Re: [PATCH] perf stat: Ignore error thread when enabling system-wide --per-thread

[Jin, Yao]

Just tested. But looks it's not OK for '--per-thread' case.

jinyao@skl:~/$ ./perf stat --per-thread
WARNING: Ignored open failure for pid 1
WARNING: Ignored open failure for pid 2
WARNING: Ignored open failure for pid 4
WARNING: Ignored open failure for pid 6
WARNING: Ignored open failure for pid 7
WARNING: Ignored open failure for pid 8
WARNING: Ignored open failure for pid 9
WARNING: Ignored open failure for pid 10
......
WARNING: Ignored open failure for pid 11783
WARNING: Ignored open failure for pid 12275
WARNING: Ignored open failure for pid 31733
WARNING: Ignored open failure for pid 31737
Error:
You may not have permission to collect system-wide stats.

Consider tweaking /proc/sys/kernel/perf_event_paranoid,
which controls use of the performance events system by
unprivileged users (without CAP_SYS_ADMIN).

The current value is 2:

  -1: Allow use of (almost) all events by all users
      Ignore mlock limit after perf_event_mlock_kb without CAP_IPC_LOCK
>= 0: Disallow ftrace function tracepoint by users without CAP_SYS_ADMIN
      Disallow raw tracepoint access by users without CAP_SYS_ADMIN
>= 1: Disallow CPU event access by users without CAP_SYS_ADMIN
>= 2: Disallow kernel profiling by users without CAP_SYS_ADMIN

To make this setting permanent, edit /etc/sysctl.conf too, e.g.:

        kernel.perf_event_paranoid = -1

Thanks
Jin Yao

On 1/16/2018 8:51 PM, Jiri Olsa wrote:
> On Tue, Jan 16, 2018 at 11:43:08PM +0800, Jin Yao wrote:
> > If we execute 'perf stat --per-thread' with non-root account
> > (even set kernel.perf_event_paranoid = -1 yet), it reports the error:
> >
> > jinyao@skl:~$ perf stat --per-thread
> > Error:
> > You may not have permission to collect system-wide stats.
> >
> > Consider tweaking /proc/sys/kernel/perf_event_paranoid,
> > which controls use of the performance events system by
> > unprivileged users (without CAP_SYS_ADMIN).
> >
> > The current value is 2:
> >
> >    -1: Allow use of (almost) all events by all users
> >        Ignore mlock limit after perf_event_mlock_kb without CAP_IPC_LOCK
> > > = 0: Disallow ftrace function tracepoint by users without CAP_SYS_ADMIN
> >        Disallow raw tracepoint access by users without CAP_SYS_ADMIN
> > > = 1: Disallow CPU event access by users without CAP_SYS_ADMIN
> > > = 2: Disallow kernel profiling by users without CAP_SYS_ADMIN
> >
> > To make this setting permanent, edit /etc/sysctl.conf too, e.g.:
> >
> >          kernel.perf_event_paranoid = -1
> >
> > Perhaps the ptrace rule doesn't allow to trace some processes. But anyway
> > the global --per-thread mode had better ignore such errors and continue
> > working on other threads.
> >
> > This patch will record the index of error thread in perf_evsel__open()
> > and remove this thread before retrying.
> >
> > For example (run with non-root, kernel.perf_event_paranoid isn't set):
> >
> > jinyao@skl:~$ perf stat --per-thread
> > ^C
> >   Performance counter stats for 'system wide':
> >
> >            vmstat-3458               6.171984      cpu-clock:u (msec)        #    0.000 CPUs utilized
> >              perf-3670               0.515599      cpu-clock:u (msec)        #    0.000 CPUs utilized
> >            vmstat-3458              1,163,643      cycles:u                  #    0.189 GHz
> >              perf-3670                 40,881      cycles:u                  #    0.079 GHz
> >            vmstat-3458              1,410,238      instructions:u            #    1.21  insn per cycle
> >              perf-3670                  3,536      instructions:u            #    0.09  insn per cycle
> >            vmstat-3458                288,937      branches:u                #   46.814 M/sec
> >              perf-3670                    936      branches:u                #    1.815 M/sec
> >            vmstat-3458                 15,195      branch-misses:u           #    5.26% of all branches
> >              perf-3670                     76      branch-misses:u           #    8.12% of all branches
> >
> >        12.651675247 seconds time elapsed
>
> could we use the existing code like in attached patch?
>
> we might also swap following warning for some generic one:
>    WARNING: Ignored open failure for pid 28392
>
> jirka
>
>
> ---
> diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c
> index 98bf9d32f222..f9d4d3ad6fc5 100644
> --- a/tools/perf/builtin-stat.c
> +++ b/tools/perf/builtin-stat.c
> @@ -276,6 +276,9 @@ static int create_perf_stat_counter(struct perf_evsel *evsel)
>   			attr->enable_on_exec = 1;
>   	}
>   
> +	if (stat_config.aggr_mode == AGGR_THREAD)
> +		evsel->ignore_missing_thread = true;
> +
>   	if (target__has_cpu(&target) && !target__has_per_thread(&target))
>   		return perf_evsel__open_per_cpu(evsel, perf_evsel__cpus(evsel));
>   
> diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
> index 8f971a2301d1..509ee175bc97 100644
> --- a/tools/perf/util/evsel.c
> +++ b/tools/perf/util/evsel.c
> @@ -1656,7 +1656,7 @@ static bool ignore_missing_thread(struct perf_evsel *evsel,
>   		return false;
>   
>   	/* The -ESRCH is perf event syscall errno for pid's not found. */
> -	if (err != -ESRCH)
> +	if (err != -ESRCH && err != -EACCES)
>   		return false;
>   
>   	/* If there's only one thread, let it fail. */