Re: [PATCH] MIPS: Remove most of the custom gpio.h

[Alban <albeu@free.fr>]

On Wed, 22 Jul 2015 19:47:18 +0200
Manuel Lauss <manuel.lauss@gmail.com> wrote:

> On Wed, Jul 22, 2015 at 7:33 PM, Alban Bedel <albeu@free.fr> wrote:
> > Currently CONFIG_ARCH_HAVE_CUSTOM_GPIO_H is defined for all MIPS
> > machines, and each machine type provides its own gpio.h. However
> > only the Alchemy machine really use the feature, all other machines
> > only use the default wrappers.
> >
> > For most machine types we can just remove the custom gpio.h, as well
> > as the custom wrappers if some exists. A few more fixes are need in
> > a few drivers as they rely on linux/gpio.h to provides some machine
> > specific definitions, or used asm/gpio.h instead of linux/gpio.h for
> > the gpio API.
> >
> > Signed-off-by: Alban Bedel <albeu@free.fr>
> > ---
> >
> > This patch is based on my previous serie:
> > "MIPS: ath79: Move the GPIO driver to drivers/gpio".
> >
> > For testing I tried to build all mips defconfig, however my
> > toolchain couldn't handle a few configs: ip28 malta_qemu_32r6
> > maltasmvp_eva sead3micro. If somebody can test these that would be
> > more than welcome.
> >
> > It might well be that some more drivers for MIPS devices that are
> > not enabled in the defconfig will break because of this change, so
> > more testing would be nice :)
> >
> > Regarding Alchemy I'm not sure what to do. It use a little more
> > complex setup, quoting arch/mips/include/asm/mach-au1x00/gpio.h:
> >
> > /* Linux gpio framework integration.
> > *
> > * 4 use cases of Alchemy GPIOS:
> > *(1) GPIOLIB=y, ALCHEMY_GPIO_INDIRECT=y:
> > *       Board must register gpiochips.
> > *(2) GPIOLIB=y, ALCHEMY_GPIO_INDIRECT=n:
> > *       A gpiochip for the 75 GPIOs is registered.
> > *
> > *(3) GPIOLIB=n, ALCHEMY_GPIO_INDIRECT=y:
> > *       the boards' gpio.h must provide the linux gpio wrapper
> > functions,
> > *
> > *(4) GPIOLIB=n, ALCHEMY_GPIO_INDIRECT=n:
> > *       inlinable gpio functions are provided which enable access
> > to the
> > *       Au1300 gpios only by using the numbers straight out of the
> > data-
> > *       sheets.
> >
> > * Cases 1 and 3 are intended for boards which want to provide their
> > own
> > * GPIO namespace and -operations (i.e. for example you have 8 GPIOs
> > * which are in part provided by spare Au1300 GPIO pins and in part
> > by
> > * an external FPGA but you still want them to be accssible in linux
> > * as gpio0-7. The board can of course use the alchemy_gpioX_*
> > functions
> > * as required).
> > */
> >
> > This sound to me like this is really not needed anymore. Is there
> > any users of this left, or can it just go?
> 
> There are no in-tree users of this, but a few out-of-tree ones (all
> made by me) Does it have to be removed?  Is it blocking anything?

It is not blocking anything, but I see little gain in it. Cases 1 and 3
should nowadays be handled using normal GPIO drivers, and not with such
platform specific constructs.

Alban