> Clearly, the MII code needs a little restructuring to get rid of these atrocities.

Kernel threads! Kernel threads! :)

Seriously, how long can it go on before we implement an actual real-time
scheduler so that most interrupts simply wake up a device driver thread
and return. It would make it a lot easier for us to fake a true prioritized
interrupt scheme on machines that have brain-dead interrupt hardware or
firmware (Alpha OSF PALcode)...

I know there are arguments against this on dog-slow machines, but lately
I've begun to wonder if the overhead is really so unacceptable.

Todd Whitesel
toddpw @ best.com