Gerhard,

Undubitably, you have SCSI problems under Linux-2.2.1 using ncr driver
3.x, but the informations your report about are undubitably not enough 
for me to understand what actually happens.

I need to know what kind of error occurs. If the SCSI layer does not 
want to tell you about, ask the driver for doing it for you:

Either, boot options: ncr53c8xx=debug:0x10
Or, run-time command: echo "setdebug result" >/proc/scsi/ncr53c8xx/0
(assuming controller is #0)
Will tell the driver to print out some messages when an error condition 
is reported to upper SCSI layers.

On the other hand, you also may want to do the same tests using the
sym53c8xx driver (a.k.a. 896) and see if it makes differences. 

ftp://ftp.tux.org/pub/roudier/README
ftp://ftp.tux.org/pub/roudier/896/

Let me know the results, and you may also want to send me the INQUIRY data
reported by the tape. 

Thanks.

Regards,
   G�rard.


On Sat, 13 Feb 1999, Gerhard Traeger wrote:

> ncr53c875 3.1f + kernel 2.2.1 + dds tape (HP35480A) are driving me crazy.
> 
> While everything seemed ok 2 weks ago, after recompiling the kernel to
> read a CD-Rom with MAC-file-system last weekend, i could again not write 
> to my dds tapes.
> 
> I did 4 cold starts, each with a different tape. The tapes were readable,
> but 
> > gt-priv [38] tar cf /dev/rmt0 .bashrc
> always gave:
> > tar: Cannot write to /dev/rmt0: Input/output error
> > tar: Error is not recoverable: exiting now
> 
> I created a new kernel source (tar xzf sunsite.doc.ic.ac.uk/Mirrors/\
> ftp.kernel.org/pub/linux/kernel/v2.2/linux-2.2.1.tar.gz) and recompiled 
> the kernel with the previous (ie. before i added module support for the 
> mac fs) ".config". 
> 
> Again 4 cold starts. Tapes are read only.
> 
> Then i reinstalled kernel v2.0.33 from suse 5.2 package. Write worked.
> 
> Tried again 2.2.1+3.1f(warm boot) -> 3 tapes are read only.
> 
> Cold start+2.2.1+3.1f -> 3 tapes are read only.
> 
> Copy ncr53c875 2.6l (from ftp.tux.org/pub/tux/roudier/Obsolete/\
> Testing-2.6/ncr53c8xx-2.6l.tar.gz) and rebuild kernel.
> 
> Warm boot to vmlinuz.old (2.2.1+3.1f) -> 3 tapes are read only.
> 
> Warm boot to vmlinuz (2.2.1+2.6l) -> 4 tapes are read/write.
> 
> 5 cold starts to vmlinuz (2.2.1+2.6l) -> 3 tapes are read/write.
> 
> Warm boot to vmlinuz.old (2.2.1+3.1f) -> last used tape can still be written
> 
> There are 2 other strange things:
> - the tape that i used to test dds access with standard kernels 2.1.132 
> and 2.2.1 produces a loud and sharp "plonck" while loading with these 
> kernels, even if i had not accessed the drive after boot. With other 
> kernels and other tapes, there is just a decent "click".
> 
> - the tape that i inserted after the first warm boot to 2.2.1+2.6l spooled
> far to long after i had written my .bashrc onto it.
> 
> And really, i never had problems to access the drive with other
> drivers than ncr53c875 3.x.
> 
> I will again compile kernel 2.1.107+3.1f. But to get significant
> results, i have to wait until 2.2.1+3.1f fails again.
> 
> Totaly confused,
> Gerhard.
> 
> BTW: This time i never entered new commands before the drive lights 
> stopped blinking and i carefully read the error messages in order
> not to call it an error of the driver when it can't read from a blank
> tape ;-/
> And of course, i did not touch the "write protection" slider, which would
> give jet an other error message anyway;-)
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to [EMAIL PROTECTED]
> 
> 


-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to [EMAIL PROTECTED]

Reply via email to