From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from mx1.redhat.com ([209.132.183.28]:24987 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753880Ab0L0OT2 (ORCPT ); Mon, 27 Dec 2010 09:19:28 -0500 Message-ID: <4D18A06B.6040902@redhat.com> Date: Mon, 27 Dec 2010 12:19:23 -0200 From: Mauro Carvalho Chehab MIME-Version: 1.0 To: Hans Verkuil CC: Linux Media Mailing List Subject: Re: [PATCH 0/6] V4L1 cleanups and videodev.h removal References: <20101227093848.324b6abd@gaivota> <201012271321.08128.hverkuil@xs4all.nl> <4D1887B0.8070709@redhat.com> <201012271345.16124.hverkuil@xs4all.nl> In-Reply-To: <201012271345.16124.hverkuil@xs4all.nl> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit List-ID: Sender: Mauro Carvalho Chehab Em 27-12-2010 10:45, Hans Verkuil escreveu: > On Monday, December 27, 2010 13:33:52 Mauro Carvalho Chehab wrote: >> Em 27-12-2010 10:21, Hans Verkuil escreveu: >>> On Monday, December 27, 2010 12:38:48 Mauro Carvalho Chehab wrote: >>>> Now that all hard work to remove V4L1 happened, it doesn't make >>>> sense on keeping videodev.h just because of two obsoleted drivers. >>> >>> Perhaps it is also time to mark the videodev2.h _OLD ioctls for removal in >>> 2.6.39? >>> >>> If we are getting rid of old stuff anyway, then this will also be a nice >>> cleanup. >>> >>> Perhaps we can even delete it without marking it for removal. After all, >>> removing it will only affect binaries compiled against a *really* old kernel >>> (I suspect 2.5.something). Anything that has been recompiled will automatically >>> use the correct ioctls. >> >> We can't just remove the _OLD, as they're used internally, in order to handle >> those old binaries. I think that not all come from 2.5 times. So, the better is to >> mark them to be removed for .39. > > I double checked and they were introduced in 2.6.2 except for CROPCAP_OLD which > was introduced in 2.6.6. Thanks for the research. > > Do you want me to mark them for removal or will you? I did it. Just sent to the ML. I'm basically with my hands tight, as kernel's Patchwork is not working today, so I have some time to do some basic cleanup, while waiting for someone to restart patchwork service. Cheers, Mauro