Re: [patch 37/53] x86/cpu: Detect real BSP on crash kernels

[Thomas Gleixner]

On Sat, Jan 13 2024 at 07:35, Zhang, Rui wrote:
> On Fri, 2024-01-12 at 16:39 +0100, Thomas Gleixner wrote:
>> It brings the effect that we can detect when we are not booting
>> (kexec
>> case) on the actual boot CPU because then the first enumerated APIC
>> ID
>> is not the same as the boot CPU APIC ID. No?
>
> Right.
> I was thinking in the way this patch series does, which just compares
> the boot CPU APIC ID and the lowest numbered APIC ID.

Yes, that's obviously not working when the lowest APIC ID is not the
actual boot CPU.

>> I agree that MADT is error prone too given the fact that not even
>> Intel
>> can get it right....
>
> For this MTL, I can raise an internal ticket to get it right.
>
> Are there quite some platforms with BSP not listed as the first entry
> in MADT?
> if so, we still have to live with the kexec boot parameter? :)

I haven't found one yet, but you might look through the P/E systems to
get a picture.