From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1161341AbWFVUjN (ORCPT ); Thu, 22 Jun 2006 16:39:13 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1161339AbWFVUjN (ORCPT ); Thu, 22 Jun 2006 16:39:13 -0400 Received: from mx1.suse.de ([195.135.220.2]:28572 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S1161341AbWFVUjM (ORCPT ); Thu, 22 Jun 2006 16:39:12 -0400 Date: Thu, 22 Jun 2006 13:35:59 -0700 From: Greg KH To: Vitaly Wool Cc: i2c@lm-sensors.org, linux-arm-kernel@lists.arm.linux.org.uk, linux-kernel@vger.kernel.org Subject: Re: [i2c] [PATCH] I2C bus driver for Philips ARM boards Message-ID: <20060622203559.GA14445@kroah.com> References: <20060622153146.2ae56e33.vitalywool@gmail.com> <20060622183955.GA6372@kroah.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.11 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Jun 23, 2006 at 12:17:55AM +0400, Vitaly Wool wrote: > On 6/22/06, Greg KH wrote: > > > >> +static inline int i2c_pnx_bus_busy(volatile struct i2c_regs *master) > >> +{ > >> + long timeout; > >> + > >> + timeout = TIMEOUT * 10000; > >> + if (timeout > 0 && (master->sts & mstatus_active)) { > > > >A big hint about this, if you have volatile in your driver, it's > >wrong... > > > Well, I do remember flames on the subject... Still __raw_readl also uses > volatile. > Anyway, I can't disagree that ioreadX-based register access are better. > Greg, is there anything else I should take care of before re-sending the > patch, except for the things that you and Ben pointed out? Not that I can think of, that would be a good starting point :) thanks, greg k-h