Blue Swirl wrote: > On Sun, Jun 6, 2010 at 8:10 AM, Jan Kiszka <jan.kis...@web.de> wrote: >> Second round, specifically adressing: >> - IRQMsg framework to refactor existing de-coalescing code >> - RTC IRQ output as GPIO pin (routed depening on HPET or -no-hpet) >> - ISA reservation for RTC IRQ >> >> If discussion around IRQMsg and de-coalescing happens to continue, I >> would suggest to merge patches 1..7 as they are likely uncontroversial >> and also fix bugs. > > Otherwise everything looks fine to me, but 10 and 11 had minor > problems. Nice work!
Thanks for the quick feedback! > > I'd suppose one possible cleanup could be to use the message payload > in place of apic_deliver_irq()? Haven't looked into such things yet, the series is already long enough. :) I could also imagine that we may avoid exporting the APIC MSI functions for HPET use and instead provide a single MSI qemu_irq object, pushing the vector information into the message payload. Jan
signature.asc
Description: OpenPGP digital signature
