[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: cdrom: lost interrupt



Erik Steffl wrote:

 system:

    hdc: TDK CDRW321040B, ATAPI CD/DVD-ROM drive
    Linux jojda 2.4.17 #1 Wed Mar 13 01:33:28 PST 2002 i686 unknown
    debian unstable

  the cdrom is a cd r/rw writer

  problem:

  when I try to rip the audio it doesn't work much (it works VERY
slowly, hours per songs) and I get the following messages in
/var/log/syslog:

Mar 29 00:47:12 localhost kernel: hdc: lost interrupt

  what works:

  reading raw cd using cat /dev/cdrom|od (it cannot read certain part of
CD but I guess that's the feature of filesystem)

  reading data CDs, when I mount data CD I can read it to my heart
content (and more)

  writing CD: so far I only tried to burn one data CD which worked
without any problems, CD is readable.

  other info:

  the same problem occurs with another CD-ROM drive and also when I
connect cdrom to ide0 (as slave, /dev/hdb). Hardrive connected to both
ide0 and ide1 works without any problems. I also tried to turn off dma
and 32 bit access, no change.

  I don't see any IRQ conflict, ide1 is on 15, nothing else is on 15 (as
listed by BIOS and /proc/interrupts).

  Is it possible that it's a HW problem considering that data cd reading
and buring works? I mean those use IRQs too (and I see the numbers
rising when I read data CD and cat /proc/interrupts). Could it be some
kernel problem? I mean how could the interrupt be lost when trying to
read audio but not when reading data? I don't really know what the
difference is, anybody can shed some light on this?


This is a "shot in the dark", but reminds me of a problem I had with my first CD-R drive. It turned out to be a DMA problem in the end. I first noticed it in Windows 98, when I couldn't burn anything, but when I switched to Linux, it worked...somewhat.

The basic problem is that some chipsets (VIA-Apollo for example) don"t handle DMAs like the Intel chipsets do. Linux was a bit better about handling this than Windows. Things REALLY started working fine in Windows when I turned off DMA, and in Linux when I re-compiled my kernel against my SPECIFIC chipset. There was a separate config option for it in the "make config".

It is entirely possible that the "lost interrupts" is a red-herring...dunno. I also could be just a end result of something else going wrong. Anyway, I would suggest you look into your BIOS and update if possible, and check out the chipset site for some info. I found out the solution to my problem at the VIA web site. My CD-R also had a BIOS update available that I could flash. All seemed to be part of the solution. Just some additional thoughts on the hw side for you to chew on...

Cheers,
-Don Spoon-




--
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org



Reply to: