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 S1752538AbbELFXf (ORCPT <rfc822;w@1wt.eu>);
	Tue, 12 May 2015 01:23:35 -0400
Received: from mail2-relais-roc.national.inria.fr ([192.134.164.83]:10789 "EHLO
	mail2-relais-roc.national.inria.fr" rhost-flags-OK-OK-OK-OK)
	by vger.kernel.org with ESMTP id S1752354AbbELFXc (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Tue, 12 May 2015 01:23:32 -0400
X-IronPort-AV: E=Sophos;i="5.13,412,1427752800"; 
   d="scan'208";a="142515375"
Date: Tue, 12 May 2015 07:23:30 +0200 (CEST)
From: Julia Lawall <julia.lawall@lip6.fr>
X-X-Sender: jll@localhost6.localdomain6
To: Joe Perches <joe@perches.com>
cc: Michal Simek <michal.simek@xilinx.com>, linux-kernel@vger.kernel.org,
        Markus Elfring <elfring@users.sourceforge.net>
Subject: Re: [PATCH v2] net: ll_temac: Use one return statement instead of
 two
In-Reply-To: <1431360053.2884.31.camel@perches.com>
Message-ID: <alpine.DEB.2.02.1505120719210.2075@localhost6.localdomain6>
References: <3a0fa74dd703db5dac4816ef15b6a512ef0f2ec0.1431353091.git.michal.simek@xilinx.com>  <1431355371.2884.11.camel@perches.com> <5550CA19.8090404@xilinx.com>  <1431358794.2884.17.camel@perches.com> <5550CE4F.2070404@xilinx.com> 
 <alpine.DEB.2.10.1505111747150.2393@hadrien> <1431360053.2884.31.camel@perches.com>
User-Agent: Alpine 2.02 (DEB 1266 2009-07-14)
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org



On Mon, 11 May 2015, Joe Perches wrote:

> On Mon, 2015-05-11 at 17:48 +0200, Julia Lawall wrote:
> > > > A coccinelle script might be rather more complicated
> > > > than the simpler grep above, but perhaps the script
> > > > could be a bit more complete as it could likely look
> > > > at more code indentation styles.
> > >
> > > Julia: Any comment?
> > 
> > Here is what I had in mind:
> > 
> > if (...) {
> >   ... when != goto l;
> >   return C;
> > }
> > return C;
> > 
> > C is a constant, to avoid that its value depends on the code in the ...
> 
> Sure but I think that would miss several instances like:
> 
> 	switch (<foo>) {
> 		...
> 	default:
> 		return <bar>;
> 	}
> 	return <bar>;

Switch Coccinelle is not very good at...

> or the similar
> 
> 	if (foo) {
> 		if (qux)
> 			return <bar>;
> 	} else {
> 		return <baz>;
> 	}
> 
> 	return <baz>;

It seems improbable, but I could look for that.  Unfortunately, I don't 
see a way to deal with arbitrarily nested ifs.  Basically, the control 
flow from one return doesn't go to the other.  It goes from the return to 
the outside of the function.  I guess something could be done by renaming 
all of the returns to function calls, but that tends to make a mess.  It 
could be done to see if such cases are worth considering though.

Another similar and popular construction is:

if (...) {
  ...
  goto l;
}
l:

julia