perf_counter, x86: protect per-cpu variables with compile barriers only

Per-cpu variables needn't to be protected with cpu barriers
(smp_wmb()). Protection is only needed for preemption on the same cpu
(rescheduling or the nmi handler). This can be done using a compiler
barrier only.

[ Impact: micro-optimization ]

Signed-off-by: Robert Richter <robert.richter@amd.com>
Cc: Paul Mackerras <paulus@samba.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1241002046-8832-6-git-send-email-robert.richter@amd.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

arch/x86/kernel/cpu/perf_counter.c

@@ -673,7 +673,7 @@ static int pmc_generic_enable(struct perf_counter *counter)
 	/*
 	 * Make it visible before enabling the hw:
 	 */
-	smp_wmb();
+	barrier();
 
 	__hw_perf_counter_set_period(counter, hwc, idx);
 	__pmc_generic_enable(counter, hwc, idx);
@@ -745,7 +745,7 @@ static void pmc_generic_disable(struct perf_counter *counter)
 	 * Make sure the cleared pointer becomes visible before we
 	 * (potentially) free the counter:
 	 */
-	smp_wmb();
+	barrier();
 
 	/*
 	 * Drain the remaining delta count out of a counter