perf: Optimize context ops
Assuming we don't mix events of different pmus onto a single context
(with the exeption of software events inside a hardware group) we can
now assume that all events on a particular context belong to the same
pmu, hence we can disable the pmu for the entire context operations.
This reduces the amount of hardware writes.
The exception for swevents comes from the fact that the sw pmu disable
is a nop.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: paulus <paulus@samba.org>
Cc: stephane eranian <eranian@googlemail.com>
Cc: Robert Richter <robert.richter@amd.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Lin Ming <ming.m.lin@intel.com>
Cc: Yanmin <yanmin_zhang@linux.intel.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
diff --git a/kernel/perf_event.c b/kernel/perf_event.c
index 357ee8d..9819a69 100644
--- a/kernel/perf_event.c
+++ b/kernel/perf_event.c
@@ -1065,6 +1065,7 @@
struct perf_event *event;
raw_spin_lock(&ctx->lock);
+ perf_pmu_disable(ctx->pmu);
ctx->is_active = 0;
if (likely(!ctx->nr_events))
goto out;
@@ -1083,6 +1084,7 @@
group_sched_out(event, cpuctx, ctx);
}
out:
+ perf_pmu_enable(ctx->pmu);
raw_spin_unlock(&ctx->lock);
}
@@ -1400,6 +1402,7 @@
if (cpuctx->task_ctx == ctx)
return;
+ perf_pmu_disable(ctx->pmu);
/*
* We want to keep the following priority order:
* cpu pinned (that don't need to move), task pinned,
@@ -1418,6 +1421,7 @@
* cpu-context we got scheduled on is actually rotating.
*/
perf_pmu_rotate_start(ctx->pmu);
+ perf_pmu_enable(ctx->pmu);
}
/*
@@ -1629,6 +1633,7 @@
rotate = 1;
}
+ perf_pmu_disable(cpuctx->ctx.pmu);
perf_ctx_adjust_freq(&cpuctx->ctx, cpuctx->timer_interval);
if (ctx)
perf_ctx_adjust_freq(ctx, cpuctx->timer_interval);
@@ -1649,6 +1654,7 @@
task_ctx_sched_in(ctx, EVENT_FLEXIBLE);
done:
+ perf_pmu_enable(cpuctx->ctx.pmu);
hrtimer_forward_now(timer, ns_to_ktime(cpuctx->timer_interval));
return restart;