Attention is currently required from: Andrey Petrov, Intel coreboot Reviewers, Julius Werner, Jérémy Compostella, Paul Menzel, Ronak Kanabar.
Subrata Banik has posted comments on this change by Subrata Banik. ( https://review.coreboot.org/c/coreboot/+/86001?usp=email )
Change subject: drivers/intel/fsp2_0: Add option to control debug log level using CBFS ......................................................................
Patch Set 8:
(1 comment)
File src/drivers/intel/fsp2_0/debug.c:
https://review.coreboot.org/c/coreboot/+/86001/comment/71c413d9_1bffdcdf?usp... : PS6, Line 185: if (!CONFIG(USE_CBFS_FILE_OPTION_BACKEND))
Why? Why shouldn't this just work for all option backends equally?
To be clear, I think this function should just be:
enum fsp_log_level fsp_get_pcd_debug_log_level(void) { return get_uint_option("fsp_pcd_debug_level", fsp_map_console_log_level()); }
[Sorry, looks like I forgot to publish this comment yesterday.]
Looks like you missed to read my comments previously (?) as I have mentioned why we can't call fsp_map_console_log_level as fallback because we are now stitching debug fsp by default with ap fw that means when we are not injecting options to control debug level (like default ap fw won't have any option included), we shouldn't see any debug log from fsp. Calling fsp_map_console_log_level api as fallback would mean that we are getting chatty console by default which is not intention of this CL.
One should see more chatty console when they inject options into cbfs. If they don't select cbfs backend kconfig then this CL doesn't change the behavior of console level.
I have tried to accommodate your feedback by adding another Kconfig (FSP_DYNAMIC_DEBUG) an option to control the debug log level of FSP blobs. This will allow existing Intel devices that are not intended to enable dynamic FSP debugging to continue to function correctly.