Re: PROBLEM: 2.4.1[89]: System hangs after "spurious 8259A

[Alan Cox]

On Tue, 2002-10-15 at 11:22, Kai Henningsen wrote:
> 
> In fact, the system hangs so soon (typically before or during fsck) 
> that I have no usefull 2.4 boot messages to offer - they scroll away 
> and then the keyboard locks. (And I can't make too extensive tests on 
> this machine, unfortunately, as it's an important server.)

Spurious IRQ 7 occurs when something asserts and deasserts the IRQ line
before the IRQ logic has time to decide whom. There is one known
"happens but harmless" case in the IDE logic where we happen to set nIEN
just as an IDE device asserts an IRQ. The second case I know about is
building with local apic support on some VIA boards. Build a kernel
without apic support