From: David Ahern <dsahern@gmail.com>
To: Adrian Hunter <adrian.hunter@intel.com>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>,
linux-kernel@vger.kernel.org,
Frederic Weisbecker <fweisbec@gmail.com>,
Jiri Olsa <jolsa@redhat.com>, Mike Galbraith <efault@gmx.de>,
Namhyung Kim <namhyung@gmail.com>,
Paul Mackerras <paulus@samba.org>,
Peter Zijlstra <peterz@infradead.org>,
Stephane Eranian <eranian@google.com>
Subject: Re: [PATCH V2 12/15] perf tools: allow non-matching sample types
Date: Thu, 27 Jun 2013 10:39:11 -0600 [thread overview]
Message-ID: <51CC6AAF.4030404@gmail.com> (raw)
In-Reply-To: <1372319707-13892-13-git-send-email-adrian.hunter@intel.com>
On 6/27/13 1:55 AM, Adrian Hunter wrote:
> Sample types need not be identical to determine
> the sample id from the event. Only the position
> of the sample id needs to be the same.
>
> Compatible sample types are ones in which the bits
> defined by PERF_COMPAT_MASK are the same.
> 'perf_evlist__config()' forces sample types to be
> compatible on that basis.
>
> Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
> ---
> tools/perf/builtin-report.c | 2 +-
> tools/perf/util/event.h | 14 +++++
> tools/perf/util/evlist.c | 135 ++++++++++++++++++++++++++++++++++++++++++--
> tools/perf/util/evlist.h | 8 ++-
> tools/perf/util/evsel.c | 64 ++++++++++++++++++++-
> tools/perf/util/evsel.h | 10 ++++
> tools/perf/util/session.c | 8 ++-
> 7 files changed, 230 insertions(+), 11 deletions(-)
>
> diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
> index ca98d34..58572ac 100644
> --- a/tools/perf/builtin-report.c
> +++ b/tools/perf/builtin-report.c
> @@ -367,7 +367,7 @@ static int process_read_event(struct perf_tool *tool,
> static int perf_report__setup_sample_type(struct perf_report *rep)
> {
> struct perf_session *self = rep->session;
> - u64 sample_type = perf_evlist__sample_type(self->evlist);
> + u64 sample_type = perf_evlist__combined_sample_type(self->evlist);
>
> if (!self->fd_pipe && !(sample_type & PERF_SAMPLE_CALLCHAIN)) {
> if (sort__has_parent) {
> diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
> index 1813895..3aef78c 100644
> --- a/tools/perf/util/event.h
> +++ b/tools/perf/util/event.h
> @@ -65,6 +65,20 @@ struct read_event {
> PERF_SAMPLE_ID | PERF_SAMPLE_STREAM_ID | \
> PERF_SAMPLE_CPU | PERF_SAMPLE_PERIOD)
>
> +/*
> + * Events have compatible sample types if the following bits all have the same
> + * value. This is because the order of sample members is fixed. For sample
> + * events the order is: PERF_SAMPLE_IP, PERF_SAMPLE_TID, PERF_SAMPLE_TIME,
> + * PERF_SAMPLE_ADDR, PERF_SAMPLE_ID. For non-sample events the sample members
> + * are accessed in reverse order. The order is: PERF_SAMPLE_ID,
> + * PERF_SAMPLE_STREAM_ID, PERF_SAMPLE_CPU.
> + */
> +#define PERF_COMPAT_MASK \
> + (PERF_SAMPLE_IP | PERF_SAMPLE_TID | \
> + PERF_SAMPLE_TIME | PERF_SAMPLE_ADDR | \
> + PERF_SAMPLE_ID | PERF_SAMPLE_STREAM_ID | \
> + PERF_SAMPLE_CPU)
> +
> struct sample_event {
> struct perf_event_header header;
> u64 array[];
> diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
> index a660f56..78331da 100644
> --- a/tools/perf/util/evlist.c
> +++ b/tools/perf/util/evlist.c
> @@ -49,6 +49,45 @@ struct perf_evlist *perf_evlist__new(void)
> return evlist;
> }
>
> +/**
> + * perf_evlist__set_id_pos - set the positions of event ids.
> + * @evlist: selected event list
> + *
> + * Events with compatible sample types all have the same id_pos
> + * and is_pos. For convenience, put a copy on evlist.
> + */
> +static void perf_evlist__set_id_pos(struct perf_evlist *evlist)
> +{
> + struct perf_evsel *first = perf_evlist__first(evlist);
> +
> + evlist->id_pos = first->id_pos;
> + evlist->is_pos = first->is_pos;
> +}
> +
> +/**
> + * perf_evlist__make_sample_types_compatible - make sample types compatible.
> + * @evlist: selected event list
> + *
> + * Events with compatible sample types all have the same id_pos and is_pos.
> + * This can be achieved by matching the bits of PERF_COMPAT_MASK.
> + */
> +void perf_evlist__make_sample_types_compatible(struct perf_evlist *evlist)
> +{
> + struct perf_evsel *evsel;
> + u64 compat = 0;
> +
> + list_for_each_entry(evsel, &evlist->entries, node)
> + compat |= evsel->attr.sample_type & PERF_COMPAT_MASK;
> +
> + list_for_each_entry(evsel, &evlist->entries, node) {
> + evsel->attr.sample_type |= compat;
> + evsel->sample_size = __perf_evsel__sample_size(evsel->attr.sample_type);
> + perf_evsel__calc_id_pos(evsel);
> + }
> +
> + perf_evlist__set_id_pos(evlist);
> +}
> +
While this works for a combined S/W and tracepoint events session, I do
not like promoting sample types to the minimum compatible level for all
events in the session. perf needs to allow each event to have its own
sample_type and not force a minimal compatibility.
David
next prev parent reply other threads:[~2013-06-27 16:39 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-27 7:54 [PATCH V2 00/15] perf tools: some fixes and tweaks Adrian Hunter
2013-06-27 7:54 ` [PATCH V2 01/15] perf tools: remove unused parameter Adrian Hunter
2013-06-27 7:54 ` [PATCH V2 02/15] perf tools: fix missing tool parameter Adrian Hunter
2013-06-27 7:54 ` [PATCH V2 03/15] perf tools: fix missing 'finished_round' Adrian Hunter
2013-06-27 7:54 ` [PATCH V2 04/15] perf tools: fix parse_events_terms() segfault on error path Adrian Hunter
2013-06-27 7:54 ` [PATCH V2 05/15] perf tools: fix new_term() missing free " Adrian Hunter
2013-06-27 7:54 ` [PATCH V2 06/15] perf tools: fix parse_events_terms() freeing local variable " Adrian Hunter
2013-06-27 16:13 ` David Ahern
2013-06-28 6:32 ` Adrian Hunter
2013-06-28 13:41 ` David Ahern
2013-06-28 13:57 ` Jiri Olsa
2013-06-27 7:54 ` [PATCH V2 07/15] perf tools: add const specifier to perf_pmu__find name parameter Adrian Hunter
2013-06-27 7:55 ` [PATCH V2 08/15] perf tools: tidy duplicated munmap code Adrian Hunter
2013-06-27 7:55 ` [PATCH V2 09/15] perf tools: validate perf event header size Adrian Hunter
2013-06-27 7:55 ` [PATCH V2 10/15] perf tools: add debug prints Adrian Hunter
2013-06-27 7:55 ` [PATCH V2 11/15] perf tools: fix symbol_conf.nr_events Adrian Hunter
2013-06-27 7:55 ` [PATCH V2 12/15] perf tools: allow non-matching sample types Adrian Hunter
2013-06-27 16:39 ` David Ahern [this message]
2013-07-01 9:32 ` Adrian Hunter
2013-07-01 18:53 ` David Ahern
2013-07-01 19:10 ` Stephane Eranian
2013-07-02 6:58 ` Adrian Hunter
2013-07-03 6:40 ` Stephane Eranian
2013-07-02 7:09 ` Adrian Hunter
2013-07-03 6:44 ` Stephane Eranian
2013-08-26 18:54 ` Arnaldo Carvalho de Melo
2013-06-27 7:55 ` [PATCH V2 13/15] perf tools: struct thread has a tid not a pid Adrian Hunter
2013-06-27 7:55 ` [PATCH V2 14/15] perf tools: add pid to struct thread Adrian Hunter
2013-06-27 16:52 ` David Ahern
2013-06-27 7:55 ` [PATCH V2 15/15] perf tools: fix ppid in thread__fork() Adrian Hunter
2013-06-27 16:57 ` David Ahern
2013-06-28 6:47 ` Adrian Hunter
2013-06-28 13:46 ` David Ahern
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=51CC6AAF.4030404@gmail.com \
--to=dsahern@gmail.com \
--cc=acme@ghostprotocols.net \
--cc=adrian.hunter@intel.com \
--cc=efault@gmx.de \
--cc=eranian@google.com \
--cc=fweisbec@gmail.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=namhyung@gmail.com \
--cc=paulus@samba.org \
--cc=peterz@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).