From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bartlomiej Zolnierkiewicz Subject: Re: [PATCH] ide: move ide_config_drive_speed() calls to upper layers Date: Sat, 4 Aug 2007 01:01:34 +0200 Message-ID: <200708040101.34574.bzolnier@gmail.com> References: <200707270222.28230.bzolnier@gmail.com> <200707272119.19438.bzolnier@gmail.com> <46AB91A2.6060303@ru.mvista.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from mu-out-0910.google.com ([209.85.134.185]:6096 "EHLO mu-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1762595AbXHCXER (ORCPT ); Fri, 3 Aug 2007 19:04:17 -0400 Received: by mu-out-0910.google.com with SMTP id i10so1186455mue for ; Fri, 03 Aug 2007 16:04:15 -0700 (PDT) In-Reply-To: <46AB91A2.6060303@ru.mvista.com> Content-Disposition: inline Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Sergei Shtylyov Cc: Alan Cox , linux-ide@vger.kernel.org Hi, On Saturday 28 July 2007, Sergei Shtylyov wrote: > Hello. > > Bartlomiej Zolnierkiewicz wrote: > > >>On Fri, 27 Jul 2007 02:22:27 +0200 > >>Bartlomiej Zolnierkiewicz wrote: > > >>>* Convert {ide_hwif_t,ide_pci_device_t}->host_flag to be u16. > > >>>* Add IDE_HFLAG_POST_SET_MODE host to indicate the need to program the > >>> host for the transfer mode after programming the device. Set it in > >>> au1xxx-ide/cs5530/cs5535/pdc202xx_new/sc1200/via82cxxx host drivers. > > >>The CS5530 at least shouldn't care what order changes are done. I don't > > And neither CS5535. And Au1200 static bus controller shouldn't care about > the order too, so au1xxx-ide hardly needs that. > > Here's the datasheet, BTW: > > http://www.razamicroelectronics.com/documents/32798e_Au1200_db.pdf > > >>think the SC1200 does either but I don't have the docs to hand. > > It seems pretty much alike CS553x except it's accessed via PCI config. > space, not MSRs... Here's the datasheet: > > http://www.amd.com/files/connectivitysolutions/geode/32579B_sc1200_ds.pdf Thanks for links to specs. > I have a feeling that only pdc202xx_new and jmicron drivers actually need > this flag (and the latter one actually doesn't care as its methods are empty > anyway). Well, AMD/VIA chips enable UltraDMA mode by snooping Set Features > command (as the drivers tell them to do so), so the order seems important > unless that is changed. Since changing of order affects the way in which hardware is accessed I prefer to keep such changes out of this patch (which is just a cleanup). Could you please handle them in pre or post patch(es)? Bart