[PATCH v7 08/13] ARM: unify MMU/!MMU addruart calls

[Arnd Bergmann]

On Tuesday 19 May 2015 23:26:53 Stefan Agner wrote:
> On 2015-05-19 20:24, Arnd Bergmann wrote:
> > On Tuesday 19 May 2015 18:06:53 Stefan Agner wrote:
> >> On 2015-05-19 13:50, Arnd Bergmann wrote:
> >> > Do you have a dependency on patch 10 (the one for EFM32) in your later
> >> > patches?
> >>
> >> Unfortunately, there is a dependency on that patch: I change the default
> >> of UNCOMPRESS_INCLUDE to debug/uncompress.h for all ARM_SINGLE_ARMV7M
> >> platforms. Without that patch, it would default to mach/uncompress.h,
> >> which does not exist for MACH_MXC/SOC_VF610.
> > 
> > An easy approach would be to list both ARM_SINGLE_ARMV7M and
> > EFM32 in the UNCOMPRESS_INCLUDE option initially, and then
> > follow-up with a patch that converts the three other platforms
> > (efm32, stm32 and lpc18xx) to ARM_SINGLE_ARMV7M and also
> > change the UNCOMPRESS_INCLUDE statement.
> > 
> > Note that I've already added both stm32 and lpc18xx to the
> > UNCOMPRESS_INCLUDE list in the next/soc branch.
> 
> Afaiks, only ARCH_LPC18XX is in UNCOMPRESS_INCLUDE, probably STM32
> doesn't support compressed image/earlyprintk?

It's actually a bug, and I have sent a patch to add the entry
(and was looking at my local tree), but haven't applied the
fix to the next/soc branch yet.

> >> > If not, you can send the other ones to Shawn, so I pull them as
> >> > a branch, and then I apply that on top of the merges. I have also
> >> > merged two other ARMv7M platforms for 4.2 now (both in next/soc),
> >> > so we should do the same change for those as well, and I'd rather
> >> > apply a patch for that, than merge a branch that is based on
> >> > next/soc.
> >>
> >> I guess, in that case, you need to take the others too? (08/09?)
> > 
> > I'm fine with that, too. If Shawn thinks the patches are ok, you can
> > just send all patches you want me to apply in arm-soc based on
> > the appropriate branches (next/soc, next/dt and next/defconfig, I
> > presume), and I'll apply them this time. In the future, you should
> > send all patches to Shawn though.
> 
> Ok, I will rebase this patchset (06-13) then, send the clocksource patch
> (07) as a single patch and send the EFM32 convertion patch (10) together
> with a conversion patch of the other two platforms as a follow-up.

Ok, sounds good.

	Arnd