From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH] libata: always use polling SETXFER Date: Fri, 25 May 2007 14:52:06 +0200 Message-ID: <4656DBF6.6060204@gmail.com> References: <20070314053338.GA15600@htj.dyndns.org> <4656C144.2080705@garzik.org> <4656CF83.5020402@gmail.com> <4656D602.3010105@garzik.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from wr-out-0506.google.com ([64.233.184.238]:32049 "EHLO wr-out-0506.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1762471AbXEYMwT (ORCPT ); Fri, 25 May 2007 08:52:19 -0400 Received: by wr-out-0506.google.com with SMTP id 76so305207wra for ; Fri, 25 May 2007 05:52:19 -0700 (PDT) In-Reply-To: <4656D602.3010105@garzik.org> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Jeff Garzik Cc: Alan Cox , linux-ide@vger.kernel.org, Linux Kernel Mailing List Jeff Garzik wrote: > Tejun Heo wrote: >> So, I don't think the problem exists for SATA in the first place. At >> least there hasn't been any report of it and doing SETXFER by polling >> can handle all the existing cases. We can and probably should deal with >> such SATA devices when and if they come up. How are we gonna verify the >> controller doesn't crap itself and ahci TF register monitoring HSM can >> work around the weirdo when we don't have any such device? Even if we >> determine that we need to do HSM over intelligent SATA controller now, I >> think we still need to push polling SETXFER first to take care of the >> existing cases. > > Doing SETXFER by polling only handles the cases where the driver > actually honors ATA_TFLAG_POLLING, which is /not/ always the case. > > If the new policy ensures that it continues to be OK to /not/ honor > ATA_TFLAG_POLLING -- thus limiting SETXFER polling assumptions to older > hardware -- that's fine, and it merely needs to be documented. Basically this flag applies to drivers which is SFF compliant, at least at TF interface level. There also are other flags/callbacks which only apply to SFF or BMDMA. It would be nice to separate them out in the long term and yeah it needs documentation. > But let us not make the assumption that this bandaid fixes all cases, > because the bandaid is not applied in all cases. It covers all the known cases but I agree that SFF specific things certainly need documentation. Thanks. -- tejun