Attention is currently required from: Felix Held, Jérémy Compostella, Nico Huber.
Patch set 2:Code-Review +2
1 comment:
Patchset:
What I remember from the cleanup times is that today's IOAPICs probably don't
care about their ID. Historically, an ID of 0 would have been wrong, because
that was usually the BSP LAPIC (and they needed a unique ID on the APIC bus,
which is gone). Not sure if anything still cares about the ID, though we have
to put it into ACPI tables (I just saw). We read the ID from hardware for this,
so we should also write it to hw I guess.
Both ioapic and lapic id's are configurable iirc. Making sure that there is no overlap, when they are on that legacy apic bus, where numbers matter, should be handled in the chipset code and not via assumptions made in common code. The commit message seems to say, that's already the case.
To view, visit change 80330. To unsubscribe, or for help writing mail filters, visit settings.