From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: [PATCH 2/3] libata: Remove excess command issue delays Date: Mon, 17 May 2010 18:21:25 -0400 Message-ID: <4BF1C165.9080907@garzik.org> References: <20100505092350.586.46344.stgit@localhost.localdomain> <20100505092503.586.89942.stgit@localhost.localdomain> <4BEE05E5.70904@garzik.org> <20100515150937.746f73ec@lxorguk.ukuu.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from mail-gy0-f174.google.com ([209.85.160.174]:55751 "EHLO mail-gy0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755265Ab0EQWV1 (ORCPT ); Mon, 17 May 2010 18:21:27 -0400 Received: by gyg13 with SMTP id 13so2469205gyg.19 for ; Mon, 17 May 2010 15:21:26 -0700 (PDT) In-Reply-To: <20100515150937.746f73ec@lxorguk.ukuu.org.uk> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Alan Cox Cc: Alan Cox , linux-ide@vger.kernel.org On 05/15/2010 10:09 AM, Alan Cox wrote: > On Fri, 14 May 2010 22:24:37 -0400 > Jeff Garzik wrote: > >> On 05/05/2010 05:25 AM, Alan Cox wrote: >>> /** >>> + * ata_sff_exec_command_nopost - issue ATA command to host controller >>> + * @ap: port to which command is being issued >>> + * @tf: ATA taskfile register set >>> + * >>> + * Issues ATA command, with proper synchronization with interrupt >>> + * handler / other threads. This version of the helper does not protect >>> + * against any delayed writes done by the underlying fabric, it must >>> + * therefore not be used for MMIO devices unless the device handles >>> + * the 400nS command delay stall internally. >>> + * >>> + * LOCKING: >>> + * spin_lock_irqsave(host lock) >>> + */ >>> +void ata_sff_exec_command_nopost(struct ata_port *ap, >>> + const struct ata_taskfile *tf) >>> +{ >>> + DPRINTK("ata%u: cmd 0x%X\n", ap->print_id, tf->command); >>> + iowrite8(tf->command, ap->ioaddr.command_addr); >>> +} >>> +EXPORT_SYMBOL_GPL(ata_sff_exec_command_nopost); >> >> where did the 400ns delay go, for the 'nopost' version? > > Good question. It should indeed have an ndelay(400) after the iowrite8 > I'll go fix that. No complaints if I go ahead and add that myself? I'd like to go ahead and get it in. Jeff