From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: [PATCH alt4 v2] libata resume fixes Date: Sat, 27 May 2006 17:06:58 -0400 Message-ID: <4478BF72.5010708@garzik.org> References: <20060527195847.GA28334@havoc.gtf.org> <20060527201024.GA29015@havoc.gtf.org> <4478B312.5040901@garzik.org> <4478B611.2030201@rtr.ca> <4478B743.2050006@garzik.org> <4478B971.3060409@rtr.ca> <4478BD09.5010309@garzik.org> <4478BDFD.60209@rtr.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Return-path: Received: from srv5.dvmed.net ([207.36.208.214]:22417 "EHLO mail.dvmed.net") by vger.kernel.org with ESMTP id S964968AbWE0VHE (ORCPT ); Sat, 27 May 2006 17:07:04 -0400 In-Reply-To: <4478BDFD.60209@rtr.ca> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Mark Lord Cc: torvalds@osdl.org, linux-ide@vger.kernel.org Mark Lord wrote: > Jeff Garzik wrote: >> >> As mentioned elsewhere, Linus's patch falls over when we start >> resuming controllers with real SATA phy registers, so that would be a >> separate sata_pci_device_resume() function. > > But does it fall over any worse than we already do with the stock kernel? Are you talking about my x86-64 box + Linus's patch, or the future implications? For the former, the previous behavior was EH spew like what you are seeing. After apply Linus's patch, it hardlocks. But I don't want that to hold up the patch... libata suspend/resume is one part luck, and one part "it's only ata_piix so far." It has a looooong way to go before it is usable outside of that domain. For the latter, it doesn't make sense to poll BSY on modern SATA controllers, particularly (a) FIS-based ones and more importantly (b) devices attached behind a Port Multiplier. Jeff