From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from moutng.kundenserver.de ([212.227.17.10]:53041 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753430Ab1FSL27 (ORCPT ); Sun, 19 Jun 2011 07:28:59 -0400 From: Arnd Bergmann To: Alan Cox Subject: Re: [PATCH 9/10 v2] Generic Watchdog Timer Driver Date: Sun, 19 Jun 2011 13:28:40 +0200 Cc: Wim Van Sebroeck , LKML , Linux Watchdog Mailing List References: <20110618172727.GJ3441@infomag.iguana.be> <201106182117.49026.arnd@arndb.de> <20110619110141.1ac221f1@lxorguk.ukuu.org.uk> In-Reply-To: <20110619110141.1ac221f1@lxorguk.ukuu.org.uk> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201106191328.40602.arnd@arndb.de> Sender: linux-watchdog-owner@vger.kernel.org List-Id: linux-watchdog@vger.kernel.org On Sunday 19 June 2011 12:01:41 Alan Cox wrote: > > Hmm, I'm not sure about this one. It does make the conversion > > of existing drivers easier but doesn't encourage doing a good > > job there. > > It's not ultimately about conversion. A lot of watchdogs have interfaces > that are not quite generic, extra requirements and custom interfaces. The > ioctl interface is needed for that, not for conversions. It's probably fine to have additional custom interfaces in some watchdogs where needed, but do you think we also need to give drivers a way to override the common ones? If a driver has a different interpretation of a common command, I would consider that a bug, not a driver specific extension. I would also like to see the proper extensions easily visible, so we can create new common interfaces in cases where multiple drivers have the same requirement. Arnd