From: Arnaldo Carvalho de Melo <acme@kernel.org>
To: Adrian Hunter <adrian.hunter@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>,
linux-kernel@vger.kernel.org, David Ahern <dsahern@gmail.com>,
Frederic Weisbecker <fweisbec@gmail.com>,
Jiri Olsa <jolsa@redhat.com>, Namhyung Kim <namhyung@gmail.com>,
Paul Mackerras <paulus@samba.org>,
Stephane Eranian <eranian@google.com>
Subject: Re: [PATCH 02/41] perf tools: Fix map groups of threads with unknown pids
Date: Mon, 14 Jul 2014 17:18:29 -0300 [thread overview]
Message-ID: <20140714201829.GA9449@kernel.org> (raw)
In-Reply-To: <1405332185-4050-3-git-send-email-adrian.hunter@intel.com>
Em Mon, Jul 14, 2014 at 01:02:26PM +0300, Adrian Hunter escreveu:
> Events like sched_switch do not provide a pid (tgid)
> which can result in threads with an unknown pid. If
> the pid is later discovered, join the map groups.
> Note the thread's map groups should be empty because
> they are populated by MMAP events which do provide the
> pid and tid.
>
> Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
> ---
> tools/perf/util/machine.c | 22 ++++++++++++++++------
> tools/perf/util/map.c | 14 ++++++++++++++
> tools/perf/util/map.h | 1 +
> tools/perf/util/thread.c | 35 +++++++++++++++++++++++++++++++++++
> tools/perf/util/thread.h | 1 +
> 5 files changed, 67 insertions(+), 6 deletions(-)
>
> diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c
> index 5b80877..0fa93c1 100644
> --- a/tools/perf/util/machine.c
> +++ b/tools/perf/util/machine.c
> @@ -272,6 +272,17 @@ void machines__set_id_hdr_size(struct machines *machines, u16 id_hdr_size)
> return;
> }
>
> +static void machine__update_thread_pid(struct machine *machine,
> + struct thread *th, pid_t pid)
> +{
> + if (pid != th->pid_ && pid != -1 && th->pid_ == -1) {
> + th->pid_ = pid;
> + if (thread__join_map_groups(th, machine))
> + pr_err("Failed to join map groups for %d:%d\n",
> + th->pid_, th->tid);
> + }
> +}
> +
I think you should combine the above function with
thread__join_map_groups, and I think that the method belongs to the
'machine' class, see below for further comments on this...
> static struct thread *__machine__findnew_thread(struct machine *machine,
> pid_t pid, pid_t tid,
> bool create)
> @@ -285,10 +296,10 @@ static struct thread *__machine__findnew_thread(struct machine *machine,
> * so most of the time we dont have to look up
> * the full rbtree:
> */
> - if (machine->last_match && machine->last_match->tid == tid) {
> - if (pid != -1 && pid != machine->last_match->pid_)
> - machine->last_match->pid_ = pid;
> - return machine->last_match;
> + th = machine->last_match;
> + if (th && th->tid == tid) {
> + machine__update_thread_pid(machine, th, pid);
> + return th;
> }
>
> while (*p != NULL) {
> @@ -297,8 +308,7 @@ static struct thread *__machine__findnew_thread(struct machine *machine,
>
> if (th->tid == tid) {
> machine->last_match = th;
> - if (pid != -1 && pid != th->pid_)
> - th->pid_ = pid;
> + machine__update_thread_pid(machine, th, pid);
> return th;
> }
>
> diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c
> index 25c571f..7af1480 100644
> --- a/tools/perf/util/map.c
> +++ b/tools/perf/util/map.c
> @@ -454,6 +454,20 @@ void map_groups__exit(struct map_groups *mg)
> }
> }
>
> +bool map_groups__empty(struct map_groups *mg)
> +{
> + int i;
> +
> + for (i = 0; i < MAP__NR_TYPES; ++i) {
> + if (maps__first(&mg->maps[i]))
> + return false;
> + if (!list_empty(&mg->removed_maps[i]))
> + return false;
> + }
> +
> + return true;
> +}
> +
> struct map_groups *map_groups__new(void)
> {
> struct map_groups *mg = malloc(sizeof(*mg));
> diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h
> index 7758c72..5806a90 100644
> --- a/tools/perf/util/map.h
> +++ b/tools/perf/util/map.h
> @@ -66,6 +66,7 @@ struct map_groups {
>
> struct map_groups *map_groups__new(void);
> void map_groups__delete(struct map_groups *mg);
> +bool map_groups__empty(struct map_groups *mg);
>
> static inline struct map_groups *map_groups__get(struct map_groups *mg)
> {
> diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c
> index 7a32f44..ca94295 100644
> --- a/tools/perf/util/thread.c
> +++ b/tools/perf/util/thread.c
> @@ -24,6 +24,41 @@ int thread__init_map_groups(struct thread *thread, struct machine *machine)
> return thread->mg ? 0 : -1;
> }
>
> +int thread__join_map_groups(struct thread *thread, struct machine *machine)
> +{
> + struct thread *leader;
> + pid_t pid = thread->pid_;
> +
> + if (pid == thread->tid)
> + return 0;
> +
> + leader = machine__findnew_thread(machine, pid, pid);
> + if (!leader)
> + return -ENOMEM;
> +
> + if (!leader->mg)
> + leader->mg = map_groups__new();
> +
> + if (!leader->mg)
> + return -ENOMEM;
> +
> + if (thread->mg) {
> + /*
> + * Maps are created from MMAP events which provide the pid and
> + * tid. Consequently there never should be any maps on a thread
> + * with an unknown pid. Just print an error if there are.
> + */
> + if (!map_groups__empty(thread->mg))
> + pr_err("Discarding thread maps for %d:%d\n",
> + thread->pid_, thread->tid);
Because there is nothing in this function (thread__join_map_groups),
checking if it is operating on a thread that has an unknown pid.
So, please consider merging this and its sole caller,
machine__update_thread_pid, so that the test is in the same function as
the assumption that it is operating on a thread with an unknown pid.
Continuing to look at the other patches...
> + map_groups__delete(thread->mg);
> + }
> +
> + thread->mg = map_groups__get(leader->mg);
> +
> + return 0;
> +}
> +
> struct thread *thread__new(pid_t pid, pid_t tid)
> {
> char *comm_str;
> diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h
> index 3c0c272..9de0629 100644
> --- a/tools/perf/util/thread.h
> +++ b/tools/perf/util/thread.h
> @@ -31,6 +31,7 @@ struct comm;
>
> struct thread *thread__new(pid_t pid, pid_t tid);
> int thread__init_map_groups(struct thread *thread, struct machine *machine);
> +int thread__join_map_groups(struct thread *thread, struct machine *machine);
> void thread__delete(struct thread *thread);
> static inline void thread__exited(struct thread *thread)
> {
> --
> 1.8.3.2
next prev parent reply other threads:[~2014-07-14 20:18 UTC|newest]
Thread overview: 122+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-07-14 10:02 [PATCH 00/41] perf tools: Preparation for call graph from Intel BTS Adrian Hunter
2014-07-14 10:02 ` [PATCH 01/41] perf tools: Fix the value used for unknown pids Adrian Hunter
2014-07-18 4:23 ` [tip:perf/core] perf machine: " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 02/41] perf tools: Fix map groups of threads with " Adrian Hunter
2014-07-14 20:18 ` Arnaldo Carvalho de Melo [this message]
2014-07-15 11:33 ` Adrian Hunter
2014-07-15 19:33 ` Arnaldo Carvalho de Melo
2014-07-16 8:07 ` [PATCH V2 " Adrian Hunter
2014-07-18 4:26 ` [tip:perf/core] perf machine: " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 03/41] perf script: Display PERF_RECORD_MISC_COMM_EXEC flag Adrian Hunter
2014-07-18 4:23 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 04/41] perf record: Select comm_exec flag if supported Adrian Hunter
2014-07-18 4:23 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 05/41] perf tools: Identify which comms are from exec Adrian Hunter
2014-07-14 20:32 ` Arnaldo Carvalho de Melo
2014-07-15 11:43 ` Adrian Hunter
2014-07-23 14:07 ` Arnaldo Carvalho de Melo
2014-07-23 14:09 ` Arnaldo Carvalho de Melo
2014-07-23 16:46 ` Adrian Hunter
2014-07-23 16:53 ` Arnaldo Carvalho de Melo
2014-07-14 10:02 ` [PATCH 06/41] perf tools: Add machine__thread_exec_comm() Adrian Hunter
2014-07-14 10:02 ` [PATCH 07/41] perf tools: Fix missing kernel map load Adrian Hunter
2014-07-18 4:23 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 08/41] perf tools: Fix missing label symbols Adrian Hunter
2014-07-14 20:38 ` Arnaldo Carvalho de Melo
2014-07-15 12:11 ` Adrian Hunter
2014-07-14 10:02 ` [PATCH 09/41] perf tools: Fix missing GNU IFUNC symbols Adrian Hunter
2014-07-18 4:24 ` [tip:perf/core] perf symbols: " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 10/41] perf inject: Fix build id injection Adrian Hunter
2014-07-18 4:24 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 11/41] perf tools: Fix appending a callchain from a previous sample Adrian Hunter
2014-07-18 4:24 ` [tip:perf/core] perf callchain: " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 12/41] perf tools: Fix leak of 'struct thread' on error path Adrian Hunter
2014-07-14 20:43 ` Arnaldo Carvalho de Melo
2014-07-16 7:19 ` [PATCH V2 0/2] perf tools: Allow deletion of a thread with no map groups Adrian Hunter
2014-07-16 7:19 ` [PATCH 1/2] " Adrian Hunter
2014-07-18 4:27 ` [tip:perf/core] perf thread: " tip-bot for Adrian Hunter
2014-07-16 7:19 ` [PATCH V2 2/2] perf tools: Fix leak of 'struct thread' on error path Adrian Hunter
2014-07-18 4:27 ` [tip:perf/core] perf machine: " tip-bot for Adrian Hunter
2014-07-16 14:02 ` [PATCH 12/41] perf tools: " Jiri Olsa
2014-07-16 14:26 ` Adrian Hunter
2014-07-14 10:02 ` [PATCH 13/41] perf tools: Add machine__kernel_ip() Adrian Hunter
2014-07-16 14:15 ` Jiri Olsa
2014-07-16 14:22 ` Adrian Hunter
2014-08-11 12:23 ` Jiri Olsa
2014-08-11 12:36 ` Adrian Hunter
2014-08-11 12:43 ` Jiri Olsa
2014-08-11 12:46 ` Adrian Hunter
2014-08-11 12:55 ` Jiri Olsa
2014-07-14 10:02 ` [PATCH 14/41] perf buildid-cache: Apply force option to copying kcore Adrian Hunter
2014-07-18 4:24 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 15/41] perf script: Improve srcline display for BTS Adrian Hunter
2014-07-15 14:16 ` Arnaldo Carvalho de Melo
2014-07-15 17:51 ` Adrian Hunter
2014-07-16 2:08 ` David Ahern
2014-07-14 10:02 ` [PATCH 16/41] perf script: Do not print dangling '=>' " Adrian Hunter
2014-07-14 10:02 ` [PATCH 17/41] perf tools: Record whether a dso is 64-bit Adrian Hunter
2014-07-18 4:24 ` [tip:perf/core] perf symbols: " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 18/41] perf tools: Record whether a dso has data Adrian Hunter
2014-07-16 14:20 ` Jiri Olsa
2014-07-17 8:43 ` [PATCH 0/2] perf tools: Fix incorrect fd error comparison Adrian Hunter
2014-07-17 8:43 ` [PATCH 1/2] " Adrian Hunter
2014-07-22 7:51 ` Jiri Olsa
2014-07-28 8:21 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-17 8:43 ` [PATCH V2 2/2] perf tools: Record whether a dso has data Adrian Hunter
2014-07-17 8:58 ` [PATCH V3] " Adrian Hunter
2014-07-22 7:55 ` Jiri Olsa
2014-07-22 13:24 ` Adrian Hunter
2014-07-14 10:02 ` [PATCH 19/41] perf tools: Do not attempt to read data from kallsyms Adrian Hunter
2014-07-18 4:25 ` [tip:perf/core] perf symbols: " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 20/41] perf tools: Let a user specify a PMU event without any config terms Adrian Hunter
2014-07-16 14:25 ` Jiri Olsa
2014-07-16 15:04 ` Adrian Hunter
2014-07-16 18:22 ` Jiri Olsa
2014-08-29 18:48 ` Adrian Hunter
2014-08-30 8:53 ` Jiri Olsa
2014-09-01 6:27 ` Adrian Hunter
2014-09-01 19:11 ` Jiri Olsa
2014-09-02 5:39 ` Adrian Hunter
2014-09-02 9:17 ` Jiri Olsa
2014-09-01 15:51 ` Arnaldo Carvalho de Melo
2014-09-01 19:15 ` Jiri Olsa
2014-07-14 10:02 ` [PATCH 21/41] perf tools: Let default config be defined for a PMU Adrian Hunter
2014-07-14 10:02 ` [PATCH 22/41] perf tools: Add perf_pmu__scan_file() Adrian Hunter
2014-07-14 10:02 ` [PATCH 23/41] perf tools: Add dsos__hit_all() Adrian Hunter
2014-07-14 10:02 ` [PATCH 24/41] perf tools: Add cpu to struct thread Adrian Hunter
2014-07-15 14:24 ` Arnaldo Carvalho de Melo
2014-07-15 17:58 ` Adrian Hunter
2014-07-14 10:02 ` [PATCH 25/41] perf tools: Add ability to record the current tid for each cpu Adrian Hunter
2014-07-14 10:02 ` [PATCH 26/41] perf tools: Add ability to iterate over a dso's symbols Adrian Hunter
2014-07-18 4:25 ` [tip:perf/core] perf symbols: Add ability to iterate over a dso' s symbols tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 27/41] perf session: Flag if the event stream is entirely in memory Adrian Hunter
2014-07-18 4:25 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 28/41] perf evlist: Pass mmap parameters in a struct Adrian Hunter
2014-07-18 4:25 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 29/41] perf tools: Add feature test for __sync_val_compare_and_swap Adrian Hunter
2014-07-18 4:26 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 30/41] perf tools: Add option macro OPT_CALLBACK_OPTARG Adrian Hunter
2014-07-18 4:26 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 31/41] perf evlist: Add perf_evlist__set_tracking_event() Adrian Hunter
2014-07-14 10:02 ` [PATCH 32/41] perf evsel: Add 'no_aux_samples' option Adrian Hunter
2014-07-18 4:26 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 33/41] perf evsel: Add 'immediate' option Adrian Hunter
2014-07-18 4:26 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:02 ` [PATCH 34/41] perf evlist: Add 'system_wide' option Adrian Hunter
2014-07-15 15:03 ` Arnaldo Carvalho de Melo
2014-07-15 18:05 ` Adrian Hunter
2014-08-11 13:12 ` Jiri Olsa
2014-08-12 5:52 ` Adrian Hunter
2014-07-14 10:02 ` [PATCH 35/41] perf tools: Add id index Adrian Hunter
2014-07-14 10:03 ` [PATCH 36/41] perf pmu: Let pmu's with no events show up on perf list Adrian Hunter
2014-07-14 10:03 ` [PATCH 37/41] perf session: Add ability to skip 4GiB or more Adrian Hunter
2014-07-14 10:03 ` [PATCH 38/41] perf session: Add perf_session__deliver_synth_event() Adrian Hunter
2014-07-15 15:17 ` Arnaldo Carvalho de Melo
2014-07-15 18:32 ` Adrian Hunter
2014-07-14 10:03 ` [PATCH 39/41] perf tools: Allow TSC conversion on any arch Adrian Hunter
2014-07-15 15:20 ` Arnaldo Carvalho de Melo
2014-07-15 16:17 ` Peter Zijlstra
2014-07-15 17:48 ` Adrian Hunter
2014-07-18 4:28 ` [tip:perf/core] " tip-bot for Adrian Hunter
2014-07-14 10:03 ` [PATCH 40/41] perf tools: Move rdtsc() function Adrian Hunter
2014-07-14 10:03 ` [PATCH 41/41] perf evlist: Add perf_evlist__enable_event_idx() Adrian Hunter
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=20140714201829.GA9449@kernel.org \
--to=acme@kernel.org \
--cc=adrian.hunter@intel.com \
--cc=dsahern@gmail.com \
--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).