Details
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Passed - Unit
No Test Coverage
Event Timeline
I think this is fine as the quick patch, but could we share the code with madt_parse_apics() and with madt_parse_local_nmi() ? I suppose some short helpers would do it.
sys/x86/acpica/madt.c | ||
---|---|---|
324 | I think this should be LocalApicId and not Uid. |
sys/x86/acpica/madt.c | ||
---|---|---|
324 | la_acpi_id is supposed to be the ID ACPI uses. It corresponds to the field present in the NMI entry (which is Uid for X2APIC and ProcessorID for plain APIC). |
FYI, I finally boot tested this on a machine with a normal MADT (no x2APIC entries) and verified that it still booted ok. I still haven't heard anything from zeising in regards to testing on the HP machine that needs this.
IMO it does not make sense to sit on the patch. On the non-buggy BIOSes, the patch should make no change. Supposedly, the patch fixes the buggy case, which is not working right now anyway. So even if, unlikely, there is some bug preventing the Dells from booting, the machines cannot boot now. But there is high chance that they would boot. Also, x2apic entries parsing is the required step.