Hi,
the -1 patch fixes some conflicting types of variables the -2 patch removes the normal/* files from the image. they're just copies of fallback/* anyway.
As for the types: Defining options as "bool" doesn't mean that the code has to use #if (CONFIG_FOO = y) or so. They're stored as 1 or 0 - but kconf (the kconfig utility) seems to be confused when an option is "both int and bool", and emits strings in that case. Or something like that anyway.
So, if you want to give the user a boolean decision, use "bool", not "int". kconf still does the right thing.
Signed-off-by: Patrick Georgi patrick.georgi@coresystems.de
Index: src/cpu/amd/model_fxx/Kconfig =================================================================== --- src/cpu/amd/model_fxx/Kconfig (revision 4540) +++ src/cpu/amd/model_fxx/Kconfig (working copy) @@ -1,10 +1,10 @@ config HAVE_INIT_TIMER - int - default 1 + bool + default y
config HAVE_MOVNTI - int - default 1 + bool + default y
config CPU_ADDR_BITS int Index: src/mainboard/amd/serengeti_cheetah/Kconfig =================================================================== --- src/mainboard/amd/serengeti_cheetah/Kconfig (revision 4540) +++ src/mainboard/amd/serengeti_cheetah/Kconfig (working copy) @@ -49,13 +49,13 @@ depends on BOARD_AMD_SERENGETI_CHEETAH
config HAVE_HARD_RESET - int - default 1 + bool + default y depends on BOARD_AMD_SERENGETI_CHEETAH
config IOAPIC - int - default 1 + bool + default y depends on BOARD_AMD_SERENGETI_CHEETAH
config SB_HT_CHAIN_ON_BUS0 @@ -164,8 +164,8 @@ depends on BOARD_AMD_SERENGETI_CHEETAH
config SERIAL_CPU_INIT - int - default 0 + bool + default n depends on BOARD_AMD_SERENGETI_CHEETAH
config AP_CODE_IN_CAR Index: src/northbridge/amd/amdk8/Kconfig =================================================================== --- src/northbridge/amd/amdk8/Kconfig (revision 4540) +++ src/northbridge/amd/amdk8/Kconfig (working copy) @@ -26,8 +26,8 @@ default 0x4000000
config HAVE_HIGH_TABLES - int - default 1 + bool + default y
-source src/northbridge/amd/amdk8/root_complex/Kconfig \ No newline at end of file +source src/northbridge/amd/amdk8/root_complex/Kconfig
Index: src/arch/i386/Makefile.inc =================================================================== --- src/arch/i386/Makefile.inc (revision 4540) +++ src/arch/i386/Makefile.inc (working copy) @@ -14,7 +14,6 @@ $(obj)/coreboot.rom: $(obj)/coreboot.bootblock $(obj)/coreboot_ram $(CBFSTOOL) $(Q)rm -f $@ $(Q)$(CBFSTOOL) $@ create $(shell expr 1024 * $(CONFIG_COREBOOT_ROMSIZE_KB)) 131072 $(obj)/coreboot.bootblock - $(Q)$(CBFSTOOL) $@ add-stage $(obj)/coreboot_ram normal/coreboot_ram $(CBFS_COMPRESS_FLAG) $(Q)if [ -f fallback/coreboot_apc ]; \ then \ $(CBFSTOOL) $@ add-stage fallback/coreboot_apc fallback/coreboot_apc $(CBFS_COMPRESS_FLAG); \ @@ -25,8 +24,6 @@ else printf " PAYLOAD $(CONFIG_FALLBACK_PAYLOAD_FILE) $(COMPRESSFLAG)\n" $(Q)$(CBFSTOOL) ./build/coreboot.rom add-payload $(CONFIG_FALLBACK_PAYLOAD_FILE) fallback/payload $(CBFS_COMPRESS_FLAG) - printf " PAYLOAD $(CONFIG_NORMAL_PAYLOAD_FILE) $(COMPRESSFLAG)\n" - $(Q)$(CBFSTOOL) ./build/coreboot.rom add-payload $(CONFIG_NORMAL_PAYLOAD_FILE) normal/payload $(CBFS_COMPRESS_FLAG) $(CBFSTOOL) ./build/coreboot.rom print endif
Thanks for the fix. Some of those incorrect Kconfig things came from my script :-)
Acked-by: Ronald G. Minnich rminnich@gmail.com
one other thing. Kontron does not build.
GEN build/build.h LINK coreboot /home/rminnich/src/bios/coreboot-v2/build/ldscript.ld:7: undefined symbol `n' referenced in expression collect2: ld returned 1 exit status make: *** [/home/rminnich/src/bios/coreboot-v2/build/coreboot] Error 1
Am 13.08.2009 03:23, schrieb ron minnich:
Thanks for the fix. Some of those incorrect Kconfig things came from my script :-)
Acked-by: Ronald G. Minnichrminnich@gmail.com
one other thing. Kontron does not build.
GEN build/build.h LINK coreboot
/home/rminnich/src/bios/coreboot-v2/build/ldscript.ld:7: undefined symbol `n' referenced in expression collect2: ld returned 1 exit status make: *** [/home/rminnich/src/bios/coreboot-v2/build/coreboot] Error 1
That's exactly what these patches should fix
Patrick
On Thu, Aug 13, 2009 at 12:27 AM, Patrick Georgipatrick@georgi-clan.de wrote:
That's exactly what these patches should fix
They fixed it.
thanks ron