On Mon, 7 Jul 2008, Ingo Molnar wrote: > > Please note two things: > > (1) The whole acpi_dmi_table[] thing originally depended on __i386__, so it > > wouldn't work on x86-64 no matter what. I removed that dependecy, but > > I have no idea why it was there and so I'm not sure if that's correct. Well spottedd -- perhaps the x86-64 was though to be perfect. ;) > > (2) The clear_IO_APIC_pin(apic1, pin1) done if > > disable_irq0_through_ioapic is true is absolutely essential. The > > symptoms are 100% reproducible without it. > > thanks, applied to tip/x86, to give this some more testing. > > the clear_IO_APIC_pin() is the most worrisome aspect of the change - but > since we are already in limited quirk mode, does it hurt? Maciej, any > preferences? It makes absolutely no sense and should be harmful to call clear_IO_APIC_pin(apic1, pin1) here, because both apic1 and pin1 should be equal to -1 here. If it has to be called, then I suppose the DMI matching did not work and the workaround has not been enabled. Rafael, can you please provide a full bootstrap log obtained with all the changes, but *without* this part? Also, can you please verify DMI IDs of your system (dmidecode or /sys/class/dmi, I am told)? I do think we should record DMI vendor and name information in the bootstrap log. It won't take a lot of space and will be more useful than some other IDs we never use for anything else, which we obtain from some other tables. Maciej -- 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
