> Short: > Is it possible to use static interrupt routing on a machine that > support configurable interrupt (via PCI Link Device)? So that kernel > will go to IOAPIC chip directly instead of using ACPI layer (such as > _CRS method) to manage the interrupt > > Long: > I am trying to resume a kernel in a machine with different hardware > which results in many hardware miss match problem. Right now I am > looking into the interrupt issue. The machines I am testing support > different interrupt routing model (Dynamic and Static). I am not sure > if a machine that support dynamic routing will also support static > routing or not. So is it possible to force the kernel to use only > static model only? Bypassing ACPI layer has better chance of success > because I think that the resume kernel does not have a correct ACPI > handler to deal with the new device. > > My understanding may be incorrect because I just read a portion of > ACPI spec and the kernel source that related to this issue. This doesn't make any sense. Interrupt routing configuration reflects wires on the motherboard. If you resume a kernel image on hardware with different wires than were you suspended, all is lost. To do something like this, you have to un-configure interrupts completely, and then re-probe your devices when you resume on changed hardware. ie. unload all your drivers. cheers, Len Brown, Intel Open Source Technoloy Center -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
