On Tue 21/8/07 7:58 pm, David Murray wrote:
I'll diff Dragonfly's acpica5 dir with FreeBSD's...
Curiously, FreeBSD 6.2.0 has an *older* Intel ACPI-CA release (20041119) than DfBSD (20061109).
However, FreeBSD's ACPI devices have drifted some way (even at 6.2.0, and then some since) since Df's were forked off. There are lots of changes. I didn't spot anything that obviously looked like it might cause interrupts to go missing, but the problem's probably much to subtle to spot by inspection.
I don't know if the philosophy here is to minimise the differences and track the FreeBSD ACPI or maintain a separate implementation. Either is non-trival, I'd say. ;-) Sorry I don't have a fix to suggest.
-- David Murray
