From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alan Cox Subject: Re: [linux-usb-devel] Re: BUG?: VIA IDE transfer-rate problem Date: Thu, 19 Jan 2006 01:05:02 +0000 Message-ID: <1137632702.1760.29.camel@localhost.localdomain> References: <200601131909.07981.pvrusb2@toplitzer.net> <200601181536.16868.pvrusb2@toplitzer.net> <58cb370e0601180813pfb6baffr953e5be26223ea1@mail.gmail.com> <200601190041.36138.pvrusb2@toplitzer.net> Mime-Version: 1.0 Content-Type: text/plain Content-Transfer-Encoding: 7bit Return-path: Received: from outpipe-village-512-1.bc.nu ([81.2.110.250]:4799 "EHLO lxorguk.ukuu.org.uk") by vger.kernel.org with ESMTP id S1161130AbWASBFt (ORCPT ); Wed, 18 Jan 2006 20:05:49 -0500 In-Reply-To: <200601190041.36138.pvrusb2@toplitzer.net> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Helmut Toplitzer Cc: Bartlomiej Zolnierkiewicz , linux-ide@vger.kernel.org, dbrownell@users.sourceforge.net, linux-usb-devel@lists.sourceforge.net On Iau, 2006-01-19 at 00:41 +0100, Helmut Toplitzer wrote: > Ok. Was just a bit digging through the sourcecode. Found a > function called disable_hlt. Some further digging. > Found in drivers/block/floppy.c a usage of disable_hlt. > It's called (if I haven't missunderstood something) > whenever a DMA transfer is setup. enable_hlt is called > when a transfer stops/is cancled etc... Yes this deals with some errata cases > Does such thing makes sense for IDE DMA-transfers too? > (Maybe I've missed something) Only for the CS5510 Cyrix as far as I know, and I really ought to fix that in the new libata driver. Generally speaking hlt is good, it takes the CPU off the bus and frees it up for the other users.