Re: [PATCH 10/10] reset: hi6220: allow to compile test driver on other architectures

From: Masahiro Yamada
Date: Wed Aug 24 2016 - 14:00:50 EST


2016-08-24 22:29 GMT+09:00 Philipp Zabel <p.zabel@xxxxxxxxxxxxxx>:
> Also remove the RESET_CONTROLLER dependency, this Kconfig file is
> included inside the menuconfig already.
>
> Cc: Chen Feng <puck.chen@xxxxxxxxxxxxx>
> Signed-off-by: Philipp Zabel <p.zabel@xxxxxxxxxxxxxx>
> ---
> drivers/reset/hisilicon/Kconfig | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/reset/hisilicon/Kconfig b/drivers/reset/hisilicon/Kconfig
> index 26bf95a..1ff8b0c 100644
> --- a/drivers/reset/hisilicon/Kconfig
> +++ b/drivers/reset/hisilicon/Kconfig
> @@ -1,5 +1,6 @@
> config COMMON_RESET_HI6220
> tristate "Hi6220 Reset Driver"
> - depends on (ARCH_HISI && RESET_CONTROLLER)
> + depends on ARCH_HISI || COMPILE_TEST
> + default ARCH_HISI
> help
> Build the Hisilicon Hi6220 reset driver.
> --
> 2.8.1
>


As you see

obj-$(CONFIG_ARCH_HISI) += hisilicon/

in drivers/reset/Makefile,
Kbuild can descend into drivers/reset/hisilicon/
only when CONFIG_ARCH_HISI is enabled.

So, this patch does not add compile-test'ability.


Currently, I only see a single file in drivers/reset/hisilicon/ directory,
but perhaps they might be planning to add more reset drivers for
other Hisilicon SoCs in the future.


Change to

obj-y += hisilicon/

or

create "config RESET_HISI" as a vendor-common option and

obj-$(CONFIG_RESET_HISI) += hisilicon/

?




--
Best Regards
Masahiro Yamada