[coreboot-gerrit] Change in coreboot[master]: SMM: Validate more user-provided pointers

[Tim Wawrzynczak (Code Review)]

Tim Wawrzynczak has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/41086 )

Change subject: SMM: Validate more user-provided pointers
......................................................................

Patch Set 15: Code-Review-1

(2 comments)

https://review.coreboot.org/c/coreboot/+/41086/12/src/soc/intel/baytrail/smi... 
File src/soc/intel/baytrail/smihandler.c:

https://review.coreboot.org/c/coreboot/+/41086/12/src/soc/intel/baytrail/smi... 
PS12, Line 337: break
> Not returning when the pointer is invalid is asking for security troubles, AFAIK. […]

It should definitely return. That was the point the authors were making in their presentation; we should detect if addresses (that are overwritable by another agent) are attempting to trick the SMI handler in to overwriting an arbitrary address.

https://review.coreboot.org/c/coreboot/+/41086/12/src/southbridge/intel/bd82... 
File src/southbridge/intel/bd82x6x/smihandler.c:

https://review.coreboot.org/c/coreboot/+/41086/12/src/southbridge/intel/bd82... 
PS12, Line 113: 		if (!smm_points_to_smram((void *)(uintptr_t)xhci_bar,
I think it may be safer to do this check at the top of the case statement. If the xhci bar has been overwritten to point to smram, then something is very wrong and it should just return.