diff options
| author | Peter Zijlstra <peterz@infradead.org> | 2020-12-10 17:14:08 +0100 |
|---|---|---|
| committer | Peter Zijlstra <peterz@infradead.org> | 2020-12-15 11:27:53 +0100 |
| commit | ae7927023243dcc7389b2d59b16c09cbbeaecc36 (patch) | |
| tree | 64bdf7a8cddcc7a2c2631a107dee1b5bbc4e7405 /scripts/recordmcount.h | |
| parent | 5b78f2dc315354c05300795064f587366a02c6ff (diff) | |
| download | linux-ae7927023243dcc7389b2d59b16c09cbbeaecc36.tar.gz | |
sched: Optimize finish_lock_switch()
The kernel test robot measured a -1.6% performance regression on
will-it-scale/sched_yield due to commit:
2558aacff858 ("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: 2558aacff858 ("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
Diffstat (limited to 'scripts/recordmcount.h')
0 files changed, 0 insertions, 0 deletions