> > When the mcd driver didnt have port accounting and had an overly-keen
> > match routine, sure, it was busted. But after the above, I see no
> > cause for especial concern, certainly not for SMC cards.
> > Am I missing something?
>
>Main difference... wt driver's default port locations were changed :-)

Sure, good move.  But from Charles' explanation, i dont get how the wt
affects the mcd problem, or vice-versa.  Could you explain, please?


I thought  mcd caused trouble back in 1.2 because 
    a) The mcd_find() probe matched too promiscuously;
    b) Back then we didn't have bus-space or port  accounting;
    c) the mcd driver would therefore attach an mcd at 0x300,
       even if some other driver (el, ed, ...) had already attached 
       an ethernet card there. 

sending mcd commands to a NIC at 0x300 let loose all sorts of hell.
But its clearly ancient history now.

And I mcd probes are, by default, no longer promiscuous.  
Am I still going to lose limbs if I suggest re-enabling mcd?