* [PATCH 1/2] perf top: Support --group-sort-idx to change the sort order
@ 2020-03-20 7:24 Jin Yao
2020-03-20 7:24 ` [PATCH 2/2] perf top: support hotkey to change " Jin Yao
0 siblings, 1 reply; 4+ messages in thread
From: Jin Yao @ 2020-03-20 7:24 UTC (permalink / raw)
To: acme, jolsa, peterz, mingo, alexander.shishkin
Cc: Linux-kernel, ak, kan.liang, yao.jin, Jin Yao
perf report has supported the option --group-sort-idx, which
sorts the output by the event at the index n in event group.
For example,
perf record -e cycles,instructions,cache-misses
perf report --group --group-sort-idx 2 --stdio
The perf-report output is sorted by cache-misses.
This patch supports --group-sort-idx in perf-top.
For example,
perf top --group -e cycles,instructions,cache-misses --group-sort-idx 2
The perf-top output is sorted by cache-misses.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
---
tools/perf/Documentation/perf-top.txt | 5 +++++
tools/perf/builtin-top.c | 4 ++++
2 files changed, 9 insertions(+)
diff --git a/tools/perf/Documentation/perf-top.txt b/tools/perf/Documentation/perf-top.txt
index 324b6b53c86b..0e97dcef794d 100644
--- a/tools/perf/Documentation/perf-top.txt
+++ b/tools/perf/Documentation/perf-top.txt
@@ -53,6 +53,11 @@ Default is to monitor all CPUS.
--group::
Put the counters into a counter group.
+--group-sort-idx::
+ Sort the output by the event at the index n in group. If n is invalid,
+ sort by the first event. It can support multiple groups with different
+ amount of events. WARNING: This should be used on grouped events.
+
-F <freq>::
--freq=<freq>::
Profile at this frequency. Use 'max' to use the currently maximum
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index f6dd1a63f159..144043637cec 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -1543,6 +1543,10 @@ int cmd_top(int argc, const char **argv)
"number of thread to run event synthesize"),
OPT_BOOLEAN(0, "namespaces", &opts->record_namespaces,
"Record namespaces events"),
+ OPT_INTEGER(0, "group-sort-idx", &symbol_conf.group_sort_idx,
+ "Sort the output by the event at the index n in group. "
+ "If n is invalid, sort by the first event. "
+ "WARNING: should be used on grouped events."),
OPTS_EVSWITCH(&top.evswitch),
OPT_END()
};
--
2.17.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [PATCH 2/2] perf top: support hotkey to change sort order
2020-03-20 7:24 [PATCH 1/2] perf top: Support --group-sort-idx to change the sort order Jin Yao
@ 2020-03-20 7:24 ` Jin Yao
2020-03-20 13:51 ` Arnaldo Carvalho de Melo
0 siblings, 1 reply; 4+ messages in thread
From: Jin Yao @ 2020-03-20 7:24 UTC (permalink / raw)
To: acme, jolsa, peterz, mingo, alexander.shishkin
Cc: Linux-kernel, ak, kan.liang, yao.jin, Jin Yao
It would be nice if we can use a hotkey in perf top browser to
select a event for sorting.
For example,
perf top --group -e cycles,instructions,cache-misses
Samples
Overhead Shared Object Symbol
40.03% 45.71% 0.03% div [.] main
20.46% 14.67% 0.21% libc-2.27.so [.] __random_r
20.01% 19.54% 0.02% libc-2.27.so [.] __random
9.68% 10.68% 0.00% div [.] compute_flag
4.32% 4.70% 0.00% libc-2.27.so [.] rand
3.84% 3.43% 0.00% div [.] rand@plt
0.05% 0.05% 2.33% libc-2.27.so [.] __strcmp_sse2_unaligned
0.04% 0.08% 2.43% perf [.] perf_hpp__is_dynamic_en
0.04% 0.02% 6.64% perf [.] rb_next
0.04% 0.01% 3.87% perf [.] dso__find_symbol
0.04% 0.04% 1.77% perf [.] sort__dso_cmp
When user press hotkey '2' (event index, starting from 0), it indicates
to sort output by the third event in group (cache-misses).
Samples
Overhead Shared Object Symbol
4.07% 1.28% 6.68% perf [.] rb_next
3.57% 3.98% 4.11% perf [.] __hists__insert_output
3.67% 11.24% 3.60% perf [.] perf_hpp__is_dynamic_e
3.67% 3.20% 3.20% perf [.] hpp__sort_overhead
0.81% 0.06% 3.01% perf [.] dso__find_symbol
1.62% 5.47% 2.51% perf [.] hists__match
2.70% 1.86% 2.47% libc-2.27.so [.] _int_malloc
0.19% 0.00% 2.29% [kernel] [k] copy_page
0.41% 0.32% 1.98% perf [.] hists__decay_entries
1.84% 3.67% 1.68% perf [.] sort__dso_cmp
0.16% 0.00% 1.63% [kernel] [k] clear_page_erms
Now the output is sorted by cache-misses.
Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
---
tools/perf/builtin-top.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index 144043637cec..b39f6ffb874e 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -616,6 +616,7 @@ static void *display_thread_tui(void *arg)
.arg = top,
.refresh = top->delay_secs,
};
+ int ret;
/* In order to read symbols from other namespaces perf to needs to call
* setns(2). This isn't permitted if the struct_fs has multiple users.
@@ -626,6 +627,7 @@ static void *display_thread_tui(void *arg)
prctl(PR_SET_NAME, "perf-top-UI", 0, 0, 0);
+repeat:
perf_top__sort_new_samples(top);
/*
@@ -638,13 +640,17 @@ static void *display_thread_tui(void *arg)
hists->uid_filter_str = top->record_opts.target.uid_str;
}
- perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
+ ret = perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
top->min_percent,
&top->session->header.env,
!top->record_opts.overwrite,
&top->annotation_opts);
- stop_top();
+ if (ret == K_RELOAD)
+ goto repeat;
+ else
+ stop_top();
+
return NULL;
}
--
2.17.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH 2/2] perf top: support hotkey to change sort order
2020-03-20 7:24 ` [PATCH 2/2] perf top: support hotkey to change " Jin Yao
@ 2020-03-20 13:51 ` Arnaldo Carvalho de Melo
2020-03-22 3:46 ` Jin, Yao
0 siblings, 1 reply; 4+ messages in thread
From: Arnaldo Carvalho de Melo @ 2020-03-20 13:51 UTC (permalink / raw)
To: Jin Yao
Cc: jolsa, peterz, mingo, alexander.shishkin, Linux-kernel, ak,
kan.liang, yao.jin
Em Fri, Mar 20, 2020 at 03:24:14PM +0800, Jin Yao escreveu:
> It would be nice if we can use a hotkey in perf top browser to
> select a event for sorting.
>
> For example,
> perf top --group -e cycles,instructions,cache-misses
>
> Samples
> Overhead Shared Object Symbol
> 40.03% 45.71% 0.03% div [.] main
> 20.46% 14.67% 0.21% libc-2.27.so [.] __random_r
> 20.01% 19.54% 0.02% libc-2.27.so [.] __random
> 9.68% 10.68% 0.00% div [.] compute_flag
> 4.32% 4.70% 0.00% libc-2.27.so [.] rand
> 3.84% 3.43% 0.00% div [.] rand@plt
> 0.05% 0.05% 2.33% libc-2.27.so [.] __strcmp_sse2_unaligned
> 0.04% 0.08% 2.43% perf [.] perf_hpp__is_dynamic_en
> 0.04% 0.02% 6.64% perf [.] rb_next
> 0.04% 0.01% 3.87% perf [.] dso__find_symbol
> 0.04% 0.04% 1.77% perf [.] sort__dso_cmp
>
> When user press hotkey '2' (event index, starting from 0), it indicates
> to sort output by the third event in group (cache-misses).
>
> Samples
> Overhead Shared Object Symbol
> 4.07% 1.28% 6.68% perf [.] rb_next
> 3.57% 3.98% 4.11% perf [.] __hists__insert_output
> 3.67% 11.24% 3.60% perf [.] perf_hpp__is_dynamic_e
> 3.67% 3.20% 3.20% perf [.] hpp__sort_overhead
> 0.81% 0.06% 3.01% perf [.] dso__find_symbol
> 1.62% 5.47% 2.51% perf [.] hists__match
> 2.70% 1.86% 2.47% libc-2.27.so [.] _int_malloc
> 0.19% 0.00% 2.29% [kernel] [k] copy_page
> 0.41% 0.32% 1.98% perf [.] hists__decay_entries
> 1.84% 3.67% 1.68% perf [.] sort__dso_cmp
> 0.16% 0.00% 1.63% [kernel] [k] clear_page_erms
>
> Now the output is sorted by cache-misses.
>
> Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
> ---
> tools/perf/builtin-top.c | 10 ++++++++--
> 1 file changed, 8 insertions(+), 2 deletions(-)
>
> diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
> index 144043637cec..b39f6ffb874e 100644
> --- a/tools/perf/builtin-top.c
> +++ b/tools/perf/builtin-top.c
> @@ -616,6 +616,7 @@ static void *display_thread_tui(void *arg)
> .arg = top,
> .refresh = top->delay_secs,
> };
> + int ret;
>
> /* In order to read symbols from other namespaces perf to needs to call
> * setns(2). This isn't permitted if the struct_fs has multiple users.
> @@ -626,6 +627,7 @@ static void *display_thread_tui(void *arg)
>
> prctl(PR_SET_NAME, "perf-top-UI", 0, 0, 0);
>
> +repeat:
> perf_top__sort_new_samples(top);
>
> /*
> @@ -638,13 +640,17 @@ static void *display_thread_tui(void *arg)
> hists->uid_filter_str = top->record_opts.target.uid_str;
> }
>
> - perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
> + ret = perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
> top->min_percent,
> &top->session->header.env,
> !top->record_opts.overwrite,
> &top->annotation_opts);
>
> - stop_top();
> + if (ret == K_RELOAD)
> + goto repeat;
> + else
> + stop_top();
> +
That is really nice and small, but shouldn't we flush all the histograms
that were in place, sorted by the previous key? I think we have a 'z'
for zeroing samples that may be what we need, take a look, please,
- Arnaldo
> return NULL;
> }
>
> --
> 2.17.1
>
--
- Arnaldo
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 2/2] perf top: support hotkey to change sort order
2020-03-20 13:51 ` Arnaldo Carvalho de Melo
@ 2020-03-22 3:46 ` Jin, Yao
0 siblings, 0 replies; 4+ messages in thread
From: Jin, Yao @ 2020-03-22 3:46 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: jolsa, peterz, mingo, alexander.shishkin, Linux-kernel, ak,
kan.liang, yao.jin
On 3/20/2020 9:51 PM, Arnaldo Carvalho de Melo wrote:
> Em Fri, Mar 20, 2020 at 03:24:14PM +0800, Jin Yao escreveu:
>> It would be nice if we can use a hotkey in perf top browser to
>> select a event for sorting.
>>
>> For example,
>> perf top --group -e cycles,instructions,cache-misses
>>
>> Samples
>> Overhead Shared Object Symbol
>> 40.03% 45.71% 0.03% div [.] main
>> 20.46% 14.67% 0.21% libc-2.27.so [.] __random_r
>> 20.01% 19.54% 0.02% libc-2.27.so [.] __random
>> 9.68% 10.68% 0.00% div [.] compute_flag
>> 4.32% 4.70% 0.00% libc-2.27.so [.] rand
>> 3.84% 3.43% 0.00% div [.] rand@plt
>> 0.05% 0.05% 2.33% libc-2.27.so [.] __strcmp_sse2_unaligned
>> 0.04% 0.08% 2.43% perf [.] perf_hpp__is_dynamic_en
>> 0.04% 0.02% 6.64% perf [.] rb_next
>> 0.04% 0.01% 3.87% perf [.] dso__find_symbol
>> 0.04% 0.04% 1.77% perf [.] sort__dso_cmp
>>
>> When user press hotkey '2' (event index, starting from 0), it indicates
>> to sort output by the third event in group (cache-misses).
>>
>> Samples
>> Overhead Shared Object Symbol
>> 4.07% 1.28% 6.68% perf [.] rb_next
>> 3.57% 3.98% 4.11% perf [.] __hists__insert_output
>> 3.67% 11.24% 3.60% perf [.] perf_hpp__is_dynamic_e
>> 3.67% 3.20% 3.20% perf [.] hpp__sort_overhead
>> 0.81% 0.06% 3.01% perf [.] dso__find_symbol
>> 1.62% 5.47% 2.51% perf [.] hists__match
>> 2.70% 1.86% 2.47% libc-2.27.so [.] _int_malloc
>> 0.19% 0.00% 2.29% [kernel] [k] copy_page
>> 0.41% 0.32% 1.98% perf [.] hists__decay_entries
>> 1.84% 3.67% 1.68% perf [.] sort__dso_cmp
>> 0.16% 0.00% 1.63% [kernel] [k] clear_page_erms
>>
>> Now the output is sorted by cache-misses.
>>
>> Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
>> ---
>> tools/perf/builtin-top.c | 10 ++++++++--
>> 1 file changed, 8 insertions(+), 2 deletions(-)
>>
>> diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
>> index 144043637cec..b39f6ffb874e 100644
>> --- a/tools/perf/builtin-top.c
>> +++ b/tools/perf/builtin-top.c
>> @@ -616,6 +616,7 @@ static void *display_thread_tui(void *arg)
>> .arg = top,
>> .refresh = top->delay_secs,
>> };
>> + int ret;
>>
>> /* In order to read symbols from other namespaces perf to needs to call
>> * setns(2). This isn't permitted if the struct_fs has multiple users.
>> @@ -626,6 +627,7 @@ static void *display_thread_tui(void *arg)
>>
>> prctl(PR_SET_NAME, "perf-top-UI", 0, 0, 0);
>>
>> +repeat:
>> perf_top__sort_new_samples(top);
>>
>> /*
>> @@ -638,13 +640,17 @@ static void *display_thread_tui(void *arg)
>> hists->uid_filter_str = top->record_opts.target.uid_str;
>> }
>>
>> - perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
>> + ret = perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
>> top->min_percent,
>> &top->session->header.env,
>> !top->record_opts.overwrite,
>> &top->annotation_opts);
>>
>> - stop_top();
>> + if (ret == K_RELOAD)
>> + goto repeat;
>> + else
>> + stop_top();
>> +
>
> That is really nice and small, but shouldn't we flush all the histograms
> that were in place, sorted by the previous key? I think we have a 'z'
> for zeroing samples that may be what we need, take a look, please,
>
> - Arnaldo
>
Set top->zero to true if return key is K_RELOAD. So that in
perf_top__resort_hists(), it will delete hists entries by calling
hists__delete_entries(hists).
+repeat:
perf_top__sort_new_samples(top);
/*
@@ -638,13 +640,18 @@ static void *display_thread_tui(void *arg)
hists->uid_filter_str = top->record_opts.target.uid_str;
}
- perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
+ ret = perf_evlist__tui_browse_hists(top->evlist, help, &hbt,
top->min_percent,
&top->session->header.env,
!top->record_opts.overwrite,
&top->annotation_opts);
- stop_top();
+ if (ret == K_RELOAD) {
+ top->zero = true;
+ goto repeat;
+ } else
+ stop_top();
+
Is this OK?
Thanks
Jin Yao
>> return NULL;
>> }
>>
>> --
>> 2.17.1
>>
>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2020-03-22 3:46 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-03-20 7:24 [PATCH 1/2] perf top: Support --group-sort-idx to change the sort order Jin Yao
2020-03-20 7:24 ` [PATCH 2/2] perf top: support hotkey to change " Jin Yao
2020-03-20 13:51 ` Arnaldo Carvalho de Melo
2020-03-22 3:46 ` Jin, Yao
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).