Felix Held has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/42339 )
Change subject: soc/amd/picasso: fix build if PICASSO_UART is unset ......................................................................
Patch Set 2:
(1 comment)
https://review.coreboot.org/c/coreboot/+/42339/2/src/soc/amd/picasso/uart.c File src/soc/amd/picasso/uart.c:
https://review.coreboot.org/c/coreboot/+/42339/2/src/soc/amd/picasso/uart.c@... PS2, Line 48: #endif /* PICASSO_UART */
Why are we guarding this w/ preprocessor macros? Can't we leave the implementations and rely upon th […]
Those two functions need to be provided by exactly one compilation unit that provides the functionality used by src/drivers/uart/ for the console UART. When an UART is used that is not the SoC's integrated one, these functions mustn't end up in the object file of this compilation unit.
Putting those two functions into a different compilation unit is definitely what should be done in the end, but since uart.c doesn't have a proper header file yet, it can't be pulled apart easily and I'm currently working on fixing things to get Mandolin working again. Basically every time I get anywhere close to have Mandolin in a state that could finally get submitted, some other change gets submitted creating significant breakage for Mandolin, which is rather frustrating for me, since Mandolin is the platform I do development. I'm planning to do that as a follow-up after fixing the APOB issue, but the further cleanup not a priority for me right now; getting this change landed however is a priority for me, since it is one of the patches required to unbreak non-zork/non-chromeos Picasso support on upstream.