Kyösti Mälkki has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/31486 )
Change subject: soc/amd/common: Refactor AmdCreateStruct() use ......................................................................
Patch Set 6:
(2 comments)
https://review.coreboot.org/#/c/31486/5/src/soc/amd/common/block/pi/agesawra... File src/soc/amd/common/block/pi/agesawrapper.c:
https://review.coreboot.org/#/c/31486/5/src/soc/amd/common/block/pi/agesawra... PS5, Line 68: aip
Unfortunately that's not true. […]
The file I referred under src/drivers is the top (or bottom, however you look at it) API common to all AGESA v8.
Yes, I know the headers and structs may differ across different PI's (MullinsPI, KaveriPI, CarrizoPI and StoneyPI). That's why there are PI dependent files (like fam15tn/state_machine.c) with open-source AGESA v8 already at place.
I encourage you to have a peek on state of the agesawrapper.c files at the time they were originally merged in [1]. And there was one copy per motherboard!
[1]: https://review.coreboot.org/c/coreboot/+/1158/10/src/mainboard/amd/parmer/ag...
https://review.coreboot.org/#/c/31486/5/src/soc/amd/common/block/pi/agesawra... PS5, Line 111: return module_dispatch(AMD_RELEASE_STRUCT, &aip->StdHeader);
The return of is internal to AGESA. That said, look agesa. […]
I don't see how AGESA.c line 58 is related to any AGESA_BOUNDS_CHK.
Please be clear when you place arguments on changes that are not in upstream coreboot tree. I believe Marshall had previously identified src/soc/amd/common did not work for quad-core (dual-compute-unit) SKUs due to cache coherency issue CB:26115.