[GIT PULL] at91: cleanup for 4.1 #4

[Olof Johansson]

On Mon, Mar 23, 2015 at 12:04:11PM +0100, Nicolas Ferre wrote:
> Arnd, Olof, Kevin,
> 
> Another cleanup pull-request for fixing some issues and stacking more PM
> cleanup.
> It is based on the previous "at91-cleanup3" tag sent last week (not inclueded
> yet).
> 
> Thanks, best regards,
> 
> The following changes since commit 8590ca655a19f9e124b52bfbf28f14eb16a05bec:
> 
>   ARM: at91: remove useless include (2015-03-16 17:03:08 +0100)
> 
> are available in the git repository at:
> 
>   git://git.kernel.org/pub/scm/linux/kernel/git/nferre/linux-at91.git tags/at91-cleanup4
> 
> for you to fetch changes up to dd905bbb09383f9f62d7a6296f11788d8c283c86:
> 
>   ARM: at91/pm: move AT91_MEMCTRL_* to pm.h (2015-03-23 11:16:00 +0100)
> 
> ----------------------------------------------------------------
> Fourth batch of cleanup for 4.1:
> - 2 issues revealed by the kbuild test robot fixed
> - move of some functions and macros into relevant files to be able to
>   streamline the at91 specific header afterwards
> 
> ----------------------------------------------------------------
> Alexandre Belloni (4):
>       ARM: at91: fix pm_suspend.S compilation when ARMv6 is selected
>       clocksource: atmel-st: select MFD_SYSCON
>       ARM: at91/pm: move the standby functions to pm.c
>       ARM: at91/pm: move AT91_MEMCTRL_* to pm.h

Having a linear cleanup branch doesn't work all that well when you mix in
driver changes like you did in the last version. I.e. I chose to merge that
branch in as a driver branch, but now you're stacking new cleanuips on top.

If the pm changes are independent of the clocksource one (which they seem to
be), then it's better to do base them on the cleanup2 branch instead of
cleanup3.

Would you mind respinning that way? We should still be able to take them
for 4.1.


-Olof