Attention is currently required from: Arthur Heymans, Kyösti Mälkki.
Nico Huber has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/74441 )
Change subject: cpu/intel/speedstep: Use acpigen_write_processor_device() ......................................................................
Patch Set 4:
(2 comments)
File src/southbridge/intel/i82801gx/fadt.c:
https://review.coreboot.org/c/coreboot/+/74441/comment/3fd026e7_3ec1cda8 PS4, Line 29: fadt->p_lvl3_lat = chip->c3_latency;
Well, the question is if _CST is a property of the installed CPU part and the location of get_cst_en […]
I wouldn't say the current implementation is right. But if we want to drop Processor(), shouldn't we first have an alternative ready?
In any case, we shouldn't leave dead code around. AFAICT, this was the only consumer of `c3_latency`.
File src/southbridge/intel/i82801jx/fadt.c:
https://review.coreboot.org/c/coreboot/+/74441/comment/76aef374_b022ed28 PS4, Line 25: fadt->p_lvl3_lat = 0; /* FIXME: Is this correct? */
Deprecation of Processor() object hides C2/C3 IO "trap" addresses (PROC_CNT or PROC_BLK +4 (C2) and […]
Ok, I get it now. ACPI spec isn't very specific about this in the FADT description. I missed that we don't advertise P_BLK anymore. Dropping these settings is fine then.