit nohz kicks
  sched/fair: Skip LLC nohz logic for asymmetric systems

 kernel/sched/fair.c | 62 ++++++++++++++++++++++++++++++++-------------
 1 file changed, 44 insertions(+), 18 deletions(-)

--
2.20.1

Powered by blists - more mailing lists