All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chris Ferron <chris.e.ferron at linux.intel.com>
To: powertop@lists.01.org
Subject: Re: [Powertop] [PATCH v1] Fix timer and work perf events timestamp tracing
Date: Thu, 06 Sep 2012 08:30:39 -0700	[thread overview]
Message-ID: <5048C19F.8010507@linux.intel.com> (raw)
In-Reply-To: 1346938764-31141-1-git-send-email-rajagopal.venkat@linaro.org

[-- Attachment #1: Type: text/plain, Size: 3308 bytes --]

On 09/06/2012 06:39 AM, Rajagopal Venkat wrote:
> Fix timer and work perf events timestamp tracing by
> - considering event exit timestamp relative to measurement first
> timestamp for events which entry timestamp is not recorded.
> Currently these events exit timestamp itself is considered as
> usage period resulting in false usage report.
> - clearing event timestamps from global map at the end of each
> measurement to avoid collision with earlier recorded timestamps.
>
> Signed-off-by: Rajagopal Venkat <rajagopal.venkat(a)linaro.org>
> ---
>   src/process/do_process.cpp | 12 ++++++++++++
>   src/process/timer.cpp      |  4 ++++
>   src/process/work.cpp       |  4 ++++
>   3 files changed, 20 insertions(+)
>
> diff --git a/src/process/do_process.cpp b/src/process/do_process.cpp
> index cb7c672..eaef7f1 100644
> --- a/src/process/do_process.cpp
> +++ b/src/process/do_process.cpp
> @@ -490,6 +490,10 @@ void perf_process_bundle::handle_trace_point(void *trace, int cpu, uint64_t time
>   		}
>   		pop_consumer(cpu);
>   		t = timer->done(time, tmr);
> +		if (t == ~0ULL) {
> +			timer->fire(first_stamp, tmr);
> +			t = timer->done(time, tmr);
> +		}
>   		consumer_child_time(cpu, t);
>   	}
>   	else if (strcmp(event->name, "hrtimer_expire_entry") == 0) {
> @@ -532,6 +536,10 @@ void perf_process_bundle::handle_trace_point(void *trace, int cpu, uint64_t time
>   
>   		pop_consumer(cpu);
>   		t = timer->done(time, tmr);
> +		if (t == ~0ULL) {
> +			timer->fire(first_stamp, tmr);
> +			t = timer->done(time, tmr);
> +		}
>   		consumer_child_time(cpu, t);
>   	}
>   	else if (strcmp(event->name, "workqueue_execute_start") == 0) {
> @@ -575,6 +583,10 @@ void perf_process_bundle::handle_trace_point(void *trace, int cpu, uint64_t time
>   		}
>   		pop_consumer(cpu);
>   		t = work->done(time, wk);
> +		if (t == ~0ULL) {
> +			work->fire(first_stamp, wk);
> +			t = work->done(time, wk);
> +		}
>   		consumer_child_time(cpu, t);
>   	}
>   	else if (strcmp(event->name, "cpu_idle") == 0) {
> diff --git a/src/process/timer.cpp b/src/process/timer.cpp
> index 8917490..1ca8c25 100644
> --- a/src/process/timer.cpp
> +++ b/src/process/timer.cpp
> @@ -79,6 +79,9 @@ uint64_t timer::done(uint64_t time, uint64_t timer_struct)
>   {
>   	int64_t delta;
>   
> +	if (running_since.find(timer_struct) == running_since.end())
> +		return ~0ULL;
> +
>   	if (running_since[timer_struct] > time)
>   		return 0;
>   
> @@ -147,6 +150,7 @@ void clear_timers(void)
>   		all_timers.erase(it);
>   		it = all_timers.begin();
>   	}
> +	running_since.clear();
>   }
>   
>   bool timer::is_deferred(void)
> diff --git a/src/process/work.cpp b/src/process/work.cpp
> index 82f13a2..e62e5d3 100644
> --- a/src/process/work.cpp
> +++ b/src/process/work.cpp
> @@ -56,6 +56,9 @@ uint64_t work::done(uint64_t time, uint64_t work_struct)
>   {
>   	int64_t delta;
>   
> +	if (running_since.find(work_struct) == running_since.end())
> +		return ~0ULL;
> +
>   	if (running_since[work_struct] > time)
>   		return 0;
>   
> @@ -102,6 +105,7 @@ void clear_work(void)
>   		all_work.erase(it);
>   		it = all_work.begin();
>   	}
> +	running_since.clear();
>   }
>   
>   
patched merged
Thank You
-Chris


             reply	other threads:[~2012-09-06 15:30 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-09-06 15:30 Chris Ferron [this message]
  -- strict thread matches above, loose matches on Subject: below --
2012-09-06 13:39 [Powertop] [PATCH v1] Fix timer and work perf events timestamp tracing Rajagopal Venkat

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=5048C19F.8010507@linux.intel.com \
    --to=powertop@lists.01.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.