[PATCH] sched: RCU domains
authorNick Piggin <nickpiggin@yahoo.com.au>
Sat, 25 Jun 2005 21:57:27 +0000 (14:57 -0700)
committerLinus Torvalds <torvalds@ppc970.osdl.org>
Sat, 25 Jun 2005 23:24:44 +0000 (16:24 -0700)
commit674311d5b411e9042df4fdf7aef0b3c8217b6240
tree4ee6b739629e7fc33b519fd087a116b4fa33a217
parent3dbd5342074a1e570ec84edf859deb9be588006d
[PATCH] sched: RCU domains

One of the problems with the multilevel balance-on-fork/exec is that it needs
to jump through hoops to satisfy sched-domain's locking semantics (that is,
you may traverse your own domain when not preemptable, and you may traverse
others' domains when holding their runqueue lock).

balance-on-exec had to potentially migrate between more than one CPU before
finding a final CPU to migrate to, and balance-on-fork needed to potentially
take multiple runqueue locks.

So bite the bullet and make sched-domains go completely RCU.  This actually
simplifies the code quite a bit.

From: Ingo Molnar <mingo@elte.hu>

schedstats RCU fix, and a nice comment on for_each_domain, from Ingo.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Nick Piggin <nickpiggin@yahoo.com.au>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
kernel/sched.c