From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755181Ab1DURcA (ORCPT ); Thu, 21 Apr 2011 13:32:00 -0400 Received: from void.printf.net ([89.145.121.20]:53873 "EHLO void.printf.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755384Ab1DURb7 (ORCPT ); Thu, 21 Apr 2011 13:31:59 -0400 From: Chris Ball To: Takashi Iwai Cc: Aries Lee , linux-mmc@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] mmc: Fix read-only detection with JMicron 388 chip References: Date: Thu, 21 Apr 2011 13:36:03 -0400 In-Reply-To: (Takashi Iwai's message of "Thu, 21 Apr 2011 19:09:36 +0200") Message-ID: User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.0.60 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Takashi, On Thu, Apr 21 2011, Takashi Iwai wrote: >> I don't like overwriting ops here -- it's too magical, and now we have >> to maintain the ops table in two places. A quirk seems justified here, >> even though we're trying to reduce them in general. > > Well, I also used quirk bit in my very first version I worked for > 2.6.32 kernel. But when I looked at 2.6.39, quirks are almost full -- > only the last one bit is left for bit 31. So I didn't want to finish > it :) > >> Can anyone find a better solution? > > One way would be to copy the ops table itself in struct sdhci_host > instead of keeping the ops table pointer. Then you can overwrite only > the specific op in each probe_slot callback. I think I'd rather not touch ops at all and keep the code simple -- if you don't mind, please post a version with quirks, and I'll work on freeing up a few bits. Thanks, - Chris. -- Chris Ball One Laptop Per Child