sched, cgroup: replace signal_struct->group_rwsem with a global percpu_rwsem
authorTejun Heo <tj@kernel.org>
Wed, 13 May 2015 20:35:17 +0000 (16:35 -0400)
committerTejun Heo <tj@kernel.org>
Wed, 27 May 2015 00:35:00 +0000 (20:35 -0400)
commitd59cfc09c32a2ae31f1c3bc2983a0cd79afb3f14
tree077533cef8f5e16c8f7fd65d7e255d75828f3820
parent7d7efec368d537226142cbe559f45797f18672f9
sched, cgroup: replace signal_struct->group_rwsem with a global percpu_rwsem

The cgroup side of threadgroup locking uses signal_struct->group_rwsem
to synchronize against threadgroup changes.  This per-process rwsem
adds small overhead to thread creation, exit and exec paths, forces
cgroup code paths to do lock-verify-unlock-retry dance in a couple
places and makes it impossible to atomically perform operations across
multiple processes.

This patch replaces signal_struct->group_rwsem with a global
percpu_rwsem cgroup_threadgroup_rwsem which is cheaper on the reader
side and contained in cgroups proper.  This patch converts one-to-one.

This does make writer side heavier and lower the granularity; however,
cgroup process migration is a fairly cold path, we do want to optimize
thread operations over it and cgroup migration operations don't take
enough time for the lower granularity to matter.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
include/linux/cgroup-defs.h
include/linux/init_task.h
include/linux/sched.h
init/Kconfig
kernel/cgroup.c
kernel/fork.c