Karmic Pull Request; UBUNTU: ARM: DOVE: All commits to add support for the Marvell dove SOC

[Tim Gardner]

Amit Kucheria wrote:
> On 09 Jul 29, Brad Figg wrote:
>> Please pull the following changes since commit
>>
>>     a11db7305c030119a9fa06a2ce6921fd142780d8 UBUNTU: ARM: IMX51: fix imx51 header change compatibility with other archs
>>
>> from the git repository at:
>>     git://kernel.ubuntu.com/bradf/ubuntu-karmic dove
>>
>> This pull request is for the 256 commits to enable support for the
>> Marvell "dove" SOC.
>>
>>
>> ------------------------------------------------------------------
>>  UBUNTU: ARM: DOVE: Fixup common configuration after addition of Marvell Dove SOC additions.
>>  ARM.dove Marvell:dove: prevent the power management unit to be registered multiple times
>>  ARM.dove Marvell:dove: add PMU type support for MPP1
>>  ARM.dove Marvell:dove: fix compilation error when building for Z0 devices
>>
> 
> Hi Brad,
> 
> One feedback I have is that these patches religiously stick to a single
> naming convention in their one-line description.
> 
> For the FSL patches that were merged and the next set to follow, I am
> using the following:
> 
> UBUNTU: ARM: IMX51: <description>
> 
> Thus all _code_ related to IMX51 can be quickly located using git log
> --pretty=one | grep "UBUNTU: ARM: IMX51".
> 
> I've done the same to your patchset from Jaunty that I forward-ported
> onto Karmic.
> 
> Can I request you to do the same? The following command on a directory
> of patches does this easily:
> 
> sed -i -e's/<old pattern>/<new pattern>/g' 0*.patch
> 
> Regards,
> Amit
> 

I've pushed all of these patches to an 'arm' Karmic topic branch. Note
that I've changed the description prefix to 'UBUNTU: ARM: DOVE:' per
Amit's request, as well as fixed some of the overly long commit subject
lines. I'll be pushing some more changes later today that rip out build
support for non-ARM arches.

At the sprint next week we're going to have to hammer out how to manage
these ARM patch sets. As it stands, they are far too invasive for the
vanilla distro kernel (even with fencing), so we may end up managing ARM
much as we do LPIA. I'm sure this will engender no small amount of
discussion.

rtg
-- 
Tim Gardner tim.gardner at canonical.com