Re: [PATCH] phy/mediatek: Make PHY_MTK_XSPHY depend on HAS_IOMEM and OF_ADDRESS to fix build errors

From: Chunfeng Yun
Date: Tue Nov 24 2020 - 21:24:14 EST


Hi Tiezhu,

On Tue, 2020-11-24 at 17:47 +0800, Tiezhu Yang wrote:
> devm_ioremap_resource() will be not built in lib/devres.c if
> CONFIG_HAS_IOMEM is not set, of_address_to_resource() will be
> not built in drivers/of/address.c if CONFIG_OF_ADDRESS is not
> set, and then there exists two build errors about undefined
> reference to "devm_ioremap_resource" and "of_address_to_resource"
> in phy-mtk-xsphy.c under COMPILE_TEST and CONFIG_PHY_MTK_XSPHY,
> make PHY_MTK_XSPHY depend on HAS_IOMEM and OF_ADDRESS to fix it.
>
> Reported-by: kernel test robot <lkp@xxxxxxxxx>
> Signed-off-by: Tiezhu Yang <yangtiezhu@xxxxxxxxxxx>
> ---
> drivers/phy/mediatek/Kconfig | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/drivers/phy/mediatek/Kconfig b/drivers/phy/mediatek/Kconfig
> index 50c5e93..66df045 100644
> --- a/drivers/phy/mediatek/Kconfig
> +++ b/drivers/phy/mediatek/Kconfig
> @@ -30,6 +30,8 @@ config PHY_MTK_XSPHY
> tristate "MediaTek XS-PHY Driver"
> depends on ARCH_MEDIATEK || COMPILE_TEST
> depends on OF
> + depends on HAS_IOMEM
> + depends on OF_ADDRESS
Why not add them into deconfig but here? In fact I don't know which way
is better and follow the kernel rule.

Vinod and Kishon, do you have any suggestion about this?

> select GENERIC_PHY
> help
> Enable this to support the SuperSpeedPlus XS-PHY transceiver for