Michael Niewöhner has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/46272 )
Change subject: soc/intel/skl + cpu/intel/common: move AES-NI locking to common cpu code ......................................................................
Patch Set 4:
(3 comments)
https://review.coreboot.org/c/coreboot/+/46272/4//COMMIT_MSG Commit Message:
https://review.coreboot.org/c/coreboot/+/46272/4//COMMIT_MSG@7 PS4, Line 7: soc/intel/skl + cpu/intel/common: move AES-NI locking to common cpu code
In case you think too, that this is much too long, how about: […]
ack
https://review.coreboot.org/c/coreboot/+/46272/4/src/include/cpu/intel/msr.h File src/include/cpu/intel/msr.h:
https://review.coreboot.org/c/coreboot/+/46272/4/src/include/cpu/intel/msr.h... PS4, Line 8: #define MSR_FEATURE_CONFIG 0x13c
Should this be removed from the other header? Users could include this […]
did you mean "the other header[s]"?
src/cpu/intel/model_2065x/model_2065x.h src/cpu/intel/model_206ax/model_206ax.h src/soc/intel/common/block/include/intelblocks/msr.h src/soc/intel/denverton_ns/include/soc/msr.h src/soc/intel/broadwell/include/soc/msr.h
https://review.coreboot.org/c/coreboot/+/46272/4/src/include/cpu/intel/msr.h... PS4, Line 9: #define AESNI_LOCK_BIT 0
As we mostly use bit masks in coreboot, this seems highly error-prone. […]
Hmm, indeed it's only used a few times in cpu/intel and once in soc/intel. I agree, that we shouldn't mix bits and bit masks. However, dropping msr_set_bit() would require dropping the simplification in CB:46275. What do you think?