From: David Ahern <dsahern@gmail.com>
To: Namhyung Kim <namhyung@kernel.org>
Cc: acme@ghostprotocols.net, linux-kernel@vger.kernel.org,
Ingo Molnar <mingo@kernel.org>,
Frederic Weisbecker <fweisbec@gmail.com>,
Peter Zijlstra <a.p.zijlstra@chello.nl>,
Mike Galbraith <efault@gmx.de>, Jiri Olsa <jolsa@redhat.com>,
Stephane Eranian <eranian@google.com>,
Pekka Enberg <penberg@kernel.org>
Subject: Re: [PATCH 6/8] perf sched: Introduce timehist command
Date: Thu, 28 Nov 2013 09:01:00 -0700 [thread overview]
Message-ID: <529768BC.2090604@gmail.com> (raw)
In-Reply-To: <87siugq2hr.fsf@sejong.aot.lge.com>
On 11/28/13, 2:50 AM, Namhyung Kim wrote:
>> +static inline void printf_nsecs(unsigned long long nsecs, int width_sec)
>> >+{
>> >+ unsigned long secs;
>> >+ unsigned long usecs;
>> >+
>> >+ secs = nsecs / NSECS_PER_SEC;
>> >+ nsecs -= secs * NSECS_PER_SEC;
>> >+ usecs = nsecs / NSECS_PER_USEC;
>> >+ printf("%*lu.%06lu ", width_sec, secs, usecs);
>> >+}
> It seems very similar to what timehist_time_str() does. Better to
> factor out?
yes and no. The above is used to dump time-based stats (time between
sched-in events, scheduling delay, and run time).
What I really want for timehist_time_str is time-of-day. And in local
variants of this code I have that. Until I can convince Peter, et al.
about the real need to pull perf_clock into userspace this is a bit of a
place holder so that local code and this version do not differ too much.
>
>
> [SNIP]
>> >+static unsigned int comm_width = 20;
>> >+
>> >+static char *timehist_get_commstr(struct thread *thread)
>> >+{
>> >+ static char str[32];
>> >+ const char *comm = thread__comm_str(thread);
>> >+ pid_t tid = thread->tid;
>> >+ pid_t pid = thread->pid_;
>> >+ unsigned int n;
>> >+
>> >+ if (pid == 0)
>> >+ scnprintf(str, sizeof(str), "%s", comm);
>> >+
>> >+ else if (tid != pid)
>> >+ scnprintf(str, sizeof(str), "%s[%d/%d]", comm, tid, pid);
>> >+
>> >+ else
>> >+ scnprintf(str, sizeof(str), "%s[%d]", comm, tid);
>> >+
>> >+ n = strlen(str);
> Why not just using return value of scnprintf()?
right. will do.
>
>
>> >+ if (n > comm_width)
>> >+ comm_width = n;
>> >+
>> >+ return str;
>> >+}
>> >+
>> >+static void timehist_header(struct perf_sched *sched)
>> >+{
>> >+ u32 max_cpus = sched->max_cpu;
>> >+ u32 i, j;
>> >+
>> >+ printf("%15s %-4s", "time", "cpu");
>> >+
>> >+ if (sched->show_cpu_visual && max_cpus) {
>> >+ printf(" ");
>> >+ for (i = 0, j = 0; i < max_cpus; ++i) {
>> >+ printf("%x", j++);
>> >+ if (j > 15)
>> >+ j = 0;
>> >+ }
>> >+ printf(" ");
>> >+ }
>> >+
>> >+ printf(" %-20s %9s %9s %9s",
>> >+ "task name[tid/pid]", "b/n time", "sch delay", "run time");
>> >+
>> >+ if (sched->show_wakeups)
>> >+ printf(" %-20s", "wakeup");
>> >+
>> >+ printf("\n");
>> >+
>> >+ printf("%15s %4s", "---------------", "----");
> You might want to use "%.15s" and graph_dotted_line.
did not know about that. will update.
>> >+static void timehist_print_sample(struct perf_sched *sched,
>> >+ union perf_event *event,
>> >+ struct perf_evsel *evsel,
>> >+ struct perf_sample *sample,
>> >+ struct thread *thread,
>> >+ struct machine *machine)
>> >+{
>> >+ struct thread_runtime *tr = thread__priv(thread);
>> >+ char tstr[64];
>> >+ u32 max_cpus = sched->max_cpu;
>> >+
>> >+ printf("%15s ", timehist_time_str(tstr, sizeof(tstr), sample->time));
>> >+
>> >+ printf("[%02d] ", sample->cpu);
>> >+
>> >+ if (sched->show_cpu_visual && max_cpus) {
>> >+ u32 i;
>> >+ char c;
>> >+ for (i = 0; i < max_cpus; ++i) {
>> >+ if (i == sample->cpu)
>> >+ c = (thread->tid == 0) ? 'i' : 's';
> It'd better explaining what the 'i' and 's' mean..
visual aid: i = idle, s = sched-event. will add a comment.
>> >+static void free_idle_threads(void)
>> >+{
>> >+ int i;
>> >+
>> >+ if (idle_threads == NULL)
>> >+ return;
>> >+
>> >+ for (i = 0; i <= idle_max_cpu; ++i)
>> >+ free(idle_threads[i]);
> Doesn't it need to be thread__delete(idle_threads[i]) ?
yes. forgot to update this to the new api.
David
next prev parent reply other threads:[~2013-11-28 16:01 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-11-18 20:32 [PATCH 0/8] perf: sched timehist command David Ahern
2013-11-18 20:32 ` [PATCH 1/8] perf tool: Skip ignored symbols while printing callchain David Ahern
2013-11-30 12:52 ` [tip:perf/core] perf evsel: " tip-bot for David Ahern
2013-11-18 20:32 ` [PATCH 2/8] perf symbols: Move idle syms check from top to generic function David Ahern
2013-11-28 8:36 ` Namhyung Kim
2013-11-30 15:58 ` David Ahern
2013-11-30 12:52 ` [tip:perf/core] " tip-bot for David Ahern
2013-11-18 20:32 ` [PATCH 3/8] perf symbol: Save vmlinux or kallsyms path loaded David Ahern
2013-11-22 18:44 ` Arnaldo Carvalho de Melo
2013-11-22 19:13 ` David Ahern
2013-11-22 19:40 ` Arnaldo Carvalho de Melo
2013-11-22 20:09 ` David Ahern
2013-11-18 20:32 ` [PATCH 4/8] perf thread: Move comm_list check into function David Ahern
2013-11-30 12:52 ` [tip:perf/core] " tip-bot for David Ahern
2013-11-18 20:32 ` [PATCH 5/8] perf tool: export setup_list David Ahern
2013-11-28 8:45 ` Namhyung Kim
2013-11-30 12:52 ` [tip:perf/core] perf tools: Export setup_list tip-bot for David Ahern
2013-11-18 20:32 ` [PATCH 6/8] perf sched: Introduce timehist command David Ahern
2013-11-28 9:50 ` Namhyung Kim
2013-11-28 16:01 ` David Ahern [this message]
2013-11-28 15:38 ` Namhyung Kim
2013-11-28 16:01 ` David Ahern
2013-11-29 0:48 ` Namhyung Kim
2013-11-29 1:58 ` David Ahern
2013-12-04 4:15 ` David Ahern
2013-11-18 20:32 ` [PATCH 7/8] perf sched timehist: Add support for context-switch event David Ahern
2013-11-18 20:32 ` [PATCH 8/8] perf sched : Add documentation for timehist options David Ahern
2013-11-28 15:42 ` Namhyung Kim
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=529768BC.2090604@gmail.com \
--to=dsahern@gmail.com \
--cc=a.p.zijlstra@chello.nl \
--cc=acme@ghostprotocols.net \
--cc=efault@gmx.de \
--cc=eranian@google.com \
--cc=fweisbec@gmail.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=namhyung@kernel.org \
--cc=penberg@kernel.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).