On Thu, Jan 13, 2005 at 10:33:00PM +0000, Richard Earnshaw wrote:
> 	Revision 1.53 if if_wm.c introduced a reset of the if_wm device using
> 	an IO map handle.  Since this time the IQ31244 port seems to have been
> 	completely unable to communicate over the wm device (the chip seems to
> 	be incompletely reset).

You'll probably find that wm1 is OK while wm0 is not.  This is
because there are no other devices and wm0 gets I/O address 0 by
default.  This appears to not work at all with the 82546--it doesn't
fault, but it also apparently doesn't do the right thing with I/O
accesses.  I thought I had committed a change to default starting
I/O space at 0x1000 or so for i80321's PCI, but this unfortunately
also broke the iyonix.  I have on my to-do list to track this down.

What version of src/sys/arch/arm/xscale/i80321_pci.c do you have?

-allen

-- 
                  Use NetBSD!  http://www.netbsd.org/