I have some problems with an 1.2 Gig Quantum Fireball disk that I installed yesterday.


When I first installed it (FAT-formatted, to disklabel and newfs it) NetBSD
stalled in findroot with:

sd1(aha0:1:0): timed out
sd1(aha0:1:0): timed out AGAIN

with the harddisk LED on continously. I got it up and running eventually by
booting on the install floppy and powering the drive off just after it had been
probed and on again before starting the initialization of the disk (before that
I even tried modifying /sys/arch/i386/i386/autoconf.c to break out immediately
when a matching disk is found (instead of checking the other disks to look for
configuration problems) thinking that findroot may have been the source of the
problem).

Now everything worked ok, until I tried:

box# dd if=/dev/rsd0d of=/dev/null bs=1048576 &
[1] 233
box# dd if=/dev/rsd1d of=/dev/null bs=1048576 &
[2] 234
box# sd0(aha0:0:0): timed out
sd1(aha0:1:0): timed out
sd1(aha0:1:0): timed out
sd0(aha0:0:0): timed out
sd1(aha0:1:0): timed out AGAIN
sd1(aha0:1:0): timed out AGAIN
sd0(aha0:0:0): timed out AGAIN
sd0(aha0:0:0): timed out AGAIN

.. notice how sd1 (the Fireball) times out twice, and right after that sd0
times out too (sd0 has never given me problems before)!

Result: System hangs with harddisk LED on.

I suspect that bad sectors on sd1 causes this, but it puzzles me that the SCSI
bus seems to hang because of this. Maybe it's simply a SCSI bus problem.

Anders