* [PATCH] perf stat: improve readability of shadow stats
@ 2021-03-01 17:24 Changbin Du
2021-03-01 19:19 ` Jiri Olsa
2021-03-02 6:53 ` kajoljain
0 siblings, 2 replies; 7+ messages in thread
From: Changbin Du @ 2021-03-01 17:24 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo, Peter Zijlstra
Cc: Ingo Molnar, Jiri Olsa, Namhyung Kim, linux-kernel, Changbin Du
This does follow two changes:
1) Select appropriate unit between K/M/G.
2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
$ sudo ./perf stat -a -- sleep 1
Before: Unit 'M' is selected even the number is very small.
Performance counter stats for 'system wide':
4,003.06 msec cpu-clock # 3.998 CPUs utilized
16,179 context-switches # 0.004 M/sec
161 cpu-migrations # 0.040 K/sec
4,699 page-faults # 0.001 M/sec
6,135,801,925 cycles # 1.533 GHz (83.21%)
5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
4,720,130,587 instructions # 0.77 insn per cycle
# 1.23 stalled cycles per insn (83.28%)
753,848,078 branches # 188.318 M/sec (83.61%)
37,457,747 branch-misses # 4.97% of all branches (83.48%)
1.001283725 seconds time elapsed
After:
$ sudo ./perf stat -a -- sleep 2
Performance counter stats for 'system wide':
8,003.20 msec cpu-clock # 3.998 CPUs utilized
9,768 context-switches # 1.221 K/cpu-sec
164 cpu-migrations # 20.492 /cpu-sec
74,146 page-faults # 9.265 K/cpu-sec
19,008,796,806 cycles # 2.375 GHz (83.21%)
14,789,443,853 stalled-cycles-frontend # 77.80% frontend cycles idle (83.29%)
11,867,812,064 stalled-cycles-backend # 62.43% backend cycles idle (66.80%)
9,898,252,603 instructions # 0.52 insn per cycle
# 1.49 stalled cycles per insn (83.41%)
2,063,251,998 branches # 257.803 M/cpu-sec (83.41%)
86,941,704 branch-misses # 4.21% of all branches (83.30%)
2.001743706 seconds time elapsed
Signed-off-by: Changbin Du <changbin.du@gmail.com>
---
tools/perf/util/stat-shadow.c | 13 +++++--------
tools/perf/util/units.c | 22 ++++++++++++++++++++++
tools/perf/util/units.h | 1 +
3 files changed, 28 insertions(+), 8 deletions(-)
diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
index 6ccf21a72f06..786b5ef512d8 100644
--- a/tools/perf/util/stat-shadow.c
+++ b/tools/perf/util/stat-shadow.c
@@ -9,6 +9,7 @@
#include "expr.h"
#include "metricgroup.h"
#include "cgroup.h"
+#include "units.h"
#include <linux/zalloc.h>
/*
@@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
} else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
- char unit = 'M';
+ char unit = ' ';
char unit_buf[10];
total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
-
if (total)
- ratio = 1000.0 * avg / total;
- if (ratio < 0.001) {
- ratio *= 1000;
- unit = 'K';
- }
- snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
+ ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
+
+ snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
} else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
print_smi_cost(config, cpu, out, st, &rsd);
diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
index a46762aec4c9..ac13b5ecde31 100644
--- a/tools/perf/util/units.c
+++ b/tools/perf/util/units.c
@@ -55,6 +55,28 @@ unsigned long convert_unit(unsigned long value, char *unit)
return value;
}
+double convert_unit_double(double value, char *unit)
+{
+ *unit = ' ';
+
+ if (value > 1000.0) {
+ value /= 1000.0;
+ *unit = 'K';
+ }
+
+ if (value > 1000.0) {
+ value /= 1000.0;
+ *unit = 'M';
+ }
+
+ if (value > 1000.0) {
+ value /= 1000.0;
+ *unit = 'G';
+ }
+
+ return value;
+}
+
int unit_number__scnprintf(char *buf, size_t size, u64 n)
{
char unit[4] = "BKMG";
diff --git a/tools/perf/util/units.h b/tools/perf/util/units.h
index 99263b6a23f7..b3ace67ac16f 100644
--- a/tools/perf/util/units.h
+++ b/tools/perf/util/units.h
@@ -13,6 +13,7 @@ struct parse_tag {
unsigned long parse_tag_value(const char *str, struct parse_tag *tags);
unsigned long convert_unit(unsigned long value, char *unit);
+double convert_unit_double(double value, char *unit);
int unit_number__scnprintf(char *buf, size_t size, u64 n);
#endif /* PERF_UNIT_H */
--
2.25.1
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH] perf stat: improve readability of shadow stats
2021-03-01 17:24 [PATCH] perf stat: improve readability of shadow stats Changbin Du
@ 2021-03-01 19:19 ` Jiri Olsa
2021-03-02 2:08 ` Namhyung Kim
2021-03-04 16:22 ` Changbin Du
2021-03-02 6:53 ` kajoljain
1 sibling, 2 replies; 7+ messages in thread
From: Jiri Olsa @ 2021-03-01 19:19 UTC (permalink / raw)
To: Changbin Du
Cc: Arnaldo Carvalho de Melo, Peter Zijlstra, Ingo Molnar,
Namhyung Kim, linux-kernel, Ian Rogers, Andi Kleen
On Tue, Mar 02, 2021 at 01:24:02AM +0800, Changbin Du wrote:
> This does follow two changes:
> 1) Select appropriate unit between K/M/G.
> 2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
>
> $ sudo ./perf stat -a -- sleep 1
>
> Before: Unit 'M' is selected even the number is very small.
> Performance counter stats for 'system wide':
>
> 4,003.06 msec cpu-clock # 3.998 CPUs utilized
> 16,179 context-switches # 0.004 M/sec
> 161 cpu-migrations # 0.040 K/sec
> 4,699 page-faults # 0.001 M/sec
> 6,135,801,925 cycles # 1.533 GHz (83.21%)
> 5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
> 4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
> 4,720,130,587 instructions # 0.77 insn per cycle
> # 1.23 stalled cycles per insn (83.28%)
> 753,848,078 branches # 188.318 M/sec (83.61%)
> 37,457,747 branch-misses # 4.97% of all branches (83.48%)
>
> 1.001283725 seconds time elapsed
>
> After:
> $ sudo ./perf stat -a -- sleep 2
>
> Performance counter stats for 'system wide':
>
> 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> 9,768 context-switches # 1.221 K/cpu-sec
> 164 cpu-migrations # 20.492 /cpu-sec
should you remove also the leading '/' in ' /cpu-sec' ?
SNIP
> @@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
> generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
> evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
> } else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
> - char unit = 'M';
> + char unit = ' ';
> char unit_buf[10];
>
> total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
> -
> if (total)
> - ratio = 1000.0 * avg / total;
> - if (ratio < 0.001) {
> - ratio *= 1000;
> - unit = 'K';
> - }
> - snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
> + ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
> +
> + snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
> print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
hum this will change -x output that people parse, so I don't think we can do that
> } else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
> print_smi_cost(config, cpu, out, st, &rsd);
> diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
> index a46762aec4c9..ac13b5ecde31 100644
> --- a/tools/perf/util/units.c
> +++ b/tools/perf/util/units.c
> @@ -55,6 +55,28 @@ unsigned long convert_unit(unsigned long value, char *unit)
> return value;
> }
>
> +double convert_unit_double(double value, char *unit)
> +{
> + *unit = ' ';
> +
> + if (value > 1000.0) {
> + value /= 1000.0;
> + *unit = 'K';
> + }
> +
> + if (value > 1000.0) {
> + value /= 1000.0;
> + *unit = 'M';
> + }
> +
> + if (value > 1000.0) {
> + value /= 1000.0;
> + *unit = 'G';
> + }
> +
> + return value;
> +}
we have convert_unit function just above doing the same only with
unsigned long.. let's have one base function with double values and
another one casting the result to unsigned long
jirka
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] perf stat: improve readability of shadow stats
2021-03-01 19:19 ` Jiri Olsa
@ 2021-03-02 2:08 ` Namhyung Kim
2021-03-02 12:25 ` Arnaldo Carvalho de Melo
2021-03-04 16:22 ` Changbin Du
1 sibling, 1 reply; 7+ messages in thread
From: Namhyung Kim @ 2021-03-02 2:08 UTC (permalink / raw)
To: Jiri Olsa
Cc: Changbin Du, Arnaldo Carvalho de Melo, Peter Zijlstra,
Ingo Molnar, linux-kernel, Ian Rogers, Andi Kleen
On Tue, Mar 2, 2021 at 4:19 AM Jiri Olsa <jolsa@redhat.com> wrote:
>
> On Tue, Mar 02, 2021 at 01:24:02AM +0800, Changbin Du wrote:
> > This does follow two changes:
> > 1) Select appropriate unit between K/M/G.
> > 2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
> >
> > $ sudo ./perf stat -a -- sleep 1
> >
> > Before: Unit 'M' is selected even the number is very small.
> > Performance counter stats for 'system wide':
> >
> > 4,003.06 msec cpu-clock # 3.998 CPUs utilized
> > 16,179 context-switches # 0.004 M/sec
> > 161 cpu-migrations # 0.040 K/sec
> > 4,699 page-faults # 0.001 M/sec
> > 6,135,801,925 cycles # 1.533 GHz (83.21%)
> > 5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
> > 4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
> > 4,720,130,587 instructions # 0.77 insn per cycle
> > # 1.23 stalled cycles per insn (83.28%)
> > 753,848,078 branches # 188.318 M/sec (83.61%)
> > 37,457,747 branch-misses # 4.97% of all branches (83.48%)
> >
> > 1.001283725 seconds time elapsed
> >
> > After:
> > $ sudo ./perf stat -a -- sleep 2
> >
> > Performance counter stats for 'system wide':
> >
> > 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> > 9,768 context-switches # 1.221 K/cpu-sec
> > 164 cpu-migrations # 20.492 /cpu-sec
>
> should you remove also the leading '/' in ' /cpu-sec' ?
The change looks good. And I think we should keep '/' otherwise it'd be
more confusing.
>
>
> SNIP
>
> > @@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
> > generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
> > evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
> > } else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
> > - char unit = 'M';
> > + char unit = ' ';
> > char unit_buf[10];
> >
> > total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
> > -
> > if (total)
> > - ratio = 1000.0 * avg / total;
> > - if (ratio < 0.001) {
> > - ratio *= 1000;
> > - unit = 'K';
> > - }
> > - snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
> > + ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
> > +
> > + snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
> > print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
>
> hum this will change -x output that people parse, so I don't think we can do that
Agreed.
>
> > } else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
> > print_smi_cost(config, cpu, out, st, &rsd);
> > diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
> > index a46762aec4c9..ac13b5ecde31 100644
> > --- a/tools/perf/util/units.c
> > +++ b/tools/perf/util/units.c
> > @@ -55,6 +55,28 @@ unsigned long convert_unit(unsigned long value, char *unit)
> > return value;
> > }
> >
> > +double convert_unit_double(double value, char *unit)
> > +{
> > + *unit = ' ';
> > +
> > + if (value > 1000.0) {
> > + value /= 1000.0;
> > + *unit = 'K';
> > + }
> > +
> > + if (value > 1000.0) {
> > + value /= 1000.0;
> > + *unit = 'M';
> > + }
> > +
> > + if (value > 1000.0) {
> > + value /= 1000.0;
> > + *unit = 'G';
> > + }
> > +
> > + return value;
> > +}
>
> we have convert_unit function just above doing the same only with
> unsigned long.. let's have one base function with double values and
> another one casting the result to unsigned long
Sounds good.
Thanks,
Namhyung
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] perf stat: improve readability of shadow stats
2021-03-01 17:24 [PATCH] perf stat: improve readability of shadow stats Changbin Du
2021-03-01 19:19 ` Jiri Olsa
@ 2021-03-02 6:53 ` kajoljain
1 sibling, 0 replies; 7+ messages in thread
From: kajoljain @ 2021-03-02 6:53 UTC (permalink / raw)
To: Changbin Du, Arnaldo Carvalho de Melo, Peter Zijlstra
Cc: Ingo Molnar, Jiri Olsa, Namhyung Kim, linux-kernel
On 3/1/21 10:54 PM, Changbin Du wrote:
> This does follow two changes:
> 1) Select appropriate unit between K/M/G.
> 2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
>
> $ sudo ./perf stat -a -- sleep 1
>
> Before: Unit 'M' is selected even the number is very small.
> Performance counter stats for 'system wide':
>
> 4,003.06 msec cpu-clock # 3.998 CPUs utilized
> 16,179 context-switches # 0.004 M/sec
> 161 cpu-migrations # 0.040 K/sec
> 4,699 page-faults # 0.001 M/sec
> 6,135,801,925 cycles # 1.533 GHz (83.21%)
> 5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
> 4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
> 4,720,130,587 instructions # 0.77 insn per cycle
> # 1.23 stalled cycles per insn (83.28%)
> 753,848,078 branches # 188.318 M/sec (83.61%)
> 37,457,747 branch-misses # 4.97% of all branches (83.48%)
>
> 1.001283725 seconds time elapsed
>
> After:
> $ sudo ./perf stat -a -- sleep 2
>
> Performance counter stats for 'system wide':
>
> 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> 9,768 context-switches # 1.221 K/cpu-sec
> 164 cpu-migrations # 20.492 /cpu-sec
> 74,146 page-faults # 9.265 K/cpu-sec
> 19,008,796,806 cycles # 2.375 GHz (83.21%)
> 14,789,443,853 stalled-cycles-frontend # 77.80% frontend cycles idle (83.29%)
> 11,867,812,064 stalled-cycles-backend # 62.43% backend cycles idle (66.80%)
> 9,898,252,603 instructions # 0.52 insn per cycle
> # 1.49 stalled cycles per insn (83.41%)
> 2,063,251,998 branches # 257.803 M/cpu-sec (83.41%)
> 86,941,704 branch-misses # 4.21% of all branches (83.30%)
>
> 2.001743706 seconds time elapsed
>
> Signed-off-by: Changbin Du <changbin.du@gmail.com>
> ---
> tools/perf/util/stat-shadow.c | 13 +++++--------
> tools/perf/util/units.c | 22 ++++++++++++++++++++++
> tools/perf/util/units.h | 1 +
> 3 files changed, 28 insertions(+), 8 deletions(-)
>
> diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c
> index 6ccf21a72f06..786b5ef512d8 100644
> --- a/tools/perf/util/stat-shadow.c
> +++ b/tools/perf/util/stat-shadow.c
> @@ -9,6 +9,7 @@
> #include "expr.h"
> #include "metricgroup.h"
> #include "cgroup.h"
> +#include "units.h"
> #include <linux/zalloc.h>
>
> /*
> @@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
> generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
> evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
> } else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
> - char unit = 'M';
> + char unit = ' ';
> char unit_buf[10];
>
> total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
> -
> if (total)
> - ratio = 1000.0 * avg / total;
> - if (ratio < 0.001) {
> - ratio *= 1000;
> - unit = 'K';
> - }
> - snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
> + ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
> +
> + snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
> print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
> } else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
> print_smi_cost(config, cpu, out, st, &rsd);
> diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
> index a46762aec4c9..ac13b5ecde31 100644
> --- a/tools/perf/util/units.c
> +++ b/tools/perf/util/units.c
> @@ -55,6 +55,28 @@ unsigned long convert_unit(unsigned long value, char *unit)
> return value;
> }
>
Hi Changbin,
Since we are using added function `convert_unit_double` just in stat-shadow.c,
I think there is no need to add it in units.h, we can directly create static func inside `stat-shadow.c`itself.
Thanks,
Kajol Jain
> +double convert_unit_double(double value, char *unit)
> +{
> + *unit = ' ';
> +
> + if (value > 1000.0) {
> + value /= 1000.0;
> + *unit = 'K';
> + }
> +
> + if (value > 1000.0) {
> + value /= 1000.0;
> + *unit = 'M';
> + }
> +
> + if (value > 1000.0) {
> + value /= 1000.0;
> + *unit = 'G';
> + }
> +
> + return value;
> +}
> +
> int unit_number__scnprintf(char *buf, size_t size, u64 n)
> {
> char unit[4] = "BKMG";
> diff --git a/tools/perf/util/units.h b/tools/perf/util/units.h
> index 99263b6a23f7..b3ace67ac16f 100644
> --- a/tools/perf/util/units.h
> +++ b/tools/perf/util/units.h
> @@ -13,6 +13,7 @@ struct parse_tag {
> unsigned long parse_tag_value(const char *str, struct parse_tag *tags);
>
> unsigned long convert_unit(unsigned long value, char *unit);
> +double convert_unit_double(double value, char *unit);
> int unit_number__scnprintf(char *buf, size_t size, u64 n);
>
> #endif /* PERF_UNIT_H */
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] perf stat: improve readability of shadow stats
2021-03-02 2:08 ` Namhyung Kim
@ 2021-03-02 12:25 ` Arnaldo Carvalho de Melo
2021-03-04 16:34 ` Changbin Du
0 siblings, 1 reply; 7+ messages in thread
From: Arnaldo Carvalho de Melo @ 2021-03-02 12:25 UTC (permalink / raw)
To: Namhyung Kim
Cc: Jiri Olsa, Changbin Du, Peter Zijlstra, Ingo Molnar,
linux-kernel, Ian Rogers, Andi Kleen
Em Tue, Mar 02, 2021 at 11:08:40AM +0900, Namhyung Kim escreveu:
> On Tue, Mar 2, 2021 at 4:19 AM Jiri Olsa <jolsa@redhat.com> wrote:
> >
> > On Tue, Mar 02, 2021 at 01:24:02AM +0800, Changbin Du wrote:
> > > This does follow two changes:
> > > 1) Select appropriate unit between K/M/G.
> > > 2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
> > >
> > > $ sudo ./perf stat -a -- sleep 1
> > >
> > > Before: Unit 'M' is selected even the number is very small.
> > > Performance counter stats for 'system wide':
> > >
> > > 4,003.06 msec cpu-clock # 3.998 CPUs utilized
> > > 16,179 context-switches # 0.004 M/sec
> > > 161 cpu-migrations # 0.040 K/sec
> > > 4,699 page-faults # 0.001 M/sec
> > > 6,135,801,925 cycles # 1.533 GHz (83.21%)
> > > 5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
> > > 4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
> > > 4,720,130,587 instructions # 0.77 insn per cycle
> > > # 1.23 stalled cycles per insn (83.28%)
> > > 753,848,078 branches # 188.318 M/sec (83.61%)
> > > 37,457,747 branch-misses # 4.97% of all branches (83.48%)
> > >
> > > 1.001283725 seconds time elapsed
> > >
> > > After:
> > > $ sudo ./perf stat -a -- sleep 2
> > >
> > > Performance counter stats for 'system wide':
> > >
> > > 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> > > 9,768 context-switches # 1.221 K/cpu-sec
> > > 164 cpu-migrations # 20.492 /cpu-sec
> >
> > should you remove also the leading '/' in ' /cpu-sec' ?
>
> The change looks good. And I think we should keep '/' otherwise it'd be
> more confusing.
Perhaps:
8,003.20 msec cpu-clock # 3.998 CPUs utilized
9,768 context-switches # 1.221 K/cpu-sec
164 cpu-migrations # 20.492 /cpu-sec
?
- Arnaldo
> >
> >
> > SNIP
> >
> > > @@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
> > > generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
> > > evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
> > > } else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
> > > - char unit = 'M';
> > > + char unit = ' ';
> > > char unit_buf[10];
> > >
> > > total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
> > > -
> > > if (total)
> > > - ratio = 1000.0 * avg / total;
> > > - if (ratio < 0.001) {
> > > - ratio *= 1000;
> > > - unit = 'K';
> > > - }
> > > - snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
> > > + ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
> > > +
> > > + snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
> > > print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
> >
> > hum this will change -x output that people parse, so I don't think we can do that
>
> Agreed.
>
> >
> > > } else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
> > > print_smi_cost(config, cpu, out, st, &rsd);
> > > diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
> > > index a46762aec4c9..ac13b5ecde31 100644
> > > --- a/tools/perf/util/units.c
> > > +++ b/tools/perf/util/units.c
> > > @@ -55,6 +55,28 @@ unsigned long convert_unit(unsigned long value, char *unit)
> > > return value;
> > > }
> > >
> > > +double convert_unit_double(double value, char *unit)
> > > +{
> > > + *unit = ' ';
> > > +
> > > + if (value > 1000.0) {
> > > + value /= 1000.0;
> > > + *unit = 'K';
> > > + }
> > > +
> > > + if (value > 1000.0) {
> > > + value /= 1000.0;
> > > + *unit = 'M';
> > > + }
> > > +
> > > + if (value > 1000.0) {
> > > + value /= 1000.0;
> > > + *unit = 'G';
> > > + }
> > > +
> > > + return value;
> > > +}
> >
> > we have convert_unit function just above doing the same only with
> > unsigned long.. let's have one base function with double values and
> > another one casting the result to unsigned long
>
> Sounds good.
>
> Thanks,
> Namhyung
--
- Arnaldo
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] perf stat: improve readability of shadow stats
2021-03-01 19:19 ` Jiri Olsa
2021-03-02 2:08 ` Namhyung Kim
@ 2021-03-04 16:22 ` Changbin Du
1 sibling, 0 replies; 7+ messages in thread
From: Changbin Du @ 2021-03-04 16:22 UTC (permalink / raw)
To: Jiri Olsa
Cc: Changbin Du, Arnaldo Carvalho de Melo, Peter Zijlstra,
Ingo Molnar, Namhyung Kim, linux-kernel, Ian Rogers, Andi Kleen
On Mon, Mar 01, 2021 at 08:19:20PM +0100, Jiri Olsa wrote:
> On Tue, Mar 02, 2021 at 01:24:02AM +0800, Changbin Du wrote:
> > This does follow two changes:
> > 1) Select appropriate unit between K/M/G.
> > 2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
> >
> > $ sudo ./perf stat -a -- sleep 1
> >
> > Before: Unit 'M' is selected even the number is very small.
> > Performance counter stats for 'system wide':
> >
> > 4,003.06 msec cpu-clock # 3.998 CPUs utilized
> > 16,179 context-switches # 0.004 M/sec
> > 161 cpu-migrations # 0.040 K/sec
> > 4,699 page-faults # 0.001 M/sec
> > 6,135,801,925 cycles # 1.533 GHz (83.21%)
> > 5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
> > 4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
> > 4,720,130,587 instructions # 0.77 insn per cycle
> > # 1.23 stalled cycles per insn (83.28%)
> > 753,848,078 branches # 188.318 M/sec (83.61%)
> > 37,457,747 branch-misses # 4.97% of all branches (83.48%)
> >
> > 1.001283725 seconds time elapsed
> >
> > After:
> > $ sudo ./perf stat -a -- sleep 2
> >
> > Performance counter stats for 'system wide':
> >
> > 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> > 9,768 context-switches # 1.221 K/cpu-sec
> > 164 cpu-migrations # 20.492 /cpu-sec
>
> should you remove also the leading '/' in ' /cpu-sec' ?
>
>
> SNIP
>
> > @@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
> > generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
> > evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
> > } else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
> > - char unit = 'M';
> > + char unit = ' ';
> > char unit_buf[10];
> >
> > total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
> > -
> > if (total)
> > - ratio = 1000.0 * avg / total;
> > - if (ratio < 0.001) {
> > - ratio *= 1000;
> > - unit = 'K';
> > - }
> > - snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
> > + ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
> > +
> > + snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
> > print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
>
> hum this will change -x output that people parse, so I don't think we can do that
okay, so skip this change and keep another.
--
Cheers,
Changbin Du
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] perf stat: improve readability of shadow stats
2021-03-02 12:25 ` Arnaldo Carvalho de Melo
@ 2021-03-04 16:34 ` Changbin Du
0 siblings, 0 replies; 7+ messages in thread
From: Changbin Du @ 2021-03-04 16:34 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: Namhyung Kim, Jiri Olsa, Changbin Du, Peter Zijlstra,
Ingo Molnar, linux-kernel, Ian Rogers, Andi Kleen
On Tue, Mar 02, 2021 at 09:25:24AM -0300, Arnaldo Carvalho de Melo wrote:
> Em Tue, Mar 02, 2021 at 11:08:40AM +0900, Namhyung Kim escreveu:
> > On Tue, Mar 2, 2021 at 4:19 AM Jiri Olsa <jolsa@redhat.com> wrote:
> > >
> > > On Tue, Mar 02, 2021 at 01:24:02AM +0800, Changbin Du wrote:
> > > > This does follow two changes:
> > > > 1) Select appropriate unit between K/M/G.
> > > > 2) Use 'cpu-sec' instead of 'sec' to state this is not the wall-time.
> > > >
> > > > $ sudo ./perf stat -a -- sleep 1
> > > >
> > > > Before: Unit 'M' is selected even the number is very small.
> > > > Performance counter stats for 'system wide':
> > > >
> > > > 4,003.06 msec cpu-clock # 3.998 CPUs utilized
> > > > 16,179 context-switches # 0.004 M/sec
> > > > 161 cpu-migrations # 0.040 K/sec
> > > > 4,699 page-faults # 0.001 M/sec
> > > > 6,135,801,925 cycles # 1.533 GHz (83.21%)
> > > > 5,783,308,491 stalled-cycles-frontend # 94.26% frontend cycles idle (83.21%)
> > > > 4,543,694,050 stalled-cycles-backend # 74.05% backend cycles idle (66.49%)
> > > > 4,720,130,587 instructions # 0.77 insn per cycle
> > > > # 1.23 stalled cycles per insn (83.28%)
> > > > 753,848,078 branches # 188.318 M/sec (83.61%)
> > > > 37,457,747 branch-misses # 4.97% of all branches (83.48%)
> > > >
> > > > 1.001283725 seconds time elapsed
> > > >
> > > > After:
> > > > $ sudo ./perf stat -a -- sleep 2
> > > >
> > > > Performance counter stats for 'system wide':
> > > >
> > > > 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> > > > 9,768 context-switches # 1.221 K/cpu-sec
> > > > 164 cpu-migrations # 20.492 /cpu-sec
> > >
> > > should you remove also the leading '/' in ' /cpu-sec' ?
> >
> > The change looks good. And I think we should keep '/' otherwise it'd be
> > more confusing.
>
> Perhaps:
>
> 8,003.20 msec cpu-clock # 3.998 CPUs utilized
> 9,768 context-switches # 1.221 K/cpu-sec
> 164 cpu-migrations # 20.492 /cpu-sec
>
> ?
>
sure. Now it is:
$ sudo ./perf stat
[sudo] password for changbin:
^C
Performance counter stats for 'system wide':
6,046.61 msec cpu-clock # 3.999 CPUs utilized
7,569 context-switches # 1.252 K/sec
587 cpu-migrations # 97.079 /sec
229 page-faults # 37.872 /sec
1,484,855,385 cycles # 0.246 GHz
2,112,863,483 stalled-cycles-frontend # 142.29% frontend cycles idle
1,980,746,623 stalled-cycles-backend # 133.40% backend cycles idle
454,477,035 instructions # 0.31 insn per cycle
# 4.65 stalled cycles per insn
89,992,521 branches # 14.883 M/sec
6,196,599 branch-misses # 6.89% of all branches
1.512029866 seconds time elapsed
> - Arnaldo
>
> > >
> > >
> > > SNIP
> > >
> > > > @@ -1270,18 +1271,14 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config,
> > > > generic_metric(config, evsel->metric_expr, evsel->metric_events, NULL,
> > > > evsel->name, evsel->metric_name, NULL, 1, cpu, out, st);
> > > > } else if (runtime_stat_n(st, STAT_NSECS, cpu, &rsd) != 0) {
> > > > - char unit = 'M';
> > > > + char unit = ' ';
> > > > char unit_buf[10];
> > > >
> > > > total = runtime_stat_avg(st, STAT_NSECS, cpu, &rsd);
> > > > -
> > > > if (total)
> > > > - ratio = 1000.0 * avg / total;
> > > > - if (ratio < 0.001) {
> > > > - ratio *= 1000;
> > > > - unit = 'K';
> > > > - }
> > > > - snprintf(unit_buf, sizeof(unit_buf), "%c/sec", unit);
> > > > + ratio = convert_unit_double(1000000000.0 * avg / total, &unit);
> > > > +
> > > > + snprintf(unit_buf, sizeof(unit_buf), "%c/cpu-sec", unit);
> > > > print_metric(config, ctxp, NULL, "%8.3f", unit_buf, ratio);
> > >
> > > hum this will change -x output that people parse, so I don't think we can do that
> >
> > Agreed.
> >
> > >
> > > > } else if (perf_stat_evsel__is(evsel, SMI_NUM)) {
> > > > print_smi_cost(config, cpu, out, st, &rsd);
> > > > diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
> > > > index a46762aec4c9..ac13b5ecde31 100644
> > > > --- a/tools/perf/util/units.c
> > > > +++ b/tools/perf/util/units.c
> > > > @@ -55,6 +55,28 @@ unsigned long convert_unit(unsigned long value, char *unit)
> > > > return value;
> > > > }
> > > >
> > > > +double convert_unit_double(double value, char *unit)
> > > > +{
> > > > + *unit = ' ';
> > > > +
> > > > + if (value > 1000.0) {
> > > > + value /= 1000.0;
> > > > + *unit = 'K';
> > > > + }
> > > > +
> > > > + if (value > 1000.0) {
> > > > + value /= 1000.0;
> > > > + *unit = 'M';
> > > > + }
> > > > +
> > > > + if (value > 1000.0) {
> > > > + value /= 1000.0;
> > > > + *unit = 'G';
> > > > + }
> > > > +
> > > > + return value;
> > > > +}
> > >
> > > we have convert_unit function just above doing the same only with
> > > unsigned long.. let's have one base function with double values and
> > > another one casting the result to unsigned long
> >
> > Sounds good.
> >
> > Thanks,
> > Namhyung
>
> --
>
> - Arnaldo
--
Cheers,
Changbin Du
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2021-03-04 16:35 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-03-01 17:24 [PATCH] perf stat: improve readability of shadow stats Changbin Du
2021-03-01 19:19 ` Jiri Olsa
2021-03-02 2:08 ` Namhyung Kim
2021-03-02 12:25 ` Arnaldo Carvalho de Melo
2021-03-04 16:34 ` Changbin Du
2021-03-04 16:22 ` Changbin Du
2021-03-02 6:53 ` kajoljain
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).