[Bug-hurd] More on block.c problem

[Brent Fulgham]

It occurs to me that I have a fairly pathalogical case for the
device code to contend with.

Because my HURD box is cobbled together out of basically
trash, I have a 1 Gig hard drive that was at one time a
"slave" or secondary drive on another machine.  The pin
setting to return this "slave" drive to "master" is long
since lost.  So this drive always identifies itself as
a secondary drive.

GRUB can handle this fine, and gnumach used to as well, but
I wonder if the IDE drive handling routines might see
that IDE 0 is missing and somehow misnumber the drives
in the gendisk chain?

Just a thought....

-Brent