Attention is currently required from: Julius Werner, Subrata Banik.
Nicholas Chin has posted comments on this change by Nicholas Chin. ( https://review.coreboot.org/c/coreboot/+/85905?usp=email )
Change subject: drivers/option: Add CBFS file based option backend ......................................................................
Patch Set 4:
(4 comments)
Patchset:
PS2:
Hi Nicholas, I'm going to address the review comments in this CL since we need this option to read t […]
Oh, I didn't notice this by the time I pushed a new patchset. I don't do coreboot stuff for work but I am able to regularly work on CL in my free time if that works.
File src/drivers/option/cbfs_file_option.c:
https://review.coreboot.org/c/coreboot/+/85905/comment/98a21f4d_a1a12e41?usp... : PS2, Line 11: cbfs_map
Note that when using boards with CONFIG_VBOOT, this will read from the current slot (RO/RW_A/RW_B). […]
Done
https://review.coreboot.org/c/coreboot/+/85905/comment/b5e14cae_684f667a?usp... : PS2, Line 11: unsigned int *p
Actually, this is also a little dicey. `cbfstool add-int` creates a u64 in little-endian in CBFS. […]
Done
https://review.coreboot.org/c/coreboot/+/85905/comment/c186ee88_a60fc6d3?usp... : PS2, Line 11: name
Since the option names are all over the place, should we group them under a prefix in CBFS (e.g. […]
Done. I used the preprocessor to add an "option/" prefix to all calls to `get/set_uint_option)` for zero runtime overhead, especially since there's no access to things like `malloc` before ramstage. Not sure if there's a better way to do that.