Kyösti Mälkki has posted comments on this change. ( https://review.coreboot.org/c/coreboot/+/42062 )
Change subject: src/*: Update makefiles to exclude x86 code from psp-verstage ......................................................................
Patch Set 9:
(2 comments)
Patch Set 8:
Patch Set 8:
Every verstage-y file with <arch/io.h> in its build dependencies is wrong.
build/verstage/drivers/pc80/rtc/mc146818rtc_boot.d
I am jumping to conclusion patchset #8 is not build-tested with enabled psp-verstage?
I expect the build to fail again with CMOS_POST=y or OPTION_TABLE=y
It's all tested.
psp_verstage supplies its own arch/io.h https://review.coreboot.org/c/coreboot/+/41816/5/src/soc/amd/picasso/psp_ver...
Right, my bad. I did not expect that because it was not mentioned in the commit message. Providing include overrides for <arch/x> looks somewhat unorthodox but if it gets you one step further. I'll comment there in CB:41816.
Also nobody brought this up in CB:42183 or CB:42416, where I put ENV_X86 around <arch/io.h>. Looks like ENV_HAS_ARCH_IO should be introduced into <rules.h> then.
https://review.coreboot.org/c/coreboot/+/42062/9/src/cpu/x86/lapic/Makefile.... File src/cpu/x86/lapic/Makefile.inc:
https://review.coreboot.org/c/coreboot/+/42062/9/src/cpu/x86/lapic/Makefile.... PS9, Line 9: verstage_x86-y += boot_cpu.c I hope the syntax is expanded, some day, such that we have "all_x86-y += " for cases like these. Even better, if every "bootblock-y +=" under cpu/x86/ would be automatically treated as "bootblock_x86-y +=" since that is ultimately what we need here.
https://review.coreboot.org/c/coreboot/+/42062/9/src/drivers/pc80/pc/Makefil... File src/drivers/pc80/pc/Makefile.inc:
https://review.coreboot.org/c/coreboot/+/42062/9/src/drivers/pc80/pc/Makefil... PS9, Line 10: verstage_x86-y += i8254.c So it's excluded because it has inline asm? Not because of <arch/io.h>. This means my reasoning for CB:42221 changes to pc80/rtc/Makefile.inc were incorrect.