Ed Swierk firstname.lastname@example.org writes:
It appears that the write to register 0x78 is attempting to switch the LPC device from subtractive decode to positive decode. Presumably the hang is due to the LPC device not being configured for some IO or memory range. Why you're seeing the hang only when mmconfig is enabled, I have no idea.
I'd vote for removing that line, but I have no means to check whether that would break any board.
Thanks for checking, Ed. The following patch implements/fixes PCI MMCONF support for AMD Fam10 systems, and includes commenting out of the offending line in the MCP55 setup code. It also converts a couple of pci_*_config accesses in amdfam10/early_ht.c to pci_io_*_config, as they run before MMCONF is ready. (There's a comment regarding this in the top of that file, but that was apparently not sufficient to prevent these calls from sneaking in.)
Signed-off-by: Arne Georg Gleditsch email@example.com