From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sun, 21 Jan 2001 07:07:17 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sun, 21 Jan 2001 07:07:00 -0500 Received: from zeus.kernel.org ([209.10.41.242]:33734 "EHLO zeus.kernel.org") by vger.kernel.org with ESMTP id ; Sun, 21 Jan 2001 06:54:45 -0500 Date: Sun, 21 Jan 2001 12:54:36 +0100 From: Vojtech Pavlik To: "David D.W. Downey" Cc: P.Flinders@ftel.co.uk, powers@b2pi.com, linux-kernel@vger.kernel.org Subject: Re: VIA chipset discussion Message-ID: <20010121125436.A870@suse.cz> In-Reply-To: <20010118081259.A694@suse.cz> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: ; from pgpkeys@hislinuxbox.com on Sun, Jan 21, 2001 at 03:04:37AM -0800 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sun, Jan 21, 2001 at 03:04:37AM -0800, David D.W. Downey wrote: > > OK, currently at work so I don't have the affected system in front of > me. I'm doing this from memory.. > > My apologies to those that responded to me. I've been unable to work on > this issue for the last few days due to work. > > My particular board is the MSI 694D Pro Dual processor board. It uses the > VT82C686A chipset and the PDC20265 Promise ATA100. > > I'm running dual PIII-733s with 1GB of Corsair RAM. The board, CPUs, and > memory all run at 133MHz. The drives are > > ide0: BM-DMA at 0xb000-0xb007, BIOS settings: hda:DMA, hdb:DMA > hda: WDC WD153AA, ATA DISK drive > hda: 30064608 sectors (15393 MB) w/2048KiB Cache, CHS=1871/255/63, UDMA(66) > ide0: BM-DMA at 0xb000-0xb007, BIOS settings: hda:DMA, hdb:DMA > hdb: WDC WD300BB-00AUA1, ATA DISK drive > hdb: 58633344 sectors (30020 MB) w/2048KiB Cache, CHS=3649/255/63, UDMA(66) > Now to recap... > Kernel command line: BOOT_IMAGE=new ro root=304 idebus=66 hdc=ignore > ide_setup: hdc=ignore -- BAD OPTION Don't use "idebus=66". It's wrong. If you want to force UDMA66/UDMA100, the correct option is ide0=ata66. If you want to use idebus=, you should use idebus=33, because that's what your PCI bus most likely operates at. Also "hdc=ignore" doesn't work either. > ide1: BM-DMA at 0xb008-0xb00f, BIOS settings: hdc:DMA, hdd:pio > hdc: SAMSUNG CD-ROM SC-148F, ATAPI CDROM drive > hdc: ATAPI 48X CD-ROM drive, 128kB Cache, DMA > > I'm experiencing consistant kernel deaths when doing any type of heavy IO > as I stated earlier. I'm using the following as my test. > > dd if=/dev/hda4 of=/tmp/testdd.img bs=1024M count=2k > > It gets about 300MB into the file and pukes the kernel. What does the 'puking' exactly do? > I get lots of errors about disabling DMA within my logs. Just this? > No perosnally, I believe it's a problem with the VIA chipset > specifically. I've seen a couple of emails regarding the same type of > problem with various setups and the commonality is the VIA chipset. How is > this affecting it? I'm not quite sure. (This may already have been covered > but for the purpose of making sure that I'm not chasing numerous > possibilities I've ignored some of them.) I **THINK** it's a problem > specifically with the DMA support. The VIA chipsets are not easy to set up and the margins for error are too thin. > However, I do still get the kernel deaths even when using ATA33. I've > specifically noticed that if you install mixed drives (ATA33 + ATA66 + > ATA100) then the system will consistantly puke as well. Take out the ATA66 > drives and things stabilize for a bit then puke. I'm not sure if dropping > an ATA33 drive on there by itself is the solution even if it does > completely stabilize the machine. There is no reason for the ATA66 support > not to work since it's a combined controller that I'm sure the VIA driver > maintainer has looked at. (He emailed me but I forgot his name. Please > accept my apologies for not remembering, no disrespect is intended.) Me, most likely. > Is anyone out there with as close a match to my setup as possible wish to > help me with this? (I'm not discounting the mismatched setups, I just want > a baseline to go from. If someone with my setup is NOT ahving trouble then > I may need to look at something different). -- Vojtech Pavlik SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/