From: Kan Liang <[email protected]>
perf top errors out on a hybrid machine
$perf top
Error:
The cycles:P event is not supported.
The perf top expects that the "cycles" is collected on all CPUs in the
system. But for hybrid there is no single "cycles" event which can cover
all CPUs. Perf has to split it into two cycles events, e.g.,
cpu_core/cycles/ and cpu_atom/cycles/. Each event has its own CPU mask.
If a event is opened on the unsupported CPU. The open fails. That's the
reason of the above error out.
Perf should only open the cycles event on the corresponding CPU. The
commit ef91871c960e ("perf evlist: Propagate user CPU maps intersecting
core PMU maps") intersect the requested CPU map with the CPU map of the
PMU. Use the evsel's cpus to replace user_requested_cpus.
The evlist's threads are also propagated to the evsel's threads in
__perf_evlist__propagate_maps(). For a system-wide event, perf appends
a dummy event and assign it to the evsel's threads. For a per-thread
event, the evlist's thread_map is assigned to the evsel's threads. The
same as the other tools, e.g., perf record, using the evsel's threads
when opening an event.
Reported-by: Arnaldo Carvalho de Melo <[email protected]>
Closes: https://lore.kernel.org/linux-perf-users/[email protected]/
Reviewed-by: Ian Rogers <[email protected]>
Signed-off-by: Kan Liang <[email protected]>
---
No changes since V2
tools/perf/builtin-top.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index ea8c7eca5eee..cce9350177e2 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -1027,8 +1027,8 @@ static int perf_top__start_counters(struct perf_top *top)
evlist__for_each_entry(evlist, counter) {
try_again:
- if (evsel__open(counter, top->evlist->core.user_requested_cpus,
- top->evlist->core.threads) < 0) {
+ if (evsel__open(counter, counter->core.cpus,
+ counter->core.threads) < 0) {
/*
* Specially handle overwrite fall back.
--
2.35.1
From: Kan Liang <[email protected]>
It's hard to distinguish the default cycles events among hybrid PMUs.
For example,
$perf top
Available samples
385 cycles:P
903 cycles:P
The other tool, e.g., perf record, uniforms the event name and adds the
hybrid PMU name before opening the event. So the events can be easily
distinguished. Apply the same methodology for the perf top as well.
The record__uniquify_name() will be invoked by both record and top.
Move it to util/record.c
With the patch
$perf top
Available samples
148 cpu_atom/cycles:P/
1K cpu_core/cycles:P/
Signed-off-by: Kan Liang <[email protected]>
---
New patch to address the display concern
https://lore.kernel.org/lkml/[email protected]/
tools/perf/builtin-record.c | 28 +---------------------------
tools/perf/builtin-top.c | 1 +
tools/perf/util/record.c | 25 +++++++++++++++++++++++++
tools/perf/util/record.h | 2 ++
4 files changed, 29 insertions(+), 27 deletions(-)
diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
index dcf288a4fb9a..a096422a4a14 100644
--- a/tools/perf/builtin-record.c
+++ b/tools/perf/builtin-record.c
@@ -2216,32 +2216,6 @@ static void hit_auxtrace_snapshot_trigger(struct record *rec)
}
}
-static void record__uniquify_name(struct record *rec)
-{
- struct evsel *pos;
- struct evlist *evlist = rec->evlist;
- char *new_name;
- int ret;
-
- if (perf_pmus__num_core_pmus() == 1)
- return;
-
- evlist__for_each_entry(evlist, pos) {
- if (!evsel__is_hybrid(pos))
- continue;
-
- if (strchr(pos->name, '/'))
- continue;
-
- ret = asprintf(&new_name, "%s/%s/",
- pos->pmu_name, pos->name);
- if (ret) {
- free(pos->name);
- pos->name = new_name;
- }
- }
-}
-
static int record__terminate_thread(struct record_thread *thread_data)
{
int err;
@@ -2475,7 +2449,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
if (data->is_pipe && rec->evlist->core.nr_entries == 1)
rec->opts.sample_id = true;
- record__uniquify_name(rec);
+ record__uniquify_name(rec->evlist);
/* Debug message used by test scripts */
pr_debug3("perf record opening and mmapping events\n");
diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
index cce9350177e2..4e8296654280 100644
--- a/tools/perf/builtin-top.c
+++ b/tools/perf/builtin-top.c
@@ -1299,6 +1299,7 @@ static int __cmd_top(struct perf_top *top)
}
}
+ record__uniquify_name(top->evlist);
ret = perf_top__start_counters(top);
if (ret)
return ret;
diff --git a/tools/perf/util/record.c b/tools/perf/util/record.c
index 9eb5c6a08999..5b4be3c72cbc 100644
--- a/tools/perf/util/record.c
+++ b/tools/perf/util/record.c
@@ -289,3 +289,28 @@ int record__parse_freq(const struct option *opt, const char *str, int unset __ma
opts->user_freq = freq;
return 0;
}
+
+void record__uniquify_name(struct evlist *evlist)
+{
+ struct evsel *pos;
+ char *new_name;
+ int ret;
+
+ if (perf_pmus__num_core_pmus() == 1)
+ return;
+
+ evlist__for_each_entry(evlist, pos) {
+ if (!evsel__is_hybrid(pos))
+ continue;
+
+ if (strchr(pos->name, '/'))
+ continue;
+
+ ret = asprintf(&new_name, "%s/%s/",
+ pos->pmu_name, pos->name);
+ if (ret) {
+ free(pos->name);
+ pos->name = new_name;
+ }
+ }
+}
diff --git a/tools/perf/util/record.h b/tools/perf/util/record.h
index a6566134e09e..9b520ab784bc 100644
--- a/tools/perf/util/record.h
+++ b/tools/perf/util/record.h
@@ -8,6 +8,7 @@
#include <linux/stddef.h>
#include <linux/perf_event.h>
#include "util/target.h"
+#include "util/evlist.h"
struct option;
@@ -85,6 +86,7 @@ extern const char * const *record_usage;
extern struct option *record_options;
int record__parse_freq(const struct option *opt, const char *str, int unset);
+void record__uniquify_name(struct evlist *evlist);
static inline bool record_opts__no_switch_events(const struct record_opts *opts)
{
--
2.35.1
On Wed, Dec 13, 2023 at 1:36 PM <[email protected]> wrote:
>
> From: Kan Liang <[email protected]>
>
> It's hard to distinguish the default cycles events among hybrid PMUs.
> For example,
>
> $perf top
> Available samples
> 385 cycles:P
> 903 cycles:P
>
> The other tool, e.g., perf record, uniforms the event name and adds the
> hybrid PMU name before opening the event. So the events can be easily
> distinguished. Apply the same methodology for the perf top as well.
>
> The record__uniquify_name() will be invoked by both record and top.
> Move it to util/record.c
>
> With the patch
> $perf top
> Available samples
> 148 cpu_atom/cycles:P/
> 1K cpu_core/cycles:P/
>
> Signed-off-by: Kan Liang <[email protected]>
Reviewed-by: Ian Rogers <[email protected]>
Thanks,
Ian
> ---
>
> New patch to address the display concern
> https://lore.kernel.org/lkml/[email protected]/
>
> tools/perf/builtin-record.c | 28 +---------------------------
> tools/perf/builtin-top.c | 1 +
> tools/perf/util/record.c | 25 +++++++++++++++++++++++++
> tools/perf/util/record.h | 2 ++
> 4 files changed, 29 insertions(+), 27 deletions(-)
>
> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
> index dcf288a4fb9a..a096422a4a14 100644
> --- a/tools/perf/builtin-record.c
> +++ b/tools/perf/builtin-record.c
> @@ -2216,32 +2216,6 @@ static void hit_auxtrace_snapshot_trigger(struct record *rec)
> }
> }
>
> -static void record__uniquify_name(struct record *rec)
> -{
> - struct evsel *pos;
> - struct evlist *evlist = rec->evlist;
> - char *new_name;
> - int ret;
> -
> - if (perf_pmus__num_core_pmus() == 1)
> - return;
> -
> - evlist__for_each_entry(evlist, pos) {
> - if (!evsel__is_hybrid(pos))
> - continue;
> -
> - if (strchr(pos->name, '/'))
> - continue;
> -
> - ret = asprintf(&new_name, "%s/%s/",
> - pos->pmu_name, pos->name);
> - if (ret) {
> - free(pos->name);
> - pos->name = new_name;
> - }
> - }
> -}
> -
> static int record__terminate_thread(struct record_thread *thread_data)
> {
> int err;
> @@ -2475,7 +2449,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
> if (data->is_pipe && rec->evlist->core.nr_entries == 1)
> rec->opts.sample_id = true;
>
> - record__uniquify_name(rec);
> + record__uniquify_name(rec->evlist);
>
> /* Debug message used by test scripts */
> pr_debug3("perf record opening and mmapping events\n");
> diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
> index cce9350177e2..4e8296654280 100644
> --- a/tools/perf/builtin-top.c
> +++ b/tools/perf/builtin-top.c
> @@ -1299,6 +1299,7 @@ static int __cmd_top(struct perf_top *top)
> }
> }
>
> + record__uniquify_name(top->evlist);
> ret = perf_top__start_counters(top);
> if (ret)
> return ret;
> diff --git a/tools/perf/util/record.c b/tools/perf/util/record.c
> index 9eb5c6a08999..5b4be3c72cbc 100644
> --- a/tools/perf/util/record.c
> +++ b/tools/perf/util/record.c
> @@ -289,3 +289,28 @@ int record__parse_freq(const struct option *opt, const char *str, int unset __ma
> opts->user_freq = freq;
> return 0;
> }
> +
> +void record__uniquify_name(struct evlist *evlist)
> +{
> + struct evsel *pos;
> + char *new_name;
> + int ret;
> +
> + if (perf_pmus__num_core_pmus() == 1)
> + return;
> +
> + evlist__for_each_entry(evlist, pos) {
> + if (!evsel__is_hybrid(pos))
> + continue;
> +
> + if (strchr(pos->name, '/'))
> + continue;
> +
> + ret = asprintf(&new_name, "%s/%s/",
> + pos->pmu_name, pos->name);
> + if (ret) {
> + free(pos->name);
> + pos->name = new_name;
> + }
> + }
> +}
> diff --git a/tools/perf/util/record.h b/tools/perf/util/record.h
> index a6566134e09e..9b520ab784bc 100644
> --- a/tools/perf/util/record.h
> +++ b/tools/perf/util/record.h
> @@ -8,6 +8,7 @@
> #include <linux/stddef.h>
> #include <linux/perf_event.h>
> #include "util/target.h"
> +#include "util/evlist.h"
>
> struct option;
>
> @@ -85,6 +86,7 @@ extern const char * const *record_usage;
> extern struct option *record_options;
>
> int record__parse_freq(const struct option *opt, const char *str, int unset);
> +void record__uniquify_name(struct evlist *evlist);
>
> static inline bool record_opts__no_switch_events(const struct record_opts *opts)
> {
> --
> 2.35.1
>
On Wed, Dec 13, 2023 at 1:36 PM <[email protected]> wrote:
>
> From: Kan Liang <[email protected]>
>
> It's hard to distinguish the default cycles events among hybrid PMUs.
> For example,
>
> $perf top
> Available samples
> 385 cycles:P
> 903 cycles:P
>
> The other tool, e.g., perf record, uniforms the event name and adds the
> hybrid PMU name before opening the event. So the events can be easily
> distinguished. Apply the same methodology for the perf top as well.
>
> The record__uniquify_name() will be invoked by both record and top.
> Move it to util/record.c
Maybe better to rename it to evlist__uniquify_name() and move it
to util/evlist.c.
Thanks,
Namhyung
>
> With the patch
> $perf top
> Available samples
> 148 cpu_atom/cycles:P/
> 1K cpu_core/cycles:P/
>
> Signed-off-by: Kan Liang <[email protected]>
> ---
>
> New patch to address the display concern
> https://lore.kernel.org/lkml/[email protected]/
>
> tools/perf/builtin-record.c | 28 +---------------------------
> tools/perf/builtin-top.c | 1 +
> tools/perf/util/record.c | 25 +++++++++++++++++++++++++
> tools/perf/util/record.h | 2 ++
> 4 files changed, 29 insertions(+), 27 deletions(-)
>
> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
> index dcf288a4fb9a..a096422a4a14 100644
> --- a/tools/perf/builtin-record.c
> +++ b/tools/perf/builtin-record.c
> @@ -2216,32 +2216,6 @@ static void hit_auxtrace_snapshot_trigger(struct record *rec)
> }
> }
>
> -static void record__uniquify_name(struct record *rec)
> -{
> - struct evsel *pos;
> - struct evlist *evlist = rec->evlist;
> - char *new_name;
> - int ret;
> -
> - if (perf_pmus__num_core_pmus() == 1)
> - return;
> -
> - evlist__for_each_entry(evlist, pos) {
> - if (!evsel__is_hybrid(pos))
> - continue;
> -
> - if (strchr(pos->name, '/'))
> - continue;
> -
> - ret = asprintf(&new_name, "%s/%s/",
> - pos->pmu_name, pos->name);
> - if (ret) {
> - free(pos->name);
> - pos->name = new_name;
> - }
> - }
> -}
> -
> static int record__terminate_thread(struct record_thread *thread_data)
> {
> int err;
> @@ -2475,7 +2449,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
> if (data->is_pipe && rec->evlist->core.nr_entries == 1)
> rec->opts.sample_id = true;
>
> - record__uniquify_name(rec);
> + record__uniquify_name(rec->evlist);
>
> /* Debug message used by test scripts */
> pr_debug3("perf record opening and mmapping events\n");
> diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
> index cce9350177e2..4e8296654280 100644
> --- a/tools/perf/builtin-top.c
> +++ b/tools/perf/builtin-top.c
> @@ -1299,6 +1299,7 @@ static int __cmd_top(struct perf_top *top)
> }
> }
>
> + record__uniquify_name(top->evlist);
> ret = perf_top__start_counters(top);
> if (ret)
> return ret;
> diff --git a/tools/perf/util/record.c b/tools/perf/util/record.c
> index 9eb5c6a08999..5b4be3c72cbc 100644
> --- a/tools/perf/util/record.c
> +++ b/tools/perf/util/record.c
> @@ -289,3 +289,28 @@ int record__parse_freq(const struct option *opt, const char *str, int unset __ma
> opts->user_freq = freq;
> return 0;
> }
> +
> +void record__uniquify_name(struct evlist *evlist)
> +{
> + struct evsel *pos;
> + char *new_name;
> + int ret;
> +
> + if (perf_pmus__num_core_pmus() == 1)
> + return;
> +
> + evlist__for_each_entry(evlist, pos) {
> + if (!evsel__is_hybrid(pos))
> + continue;
> +
> + if (strchr(pos->name, '/'))
> + continue;
> +
> + ret = asprintf(&new_name, "%s/%s/",
> + pos->pmu_name, pos->name);
> + if (ret) {
> + free(pos->name);
> + pos->name = new_name;
> + }
> + }
> +}
> diff --git a/tools/perf/util/record.h b/tools/perf/util/record.h
> index a6566134e09e..9b520ab784bc 100644
> --- a/tools/perf/util/record.h
> +++ b/tools/perf/util/record.h
> @@ -8,6 +8,7 @@
> #include <linux/stddef.h>
> #include <linux/perf_event.h>
> #include "util/target.h"
> +#include "util/evlist.h"
>
> struct option;
>
> @@ -85,6 +86,7 @@ extern const char * const *record_usage;
> extern struct option *record_options;
>
> int record__parse_freq(const struct option *opt, const char *str, int unset);
> +void record__uniquify_name(struct evlist *evlist);
>
> static inline bool record_opts__no_switch_events(const struct record_opts *opts)
> {
> --
> 2.35.1
>
On 2023-12-13 4:57 p.m., Namhyung Kim wrote:
> On Wed, Dec 13, 2023 at 1:36 PM <[email protected]> wrote:
>>
>> From: Kan Liang <[email protected]>
>>
>> It's hard to distinguish the default cycles events among hybrid PMUs.
>> For example,
>>
>> $perf top
>> Available samples
>> 385 cycles:P
>> 903 cycles:P
>>
>> The other tool, e.g., perf record, uniforms the event name and adds the
>> hybrid PMU name before opening the event. So the events can be easily
>> distinguished. Apply the same methodology for the perf top as well.
>>
>> The record__uniquify_name() will be invoked by both record and top.
>> Move it to util/record.c
>
> Maybe better to rename it to evlist__uniquify_name() and move it
> to util/evlist.c.
Sure, I will change it in V4.
Thanks,
Kan
>
> Thanks,
> Namhyung
>
>>
>> With the patch
>> $perf top
>> Available samples
>> 148 cpu_atom/cycles:P/
>> 1K cpu_core/cycles:P/
>>
>> Signed-off-by: Kan Liang <[email protected]>
>> ---
>>
>> New patch to address the display concern
>> https://lore.kernel.org/lkml/[email protected]/
>>
>> tools/perf/builtin-record.c | 28 +---------------------------
>> tools/perf/builtin-top.c | 1 +
>> tools/perf/util/record.c | 25 +++++++++++++++++++++++++
>> tools/perf/util/record.h | 2 ++
>> 4 files changed, 29 insertions(+), 27 deletions(-)
>>
>> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
>> index dcf288a4fb9a..a096422a4a14 100644
>> --- a/tools/perf/builtin-record.c
>> +++ b/tools/perf/builtin-record.c
>> @@ -2216,32 +2216,6 @@ static void hit_auxtrace_snapshot_trigger(struct record *rec)
>> }
>> }
>>
>> -static void record__uniquify_name(struct record *rec)
>> -{
>> - struct evsel *pos;
>> - struct evlist *evlist = rec->evlist;
>> - char *new_name;
>> - int ret;
>> -
>> - if (perf_pmus__num_core_pmus() == 1)
>> - return;
>> -
>> - evlist__for_each_entry(evlist, pos) {
>> - if (!evsel__is_hybrid(pos))
>> - continue;
>> -
>> - if (strchr(pos->name, '/'))
>> - continue;
>> -
>> - ret = asprintf(&new_name, "%s/%s/",
>> - pos->pmu_name, pos->name);
>> - if (ret) {
>> - free(pos->name);
>> - pos->name = new_name;
>> - }
>> - }
>> -}
>> -
>> static int record__terminate_thread(struct record_thread *thread_data)
>> {
>> int err;
>> @@ -2475,7 +2449,7 @@ static int __cmd_record(struct record *rec, int argc, const char **argv)
>> if (data->is_pipe && rec->evlist->core.nr_entries == 1)
>> rec->opts.sample_id = true;
>>
>> - record__uniquify_name(rec);
>> + record__uniquify_name(rec->evlist);
>>
>> /* Debug message used by test scripts */
>> pr_debug3("perf record opening and mmapping events\n");
>> diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c
>> index cce9350177e2..4e8296654280 100644
>> --- a/tools/perf/builtin-top.c
>> +++ b/tools/perf/builtin-top.c
>> @@ -1299,6 +1299,7 @@ static int __cmd_top(struct perf_top *top)
>> }
>> }
>>
>> + record__uniquify_name(top->evlist);
>> ret = perf_top__start_counters(top);
>> if (ret)
>> return ret;
>> diff --git a/tools/perf/util/record.c b/tools/perf/util/record.c
>> index 9eb5c6a08999..5b4be3c72cbc 100644
>> --- a/tools/perf/util/record.c
>> +++ b/tools/perf/util/record.c
>> @@ -289,3 +289,28 @@ int record__parse_freq(const struct option *opt, const char *str, int unset __ma
>> opts->user_freq = freq;
>> return 0;
>> }
>> +
>> +void record__uniquify_name(struct evlist *evlist)
>> +{
>> + struct evsel *pos;
>> + char *new_name;
>> + int ret;
>> +
>> + if (perf_pmus__num_core_pmus() == 1)
>> + return;
>> +
>> + evlist__for_each_entry(evlist, pos) {
>> + if (!evsel__is_hybrid(pos))
>> + continue;
>> +
>> + if (strchr(pos->name, '/'))
>> + continue;
>> +
>> + ret = asprintf(&new_name, "%s/%s/",
>> + pos->pmu_name, pos->name);
>> + if (ret) {
>> + free(pos->name);
>> + pos->name = new_name;
>> + }
>> + }
>> +}
>> diff --git a/tools/perf/util/record.h b/tools/perf/util/record.h
>> index a6566134e09e..9b520ab784bc 100644
>> --- a/tools/perf/util/record.h
>> +++ b/tools/perf/util/record.h
>> @@ -8,6 +8,7 @@
>> #include <linux/stddef.h>
>> #include <linux/perf_event.h>
>> #include "util/target.h"
>> +#include "util/evlist.h"
>>
>> struct option;
>>
>> @@ -85,6 +86,7 @@ extern const char * const *record_usage;
>> extern struct option *record_options;
>>
>> int record__parse_freq(const struct option *opt, const char *str, int unset);
>> +void record__uniquify_name(struct evlist *evlist);
>>
>> static inline bool record_opts__no_switch_events(const struct record_opts *opts)
>> {
>> --
>> 2.35.1
>>