On Wed, Apr 6, 2011 at 9:55 AM, Claudio Jeker <cje...@diehard.n-r-g.com> wrote: > Here is a better version that may get commited if it works for you. > Currently only amd64 is fixed, we're looking into i386 to do the same > dance with the interrupt return values. > So the idea is to establish the interrupt handler for the ppb as last and > jump out of interrupt processing if the handler returns 1 (HW was the > source of interrupt). So we should not end up in the slow ppb interrupt > handler unless it is actually a hotplug interrupt.
I think this is a very bad idea. Not running ppb sounds like a good idea, but jumping out is bad. If you have two devices sharing, and the first one is busy, the second will be starved. Drivers don't even return 1 accurately, wi is a good example of one that doesn't. There are many more.