Skip to contentSkip to conversation

perf tool: don't print bogus data on -e instructions

PATCH In review Ramkumar Ramachandra 3 participants 3 messages Oct 1, 2013 – Oct 15, 2013 perf events

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 perf tool: don't print bogus data on -e instructions

When only the instructions event is requested:

$ perf stat -e instructions git s
M  builtin-stat.c

Performance counter stats for 'git s':

917,453,420 instructions # 0.00 insns per cycle

0.213002926 seconds time elapsed

The 0.00 insns per cycle comment in the output is totally bogus and misleading. It happens because update_shadow_stats() doesn't touch runtime_cycles_stats when only the instructions event is requested. So, omit printing the bogus data altogether.

Cc Ingo Molnar <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Paul Mackerras <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>
---
 tools/perf/builtin-stat.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
tools/perf/builtin-stat.c +3 -4
index cc167ae..0b747a5 100644
@@ -852,11 +852,10 @@ static void abs_printout(int cpu, int nr, struct perf_evsel *evsel, double avg)
 
 	if (perf_evsel__match(evsel, HARDWARE, HW_INSTRUCTIONS)) {
 		total = avg_stats(&runtime_cycles_stats[cpu]);
-		if (total)
+		if (total) {
 			ratio = avg / total;
-
-		fprintf(output, " #   %5.2f  insns per cycle        ", ratio);
-
+			fprintf(output, " #   %5.2f  insns per cycle        ", ratio);
+		}
 		total = avg_stats(&runtime_stalled_cycles_front_stats[cpu]);
 		total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[cpu]));
 
Signature · 1 line
1.8.4.477.g5d89aa9
Acked-by on v1 1/1
#1 lore raw
Quoted 16 lines · Ramkumar Ramachandra

When only the instructions event is requested:

$ perf stat -e instructions git s
M  builtin-stat.c

Performance counter stats for 'git s':

917,453,420 instructions # 0.00 insns per cycle

0.213002926 seconds time elapsed

The 0.00 insns per cycle comment in the output is totally bogus and misleading. It happens because update_shadow_stats() doesn't touch runtime_cycles_stats when only the instructions event is requested. So, omit printing the bogus data altogether.

Quoted 5 lines
Cc Ingo Molnar <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Paul Mackerras <[email protected]>
Cc Arnaldo Carvalho de Melo <[email protected]>

Thanks,

Ingo

#2 lore raw

[tip:perf/core] perf stat: Don' t print bogus data on -e instructions

In reply to Ramkumar Ramachandra
perf stat: Don't print bogus data on -e instructions
Commit-ID:  3e7a081796146f97f166d77a655c0eb585065077
Gitweb:     http://git.kernel.org/tip/3e7a081796146f97f166d77a655c0eb585065077
Author:     Ramkumar Ramachandra <[email protected]>
AuthorDate: Tue, 1 Oct 2013 14:06:44 +0530
Committer:  Arnaldo Carvalho de Melo <[email protected]>
CommitDate: Fri, 11 Oct 2013 12:17:35 -0300

perf stat: Don't print bogus data on -e instructions

When only the instructions event is requested:

$ perf stat -e instructions git s
M  builtin-stat.c

Performance counter stats for 'git s':

917,453,420 instructions # 0.00 insns per cycle

0.213002926 seconds time elapsed

The 0.00 insns per cycle comment in the output is totally bogus and misleading. It happens because update_shadow_stats() doesn't touch runtime_cycles_stats when only the instructions event is requested. So, omit printing the bogus data altogether.

---
 tools/perf/builtin-stat.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
tools/perf/builtin-stat.c +3 -4
index ce2266c..fb02b53 100644
@@ -930,11 +930,10 @@ static void abs_printout(int cpu, int nr, struct perf_evsel *evsel, double avg)
 
 	if (perf_evsel__match(evsel, HARDWARE, HW_INSTRUCTIONS)) {
 		total = avg_stats(&runtime_cycles_stats[cpu]);
-		if (total)
+		if (total) {
 			ratio = avg / total;
-
-		fprintf(output, " #   %5.2f  insns per cycle        ", ratio);
-
+			fprintf(output, " #   %5.2f  insns per cycle        ", ratio);
+		}
 		total = avg_stats(&runtime_stalled_cycles_front_stats[cpu]);
 		total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[cpu]));