Hi Bryan,

On 2026-08-04T21:55:02, Bryan Brattlof <[email protected]> wrote:
> boot/Kconfig: disable fit post processing for AM62L
>
> The AM62L family of SoCs utilize a 2 stage ROM boot which uses x509
> certificates for booting securely. There is no need for processing of a
> fitImage. Disable it for the AM62L SoC family.

Just to check - the AM62L defconfig added later in this series does
not enable TI_SECURE_DEVICE, so this change has no effect on the
boards being added here. Presumably it is aimed at a future HS
variant. Please say that in the commit message so it is clear why the
change is needed now.

>
> Tested-by: Anshul Dalal <[email protected]>
> Reviewed-by: Anshul Dalal <[email protected]>
> Signed-off-by: Bryan Brattlof <[email protected]>
>
> boot/Kconfig | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)

> diff --git a/boot/Kconfig b/boot/Kconfig
> @@ -333,7 +333,7 @@ config SPL_LOAD_FIT_FULL
>  config SPL_FIT_IMAGE_POST_PROCESS
>       bool "Enable post-processing of FIT artifacts after loading by the SPL"
>       depends on SPL_LOAD_FIT
> -     default y if TI_SECURE_DEVICE
> +     default y if (TI_SECURE_DEVICE && !SOC_K3_AM62L3)

Adding an SoC-specific negation to a generic boot/Kconfig option does
not scale - the next K3 SoC with the same boot flow will need another
'&& !SOC_K3_xxx' term. I think this would be better decided in the SoC
layer, e.g. by adding a 'default n if SOC_K3_AM62L3' override in
arch/arm/mach-k3/Kconfig alongside the other K3-specific defaults, or
by having the AM62L Kconfig not select whatever pulls TI_SECURE_DEVICE
in. That keeps boot/Kconfig free of per-SoC knowledge.

Regards,
Simon

Reply via email to