Since commit bb30acae4c4d ("perf report: Bail out --mem-mode if mem info
is not available") "perf mem report" and "perf report --mem-mode"
don't report result if the PERF_SAMPLE_DATA_SRC bit is missed in sample
type.

The commit ffab48705205 ("perf: arm-spe: Fix perf report --mem-mode")
partially fixes the issue.  It adds PERF_SAMPLE_DATA_SRC bit for Arm SPE
event, this allows the perf data file generated by kernel v5.18-rc1 or
later version can be reported properly.

On the other hand, perf tool still fails to be backward compatibility
for a data file recorded by an older version's perf which contains Arm
SPE trace data.  This patch is a workaround in reporting phase, when
detects ARM SPE PMU event and without PERF_SAMPLE_DATA_SRC bit, it will
force to set the bit in the sample type and give a warning info.

Fixes: bb30acae4c4d ("perf report: Bail out --mem-mode if mem info is not available")
Signed-off-by: Leo Yan <leo.yan@linaro.org>
---
 tools/perf/builtin-report.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 1ad75c7ba074..f26dd14eb852 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -353,6 +353,7 @@ static int report__setup_sample_type(struct report *rep)
 	struct perf_session *session = rep->session;
 	u64 sample_type = evlist__combined_sample_type(session->evlist);
 	bool is_pipe = perf_data__is_pipe(session->data);
+	struct evsel *evsel;
 
 	if (session->itrace_synth_opts->callchain ||
 	    session->itrace_synth_opts->add_callchain ||
@@ -407,6 +408,21 @@ static int report__setup_sample_type(struct report *rep)
 	}
 
 	if (sort__mode == SORT_MODE__MEMORY) {
+		/*
+		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
+		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
+		 * compatibility, set the bit if it's an old perf data file.
+		 */
+		evlist__for_each_entry(session->evlist, evsel) {
+			if (strstr(evsel->name, "arm_spe_") &&
+				!(sample_type & PERF_SAMPLE_DATA_SRC)) {
+				ui__warning("PERF_SAMPLE_DATA_SRC bit is not set "
+					    "for Arm SPE event.\n");
+				evsel->core.attr.sample_type |= PERF_SAMPLE_DATA_SRC;
+				sample_type |= PERF_SAMPLE_DATA_SRC;
+			}
+		}
+
 		if (!is_pipe && !(sample_type & PERF_SAMPLE_DATA_SRC)) {
 			ui__error("Selected --mem-mode but no mem data. "
 				  "Did you call perf record without -d?\n");
-- 
2.25.1

Hi Leo,

On 13/04/2022 08:51, Leo Yan wrote:
> Since commit bb30acae4c4d ("perf report: Bail out --mem-mode if mem info
> is not available") "perf mem report" and "perf report --mem-mode"
> don't report result if the PERF_SAMPLE_DATA_SRC bit is missed in sample
> type.
>
> The commit ffab48705205 ("perf: arm-spe: Fix perf report --mem-mode")
> partially fixes the issue.  It adds PERF_SAMPLE_DATA_SRC bit for Arm SPE
> event, this allows the perf data file generated by kernel v5.18-rc1 or
> later version can be reported properly.
>
> On the other hand, perf tool still fails to be backward compatibility
> for a data file recorded by an older version's perf which contains Arm
> SPE trace data.  This patch is a workaround in reporting phase, when
> detects ARM SPE PMU event and without PERF_SAMPLE_DATA_SRC bit, it will
> force to set the bit in the sample type and give a warning info.
>
> Fixes: bb30acae4c4d ("perf report: Bail out --mem-mode if mem info is not available")
> Signed-off-by: Leo Yan <leo.yan@linaro.org>
> ---
>  tools/perf/builtin-report.c | 16 ++++++++++++++++
>  1 file changed, 16 insertions(+)
>
> diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
> index 1ad75c7ba074..f26dd14eb852 100644
> --- a/tools/perf/builtin-report.c
> +++ b/tools/perf/builtin-report.c
> @@ -353,6 +353,7 @@ static int report__setup_sample_type(struct report *rep)
>  	struct perf_session *session = rep->session;
>  	u64 sample_type = evlist__combined_sample_type(session->evlist);
>  	bool is_pipe = perf_data__is_pipe(session->data);
> +	struct evsel *evsel;
>  
>  	if (session->itrace_synth_opts->callchain ||
>  	    session->itrace_synth_opts->add_callchain ||
> @@ -407,6 +408,21 @@ static int report__setup_sample_type(struct report *rep)
>  	}
>  
>  	if (sort__mode == SORT_MODE__MEMORY) {
> +		/*
> +		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
> +		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
> +		 * compatibility, set the bit if it's an old perf data file.
> +		 */
> +		evlist__for_each_entry(session->evlist, evsel) {
> +			if (strstr(evsel->name, "arm_spe_") &&

This didn't work for me when the file recorded "-e arm_spe//" instead of
"-e arm_spe_0//". Could you remove the trailing _? With that:

Tested-by: German Gomez <german.gomez@arm.com>

> +				!(sample_type & PERF_SAMPLE_DATA_SRC)) {
> +				ui__warning("PERF_SAMPLE_DATA_SRC bit is not set "
> +					    "for Arm SPE event.\n");
> +				evsel->core.attr.sample_type |= PERF_SAMPLE_DATA_SRC;
> +				sample_type |= PERF_SAMPLE_DATA_SRC;
> +			}
> +		}
> +
>  		if (!is_pipe && !(sample_type & PERF_SAMPLE_DATA_SRC)) {
>  			ui__error("Selected --mem-mode but no mem data. "
>  				  "Did you call perf record without -d?\n");

On Wed, Apr 13, 2022 at 09:15:40AM +0100, German Gomez wrote:

[...]

> >  	if (sort__mode == SORT_MODE__MEMORY) {
> > +		/*
> > +		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
> > +		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
> > +		 * compatibility, set the bit if it's an old perf data file.
> > +		 */
> > +		evlist__for_each_entry(session->evlist, evsel) {
> > +			if (strstr(evsel->name, "arm_spe_") &&
> 
> This didn't work for me when the file recorded "-e arm_spe//" instead of
> "-e arm_spe_0//". Could you remove the trailing _? With that:

Sure, will change to "arm_spe".  Just curious, if there any local
change at your side so we have the different event name?

> Tested-by: German Gomez <german.gomez@arm.com>

Thanks a lot, German!

Leo

Em Wed, Apr 13, 2022 at 04:49:41PM +0800, Leo Yan escreveu:
> On Wed, Apr 13, 2022 at 09:15:40AM +0100, German Gomez wrote:
> 
> [...]
> 
> > >  	if (sort__mode == SORT_MODE__MEMORY) {
> > > +		/*
> > > +		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
> > > +		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
> > > +		 * compatibility, set the bit if it's an old perf data file.
> > > +		 */
> > > +		evlist__for_each_entry(session->evlist, evsel) {
> > > +			if (strstr(evsel->name, "arm_spe_") &&
> > 
> > This didn't work for me when the file recorded "-e arm_spe//" instead of
> > "-e arm_spe_0//". Could you remove the trailing _? With that:
> 
> Sure, will change to "arm_spe".  Just curious, if there any local
> change at your side so we have the different event name?

Ok, waiting for v2
 
> > Tested-by: German Gomez <german.gomez@arm.com>
> 
> Thanks a lot, German!
> 
> Leo

-- 

- Arnaldo

On 13/04/2022 09:49, Leo Yan wrote:
> On Wed, Apr 13, 2022 at 09:15:40AM +0100, German Gomez wrote:
>
> [...]
>
>>>  	if (sort__mode == SORT_MODE__MEMORY) {
>>> +		/*
>>> +		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
>>> +		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
>>> +		 * compatibility, set the bit if it's an old perf data file.
>>> +		 */
>>> +		evlist__for_each_entry(session->evlist, evsel) {
>>> +			if (strstr(evsel->name, "arm_spe_") &&
>> This didn't work for me when the file recorded "-e arm_spe//" instead of
>> "-e arm_spe_0//". Could you remove the trailing _? With that:
> Sure, will change to "arm_spe".  Just curious, if there any local
> change at your side so we have the different event name?

No local changes. I've always used "arm_spe//" and it always defaults to
"arm_spe_0//". But it's still stored as the former in the data file. I
haven't checked where this default happens though. Isn't it the same for
you?

Thanks,
German
>
>> Tested-by: German Gomez <german.gomez@arm.com>
> Thanks a lot, German!
>
> Leo

On Wed, Apr 13, 2022 at 10:14:00AM +0100, German Gomez wrote:
> 
> On 13/04/2022 09:49, Leo Yan wrote:
> > On Wed, Apr 13, 2022 at 09:15:40AM +0100, German Gomez wrote:
> >
> > [...]
> >
> >>>  	if (sort__mode == SORT_MODE__MEMORY) {
> >>> +		/*
> >>> +		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
> >>> +		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
> >>> +		 * compatibility, set the bit if it's an old perf data file.
> >>> +		 */
> >>> +		evlist__for_each_entry(session->evlist, evsel) {
> >>> +			if (strstr(evsel->name, "arm_spe_") &&
> >> This didn't work for me when the file recorded "-e arm_spe//" instead of
> >> "-e arm_spe_0//". Could you remove the trailing _? With that:
> > Sure, will change to "arm_spe".  Just curious, if there any local
> > change at your side so we have the different event name?
> 
> No local changes. I've always used "arm_spe//" and it always defaults to
> "arm_spe_0//". But it's still stored as the former in the data file. I
> haven't checked where this default happens though. Isn't it the same for
> you?

Yeah, this is same with me.  Thanks for explanation.

Leo

On Wed, Apr 13, 2022 at 04:49:41PM +0800, Leo Yan wrote:
> On Wed, Apr 13, 2022 at 09:15:40AM +0100, German Gomez wrote:
> 
> [...]
> 
> > >  	if (sort__mode == SORT_MODE__MEMORY) {
> > > +		/*
> > > +		 * FIXUP: prior to kernel 5.18, Arm SPE missed to set
> > > +		 * PERF_SAMPLE_DATA_SRC bit in sample type.  For backward
> > > +		 * compatibility, set the bit if it's an old perf data file.
> > > +		 */
> > > +		evlist__for_each_entry(session->evlist, evsel) {
> > > +			if (strstr(evsel->name, "arm_spe_") &&
> > 
> > This didn't work for me when the file recorded "-e arm_spe//" instead of
> > "-e arm_spe_0//". Could you remove the trailing _? With that:
> 
> Sure, will change to "arm_spe".  Just curious, if there any local
> change at your side so we have the different event name?

I think I know the reason now, though the PMU event is 'arm_spe_0',
but we could use '-e arm_spe//' to record, so finally the event name
is 'arm_spe'.

Thanks,
Leo