Folks who'll be using -current after the 1.2 release:

I just checked in some mostly mechanical changes to the eg (3c505) 
driver, mostly conversion to use <machine/bus.h> (so that the driver can 
participate in the i/o port accounting process, mostly).

While converting the driver, I found two bugs by inspection:

	- In egstart(), a missing call to m_freem() if the
	  Send Packet command failed, which would result in
	  a memory/mbuf leak.

	- In egintr(), the driver always returned 0, which, by
	  convention, indicates "interrupt wasn't mine".  I changed
	  egintr() to return 1 (which indicates "I serviced it")
	  if a valid transmit, receive, or status interrupt was
	  processed.  The driver still returns 0 if the card
	  didn't appear to generate the interrupt, or if an unknown
	  interrupt was generated by the card.

If any 3c505 users out there experience problems with the driver once 
1.2A is available, please let me know.  I have a 3c505 in my (currently 
turned-off) Apollo at home; I'll try to remember to bring it into work so 
I can stress out the driver myself.

 -- save the ancient forests - http://www.bayarea.net/~thorpej/forest/ -- 
Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939