sched: for_each_domain() vs RCU
authorPeter Zijlstra <a.p.zijlstra@chello.nl>
Thu, 10 Sep 2009 16:18:47 +0000 (18:18 +0200)
committerIngo Molnar <mingo@elte.hu>
Tue, 15 Sep 2009 14:01:06 +0000 (16:01 +0200)
commit83f54960c11a14942ab00b54c51e91906b9d8235
tree6c7df778777568bb0c7375b4e7ec84ed532809df
parentae154be1f34a674e6cbb43ccf6e442f56acd7a70
sched: for_each_domain() vs RCU

for_each_domain() uses RCU to serialize the sched_domains, except
it doesn't actually use rcu_read_lock() and instead relies on
disabling preemption -> FAIL.

XXX: audit other sched_domain code.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
kernel/sched_fair.c