SparQ drive problems.

["A. J. Clark" <ajclark usa net>]

Howdy,

I'm running RedHat 4.2 on a Pentium-based PC.  I'm running Kernel
V2.0.34 with the paride-2.0.35 patch (which, according to the website,
patches nicely with 2.0.34, and I didn't have any patching problems
here).

At any rate, I have the kernel compiled with the paride-epat patch
compiled as a module, and the drive seems to work - to a point.

For some reason in the middle of operation it will stop what it's doing,
the Sparq's drive activity light will go full green, and I'll start
receiving errors.

For example, when running e2fsck -c, I get the following after a period
of time:

pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
end_request: I/O error, dev 2d:01, sector 881470
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
pda: do_pd_read: status = 0x4051 = ERR SEEK READY UNC
end_request: I/O error, dev 2d:01, sector 881472
                                                    
Similar things will happen in the middle of a regular e2fsck as well.

Once this problem happens, the drive becomes inaccessable.  I have to
rmmod pd and then insmod pd again to be able to access it.

Finally, I do have the parallel port in EPP mode, and have tried
Standard mode, which behaves the same, just slower.

Anybody have any ideas on this?  I'm at a loss.  Thanks in advance!

-Adam