[PATCH] arm/dt: Add basic device tree support for mx53 loco board

[Grant Likely]

On Thu, Mar 31, 2011 at 10:50 AM, Grant Likely
<grant.likely at secretlab.ca> wrote:
> On Fri, Apr 01, 2011 at 12:36:16AM +0800, Shawn Guo wrote:
>> Hi Grant,
>>
>> On Wed, Mar 30, 2011 at 09:52:15PM -0600, Grant Likely wrote:
>> > On Tue, Mar 29, 2011 at 10:34:12AM +0000, Liu Hui-R64343 wrote:
>> > > Hi, Grant,
>> > > The two patches for mx51/mx53 DT support have the same issue, which
>> > > is the S-O-B will be missed when you git am. Let me know if you want me
>> > > re-send the two patches or you would take care when you am it? Thanks,
>> >
>> > I fixed it up.  Don't worry about it.
>> >
>> I'm little confused by the Kconfig changes in this patch.  I saw your
>> comments on SMDKV310 patch as below.  But this patch was accepted with
>> MX5_DT_COMMON selected, which in turn selects USE_OF and
>> PROC_DEVICETREE.
>
> I picked up a bunch of these patches even though there were little
> things that should probably be fixed up.  I'll get it all resolved
> before I ask Nicolas to pull.
>
>>
>> --- quota begins ---
>> > diff --git a/arch/arm/mach-exynos4/Kconfig b/arch/arm/mach-exynos4/Kconfig
>> > index a021b52..78f5924 100644
>> > --- a/arch/arm/mach-exynos4/Kconfig
>> > +++ b/arch/arm/mach-exynos4/Kconfig
>> > @@ -123,6 +123,7 @@ config MACH_SMDKV310
>> >     select EXYNOS4_SETUP_I2C1
>> >     select EXYNOS4_SETUP_KEYPAD
>> >     select EXYNOS4_SETUP_SDHCI
>> > +   select USE_OF
>>
>> I would drop this bit (I dropped it from my branch when I applied the
>> patch).  Device tree support remains an optional feature.  Instead it
>> can be enabled by default in the defconfig (or however the config is
>> managed for the kernel package)
>> --- quota ends ---
>>
>> BTW, are you silently renaming the original devicetree/test branch
>> to devicetree/test-2.6.38 and using new devicetree/test for Nicolas
>> to pull Linaro bits?  If that is the case, what branch
>> devicetree/arm-linaro-2.6.38-rebuilt will be for?
>
> devicetree/test is /always/ an unstable branch that I try to keep as
> close to Linus' tree as possible.  I usually rebase for every -rc
> release.  devicetree/arm is based on devicetree/test, but it is
> published in a way that means it can always be merged.
>
> Also, which devicetree/test is pretty much "anything goes", I'm a lot
> more careful about which patches I actually publish to devicetree/arm.
>
> devicetree/arm-linaro-2.6.38 is the tree that I used previously for
> asking Nicolas to pull.  I created
> devicetree/arm-linaro-2.6.38-rebuilt solely for the purpose of
> Nicolas' new rebuilt branch, but now I hear that he isn't going to use
> that tree, so I'll push the latest changes into the
> devicetree/arm-linaro-2.6.38 branch too.

Actually, it looks like linaro-2.6.38 master and rebuilt are merged
now, so I'll go back to using devicetree/arm-linaro-2.6.38 for stuff I
push to Nicolas.

g.