2015-02-13 18:41:09

by Suzuki K Poulose

Subject: [PATCHv2] perf/stat: Report unsupported events properly

From: "Suzuki K. Poulose" <[email protected]>

Commit 1971f59 (perf stat: Use read_counter in read_counter_aggr )
broke the perf stat output for unsupported counters.

$ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
Warning:
CCI_400/config=24/ event is not supported by the kernel.

Performance counter stats for 'system wide':

0 CCI_400/config=24/

1.080265400 seconds time elapsed

Where it used to be :

$ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
Warning:
CCI_400/config=24/ event is not supported by the kernel.

Performance counter stats for 'system wide':

<not supported> CCI_400/config=24/

1.083840675 seconds time elapsed

This patch fixes the issues by checking if the counter is supported,
before reading and logging the counter value.

Cc: Jiri Olsa <[email protected]>
Cc: Arnaldo Carvalho de Melo <[email protected]>
Signed-off-by: Suzuki K. Poulose <[email protected]>
---
tools/perf/builtin-stat.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
index e598e4e..d28949d 100644
--- a/tools/perf/builtin-stat.c
+++ b/tools/perf/builtin-stat.c
@@ -510,6 +510,9 @@ static int read_counter(struct perf_evsel *counter)
int ncpus = perf_evsel__nr_cpus(counter);
int cpu, thread;

+ if (!counter->supported)
+ return -ENOENT;
+
if (counter->system_wide)
nthreads = 1;

@@ -1285,7 +1288,7 @@ static void print_counter_aggr(struct perf_evsel *counter, char *prefix)
if (prefix)
fprintf(output, "%s", prefix);

- if (scaled == -1) {
+ if (scaled == -1 || !counter->supported) {
fprintf(output, "%*s%s",
csv_output ? 0 : 18,
counter->supported ? CNTR_NOT_COUNTED : CNTR_NOT_SUPPORTED,
--
1.7.9.5

2015-02-13 19:39:27

by David Ahern

Subject: Re: [PATCHv2] perf/stat: Report unsupported events properly

On 02/13/2015 11:40 AM, Suzuki K. Poulose wrote:
> From: "Suzuki K. Poulose" <[email protected]>
>
> Commit 1971f59 (perf stat: Use read_counter in read_counter_aggr )
> broke the perf stat output for unsupported counters.
>
> $ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
> Warning:
> CCI_400/config=24/ event is not supported by the kernel.
>
> Performance counter stats for 'system wide':
>
> 0 CCI_400/config=24/
>
> 1.080265400 seconds time elapsed
>
> Where it used to be :
>
> $ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
> Warning:
> CCI_400/config=24/ event is not supported by the kernel.
>
> Performance counter stats for 'system wide':
>
> <not supported> CCI_400/config=24/
>
> 1.083840675 seconds time elapsed
>
> This patch fixes the issues by checking if the counter is supported,
> before reading and logging the counter value.
>
> Cc: Jiri Olsa <[email protected]>
> Cc: Arnaldo Carvalho de Melo <[email protected]>
> Signed-off-by: Suzuki K. Poulose <[email protected]>
> ---
> tools/perf/builtin-stat.c | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
> index e598e4e..d28949d 100644
> --- a/tools/perf/builtin-stat.c
> +++ b/tools/perf/builtin-stat.c
> @@ -510,6 +510,9 @@ static int read_counter(struct perf_evsel *counter)
> int ncpus = perf_evsel__nr_cpus(counter);
> int cpu, thread;
>
> + if (!counter->supported)
> + return -ENOENT;
> +
> if (counter->system_wide)
> nthreads = 1;
>
> @@ -1285,7 +1288,7 @@ static void print_counter_aggr(struct perf_evsel *counter, char *prefix)
> if (prefix)
> fprintf(output, "%s", prefix);
>
> - if (scaled == -1) {
> + if (scaled == -1 || !counter->supported) {
> fprintf(output, "%*s%s",
> csv_output ? 0 : 18,
> counter->supported ? CNTR_NOT_COUNTED : CNTR_NOT_SUPPORTED,
>

Just hit this as well.

Acked-and-Tested-by: David Ahern <[email protected]>

2015-02-13 20:50:41

by Arnaldo Carvalho de Melo

Subject: Re: [PATCHv2] perf/stat: Report unsupported events properly

Em Fri, Feb 13, 2015 at 12:39:24PM -0700, David Ahern escreveu:
>
>
> On 02/13/2015 11:40 AM, Suzuki K. Poulose wrote:
> >From: "Suzuki K. Poulose" <[email protected]>
> >
> >Commit 1971f59 (perf stat: Use read_counter in read_counter_aggr )
> >broke the perf stat output for unsupported counters.
> >
> > $ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
> > Warning:
> > CCI_400/config=24/ event is not supported by the kernel.
> >
> > Performance counter stats for 'system wide':
> >
> > 0 CCI_400/config=24/
> >
> > 1.080265400 seconds time elapsed
> >
> >Where it used to be :
> >
> >$ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
> > Warning:
> > CCI_400/config=24/ event is not supported by the kernel.
> >
> > Performance counter stats for 'system wide':
> >
> > <not supported> CCI_400/config=24/
> >
> > 1.083840675 seconds time elapsed
> >
> >This patch fixes the issues by checking if the counter is supported,
> >before reading and logging the counter value.
> >
> >Cc: Jiri Olsa <[email protected]>
> >Cc: Arnaldo Carvalho de Melo <[email protected]>
> >Signed-off-by: Suzuki K. Poulose <[email protected]>
> >---
> > tools/perf/builtin-stat.c | 5 ++++-
> > 1 file changed, 4 insertions(+), 1 deletion(-)
> >
> >diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
> >index e598e4e..d28949d 100644
> >--- a/tools/perf/builtin-stat.c
> >+++ b/tools/perf/builtin-stat.c
> >@@ -510,6 +510,9 @@ static int read_counter(struct perf_evsel *counter)
> > int ncpus = perf_evsel__nr_cpus(counter);
> > int cpu, thread;
> >
> >+ if (!counter->supported)
> >+ return -ENOENT;
> >+
> > if (counter->system_wide)
> > nthreads = 1;
> >
> >@@ -1285,7 +1288,7 @@ static void print_counter_aggr(struct perf_evsel *counter, char *prefix)
> > if (prefix)
> > fprintf(output, "%s", prefix);
> >
> >- if (scaled == -1) {
> >+ if (scaled == -1 || !counter->supported) {
> > fprintf(output, "%*s%s",
> > csv_output ? 0 : 18,
> > counter->supported ? CNTR_NOT_COUNTED : CNTR_NOT_SUPPORTED,
> >
>
> Just hit this as well.
>
> Acked-and-Tested-by: David Ahern <[email protected]>

Gack, 10ms after I sent my pre-California pull req :-\

Will go in the next one :-)

- Arnaldo

2015-02-17 10:13:49

by Suzuki K Poulose

Subject: Re: [PATCHv2] perf/stat: Report unsupported events properly

On 13/02/15 20:50, Arnaldo Carvalho de Melo wrote:
> Em Fri, Feb 13, 2015 at 12:39:24PM -0700, David Ahern escreveu:
>>
>>
>> On 02/13/2015 11:40 AM, Suzuki K. Poulose wrote:
>>> From: "Suzuki K. Poulose" <[email protected]>
>>>
>>> Commit 1971f59 (perf stat: Use read_counter in read_counter_aggr )
>>> broke the perf stat output for unsupported counters.
>>>
>>> $ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
>>> Warning:
>>> CCI_400/config=24/ event is not supported by the kernel.
>>>
>>> Performance counter stats for 'system wide':
>>>
>>> 0 CCI_400/config=24/
>>>
>>> 1.080265400 seconds time elapsed
>>>
>>> Where it used to be :
>>>
>>> $ perf stat -v -a -C 0 -e CCI_400/config=24/ sleep 1
>>> Warning:
>>> CCI_400/config=24/ event is not supported by the kernel.
>>>
>>> Performance counter stats for 'system wide':
>>>
>>> <not supported> CCI_400/config=24/
>>>
>>> 1.083840675 seconds time elapsed
>>>
>>> This patch fixes the issues by checking if the counter is supported,
>>> before reading and logging the counter value.
>>>
>>> Cc: Jiri Olsa <[email protected]>
>>> Cc: Arnaldo Carvalho de Melo <[email protected]>
>>> Signed-off-by: Suzuki K. Poulose <[email protected]>
>>> ---
>>> tools/perf/builtin-stat.c | 5 ++++-
>>> 1 file changed, 4 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
>>> index e598e4e..d28949d 100644
>>> --- a/tools/perf/builtin-stat.c
>>> +++ b/tools/perf/builtin-stat.c
>>> @@ -510,6 +510,9 @@ static int read_counter(struct perf_evsel *counter)
>>> int ncpus = perf_evsel__nr_cpus(counter);
>>> int cpu, thread;
>>>
>>> + if (!counter->supported)
>>> + return -ENOENT;
>>> +
>>> if (counter->system_wide)
>>> nthreads = 1;
>>>
>>> @@ -1285,7 +1288,7 @@ static void print_counter_aggr(struct perf_evsel *counter, char *prefix)
>>> if (prefix)
>>> fprintf(output, "%s", prefix);
>>>
>>> - if (scaled == -1) {
>>> + if (scaled == -1 || !counter->supported) {
>>> fprintf(output, "%*s%s",
>>> csv_output ? 0 : 18,
>>> counter->supported ? CNTR_NOT_COUNTED : CNTR_NOT_SUPPORTED,
>>>
>>
>> Just hit this as well.
>>
>> Acked-and-Tested-by: David Ahern <[email protected]>
>
> Gack, 10ms after I sent my pre-California pull req :-\
>
> Will go in the next one :-)
I think this may need to goto 3.19 stable series as well. Missed that.
Do you want me to send it there ?

Cheers
Suzuki

2015-02-27 19:40:21

by David Ahern

Subject: Re: [PATCHv2] perf/stat: Report unsupported events properly

On 2/13/15 1:50 PM, Arnaldo Carvalho de Melo wrote:
>
> Gack, 10ms after I sent my pre-California pull req :-\
>
> Will go in the next one :-)
>

still not in your queues.