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, 11 May 2002 09:27:19 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Sat, 11 May 2002 09:27:18 -0400 Received: from netmail.netcologne.de ([194.8.194.109]:49931 "EHLO netmail.netcologne.de") by vger.kernel.org with ESMTP id ; Sat, 11 May 2002 09:27:17 -0400 Message-Id: <200205111327.AWB48806@netmail.netcologne.de> Content-Type: text/plain; charset="iso-8859-15" From: =?iso-8859-15?q?J=F6rg=20Prante?= Reply-To: joergprante@gmx.de Organization: Linux jungle 2.4.19-pre8 #4 Don Mai 9 23:37:47 CEST 2002 i686 unknown To: linux-kernel@vger.kernel.org Subject: Re: [PATCHSET] 2.4.19pre8-jp11 Date: Sat, 11 May 2002 15:25:07 +0200 X-Mailer: KMail [version 1.3.1] In-Reply-To: <200205102107.AWB17222@netmail.netcologne.de> <20020511010955.GB23320@opeth.ath.cx> Cc: David Nielsen MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I just discovered that something in jp11 IDE code is broken because of supermount. The supermount patches were originally for 2.4.18 and conflict with other parts of the jp patch set. That has been already the case in jp10, when supermount was suspended because it caused oopses at boot time. IDE Disk seems to work, but IDE CD is unusable, for data and audio. "HDIO_GETGEO" failed, and user apps get "no valid block device found" even if there is a CD inserted. Audio access with KDE gives me "no CD", and cdfs give me an oops in sys_lstat64 when trying 'ls' on the mounted device. The reason must be in 28_new-stat, which comes with supermount patch. Maybe 28_new-stat is buggy, or it causes unwanted interaction between the IDE code, the stat changes in 2.4.19pre4-pre5, and supermount. I will try upgrading the supermount patch. It's not trivial for me and might take some time to investigate, and maybe Juan Jose Quintela can point me to the right direction (http://people.mandrakesoft.com/~quintela/supermount/2.4.18) For some "funny" effects see below. Strong recommendation: please remove the supermount patches from jp11. The patches involved are 26_autofs4 27_isrdonly 28_new-stat 29_mediactl 30_llseek 31_mount 32_device 33_supermount Sorry folks. Any help is strongly appreciated to fix that problem quickly. Jörg ------- snip ------------------- jungle:/home/joerg # hdparm -I /dev/hdb /dev/hdb: HDIO_GETGEO failed: Invalid argument Model=OTHSBI AVD-DOR MDSC-5220 , FwRev=D131 , SerialNo=170003 6335 Config={ Fixed Removeable DTR<=5Mbs DTR>10Mbs nonMagnetic } RawCHS=0/0/0, TrkSize=0, SectSize=0, ECCbytes=0 BuffType=unknown, BuffSize=128kB, MaxMultSect=0 (maybe): CurCHS=0/0/0, CurSects=0, LBA=yes, LBAsects=0 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio1 pio2 pio3 pio4 DMA modes: sdma0 sdma1 sdma2 mdma0 mdma1 mdma2 udma0 udma1 *udma2 AdvancedPM=no Drive Supports : ATA/ATAPI-5 T13 1321D revision 3 : ATA-2 ATA-3 ATA-4 ATA-5 [few minutes later and invoked 'hdparm -i /dev/hda' with ok results] jungle:/home/joerg # hdparm -i /dev/hdb /dev/hdb: HDIO_GETGEO failed: Invalid argument Model=TOSHIBA DVD-ROM SD-C2502, FwRev=1D13, SerialNo=7100303653 Config={ Fixed Removeable DTR<=5Mbs DTR>10Mbs nonMagnetic } RawCHS=0/0/0, TrkSize=0, SectSize=0, ECCbytes=0 BuffType=unknown, BuffSize=128kB, MaxMultSect=0 (maybe): CurCHS=0/0/0, CurSects=0, LBA=yes, LBAsects=0 IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO modes: pio0 pio1 pio2 pio3 pio4 DMA modes: sdma0 sdma1 sdma2 mdma0 mdma1 mdma2 udma0 udma1 *udma2 AdvancedPM=no Drive Supports : ATA/ATAPI-5 T13 1321D revision 3 : ATA-2 ATA-3 ATA-4 ATA-5 ---- snip ---- jungle:/home/joerg # mount -t cdfs -o ro /dev/hdb /mnt/cdfs May 11 12:33:55 jungle kernel: cdfs 0.5b loaded. jungle:/home/joerg # ls /mnt/cdfs May 11 12:34:02 jungle kernel: Unable to handle kernel NULL pointer dereference at virtual address 0000003c May 11 12:34:02 jungle kernel:  printing eip: May 11 12:34:02 jungle kernel: c0141713 May 11 12:34:02 jungle kernel: *pde = 00000000 May 11 12:34:02 jungle kernel: Oops: 0000 May 11 12:34:02 jungle kernel: CPU:    0 May 11 12:34:02 jungle kernel: EIP:    0010:[vfs_lstat+91/148]    Not tainted May 11 12:34:02 jungle kernel: EFLAGS: 00210246 May 11 12:34:02 jungle kernel: eax: ccd19400   ebx: 00000000   ecx: 00000000   edx: cb771bc0 May 11 12:34:02 jungle kernel: esi: cbe41f5c   edi: cbe41f88   ebp: bfffec8c   esp: cbe41f50 May 11 12:34:02 jungle kernel: ds: 0018   es: 0018   ss: 0018 May 11 12:34:02 jungle kernel: Process ls (pid: 920, stackpage=cbe41000) May 11 12:34:02 jungle kernel: Stack: cbe41f88 40164f4c bfffeca4 cb771bc0 cffe84a0 00000000 3c507f10 00000000 May 11 12:34:02 jungle kernel:        00000008 00000001 c0141e57 bfffeca4 cbe41f88 cbe40000 00117741 cbe40000 May 11 12:34:02 jungle kernel:        41ed0307 0000000a 00000000 00000000 00200000 00001000 00000000 3cdcf1ed May 11 12:34:02 jungle kernel: Call Trace: [sys_lstat64+19/48] [system_call+51/56] May 11 12:34:02 jungle kernel: May 11 12:34:02 jungle kernel: Code: 83 79 3c 00 74 0e 57 52 8b 44 24 18 50 8b 41 3c ff d0 eb 0c ---- snip---- On Duron 1Ghz K7S5A board I get the name right, but HDIO_GETGEO failed too. mountain:/home/joerg # hdparm -I /dev/hdb /dev/hdb: HDIO_GETGEO failed: Invalid argument Model=LG DVD-ROM DRD-8160B, FwRev=1.01, SerialNo= Config={ Fixed Removeable DTR<=5Mbs DTR>10Mbs nonMagnetic } RawCHS=0/0/0, TrkSize=0, SectSize=0, ECCbytes=0 BuffType=unknown, BuffSize=0kB, MaxMultSect=0 (maybe): CurCHS=0/0/0, CurSects=0, LBA=yes, LBAsects=0 IORDY=on/off, tPIO={min:227,w/IORDY:120}, tDMA={min:120,rec:150} PIO modes: pio0 pio1 pio2 pio3 pio4 DMA modes: sdma0 sdma1 sdma2 mdma0 mdma1 mdma2 udma0 udma1 *udma2 AdvancedPM=no