On 26 xxx -1, Peter Teichmann wrote:

> when reading from floppys under RiscBSD I always get errors like this:
> 
> Feb 25 22:49:54 Aldebaran /netbsd: fd0a: hard error reading fsbn 63 (st0
> 44<abnrml,top_head> st1 10<data_overrun> st2 0 cyl 1 head 1 sec 10)

  I think this is due to RiscBSD not processing interrupts from the 
floppy drive quickly enough when the CPU is loaded (loading seems to 
include moving the mouse under X!)  This appears to be significantly 
worse on SA systems due to the non optimised cache flushing.

  The solution is to make sure you don't touch your machine while 
reading/writing to the floppy.

Ale.