ide CDRW problems under ide-scsi

jwbrownuk · 04-30-2003, 11:04 AM

I have searched google / usenet / LinuxQuestions etc for days and while
I have seen this problem (or ones similar) nothing seems to resolve it...

I cannot get my CDRW to work properly via ide-scsi... When I started I was
looking at it with cdrecord -scanbus, but I have found
that the issue and errors happen with a simple mount so I will use
that as my example... Obviously, if all I wanted to do was read it
I would not be doing the ide-scsi, but is is the better example

First the basics... 4 ide devices
ide0
hda = hard disk
hdb = hi-val 2242 CD-RW
ide1
hdc = DVD ROM
hdd = hard disk

having added to /etc/modules.conf
<snip>
#alias scsi_hostadapter ide-scsi
options ide-cd ignore=hdb
alias scd0 sr_mod # load sr_mod upon access of scd0
pre-install sg modprobe ide-scsi # load ide-scsi before sg
pre-install sr_mod modprobe ide-scsi # load ide-scsi before sr_mod
pre-install ide-scsi modprobe ide-cd # load ide-cd before ide-scsi
<snip>

and in my grub.conf
<snip>
title Red Hat Linux (2.4.18-27.8.0)
root (hd0,1)
kernel /vmlinuz-2.4.18-27.8.0 ro root=LABEL=/ hdb=ide-scsi
initrd /initrd-2.4.18-27.8.0.img
<snip>

i get the following in my boot logs
<snip>
Apr 30 16:23:40 HomeSys kernel: Kernel command line: ro root=LABEL=/
hdb=ide-scsi
Apr 30 16:23:40 HomeSys kernel: ide_setup: hdb=ide-scsi
...
Apr 30 16:23:41 HomeSys kernel: ide0: BM-DMA at 0x2020-0x2027, BIOS
settings: hda:DMA, hdb:pio
Apr 30 16:23:41 HomeSys kernel: ide1: BM-DMA at 0x2028-0x202f, BIOS
settings: hdc:DMA, hdd:DMA
Apr 30 16:23:41 HomeSys kernel: hda: WDC WD100AA, ATA DISK drive
Apr 30 16:23:41 HomeSys kernel: hdb: HI-VAL H2224, ATAPI CD/DVD-ROM drive
Apr 30 16:23:41 HomeSys kernel: hdc: Compaq DVD-ROM DVD-113 0114, ATAPI
CD/DVD-ROM drive
Apr 30 16:23:41 HomeSys kernel: hdd: WDC WD600AB-32BVA0, ATA DISK drive
Apr 30 16:23:41 HomeSys kernel: ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
Apr 30 16:23:41 HomeSys kernel: ide1 at 0x170-0x177,0x376 on irq 15
...
Apr 30 16:23:41 HomeSys kernel: hdc: ATAPI DVD-ROM drive, 512kB Cache
Apr 30 16:23:41 HomeSys kernel: Uniform CD-ROM driver Revision: 3.12
Apr 30 16:23:41 HomeSys kernel: SCSI subsystem driver Revision: 1.00
Apr 30 16:23:41 HomeSys kernel: scsi0 : SCSI host adapter emulation for IDE
ATAPI devices
Apr 30 16:23:41 HomeSys kernel: scsi : aborting command due to timeout : pid
0,
scsi0, channel 0, id 0, lun 0 Inquiry 00 00 00 ff 00
Apr 30 16:23:41 HomeSys kernel: hdb: lost interrupt
Apr 30 16:23:41 HomeSys kernel: Vendor: HI-VAL Model: H2242
Rev: 1.06
Apr 30 16:23:41 HomeSys kernel: Type: CD-ROM
ANSI SCSI revision: 03
...
<<< At this point I let the system sit for about 23 mins and then tried to
log in>>
Apr 30 16:47:04 HomeSys gdm(pam_unix)[1167]: session opened for user root by
(uid=0)
Apr 30 16:47:06 HomeSys gconfd (root-1330): starting (version 1.2.1), pid 1330
user 'root'
Apr 30 16:47:06 HomeSys gconfd (root-1330): Resolved address
"xml:readonly:/etc/gconf/gconf.xml.mandatory" to a read-only config source at
position 0
Apr 30 16:47:06 HomeSys gconfd (root-1330): Resolved address
"xml:readwrite:/root/.gconf" to a writable config source at position 1
Apr 30 16:47:06 HomeSys gconfd (root-1330): Resolved address
"xml:readonly:/etc/gconf/gconf.xml.defaults" to a read-only config source at
position 2
Apr 30 16:47:10 HomeSys kernel: solo1: version v0.19 time 07:02:06 Mar 14 2003
Apr 30 16:47:10 HomeSys kernel: PCI: Found IRQ 5 for device 01:05.0
Apr 30 16:47:10 HomeSys kernel: PCI: Sharing IRQ 5 with 00:1f.3
Apr 30 16:47:10 HomeSys kernel: solo1: joystick port at 0x108d
Apr 30 16:47:10 HomeSys modprobe: modprobe: Can't locate module
sound-service-0-3
Apr 30 16:47:15 HomeSys kernel: Attached scsi CD-ROM sr0 at scsi0, channel 0,
id 0, lun 0
Apr 30 16:47:45 HomeSys kernel: scsi : aborting command due to timeout : pid
2,
scsi0, channel 0, id 0, lun 0 Request Sense 00 00 00 40 00
Apr 30 16:47:45 HomeSys kernel: hdb: lost interrupt
Apr 30 16:48:15 HomeSys kernel: scsi : aborting command due to timeout : pid
3,
scsi0, channel 0, id 0, lun 0 Mode Sense 00 2a 00 80 00
Apr 30 16:48:15 HomeSys kernel: hdb: lost interrupt
Apr 30 16:48:15 HomeSys kernel: sr0: scsi3-mmc drive: 24x/24x writer cd/rw
xa/form2 cdda tray
Apr 30 16:48:45 HomeSys kernel: scsi : aborting command due to timeout : pid
6,
scsi0, channel 0, id 0, lun 0 Request Sense 00 00 00 40 00
Apr 30 16:48:45 HomeSys kernel: hdb: lost interrupt
Apr 30 16:49:16 HomeSys kernel: scsi : aborting command due to timeout : pid
8,
scsi0, channel 0, id 0, lun 0 Request Sense 00 00 00 40 00
Apr 30 16:49:16 HomeSys kernel: hdb: lost interrupt
Apr 30 16:49:46 HomeSys kernel: scsi : aborting command due to timeout : pid
9,
scsi0, channel 0, id 0, lun 0 Read Capacity 00 00 00 00 00 00 00 00 00
Apr 30 16:49:46 HomeSys kernel: hdb: lost interrupt
Apr 30 16:50:16 HomeSys kernel: scsi : aborting command due to timeout : pid
12, scsi0, channel 0, id 0, lun 0 Request Sense 00 00 00 40 00
Apr 30 16:50:16 HomeSys kernel: hdb: lost interrupt
<<At this point the login completed and the icons appeared>>
<snip>

Any attempt to touch the device generates similar errors

the fstab has
<snip>
/dev/scd0 /mnt/scdrom iso9660 noauto,owner,kudzu,ro 0 0
<snip>
mount /dev/scd0
generates messages (after about 30 seconds the messages start) as
<snip>
Apr 30 17:13:01 HomeSys kernel: scsi : aborting command due to timeout : pid
26, scsi0, channel 0, id 0, lun 0 Request Sense 00 00 00 40 00
Apr 30 17:13:01 HomeSys kernel: hdb: lost interrupt
Apr 30 17:13:31 HomeSys kernel: scsi : aborting command due to timeout : pid
28, scsi0, channel 0, id 0, lun 0 Read TOC 00 00 00 00 00 00 00 0c 00
Apr 30 17:13:31 HomeSys kernel: hdb: lost interrupt
Apr 30 17:14:01 HomeSys kernel: scsi : aborting command due to timeout : pid
29, scsi0, channel 0, id 0, lun 0 Read TOC 02 00 00 00 00 01 00 0c 00
Apr 30 17:14:01 HomeSys kernel: hdb: lost interrupt
Apr 30 17:14:31 HomeSys kernel: scsi : aborting command due to timeout : pid
30, scsi0, channel 0, id 0, lun 0 Read TOC 02 00 00 00 00 02 00 0c 00
Apr 30 17:14:31 HomeSys kernel: hdb: lost interrupt
Apr 30 17:15:01 HomeSys kernel: scsi : aborting command due to timeout : pid
31, scsi0, channel 0, id 0, lun 0 Read TOC 02 00 00 00 00 03 00 0c 00
Apr 30 17:15:01 HomeSys kernel: hdb: lost interrupt
Apr 30 17:15:31 HomeSys kernel: scsi : aborting command due to timeout : pid
33, scsi0, channel 0, id 0, lun 0 Read TOC 00 00 00 00 00 00 00 0c 40
Apr 30 17:15:31 HomeSys kernel: hdb: lost interrupt
Apr 30 17:16:01 HomeSys kernel: scsi : aborting command due to timeout : pid
34, scsi0, channel 0, id 0, lun 0 Read TOC 00 00 00 00 00 00 00 0c 00
Apr 30 17:16:01 HomeSys kernel: hdb: lost interrupt
Apr 30 17:16:31 HomeSys kernel: scsi : aborting command due to timeout : pid
35, scsi0, channel 0, id 0, lun 0 Read TOC 00 00 00 00 00 01 00 0c 00
Apr 30 17:16:31 HomeSys kernel: hdb: lost interrupt
Apr 30 17:17:01 HomeSys kernel: scsi : aborting command due to timeout : pid
36, scsi0, channel 0, id 0, lun 0 Read Capacity 00 00 00 00 00 00 00 00 00
Apr 30 17:17:01 HomeSys kernel: hdb: lost interrupt
..... The command finally completed
..... then about 30 secs later
Apr 30 17:17:35 HomeSys kernel: scsi : aborting command due to timeout : pid
46, scsi0, channel 0, id 0, lun 0 Read TOC 00 00 00 00 00 00 00 0c 00
Apr 30 17:17:35 HomeSys kernel: hdb: lost interrupt
Apr 30 17:19:05 HomeSys kernel: scsi : aborting command due to timeout : pid
scsi0, channel 0, id 0, lun 0 Read TOC 02 00 00 00 00 03 00 0c 00
Apr 30 17:19:05 HomeSys kernel: hdb: lost interrupt
.... then the 'nautalis page comes up'
<snip>

It looks like the last two set of message are generated by the loading of the
nautalius page.. Which bizzarly works... as do all other operations on the
device..

The mount worked...
# df -m
Filesystem 1M-blocks Used Available Use% Mounted on
/dev/hdd2 11838 6622 4615 59% /
/dev/hda2 100 32 63 33% /boot
none 61 0 61 0% /dev/shm
/dev/hda5 407 9 378 3% /spare
/dev/hda1 8497 6622 1875 78% /c
/dev/hdd1 35073 27489 7584 79% /d
/dev/hdd3 9963 7029 2427 75% /export
/dev/scd0 341 342 0 100% /mnt/scdrom

and browsing around the disk etc... also has no messages generated...

umount /dev/scd0

No messages

mount /dev/scd0
Same errors as above in the messages file...

lsmod shows
<snip>
Module Size Used by Not tainted
sr_mod 18168 1 (autoclean)
esssolo1 28584 0 (autoclean)
gameport 3412 0 (autoclean) [esssolo1]
soundcore 6532 4 (autoclean) [esssolo1]
i810 72736 1
agpgart 43136 7 (autoclean)
binfmt_misc 7528 1
autofs 13348 0 (autoclean) (unused)
ne2k-pci 7296 1
8390 8364 0 [ne2k-pci]
ipt_REJECT 3736 2 (autoclean)
ipt_MASQUERADE 2200 1 (autoclean)
iptable_nat 19960 1 (autoclean) [ipt_MASQUERADE]
ip_conntrack 21244 1 (autoclean) [ipt_MASQUERADE iptable_nat]
iptable_filter 2412 1 (autoclean)
ip_tables 15224 6 [ipt_REJECT ipt_MASQUERADE iptable_nat
iptable_filter]
microcode 4668 0 (autoclean)
ide-scsi 10512 0
scsi_mod 107240 2 [sr_mod ide-scsi]
ide-cd 33608 0
cdrom 33696 0 [sr_mod ide-cd]
vicam 14048 0 (unused)
videodev 8288 1 [vicam]
nls_iso8859-1 3516 2 (autoclean)
nls_cp437 5148 2 (autoclean)
vfat 13084 2 (autoclean)
fat 38712 0 (autoclean) [vfat]
scanner 10420 0 (unused)
mousedev 5524 1
keybdev 2976 0 (unused)
hid 22244 0 (unused)
input 5920 0 [mousedev keybdev hid]
usb-uhci 26188 0 (unused)
usbcore 77024 1 [vicam scanner hid usb-uhci]
ext3 70336 4
jbd 52212 4 [ext3]
<snip>

Final Notes:
The delay in logging is probably due to gnome checking out the devices and
hitting the scsi bus

I have tried without the hdb=ide-scsi kernel argument
no delay on login - no scsi log messages at boot
hdb is ignored by the ide-cd drivers thanks to
options ide-cd ignore=hdb # in modules.conf
as soon as I touch /dev/scd0 the scsi modules get loaded by modrpobe
exact same results and delays from that point on
I have tried manually loading the modules for scsi
(used the script in the CD Writer howto)
same results
I have tested with cdrecord -scanbus
get the same time out messages, just
a few more
I have tried letting xcdroast attempt to write a cd
attempts to write to the device and finally gives up...
same errors in the logs

What seems the most curious to me is that the mount worked in the end... so I
have been treating this as a configuration problem... The ide controllers
seem to be the only devices playing on interrupt 14 and 15 - and as I have a
hard disk on each as well (hda / hdd) I would expect worse problems if I were
really dropping interupts...

I would not discount that the drive may have a problem - except that for
reading I get the errors - and then it works anyway (mount and then use the
drive for ROM)...

I am not sure what to do next... One suggestion I saw in a board indicated
that I should disable atapi cdrom AND turn off SCSI CDROM support and rebuild
the kernel... At the moment I am using the binary distribution of Redhat
2.4.18-27.8.0 and have not rebuilt (turning these off seems wrong and some
repies the the person posting the suggestion also say this)

So I am lost at this point and looking for help...

Applogies for the very long post, but it seemed best to get all the details
in...

Thanks in advance,

jbrown

fancypiper · 04-30-2003, 11:44 AM

I can't remember is I set up my cdrw or if the install did it, but here are a couple of configuration files in my system for your comparison:

My /boot/grub/grub.conf entry:

Code:

title Red Hat Linux 7.3
        root (hd0,2)
        kernel /boot/vmlinuz-2.4.18-3 ro root=/dev/hda3 hdd=ide-scsi
        initrd /boot/initrd-2.4.18-3.img

My /etc/fstab:

Code:

# Redhat 7.3 Celeron 850 box tinwhistle /etc/fstab file
/dev/hda3       /                       ext3    defaults        1 1
none            /dev/pts                devpts  gid=5,mode=620  0 0
/dev/hda8       /home                   ext3    defaults        1 2
/dev/hda6       /mnt/gentoo             reiserfs defaults       1 2
/dev/hda2       /mnt/gentoo/boot        ext3    defaults        1 1
/dev/hda7       /mnt/mandrake           ext3    defaults        1 2
/dev/hda1       /mnt/winc               vfat    defaults        0 0
/dev/hdc1       /mnt/wind               vfat    defaults        0 0
none            /proc                   proc    defaults        0 0
none            /dev/shm                tmpfs   defaults        0 0
/dev/hda9       /snd                    reiserfs defaults       1 2
/dev/hdc2       /pub                    ext3    defaults        1 2
/dev/hda5       swap                    swap    defaults        0 0
/dev/fd0        /mnt/floppy             auto    noauto,owner,kudzu 0 0
/dev/cdrom      /mnt/cdrom              iso9660 noauto,owner,kudzu,ro 0 0

/dev/cdrom is a symbolic link to /dev/scd0
I have nothing in /etc/modules.conf or /etc/rc.d/rc.local for loading modules, so I guess that kudzu did it for me when I installed.

HTH

# CD burning info
Adding an IDE CD-Writer to Linux
CD Writing HOWTO
# Burn an ISO to disk
cdrecord -v speed=5 dev=0,0,0 /path/to/foo.iso
# Burn from disk to disk
cdrecord -v dev=0,0,0 speed=5 -isosize /dev/cdrom
# Generate an ISO from a directory.
mkisofs -Jr -o foo.iso /path/to/directory
mkisofs -vrTJUV "Label" -o foo.iso /path/to/directory
# Generate an ISO from a CD
dd if=/dev/cdrom of=foo.iso
Linux MP3 CD Burning mini-HOWTO
# Convert mp3 to wav with lame
for i in *.mp3; do lame --decode $i `basename $i .mp3`.wav; done
# Burn a CD from wav files
cdrecord -v -audio -pad speed=5 dev=0,0,0 /path/to/*.wav
# Erase a CDRW
cdrecord -v dev=0,0,0 speed=5 blank=fast

fancypiper · 04-30-2003, 11:56 AM

Perhaps your hard drive parameters aren't set right.

What is the output of this command?

hdparm hdb

My settings for my burner:

Code:

[root@tinwhistle root]# hdparm /dev/hdd

/dev/hdd:
 HDIO_GET_MULTCOUNT failed: Input/output error
 I/O support  =  0 (default 16-bit)
 unmaskirq    =  0 (off)
 using_dma    =  0 (off)
 keepsettings =  0 (off)
 HDIO_GET_NOWERR failed: Input/output error
 readonly     =  0 (off)
 BLKRAGET failed: Input/output error
 HDIO_GETGEO failed: Invalid argument
 busstate     =  1 (on)
[root@tinwhistle root]#

jwbrownuk · 04-30-2003, 12:08 PM

for hdparm on hdb I have

[root@HomeSys root]# hdparm /dev/hdb

/dev/hdb:
HDIO_GET_MULTCOUNT failed: Input/output error
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
BLKRAGET failed: Input/output error
HDIO_GETGEO failed: Invalid argument

setting DMA to off

[root@HomeSys root]# hdparm -d0 /dev/hdb

/dev/hdb:
setting using_dma to 0 (off)
using_dma = 0 (off)

gives the same resultwith the lost interrupt

Thanks though....

jbrown