Comment 183 for bug 272247

The bisected merge seems also to touch timer interrupt issues through the io apic and in another upstream kernel bug I found some hints suspecting something there. Also there was a comment claiming the combination of "noapic lapic" would prevent the stops. This would be more proof to that theory. Also is there anything seen when looking at the output of /proc/interrupts (especially IRQ0/timer), which always should increase.