powerpc/perf: Enable PMU counters post partition migration if PMU is …

…active

During Live Partition Migration (LPM), it is observed that after
migration completion, perf counter values reports 0 incase of
system/cpu wide monitoring. However 'perf stat' with workload
continues to show counts post migration since PMU gets disabled/enabled
during sched switches.

Example:
 ./perf stat -e r1001e -I 1000
           time             counts unit events
     1.001010437         22,137,414      r1001e
     2.002495447         15,455,821      r1001e
<<>> As seen in next below logs, the counter values shows zero
        after migration is completed.
<<>>
    86.142535370    129,392,333,440      r1001e
    87.144714617                  0      r1001e
    88.146526636                  0      r1001e
    89.148085029                  0      r1001e

Here PMU is enabled during start of perf session and counter
values are read at intervals. Counters are only disabled at the
end of session. The powerpc mobility code presently does not handle
disabling and enabling back of PMU counters during partition
migration. Also since the PMU register values are not saved/restored
during migration, PMU registers like Monitor Mode Control Register 0
(MMCR0), Monitor Mode Control Register 1 (MMCR1) will not contain
the value it was programmed with. Hence PMU counters will not be
enabled correctly post migration.

Fix this in mobility code by handling disabling and enabling of
PMU in all cpu's before and after migration. Patch introduces two
functions 'mobility_pmu_disable' and 'mobility_pmu_enable'.
mobility_pmu_disable() is called before the processor threads goes
to suspend state so as to disable the PMU counters. And disable is
done only if there are any active events running on that cpu.
mobility_pmu_enable() is called after the processor threads are
back online to enable back the PMU counters.

Since the performance Monitor counters ( PMCs) are not
saved/restored during LPM, results in PMC value being zero and the
'event->hw.prev_count' being non-zero value. This causes problem
during updation of event->count since we always accumulate
(event->hw.prev_count - PMC value) in event->count.  If
event->hw.prev_count is greater PMC value, event->count becomes
negative. Fix this by re-initialising 'prev_count' also for all
events while enabling back the events. A new variable 'migrate' is
introduced in 'struct cpu_hw_event' to achieve this for LPM cases
in power_pmu_enable. Use the 'migrate' value to clear the PMC
index (stored in event->hw.idx) for all events so that event
count settings will get re-initialised correctly.

Signed-off-by: Athira Rajeev <atrajeev@linux.vnet.ibm.com>

arch/powerpc/include/asm/rtas.h

@@ -376,8 +376,12 @@ extern int call_rtas(const char *, int, int, unsigned long *, ...);
 
 #ifdef CONFIG_HV_PERF_CTRS
 void read_24x7_sys_info(void);
+void mobility_pmu_disable(void);
+void mobility_pmu_enable(void);
 #else
 static inline void read_24x7_sys_info(void) { }
+static inline void mobility_pmu_disable(void) { }
+static inline void mobility_pmu_enable(void) { }
 #endif
 
 #endif /* __KERNEL__ */

arch/powerpc/perf/core-book3s.c

@@ -58,6 +58,7 @@ struct cpu_hw_events {
 
 	/* Store the PMC values */
 	unsigned long pmcs[MAX_HWEVENTS];
+	int migrate;
 };
 
 static DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events);
@@ -1334,6 +1335,22 @@ static void power_pmu_disable(struct pmu *pmu)
 	local_irq_restore(flags);
 }
 
+/*
+ * Called from powerpc mobility code
+ * before migration to disable counters
+ * if the PMU is active.
+ */
+void mobility_pmu_disable(void)
+{
+	struct cpu_hw_events *cpuhw;
+
+	cpuhw = this_cpu_ptr(&cpu_hw_events);
+	if (cpuhw->n_events != 0) {
+		power_pmu_disable(NULL);
+		cpuhw->migrate = 1;
+	}
+}
+
 /*
  * Re-enable all events if disable == 0.
  * If we were previously disabled and events were added, then
@@ -1379,8 +1396,10 @@ static void power_pmu_enable(struct pmu *pmu)
 	 * no need to recalculate MMCR* settings and reset the PMCs.
 	 * Just reenable the PMU with the current MMCR* settings
 	 * (possibly updated for removal of events).
+	 * While reenabling PMU during partition migration, continue
+	 * with normal flow.
 	 */
-	if (!cpuhw->n_added) {
+	if (!cpuhw->n_added && !cpuhw->migrate) {
 		mtspr(SPRN_MMCRA, cpuhw->mmcr.mmcra & ~MMCRA_SAMPLE_ENABLE);
 		mtspr(SPRN_MMCR1, cpuhw->mmcr.mmcr1);
 		if (ppmu->flags & PPMU_ARCH_31)
@@ -1434,11 +1453,15 @@ static void power_pmu_enable(struct pmu *pmu)
 	/*
 	 * Read off any pre-existing events that need to move
 	 * to another PMC.
+	 * While enabling PMU during partition migration,
+	 * skip power_pmu_read since all event count settings
+	 * needs to be re-initialised after migration.
 	 */
 	for (i = 0; i < cpuhw->n_events; ++i) {
 		event = cpuhw->event[i];
-		if (event->hw.idx && event->hw.idx != hwc_index[i] + 1) {
-			power_pmu_read(event);
+		if ((event->hw.idx && event->hw.idx != hwc_index[i] + 1) || (cpuhw->migrate)) {
+			if (!cpuhw->migrate)
+				power_pmu_read(event);
 			write_pmc(event->hw.idx, 0);
 			event->hw.idx = 0;
 		}
@@ -1506,6 +1529,20 @@ static void power_pmu_enable(struct pmu *pmu)
 	local_irq_restore(flags);
 }
 
+/*
+ * Called from powerpc mobility code
+ * during migration completion to
+ * enable back PMU counters.
+ */
+void mobility_pmu_enable(void)
+{
+	struct cpu_hw_events *cpuhw;
+
+	cpuhw = this_cpu_ptr(&cpu_hw_events);
+	power_pmu_enable(NULL);
+	cpuhw->migrate = 0;
+}
+
 static int collect_events(struct perf_event *group, int max_count,
 			  struct perf_event *ctrs[], u64 *events,
 			  unsigned int *flags)

arch/powerpc/platforms/pseries/mobility.c

@@ -476,6 +476,8 @@ static int do_join(void *arg)
 retry:
 	/* Must ensure MSR.EE off for H_JOIN. */
 	hard_irq_disable();
+	/* Disable PMU before suspend */
+	mobility_pmu_disable();
 	hvrc = plpar_hcall_norets(H_JOIN);
 
 	switch (hvrc) {
@@ -530,6 +532,8 @@ static int do_join(void *arg)
 	 * reset the watchdog.
 	 */
 	touch_nmi_watchdog();
+	/* Enable PMU after resuming */
+	mobility_pmu_enable();
 	return ret;
 }