From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: + pata_viac-add-flag-for-vx800-and-add-a-function-for-fixing-internal-bugs-for-via-chipsets.patch added to -mm tree Date: Mon, 28 Jul 2008 22:14:40 +0900 Message-ID: <488DC640.3070406@kernel.org> References: <200807220814.m6M8EYoC017989@imap1.linux-foundation.org> <488DC365.1020806@kernel.org> <20080728134847.612106ec@lxorguk.ukuu.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from hera.kernel.org ([140.211.167.34]:49827 "EHLO hera.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751070AbYG1NP7 (ORCPT ); Mon, 28 Jul 2008 09:15:59 -0400 In-Reply-To: <20080728134847.612106ec@lxorguk.ukuu.org.uk> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Alan Cox Cc: akpm@linux-foundation.org, mm-commits@vger.kernel.org, JosephChan@via.com.tw, jeff@garzik.org, IDE/ATA development list Alan Cox wrote: >> And, likewise, I think >> >> if (tf->ctl != tf->last_ctl) >> tf->flags |= ATA_TFLAG_DEVICE; >> ata_sff_tf_load(ap, tf); >> >> would be better. > > Are we actually guranteed that tf->device is valid in every case where > ATA_TFLAG_DEVICE is not set ? I don't think we actually are, in which > case we should cache the last tf->device in the ap->private_data > structures Yeap, tf->device is setup during ata_tf_init(). -- tejun