perf: Avoid RCU vs preemption assumptions
authorPeter Zijlstra <a.p.zijlstra@chello.nl>
Thu, 16 Sep 2010 17:17:24 +0000 (19:17 +0200)
committerIngo Molnar <mingo@elte.hu>
Tue, 21 Sep 2010 11:55:44 +0000 (13:55 +0200)
commit41945f6ccf1e86f87fddf6b32db9cf431c05fb54
treecf64d47bfff06ff4560efcaa23a067fa99db34dd
parent8b8e2ec1eeca7f6941bc81cefc9663018d6ceb57
perf: Avoid RCU vs preemption assumptions

The per-pmu per-cpu context patch converted things from
get_cpu_var() to this_cpu_ptr(), but that only works if
rcu_read_lock() actually disables preemption, and since
there is no such guarantee, we need to fix that.

Use the newly introduced {get,put}_cpu_ptr().

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Tejun Heo <tj@kernel.org>
LKML-Reference: <20100917093009.308453028@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
kernel/perf_event.c