Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1600817
| From | Arnaldo Carvalho de Melo <acme@kernel.org> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | [PATCH 14/19] perf sched timehist: Add --next option |
| Date | 2017-03-14 20:00 +0100 |
| Message-ID | <tkZPQ-4ju-41@gated-at.bofh.it> (permalink) |
| References | <tkZPP-4ju-5@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
From: Brendan Gregg <bgregg@netflix.com>
The --next option shows the next task for each context switch, providing
more context for the sequence of scheduler events.
$ perf sched timehist --next | head
Samples do not have callchains.
time cpu task name waittime schdelay run time
[tid/pid] (msec) (msec) (msec)
---------- --- ---------- --------- ------ -----
374.793792 [0] <idle> 0.000 0.000 0.000 next: rngd[1524]
374.793801 [0] rngd[1524] 0.000 0.000 0.009 next: swapper/0[0]
374.794048 [7] <idle> 0.000 0.000 0.000 next: yes[30884]
374.794066 [7] yes[30884] 0.000 0.000 0.018 next: swapper/7[0]
374.794126 [2] <idle> 0.000 0.000 0.000 next: rngd[1524]
374.794140 [2] rngd[1524] 0.325 0.006 0.013 next: swapper/2[0]
374.794281 [3] <idle> 0.000 0.000 0.000 next: perf[31070]
Signed-off-by: Brendan Gregg <bgregg@netflix.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1489456589-32555-1-git-send-email-bgregg@netflix.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
tools/perf/Documentation/perf-sched.txt | 4 ++++
tools/perf/builtin-sched.c | 25 ++++++++++++++++++++-----
2 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/tools/perf/Documentation/perf-sched.txt b/tools/perf/Documentation/perf-sched.txt
index d33deddb0146..a092a2499e8f 100644
--- a/tools/perf/Documentation/perf-sched.txt
+++ b/tools/perf/Documentation/perf-sched.txt
@@ -132,6 +132,10 @@ OPTIONS for 'perf sched timehist'
--migrations::
Show migration events.
+-n::
+--next::
+ Show next task.
+
-I::
--idle-hist::
Show idle-related events only.
diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 16170e9b47e6..b92c4d97192c 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -221,6 +221,7 @@ struct perf_sched {
unsigned int max_stack;
bool show_cpu_visual;
bool show_wakeups;
+ bool show_next;
bool show_migrations;
bool show_state;
u64 skipped_samples;
@@ -1897,14 +1898,18 @@ static char task_state_char(struct thread *thread, int state)
}
static void timehist_print_sample(struct perf_sched *sched,
+ struct perf_evsel *evsel,
struct perf_sample *sample,
struct addr_location *al,
struct thread *thread,
u64 t, int state)
{
struct thread_runtime *tr = thread__priv(thread);
+ const char *next_comm = perf_evsel__strval(evsel, sample, "next_comm");
+ const u32 next_pid = perf_evsel__intval(evsel, sample, "next_pid");
u32 max_cpus = sched->max_cpu + 1;
char tstr[64];
+ char nstr[30];
u64 wait_time;
timestamp__scnprintf_usec(t, tstr, sizeof(tstr));
@@ -1937,7 +1942,12 @@ static void timehist_print_sample(struct perf_sched *sched,
if (sched->show_state)
printf(" %5c ", task_state_char(thread, state));
- if (sched->show_wakeups)
+ if (sched->show_next) {
+ snprintf(nstr, sizeof(nstr), "next: %s[%d]", next_comm, next_pid);
+ printf(" %-*s", comm_width, nstr);
+ }
+
+ if (sched->show_wakeups && !sched->show_next)
printf(" %-*s", comm_width, "");
if (thread->tid == 0)
@@ -2531,7 +2541,7 @@ static int timehist_sched_change_event(struct perf_tool *tool,
}
if (!sched->summary_only)
- timehist_print_sample(sched, sample, &al, thread, t, state);
+ timehist_print_sample(sched, evsel, sample, &al, thread, t, state);
out:
if (sched->hist_time.start == 0 && t >= ptime->start)
@@ -3341,6 +3351,7 @@ int cmd_sched(int argc, const char **argv, const char *prefix __maybe_unused)
OPT_BOOLEAN('S', "with-summary", &sched.summary,
"Show all syscalls and summary with statistics"),
OPT_BOOLEAN('w', "wakeups", &sched.show_wakeups, "Show wakeup events"),
+ OPT_BOOLEAN('n', "next", &sched.show_next, "Show next task"),
OPT_BOOLEAN('M', "migrations", &sched.show_migrations, "Show migration events"),
OPT_BOOLEAN('V', "cpu-visual", &sched.show_cpu_visual, "Add CPU visual"),
OPT_BOOLEAN('I', "idle-hist", &sched.idle_hist, "Show idle events only"),
@@ -3438,10 +3449,14 @@ int cmd_sched(int argc, const char **argv, const char *prefix __maybe_unused)
if (argc)
usage_with_options(timehist_usage, timehist_options);
}
- if (sched.show_wakeups && sched.summary_only) {
- pr_err(" Error: -s and -w are mutually exclusive.\n");
+ if ((sched.show_wakeups || sched.show_next) &&
+ sched.summary_only) {
+ pr_err(" Error: -s and -[n|w] are mutually exclusive.\n");
parse_options_usage(timehist_usage, timehist_options, "s", true);
- parse_options_usage(NULL, timehist_options, "w", true);
+ if (sched.show_wakeups)
+ parse_options_usage(NULL, timehist_options, "w", true);
+ if (sched.show_next)
+ parse_options_usage(NULL, timehist_options, "n", true);
return -EINVAL;
}
--
2.9.3
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
[GIT PULL 00/19] perf/core improvements and fixes Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 17/19] perf powerpc: Choose local entry point with kretprobes Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 12/19] perf script: Add script print support for namespace events Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 16/19] perf kretprobes: Offset from reloc_sym if kernel supports it Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 04/19] perf tools: Missing c2c command in command-list Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 07/19] perf report: Document +field style argument support for --field option Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 05/19] perf tools: Ignore generated files pmu-events/{jevents,pmu-events.c} for git Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 06/19] perf sort: Fix segfault with basic block 'cycles' sort dimension Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 19/19] kprobes: Convert kprobe_exceptions_notify to use NOKPROBE_SYMBOL Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 14/19] perf sched timehist: Add --next option Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 13/19] perf tools: Add 'cgroup_id' sort order keyword Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 03/19] perf c2c: Fix display bug when using pipe Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 01/19] perf report: Hide tip message when -q option is given Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 02/19] perf c2c: Clarify help message of --stats option Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 18/19] doc: trace/kprobes: add information about NOKPROBE_SYMBOL Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 10/19] perf tools: Add PERF_RECORD_NAMESPACES to include namespaces related info Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 08/19] perf hists browser: Fix typo in function switch_data_file Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
[PATCH 15/19] perf probe: Factor out the ftrace README scanning Arnaldo Carvalho de Melo <acme@kernel.org> - 2017-03-14 20:00 +0100
Re: [GIT PULL 00/19] perf/core improvements and fixes Ingo Molnar <mingo@kernel.org> - 2017-03-15 19:40 +0100
csiph-web