[PATCH 1/3] usb: chipidea: add CSR SiRFSoC ci13xxx usb driver

[Barry Song]

2013/6/14 Alexander Shishkin <alexander.shishkin at linux.intel.com>:
> Barry Song <21cnbao at gmail.com> writes:
>
>> 2013/6/14 Alexander Shishkin <alexander.shishkin at linux.intel.com>:
>>> Barry Song <21cnbao at gmail.com> writes:
>>>
>>>> +if USB_CHIPIDEA_UDC && USB_CHIPIDEA_HOST
>>>> +
>>>> +config USB_CHIPIDEA_SIRF
>>>> +     depends on ARCH_SIRF
>>>> +     bool "SiRF USB controller ChipIdea driver binding"
>>>> +     default y
>>>> +     help
>>>> +       Say Y here to enable sirf usb ChipIdea driver binding.
>>>> +
>>>> +config USB_CHIPIDEA_MSM
>>>> +     depends on ARCH_MSM
>>>> +     bool "MSM USB controller ChipIdea driver binding"
>>>> +     default y
>>>> +     help
>>>> +       Say Y here to enable msm usb ChipIdea driver binding.
>>>> +
>>>> +config USB_CHIPIDEA_IMX
>>>> +     depends on ARCH_MXC || ARCH_MXS
>>>> +     bool "i.MX USB controller ChipIdea driver binding"
>>>> +     default y
>>>> +     help
>>>> +       Say Y here to enable imx usb ChipIdea driver binding.
>>>> +
>>>> +endif
>>>
>>> No. This is wrong on at least 3 different levels.
>>
>> the point is we don't directly include MSM, IMX, SiRF and all other
>> drivers once we enable chipidea. that makes these drivers optional,
>> not always be built-in.
>
> This has been discussed several times already. Just set the
> USB_CHIPIDEA=m and then everything will be built as modules and nothing
> will be built in. What exactly is the problem you're trying to solve
> with this?
>

Alex, it seems it is not much a generic way to use building-module to
avoid all involved components built-in.
but since you have discussed several times and got agreement, i'll
follow it and don't re-open it.

> The only way something like this will go in is with Jiri's COMPILE_TEST
> patchset. This snippet above is just buggy.
>
> Regards,
> --
> Alex

-barry