sched/cpufreq: Align trace event behavior of fast switching
authorDouglas RAILLARD <douglas.raillard@arm.com>
Wed, 7 Aug 2019 15:33:40 +0000 (16:33 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Sat, 5 Oct 2019 11:09:51 +0000 (13:09 +0200)
[ Upstream commit 77c84dd1881d0f0176cb678d770bfbda26c54390 ]

Fast switching path only emits an event for the CPU of interest, whereas the
regular path emits an event for all the CPUs that had their frequency changed,
i.e. all the CPUs sharing the same policy.

With the current behavior, looking at cpu_frequency event for a given CPU that
is using the fast switching path will not give the correct frequency signal.

Signed-off-by: Douglas RAILLARD <douglas.raillard@arm.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
kernel/sched/cpufreq_schedutil.c

index 64d54acc992829d98c65e7371b9967203e1b08c8..54fcff656ecd7f250f5770ae8973bc0c3d48cd8d 100644 (file)
@@ -118,6 +118,7 @@ static void sugov_fast_switch(struct sugov_policy *sg_policy, u64 time,
                              unsigned int next_freq)
 {
        struct cpufreq_policy *policy = sg_policy->policy;
+       int cpu;
 
        if (!sugov_update_next_freq(sg_policy, time, next_freq))
                return;
@@ -127,7 +128,11 @@ static void sugov_fast_switch(struct sugov_policy *sg_policy, u64 time,
                return;
 
        policy->cur = next_freq;
-       trace_cpu_frequency(next_freq, smp_processor_id());
+
+       if (trace_cpu_frequency_enabled()) {
+               for_each_cpu(cpu, policy->cpus)
+                       trace_cpu_frequency(next_freq, cpu);
+       }
 }
 
 static void sugov_deferred_update(struct sugov_policy *sg_policy, u64 time,