From mboxrd@z Thu Jan  1 00:00:00 1970
From: arnd@arndb.de (Arnd Bergmann)
Date: Sat, 26 Jul 2014 11:46:46 +0200
Subject: [GIT PULL] ARM: mvebu: SoC changes for v3.17 (round 4)
In-Reply-To: <20140725124926.GH23220@titan.lakedaemon.net>
References: <20140725124926.GH23220@titan.lakedaemon.net>
Message-ID: <4997815.jasXMfNc8s@wuerfel>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Friday 25 July 2014 08:49:26 Jason Cooper wrote:
> Here's the (late) last round of changes for mvebu.  Daniel and Thomas
> did a great job reviewing and resubmitting the cpuidle series, so that
> is the primary content of this pull request.  There are also a few small
> changes and Thomas added the URLs for publicly released datasheets for
> the Armada XP SoCs!
> 
> Note that the cpuidle series depends on some fixes in mvebu/fixes,
> therefore, mvebu/soc now depends on mvebu/fixes.  No matter how I did a
> temporary merge branch, the commit list is correct (no commits from
> mvebu/fixes), however, the diffstat includes the changes from
> mvebu/fixes (the dts changes give it away).

I suspect this is the 'backmerge' problem: You pulled the fixes branch
into the soc-cpuidle branch, so git now sees the fixes as part of your
branch. IIRC you can avoid this if you do the merge the other way round
and start a new branch on top of fixes and then pull the soc-cpuidle
branch into that. That in turn will give you an odd-looking commit
message though, so you can't do it right until someone adds support into
git-merge to handle it better.

> To remedy this, I've included the diffstat of v3.16-rc1..mvebu/fixes at
> the bottom so that you can compare.  Everything in /fixes has already
> been pulled.
> 
> Based on tags/mvebu-soc-3.17-3, please pull.

Pulled into next/soc, thanks!

	Arnd