[coreboot-gerrit] Change in coreboot[master]: [TESTONLY] arch/x86: experiment with SMM defaults

[Nico Huber (Code Review)]

Nico Huber has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/34128 )

Change subject: [TESTONLY] arch/x86: experiment with SMM defaults
......................................................................

Patch Set 6:
> Patch Set 6:
>
> So, we have all boards passing here with HAVE_SMI_HANDLER=n. (OK I broke one)
>
> Now, can we write simplified smmstub such that PARALLEL_MP could actually work without the smm-class object and exit with empty (optionally locked) TSEG?

Sounds like a plan. I love this work btw. :)

I'm a bit too puzzled atm to look into it. Isn't it that we
only have to change the pointers passed to parallel-mp? Or
are you suggesting that we pass an alternative function that
is called when HAVE_SMI_HANDLER is disabled?