On Sat, 4 Aug 2007, Mark Davies wrote:

> Yes disabling azalia _or_ disabling ehci gets a system that doesn't
> hang so its presumably some interaction between the two - I suspect
> its more likely to be ehci thats at fault.  Would be nice to get that
> sorted.

   I have tracked this down more.  When ehci is enabled, something appears 
to halt the uhci0 and uhci1 controllers.  When their interrupt handler 
runs (which it will for uhci0 for each azalia0 and bge0 interrupt since 
they share the same interrupt), the handler prints the controller halted 
message.  In addition if azalia is enabled, uhci0 get disabled before it
has completed the configuration (the intial discovery run in the event 
process for uhci0) and the system hangs on boot waiting for all the 
connfiguration tasks to complete.

   I'm trying to figure out why uhci0 is getting halted, but haven't 
figured out anything so far (other than it has something to do with 
enabling just ehci0).

   I've gotten around the problem for now by changing the uhci interrupt
routine to not consider UHCI_STS_HCH as an interrupt condition (so when 
the interrupt routine gets run by an azalia0 or bge0 interrupt it doesn't 
disable access to the controller).  The real fix will be to figure out
what is shutting down the uhci0 and uhci1 controllers and fix that 
problem.

--
Michael L. Hitch			mhitch@montana.edu
Computer Consultant
Information Technology Center
Montana State University	Bozeman, MT	USA