Re: [PATCH V4 1/2] perf/x86/intel/ds: Flush the PEBS buffer in PEBS enable

[Peter Zijlstra <peterz@infradead.org>]

On Fri, Apr 21, 2023 at 11:45:28AM -0700, kan.liang@linux.intel.com wrote:
> From: Kan Liang <kan.liang@linux.intel.com>
> 
> Several similar kernel warnings can be triggered,
> 
>   [56605.607840] CPU0 PEBS record size 0, expected 32, config 0
>   cpuc->record_size=208
> 
> when the below commands are running in parallel for a while on SPR.
> 
>   while true; do perf record --no-buildid -a --intr-regs=AX -e
>   cpu/event=0xd0,umask=0x81/pp -c 10003 -o /dev/null ./triad; done &
> 
>   while true; do perf record -o /tmp/out -W -d -e
>   '{ld_blocks.store_forward:period=1000000,
>   MEM_TRANS_RETIRED.LOAD_LATENCY:u:precise=2:ldlat=4}'
>   -c 1037 ./triad; done
>   *The triad program is just the generation of loads/stores.
> 
> The warnings are triggered when an unexpected PEBS record (with a
> different config and size) is found.
> 
> A system-wide PEBS event with the large PEBS config may be enabled
> during a context switch. Some PEBS records for the system-wide PEBS may
> be generated while the old task is sched out but the new one hasn't been
> sched in yet. When the new task is sched in, the cpuc->pebs_record_size
> may be updated for the per-task PEBS events. So the existing system-wide
> PEBS records have a different size from the later PEBS records.
> 
> The PEBS buffer should be flushed right before the hardware is
> reprogrammed. The new size and threshold should be updated after the old
> buffer has been flushed.
> 
> Reported-by: Stephane Eranian <eranian@google.com>
> Suggested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> Signed-off-by: Kan Liang <kan.liang@linux.intel.com>
> ---

So I find it much easier to read the whole thing when collapsed.
Something like the below; that ok with you?

---
 arch/x86/events/intel/ds.c        | 56 ++++++++++++++++++++++-----------------
 arch/x86/include/asm/perf_event.h |  3 +++
 2 files changed, 35 insertions(+), 24 deletions(-)

diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c
index a2e566e53076..df88576d6b2a 100644
--- a/arch/x86/events/intel/ds.c
+++ b/arch/x86/events/intel/ds.c
@@ -1229,12 +1229,14 @@ pebs_update_state(bool needed_cb, struct cpu_hw_events *cpuc,
 		  struct perf_event *event, bool add)
 {
 	struct pmu *pmu = event->pmu;
+
 	/*
 	 * Make sure we get updated with the first PEBS
 	 * event. It will trigger also during removal, but
 	 * that does not hurt:
 	 */
-	bool update = cpuc->n_pebs == 1;
+	if (cpuc->n_pebs == 1)
+		cpuc->pebs_data_cfg = PEBS_UPDATE_DS_SW;
 
 	if (needed_cb != pebs_needs_sched_cb(cpuc)) {
 		if (!needed_cb)
@@ -1242,7 +1244,7 @@ pebs_update_state(bool needed_cb, struct cpu_hw_events *cpuc,
 		else
 			perf_sched_cb_dec(pmu);
 
-		update = true;
+		cpuc->pebs_data_cfg |= PEBS_UPDATE_DS_SW;
 	}
 
 	/*
@@ -1252,24 +1254,13 @@ pebs_update_state(bool needed_cb, struct cpu_hw_events *cpuc,
 	if (x86_pmu.intel_cap.pebs_baseline && add) {
 		u64 pebs_data_cfg;
 
-		/* Clear pebs_data_cfg and pebs_record_size for first PEBS. */
-		if (cpuc->n_pebs == 1) {
-			cpuc->pebs_data_cfg = 0;
-			cpuc->pebs_record_size = sizeof(struct pebs_basic);
-		}
-
 		pebs_data_cfg = pebs_update_adaptive_cfg(event);
-
-		/* Update pebs_record_size if new event requires more data. */
-		if (pebs_data_cfg & ~cpuc->pebs_data_cfg) {
-			cpuc->pebs_data_cfg |= pebs_data_cfg;
-			adaptive_pebs_record_size_update();
-			update = true;
-		}
+		/*
+		 * Be sure to update the thresholds when we change the record.
+		 */
+		if (pebs_data_cfg & ~cpuc->pebs_data_cfg)
+			cpuc->pebs_data_cfg |= pebs_data_cfg | PEBS_UPDATE_DS_SW;
 	}
-
-	if (update)
-		pebs_update_threshold(cpuc);
 }
 
 void intel_pmu_pebs_add(struct perf_event *event)
@@ -1326,9 +1317,17 @@ static void intel_pmu_pebs_via_pt_enable(struct perf_event *event)
 	wrmsrl(base + idx, value);
 }
 
+static inline void intel_pmu_drain_large_pebs(struct cpu_hw_events *cpuc)
+{
+	if (cpuc->n_pebs == cpuc->n_large_pebs &&
+	    cpuc->n_pebs != cpuc->n_pebs_via_pt)
+		intel_pmu_drain_pebs_buffer();
+}
+
 void intel_pmu_pebs_enable(struct perf_event *event)
 {
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
+	u64 pebs_data_cfg = cpuc->pebs_data_cfg & ~PEBS_UPDATE_DS_SW;
 	struct hw_perf_event *hwc = &event->hw;
 	struct debug_store *ds = cpuc->ds;
 	unsigned int idx = hwc->idx;
@@ -1344,11 +1343,22 @@ void intel_pmu_pebs_enable(struct perf_event *event)
 
 	if (x86_pmu.intel_cap.pebs_baseline) {
 		hwc->config |= ICL_EVENTSEL_ADAPTIVE;
-		if (cpuc->pebs_data_cfg != cpuc->active_pebs_data_cfg) {
-			wrmsrl(MSR_PEBS_DATA_CFG, cpuc->pebs_data_cfg);
-			cpuc->active_pebs_data_cfg = cpuc->pebs_data_cfg;
+		if (pebs_data_cfg != cpuc->active_pebs_data_cfg) {
+			/*
+			 * drain_pebs() assumes uniform record size;
+			 * hence we need to drain when changing said
+			 * size.
+			 */
+			intel_pmu_drain_large_pebs(cpuc);
+			adaptive_pebs_record_size_update();
+			wrmsrl(MSR_PEBS_DATA_CFG, pebs_data_cfg);
+			cpuc->active_pebs_data_cfg = pebs_data_cfg;
 		}
 	}
+	if (cpuc->pebs_data_cfg & PEBS_UPDATE_DS_SW) {
+		cpuc->pebs_data_cfg = pebs_data_cfg;
+		pebs_update_threshold(cpuc);
+	}
 
 	if (idx >= INTEL_PMC_IDX_FIXED) {
 		if (x86_pmu.intel_cap.pebs_format < 5)
@@ -1391,9 +1401,7 @@ void intel_pmu_pebs_disable(struct perf_event *event)
 	struct cpu_hw_events *cpuc = this_cpu_ptr(&cpu_hw_events);
 	struct hw_perf_event *hwc = &event->hw;
 
-	if (cpuc->n_pebs == cpuc->n_large_pebs &&
-	    cpuc->n_pebs != cpuc->n_pebs_via_pt)
-		intel_pmu_drain_pebs_buffer();
+	intel_pmu_drain_large_pebs(cpuc);
 
 	cpuc->pebs_enabled &= ~(1ULL << hwc->idx);
 
diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h
index 8fc15ed5e60b..abf09882f58b 100644
--- a/arch/x86/include/asm/perf_event.h
+++ b/arch/x86/include/asm/perf_event.h
@@ -121,6 +121,9 @@
 #define PEBS_DATACFG_LBRS	BIT_ULL(3)
 #define PEBS_DATACFG_LBR_SHIFT	24
 
+/* Steal the highest bit of pebs_data_cfg for SW usage */
+#define PEBS_UPDATE_DS_SW	BIT_ULL(63)
+
 /*
  * Intel "Architectural Performance Monitoring" CPUID
  * detection/enumeration details: