[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20090320092714.GA6256@elte.hu>
Date: Fri, 20 Mar 2009 10:27:14 +0100
From: Ingo Molnar <mingo@...e.hu>
To: Peter Zijlstra <a.p.zijlstra@...llo.nl>
Cc: paulus@...ba.org, rostedt@...dmis.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 00/10] tracepoint perf counter events and new ABI
* Peter Zijlstra <a.p.zijlstra@...llo.nl> wrote:
> These patches change the event tracer to provide id files:
>
> [root@...eron ~]# for i in /debug/tracing/events/irq/*/id; do echo `cat $i` - $i; done
> 28 - /debug/tracing/events/irq/irq_handler_entry/id
> 29 - /debug/tracing/events/irq/irq_handler_exit/id
> 27 - /debug/tracing/events/irq/softirq_entry/id
> 26 - /debug/tracing/events/irq/softirq_exit/id
>
> These can then be used in profiling as:
>
> [root@...eron linux-2.6]# ~/perfstat -e 2:28 make O=native-build/ -j4 lib
> ...
>
> Performance counter stats for 'make':
>
> 18280.657602 task clock ticks (msecs)
>
> 465 tp: 28 (events)
>
> Wall-clock time elapsed: 11265.199196 msecs
>
> Showing that we had 465 irq_handler_entry events during that workload.
nice! I suspect perfstat could use "-e irq_handler_entry" type of
symbolic input as well.
I've started picking up the patches and started testing them.
Thanks Peter,
Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists