From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+w=401wt.eu-S1753191AbZHIXKL@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1753191AbZHIXKL (ORCPT <rfc822;w@1wt.eu>);
	Sun, 9 Aug 2009 19:10:11 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751857AbZHIXKK
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Sun, 9 Aug 2009 19:10:10 -0400
Received: from mga06.intel.com ([134.134.136.21]:39095 "EHLO
	orsmga101.jf.intel.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org
	with ESMTP id S1751545AbZHIXKJ (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Sun, 9 Aug 2009 19:10:09 -0400
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="4.43,349,1246863600"; 
   d="scan'208";a="539876821"
Message-ID: <4A7F5754.3040703@linux.intel.com>
Date: Sun, 09 Aug 2009 16:10:12 -0700
From: Arjan van de Ven <arjan@linux.intel.com>
User-Agent: Thunderbird 2.0.0.22 (Windows/20090605)
MIME-Version: 1.0
To: Frederic Weisbecker <fweisbec@gmail.com>
CC: Ingo Molnar <mingo@elte.hu>,
       Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: perf: store and retrieve trace event names in the perf.data file
References: <4A7F4463.3050508@linux.intel.com> <20090809224400.GA6089@nowhere> <4A7F554A.8080402@linux.intel.com> <20090809230430.GB6089@nowhere>
In-Reply-To: <20090809230430.GB6089@nowhere>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

Frederic Weisbecker wrote:
> On Sun, Aug 09, 2009 at 04:01:30PM -0700, Arjan van de Ven wrote:
>> Frederic Weisbecker wrote:
>>> On Sun, Aug 09, 2009 at 02:49:23PM -0700, Arjan van de Ven wrote:
>>>> In order to be able to use trace events, a key aspect is the "type" field;
>>>> this is the id of the event that you recorded, as stored in the /sys/kernel/debug/events/*/*/id file.
>>>> What is currently missing is a way to map the id number (which is not abi stable)
>>>> to the event that was recorded.
>>>
>>>
>>> We already have that.
>>> See tools/perf/util/parse-events.c: char *tracepoint_id_to_name(u64 id)
>>
>> sadly... we don't.
>>
>> You do it *on the same boot and machine* as you created the perf.data file.
>> You don't have it when you reboot into a new kernel, or move the perf.data file
>> to a new machine........
>>
>> Which is to be honest a very common usecase.
> 
> 
> Ah right... I always forget the offline report case.
> But you can at least reuse this existing helper to resolve
> the id to name.
> 

ok I'll unstatic it, move it to a different file and ..

eh how about not?

this helper looks up an on disk id/name pair, and is private to it's existing file.
Fine.

The other one is based on the perf data, and all that is in header.c instead....