As far as I can tell, the IRQ sharing problem I found is related to PCI quirks on my motherboard/BIOS or possibly Unicorn hardware and not caused by the Unicorn driver.
It only arises when the BIOS allocates Unicorn a shared IRQ with the motherboard VT8233 AC97 audio chip. The kernel ACPI code later re-allocates the Unicorn to a separate IRQ but something still keeps raising (and failing to clear) spurious interrupts on the original IRQ. Possibly a signal level is not being reset on the original IRQ allocation. Causing the Unicorn to answer both interrupts does not help, the original IRQ seems to be stuck in a triggered state after the Unicorn has been moved away from it. Sharing with other devices apart from the VT8233 doesn't seem to be a problem, hence my suspicion that this is a PCI quirk and not a driver error. There may be a driver related solution but it's beyond my current skills to find out. I'll open a separate bug for it if I can narrow things down further. I'll therefore go ahead with asking for an upload as Philippe suggests. Thanks very much for all your help ! Nick -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
