> On Sun, 13 Aug 2006 13:30:26 -0700 > bugme-daemon@xxxxxxxxxxxxxxxxxxx wrote: > >> http://bugzilla.kernel.org/show_bug.cgi?id=6997 > > Could someone please take a look at this bug from an ACPI POV? I _think_ > what's happened is that APCI decided to move his tg3 NIC onto a different > interrupt, causing it to collide with an i2c driver which doesn't share > interrupts. i2c-pca-isa requests IRQ 10 exclusively (non-shared). It has no mechanism for detecting whether its hardware is actually present, so I see no path where it will ever release IRQ 10. ACPI might have moved the tg3 IRQ to collide with the i2c-pca-isa one, but I don't know what it could do differently. If the i2c-pca-isa thing were described as an ACPI device (which it should be, if it appears on modern boxes), the driver could be fixed to pay attention to what the BIOS is trying to tell it. 8250_pnp.c is a good example of how to do that. If there's no such thing as i2c-pca-isa on ACPI boxes, maybe its driver should just exit early if ACPI is enabled, instead of claiming a random IRQ that isn't connected to anything and nobody else in the system knows about. - 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
