Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Fri, 6 Sep 2002 01:27:18 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Fri, 6 Sep 2002 01:27:18 -0400 Received: from rom.cscaper.com ([216.19.195.129]:60865 "HELO mail.cscaper.com") by vger.kernel.org with SMTP id ; Fri, 6 Sep 2002 01:27:17 -0400 Subject: Summary of kt333a woes, with happy ALi ending Content-Transfer-Encoding: 7BIT To: linux-kernel@vger.kernel.org From: "Joseph N. Hall" Content-Type: text/plain; charset=US-ASCII Mime-version: 1.0 Date: Thu, 5 Sep 2002 22:33 -0700 X-mailer: Mailer from Hell v1.0 Message-Id: <20020906052717Z318314-685+43734@vger.kernel.org> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2354 Lines: 62 Also FYI: the Soyo Dragon Ultra Platinum m/b never did give me proper DMA performance on my ATAPI devices. I tried: stock RH 7.3 kernel 2.4.19 2.4.20-pre5-ac1 The stock kernel might not have had proper DMA support at all; I upgraded to 2.4.19 quickly for other reasons. The most obvious problem with 2.4.19 and 2.4.20-pre5-ac1 was that my IDE DVD-R/RAM/ROM device did not do DMA, ever, and when forced to attempt DMA the machine eventually locked, with no error messages that I ever saw. It works fine with my new Iwill m/b: [joseph@dhcppc5 ticket]# hdparm -I /dev/hdd /dev/hdd: ATAPI CD-ROM, with removable media Model Number: MATSHITADVD-RAM LF-D310 Serial Number: Firmware Revision: A123 Standards: Configuration: DRQ response: 50us. Packet size: 12 bytes Capabilities: LBA, IORDY(can be disabled) DMA: sdma0 sdma1 sdma2 mdma0 mdma1 mdma2 udma0 udma1 *udma2 Cycle time: min=120ns recommended=120ns PIO: pio0 pio1 pio2 pio3 pio4 Cycle time: no flow control=180ns IORDY flow control=120ns Nor did other ATAPI devices do DMA--I tried a generic CDROM and an ASUS CDR/RW/ROM. I saw the same behavior when the ATAPI devices were plugged into the Highpoint controller. Also HD access seemed slow at times, and there would occasionally be mysteriously long pauses of a few sec during kernel compiles etc. where it seemed that some sort of I/O blocking was going on. I observed similar behavior in Win2k but ... I don't know that I ever got a proper set of drivers installed. All I know is that with the IWill m/b (ALi chipset) everything is working OK. I did not get proper DVD playback on Win2k--it seemed exactly the same slowness as with Linux. OTOH my DVDs play fine off the drive from mplayer now, and I can read and write at appropriate speeds. I would like to get the kt333a m/b working eventually, because I don't think my IWill m/b will support the new core Athlons. I suspect that the board "works" but there must be a BIOS or chipset issue that isn't resolved. -joseph - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/