From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Sat, 1 Dec 2001 00:29:06 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sat, 1 Dec 2001 00:28:56 -0500 Received: from mail6.speakeasy.net ([216.254.0.206]:51647 "EHLO mail6.speakeasy.net") by vger.kernel.org with ESMTP id ; Sat, 1 Dec 2001 00:28:41 -0500 Subject: 2.4.17-pre2 having dma problems with atapi devices From: safemode To: linux-kernel@vger.kernel.org Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Evolution/0.99.2 (Preview Release) Date: 01 Dec 2001 00:28:39 -0500 Message-Id: <1007184520.408.0.camel@psuedomode> Mime-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I'm trying to write to a cd at 8x, which i could do in prior kernels (2.4.15-prex and before), but now i'm getting cdrecord errors and sometimes dma timeouts. Seems that data is being corrupted somewhere again. I got this kind of performance when i had it on the promise controller and when i moved it back to my via controller it worked perfectly. Ide controller: VIA vt82c686a (rev 22) IDE UDMA66 controller CDR : CREATIVE CD-RW RW8438E, ATAPI CD/DVD-ROM drive (set as master and DMA set by kernel) setting to a lower write speed allows successful writes. cdrecord error: /usr/bin/cdrecord: Input/output error. send opc: scsi sendcmd: no error CDB: 54 01 00 00 00 00 00 00 00 00 status: 0x2 (CHECK CONDITION) Sense Bytes: 70 00 03 00 00 00 00 0A 00 00 00 00 73 03 00 00 Sense Key: 0x3 Medium Error, Segment 0 Sense Code: 0x73 Qual 0x03 (power calibration area error) Fru 0x0 Sense flags: Blk 0 (not valid) cmd finished after 17.724s timeout 60s /usr/bin/cdrecord: Resource temporarily unavailable. OPC failed. /usr/bin/cdrecord: fifo had 64 puts and 0 gets. /usr/bin/cdrecord: fifo was 0 times empty and 0 times full, min fill was 100%. The dma errors weren't very special, same old dma timeout and reset messages. The cds are perfectly fine, i use the same one that errored at 4x and it reads fine after burn.