Re: [PATCH 1/2] misc: rtsx: Move Realtek Card Reader Driver to misc

[Lee Jones]

On Tue, 24 Oct 2017, Arnd Bergmann wrote:

> On Mon, Oct 23, 2017 at 11:35 AM, Lee Jones <lee.jones@xxxxxxxxxx> wrote:
> > On Thu, 19 Oct 2017, Arnd Bergmann wrote:
> >
> >> On Thu, Oct 19, 2017 at 5:22 AM,  <rui_feng@xxxxxxxxxxxxxx> wrote:
> >> > From: Rui Feng <rui_feng@xxxxxxxxxxxxxx>
> >> >
> >> > Because Realtek PCIE card reader driver is a pcie driver,
> >> > and it bridges mmc subsystem and memstick subsystem, it's
> >> > not a mfd driver. Greg and Lee Jones had a discussion about
> >> > where to put the driver, the result is that misc is a good
> >> > place for it, so I move all files to misc. If I don't move
> >> > it to a right place, I can't add any patch for this driver.
> >> >
> >> > Signed-off-by: Rui Feng <rui_feng@xxxxxxxxxxxxxx>
> >> > ---
> >> >  drivers/memstick/host/rtsx_pci_ms.c      |  2 +-
> >> >  drivers/mfd/Kconfig                      | 11 -----------
> >> >  drivers/mfd/Makefile                     |  2 --
> >> >  drivers/misc/Kconfig                     |  1 +
> >> >  drivers/misc/Makefile                    |  1 +
> >> >  drivers/misc/realtek/Kconfig             | 10 ++++++++++
> >> >  drivers/misc/realtek/Makefile            |  3 +++
> >> >  drivers/{mfd => misc/realtek}/rtl8411.c  |  1 -
> >> >  drivers/{mfd => misc/realtek}/rts5209.c  |  1 -
> >> >  drivers/{mfd => misc/realtek}/rts5227.c  |  1 -
> >>
> >> Thanks for redoing the patch. If Lee and Greg want to see it moved
> >> here, I won't object, but I'd still like to understand the distinction better,
> >> since there are likely going to be other drivers in the same situation
> >> (either adding into misc or mfd) and it would be good if we can all
> >> give consistent advice.
> >>
> >> Greg, Lee: can you come up with a simple definition of what an
> >> MFD is that this driver does not fit into? What about other
> >> PCI drivers like intel-lpss-pci and lpc_ich? Are they different
> >> from this one or should they also be moved out of MFD?
> >
> > MFDs drivers don't *do* anything.  They handle shared resources,
> > register featureful devices residing in other subsystems and provide
> > simple shared functionality pertaining to the chip as a whole.
> >
> > These card reader devices *are* the featureful drivers and brim with
> > useful functionality.
> >
> > intel-lpss-pci and lpc_ich conform to the aforementioned definition.
> 
> Ok. I think we have some other drivers in MFD that fall into similar
> categories as rtsx, maybe the db8500 prcmu for instance, but it

Yes, the rts drivers should be moved as they do not conform and are
still in active development.  And I agree that there is an argument
for the x8500 set to be considered as platform drivers.  However,
since they are now stable I see no reason to move them unnecessarily.

> does sound like a reasonable definition. If we ever get too many
> drivers like rtsx in drivers/misc, we may have to come up with a
> new place for them.

Agreed.

> For now
> 
> Acked-by: Arnd Bergmann <arnd@xxxxxxxx>

Thanks.

Rui, please add this to your set.

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org â Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog