Attention is currently required from: Arthur Heymans, Felix Held, Felix Singer, Jérémy Compostella, Shuo Liu.
yuchi.chen@intel.com has posted comments on this change by yuchi.chen@intel.com. ( https://review.coreboot.org/c/coreboot/+/83318?usp=email )
Change subject: soc/intel/common/systemagent: Fixup systemagent address ......................................................................
Patch Set 16:
(4 comments)
Commit Message:
https://review.coreboot.org/c/coreboot/+/83318/comment/926027f5_79d08e32?usp... : PS13, Line 9: System agent in Intel common block (1) assumes TOLUD and TOUUD : registers hold the max available address plus 1, but on some SoC like : Snow Ridge, it holds the max available address; (2) aligns TOLUD, TOUUD : and TSEG registers to 1 MiB default, but some SoC may have different : alignments. This patch add a new weak function : `soc_systemagent_fixup_address()` to improve it.
Can you not capture the different semantics of the registers in Kconfig, have SOC select it and keep […]
Now I'm using a new Kconfig item SA_FIXUP_ADDRESS to select if SoC need to fixup the address base and limit, please review it again.
Commit Message:
https://review.coreboot.org/c/coreboot/+/83318/comment/69bb3605_36b4d7c1?usp... : PS15, Line 14: soc_systemagent_fixup_address() to improve it.
commit message need to be updated accordingly
Commit message doesn't describe the TSEG limit changes.
File src/soc/intel/common/block/systemagent/Kconfig:
https://review.coreboot.org/c/coreboot/+/83318/comment/5ea63fbb_a6186d21?usp... : PS15, Line 65: config HAVE_TSEG_LIMIT_REGISTER
can this be split to a separate patch?
Done, please see https://review.coreboot.org/c/coreboot/+/84200
File src/soc/intel/common/block/systemagent/systemagent_early.c:
https://review.coreboot.org/c/coreboot/+/83318/comment/7c41bd6e_78b6e32f?usp... : PS15, Line 178: __weak uint64_t soc_systemagent_fixup_address(unsigned int reg, uint64_t value)
no need for this weak
Done
-
Name of user not set (Code Review)