bwn + BETA3/ppc = very unstable

[Adrian Chadd]

There've been a few comments about bwn stability and performance when
running on ppc.
Unfortunately I don't have any ppc hardware that's suitable.

The performance issue is likely due to endianness in some eeprom fiddling code.

The illegal instruction? i'm not sure; can you please get a backtrace
from the debugger so we can see exactly where it's dying?

THanks,


Adrian