Re: [PATCH] perf, record: Add clockid parameter

[David Ahern <dsahern@gmail.com>]

On 3/27/15 8:32 AM, Peter Zijlstra wrote:
> On Fri, Mar 27, 2015 at 04:48:08AM -0700, tip-bot for Peter Zijlstra wrote:
>> perf: Add per event clockid support
>
> And here the accompanying userspace; which I'd totally forgotten about.
>
> XXX: do we want to store the clockid in the data file as well, such that
> we can verify at perf-inject time the clocks match with our
> expectations?
>
> ---
> Subject: perf, record: Add clockid parameter
>
> Teach perf-record about the new perf_event_attr::{use_clockid, clockid}
> fields. Add a simple parameter to set the clock (if any) to be used for
> the events to be recorded into the data file.
>
> Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> ---
>   tools/perf/builtin-record.c | 3 +++
>   tools/perf/perf.h           | 1 +
>   tools/perf/util/evsel.c     | 5 +++++
>   3 files changed, 9 insertions(+)

missing Documentation/perf-record.txt update

>
> diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c
> index 18aad239b401..9d4ed884b1c8 100644
> --- a/tools/perf/builtin-record.c
> +++ b/tools/perf/builtin-record.c
> @@ -739,6 +739,7 @@ static struct record record = {
>   			.uses_mmap   = true,
>   			.default_per_cpu = true,
>   		},
> +		.clockid             = -1,
>   	},
>   	.tool = {
>   		.sample		= process_sample_event,
> @@ -842,6 +843,8 @@ struct option __record_options[] = {
>   		    "Sample machine registers on interrupt"),
>   	OPT_BOOLEAN(0, "running-time", &record.opts.running_time,
>   		    "Record running/enabled time of read (:S) events"),
> +	OPT_INTEGER('k', "clockid", &record.opts.clockid,
> +		    "clockid to use for events"),
>   	OPT_END()
>   };
>
> diff --git a/tools/perf/perf.h b/tools/perf/perf.h
> index c38a085a5571..275c0c58fbbe 100644
> --- a/tools/perf/perf.h
> +++ b/tools/perf/perf.h
> @@ -62,6 +62,7 @@ struct record_opts {
>   	u64	     user_interval;
>   	bool	     sample_transaction;
>   	unsigned     initial_delay;
> +	clockid_t    clockid;
>   };
>
>   struct option;
> diff --git a/tools/perf/util/evsel.c b/tools/perf/util/evsel.c
> index 358e5954baa8..309208b16632 100644
> --- a/tools/perf/util/evsel.c
> +++ b/tools/perf/util/evsel.c
> @@ -761,6 +761,11 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
>   		attr->disabled = 0;
>   		attr->enable_on_exec = 0;
>   	}
> +
> +	if (opts->clockid >= 0) {
> +		attr->use_clockid = 1;
> +		attr->clockid = opts->clockid;
> +	}
>   }
>
>   static int perf_evsel__alloc_fd(struct perf_evsel *evsel, int ncpus, int nthreads)
>

This is a new feature which means use_clockid on older kernels will 
fail. So need to catch that and throw an error -- perhaps yet another 
probe function.

Also, if the intent is to allow clock selection per event should there 
be an event modifier as well (see get_event_modifier())?