From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1752318AbbLVG5Q (ORCPT <rfc822;w@1wt.eu>);
	Tue, 22 Dec 2015 01:57:16 -0500
Received: from mx1.redhat.com ([209.132.183.28]:50050 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1751237AbbLVG5P (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Tue, 22 Dec 2015 01:57:15 -0500
Date: Tue, 22 Dec 2015 07:57:12 +0100
From: Jiri Olsa <jolsa@redhat.com>
To: Namhyung Kim <namhyung@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>, Ingo Molnar <mingo@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        LKML <linux-kernel@vger.kernel.org>, David Ahern <dsahern@gmail.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Frederic Weisbecker <fweisbec@gmail.com>,
        Andi Kleen <andi@firstfloor.org>, Wang Nan <wangnan0@huawei.com>
Subject: Re: [PATCH 08/10] perf tools: Add --raw-trace option
Message-ID: <20151222065712.GB13404@krava.brq.redhat.com>
References: <1450193743-4409-1-git-send-email-namhyung@kernel.org>
 <1450193743-4409-9-git-send-email-namhyung@kernel.org>
 <20151220145842.GA18893@krava.brq.redhat.com>
 <20151221084441.GB22464@danjae.kornet>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20151221084441.GB22464@danjae.kornet>
User-Agent: Mutt/1.5.24 (2015-08-30)
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Mon, Dec 21, 2015 at 05:44:41PM +0900, Namhyung Kim wrote:

SNIP

> > 
> > I think the default is good with the 'trace' sort key as it is now:
> > 
> > 	$ ./perf report -s trace 
> > 	  27.01%  offlineimap:17701 [120] S ==> swapper/1:0 [120]
> > 	  27.00%  swapper/1:0 [120] R ==> offlineimap:17701 [120]
> > 
> > 
> > But for '--raw-trace' option rather than displaying fields in the '=' notation:
> > 
> > 	$ ./perf report -s trace --raw-trace
> > 	  27.01%   prev_comm=offlineimap prev_pid=17701 prev_prio=120 prev_state=1 next_comm=swapper/1 next_pid=0 next_prio=120
> > 	  27.00%   prev_comm=swapper/1 prev_pid=0 prev_prio=120 prev_state=0 next_comm=offlineimap next_pid=17701 next_prio=120
> > 
> > 
> > it would be more readable to register all available columns as is possible to do now with:
> > 
> > 	$ ./perf report -s prev_comm,prev_pid,prev_prio,prev_state,next_comm,next_pid,next_prio
> > 	Overhead         prev_comm    prev_pid   prev_prio          prev_state         next_comm    next_pid   next_prio
> > 	  27.01%       offlineimap       17701         120                   1         swapper/1           0         120
> > 	  27.00%         swapper/1           0         120                   0       offlineimap       17701         120
> > 
> > 
> > it would be default output for 'perf report -s trace --raw-trace' command
> > 
> > how about that?
> 
> 
> I agree with the latter is much more readable.  But the dynamic sort
> keys are event specific while 'trace' is generic.  So if there are
> more than one events, dynamic sort keys should be generated for each
> event and they'll show "N/A" for other events.

I see.. haven't realized that, maybe we should add
columns based in the hists object.. but I guess
that's out of scope of this patchset ;-)

> 
> Also I don't like the --raw-trace option affects number of sort keys.
> Maybe it'd be better adding 'EVENT.*' dynamic sort key for your case?

that's good workaround, thanks

jirka