Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1411781 > unrolled thread

[PATCH 3/3] perf, perf/tool: trigger lbr_from signext bug

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [PATCH 3/3] perf, perf/tool: trigger lbr_from signext bug David Carrillo-Cisneros <davidcc@google.com> - 2016-06-02 04:50 +0200

#1411781 — [PATCH 3/3] perf, perf/tool: trigger lbr_from signext bug

Change kernel and perf tool to activate tracking and context
switch for kernel branches.

Signed-off-by: David Carrillo-Cisneros <davidcc@google.com>
---
 arch/x86/events/intel/lbr.c |  3 ++-
 tools/perf/util/evsel.c     | 17 ++++++++++++++---
 2 files changed, 16 insertions(+), 4 deletions(-)

diff --git a/arch/x86/events/intel/lbr.c b/arch/x86/events/intel/lbr.c
index 6aa2d8a..d808e54 100644
--- a/arch/x86/events/intel/lbr.c
+++ b/arch/x86/events/intel/lbr.c
@@ -384,7 +384,8 @@ void intel_pmu_lbr_sched_task(struct perf_event_context *ctx, bool sched_in)
 
 static inline bool branch_user_callstack(unsigned br_sel)
 {
-	return (br_sel & X86_BR_USER) && (br_sel & X86_BR_CALL_STACK);
+	return (br_sel & (X86_BR_USER | X86_BR_KERNEL)) &&
+	       (br_sel & X86_BR_CALL_STACK);
 }
 
 void intel_pmu_lbr_enable(struct perf_event *event)
diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
index 5d7037e..b39aafc 100644
--- a/tools/perf/util/evsel.c
+++ b/tools/perf/util/evsel.c
@@ -575,9 +575,20 @@ void perf_evsel__config_callchain(struct perf_evsel *evsel,
 	if (param->record_mode == CALLCHAIN_LBR) {
 		if (!opts->branch_stack) {
 			if (attr->exclude_user) {
-				pr_warning("LBR callstack option is only available "
-					   "to get user callchain information. "
-					   "Falling back to framepointers.\n");
+				/*
+				 * Sample kernel AND user branches. Do not exclude user
+				 * branches because task_ctx->lbr_callstack_users in
+				 * arch/x86/events/intel/lbr.c only count users for
+				 * user branches.
+				 */
+				pr_warning("BRANCH_STACK with kernel and user\n");
+				perf_evsel__set_sample_bit(evsel, BRANCH_STACK);
+				attr->branch_sample_type = PERF_SAMPLE_BRANCH_KERNEL |
+							PERF_SAMPLE_BRANCH_USER |
+							PERF_SAMPLE_BRANCH_CALL_STACK |
+							PERF_SAMPLE_BRANCH_NO_CYCLES |
+							PERF_SAMPLE_BRANCH_NO_FLAGS;
+
 			} else {
 				perf_evsel__set_sample_bit(evsel, BRANCH_STACK);
 				attr->branch_sample_type = PERF_SAMPLE_BRANCH_USER |
-- 
2.8.0.rc3.226.g39d4020

[toc] | [standalone]

Back to top | Article view | linux.kernel

csiph-web