redonkable
/
remarkable-linux
1
0
Fork 0
Code Releases Activity
remarkable-linux/kernel/sched
History
Linus Torvalds 37407ea7f9 Revert "sched: Improve scalability via 'CPU buddies', which withstand random perturbations" 
This reverts commit 970e178985.

Nikolay Ulyanitsky reported thatthe 3.6-rc5 kernel has a 15-20%
performance drop on PostgreSQL 9.2 on his machine (running "pgbench").

Borislav Petkov was able to reproduce this, and bisected it to this
commit 970e178985 ("sched: Improve scalability via 'CPU buddies' ...")
apparently because the new single-idle-buddy model simply doesn't find
idle CPU's to reschedule on aggressively enough.

Mike Galbraith suspects that it is likely due to the user-mode spinlocks
in PostgreSQL not reacting well to preemption, but we don't really know
the details - I'll just revert the commit for now.

There are hopefully other approaches to improve scheduler scalability
without it causing these kinds of downsides.

Reported-by: Nikolay Ulyanitsky <lystor@gmail.com>
Bisected-by: Borislav Petkov <bp@alien8.de>
Acked-by: Mike Galbraith <efault@gmx.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
 		2012-09-16 12:29:43 -07:00
..
Makefile init_task: Create generic init_task instance 2012-05-05 13:00:21 +02:00
auto_group.c sched: Clean up parameter passing of proc_sched_autogroup_set_nice() 2012-03-02 12:23:49 +01:00
auto_group.h
…
clock.c
…
core.c Revert "sched: Improve scalability via 'CPU buddies', which withstand random perturbations" 2012-09-16 12:29:43 -07:00
cpupri.c sched: Fix minor code style issues 2012-07-26 11:47:00 +02:00
cpupri.h
…
debug.c sched/debug: Fix printing large integers on 32-bit platforms 2012-05-14 15:05:28 +02:00
fair.c Revert "sched: Improve scalability via 'CPU buddies', which withstand random perturbations" 2012-09-16 12:29:43 -07:00
features.h sched: Fix more load-balancing fallout 2012-04-26 12:54:52 +02:00
idle_task.c sched/nohz: Rewrite and fix load-avg computation -- again 2012-07-05 20:58:13 +02:00
rt.c sched: Unthrottle rt runqueues in __disable_runtime() 2012-09-04 14:30:30 +02:00
sched.h sched: Unthrottle rt runqueues in __disable_runtime() 2012-09-04 14:30:30 +02:00
stats.c sched: Remove sched_switch 2012-01-27 13:28:53 +01:00
stats.h
…
stop_task.c sched: Fix migration thread runtime bogosity 2012-08-13 18:41:55 +02:00