On 18 July 2017 at 22:43, Tom Rini trini@konsulko.com wrote:
On Mon, Jul 17, 2017 at 04:44:55PM +0300, Sam Protsenko wrote:
Also introduce CONFIG_USE_BOOTARGS option so we can control if CONFIG_BOOTARGS defined at all.
Signed-off-by: Sam Protsenko semen.protsenko@linaro.org
[snip]
+config USE_BOOTARGS
bool "Enable boot arguments"default nn is default, we don't need this.
Noted. Will fix it in PATCH v2.
[snip]
diff --git a/configs/P1020MBG-PC_36BIT_SDCARD_defconfig b/configs/P1020MBG-PC_36BIT_SDCARD_defconfig index e297a0da5b..88a795b42e 100644 --- a/configs/P1020MBG-PC_36BIT_SDCARD_defconfig +++ b/configs/P1020MBG-PC_36BIT_SDCARD_defconfig @@ -13,6 +13,7 @@ CONFIG_OF_BOARD_SETUP=y CONFIG_OF_STDOUT_VIA_ALIAS=y CONFIG_SYS_EXTRA_OPTIONS="SDCARD" CONFIG_BOOTDELAY=10 +CONFIG_USE_BOOTARGS=y CONFIG_BOARD_EARLY_INIT_F=y CONFIG_SPL=y CONFIG_SPL_ENV_SUPPORT=y
This is the first of a number of examples where USE_BOOTARGS is set to y but there's no BOOTARGS set. That doesn't make sense.
By doing this I tried to achieve exactly the same behavior like it was done before this patch. I.e. having empty "bootargs=" variable defined in the environment (just so buildman doesn't bark a lot). I can remove those lines, if you like.
Looking at it a bit more in detail, the conversion didn't happen right in cases where we have: #define CONFIG_BOOTARGS SOME_OTHER_DEFINE.
Why exactly? Can you show some corresponding example?
That may become tricky. I'm going to post a horrible hack to help, shortly.
Please link it here when it's done. Also I'd like to hear some details about it (what is not right and how to use your hack to overcome it).
Thanks.
-- Tom