Re: [PATCH] gpio: Fix OF build problem on UM

[Linus Walleij <linus.walleij@linaro.org>]

On Fri, Aug 19, 2016 at 11:16 AM, Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
> On Fri, Aug 19, 2016 at 10:53 AM, Linus Walleij
> <linus.walleij@linaro.org> wrote:
>> On Wed, Aug 17, 2016 at 11:03 AM, Geert Uytterhoeven
>> <geert@linux-m68k.org> wrote:
>>> On Tue, Aug 16, 2016 at 10:06 AM, Linus Walleij
>>> <linus.walleij@linaro.org> wrote:
>>>> The UserMode (UM) Linux build was failing in gpiolib-of as it requires
>>>> ioremap()/iounmap() to exist, which is absent from UM. The non-existence
>>>> of IO memory is negatively defined as CONFIG_NO_IOMEM so put a reverse
>>>> dependency on !NO_IOMEM for OF_GPIO to fix the build.
>>>>
>>>> Cc: stable@vger.kernel.org
>>>> Reported-by: kbuild test robot <fengguang.wu@intel.com>
>>>> Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
>>>> ---
>>>>  drivers/gpio/Kconfig | 1 +
>>>>  1 file changed, 1 insertion(+)
>>>>
>>>> diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig
>>>> index 98dd47a30fc7..a6026f2fe7f5 100644
>>>> --- a/drivers/gpio/Kconfig
>>>> +++ b/drivers/gpio/Kconfig
>>>> @@ -50,6 +50,7 @@ config GPIO_DEVRES
>>>>  config OF_GPIO
>>>>         def_bool y
>>>>         depends on OF
>>>> +       depends on !NO_IOMEM
>>>
>>> depends on HAS_IOMEM?
>>
>> Apparently not, and that is the scary thing.
>>
>> HAS_IOMEM depends on !NO_IOMEM indeed but
>> it also selects GENERIC_IO, and AFAICT there are
>> some configs out there in the intersection between,
>> that have IOMEM but are not using GENERIC_IO,
>> tell me if I'm wrong :/
>
> lib/Kconfig:
>
>     config HAS_IOMEM
>             bool
>             depends on !NO_IOMEM
>             select GENERIC_IO
>             default y
>
> Hence if NO_IOMEM is not set, HAS_IOMEM and GENERIC_IO are always set.
>
> UML selects GENERIC_IO, but not HAS_IOMEM (it sets NO_IOMEM).

Ah you're right.

OK sent a new version!

Yours,
Linus Walleij