[LUNI] Update: IDE CD burner frustrations -- Black Screen of Death

[paulr]

Shouldn't a CD burner *in write mode* be using the
sg[0-7] devices? In write mode, it needs to appear
as a character device.  This is for 2.4. kernels.

The libschily transport layer should turn the IDE
burner protocol into a SCSI generic character device
(bit-serial protocol, just like mag tape.)

To read the disk back, it needs to be mounted using
(probably) the IDE hd* designation.  It's a block
device when reading.   Clear as mud?  Or am I missing
something?

As for the kernel panics, how much RAM do you have?
Can you run ksymoops against the panic?  Perhaps it's
time to go to www.kernel.org and get the latest and
greatest 2.4.19 and the pre-empt patch.  It works
quite well to eliminate the skips that sometimes
occur when burning....


Paul


> On Mon, Nov 18, 2002 at 01:55:06PM -0600, Steve Johnson wrote:
> 
>> ls -l /dev/scd*
>> show us the output, 
> 
> 
> cat /proc/scsi/scsi might be more to the point.  Besides, aren't the
> SCSI CD devices /dev/sr* ?
> 
> 
>> try using the cdrecord command line for now, as it'll be easier to
>> debug.
>> send us the cdrecord command your typing in and it's output...
> 
> 
> And use -v at least once to make it show more.  Of course that could
> complicate things if the useful bit then scolls off the screen.  :-(
> 
>> On Mon, 2002-11-18 at 10:59, Jirsa, Jeff wrote:
> 
>> > Anybody have any ideas on how to prevent the kernel from freaking out?  :-(
> 
> 
> This is just generic advice, but... have you ever left this machine
> running memtest86 overnight?  Was that since this problem started?  :-/
> 
> Luck!


-- 
*********************************************
Paul Reich              RF/Microwave Engineer

     Support the "Freedom To Innovate"...
                 Just say "No".

*********************************************