Revert of Force calling PostDisplayConfigurationChange() even if earyling out in UpdateDisplays

Revert of Force calling PostDisplayConfigurationChange() even if earyling out in UpdateDisplays (patchset #4 id:60001 of https://codereview.chromium.org/559213002/)

Reason for revert:
I am reverting this as it resulted in test VirtualKeyboardUsabilityExperimentTest.VirtualKeyboardWindowTest consistently crashing on Linux ChromiumOS Tests, see:

http://build.chromium.org/p/chromium.chromiumos/builders/Linux%20ChromiumOS%20Tests%20%28dbg%29%281%29/builds/31731
http://build.chromium.org/p/chromium.memory/builders/Linux%20Chromium%20OS%20ASan%20LSan%20Tests%20(1)/builds/2914

Stack trace:

Objects involved in the operation:
sequence "this" @ 0x0x3e568fbdcc90 {
}
Received signal 6
#0 0x7fcb9c8936fe base::debug::StackTrace::StackTrace()
#1 0x7fcb9c893230 base::debug::(anonymous namespace)::StackDumpSignalHandler()
#2 0x7fcb8c474cb0 \u003Cunknown>
#3 0x7fcb8bbc4425 gsignal
#4 0x7fcb8bbc7b8b abort
#5 0x7fcb8c1c75ad __gnu_debug::_Error_formatter::_M_error()
#6 0x7fcb9a306513 std::__debug::vector\u003C>::operator[]()
#7 0x7fcb9a2fbba9 ash::DisplayManager::GetCurrentDisplayIdPair()
#8 0x7fcb9a2d96cf ash::DisplayController::PostDisplayConfigurationChange()
#9 0x7fcb9a2d98bc ash::DisplayController::PostDisplayConfigurationChange()
#10 0x7fcb9a2fd22a ash::DisplayManager::UpdateDisplays()
#11 0x7fcb9a2fdbe8 ash::DisplayManager::SetDisplayRotation()
#12 0x7fcb9a33425b ash::VirtualKeyboardWindowController::FlipDisplay()
#13 0x7fcb9a3340d7 ash::VirtualKeyboardWindowController::UpdateWindow()
#14 0x7fcb9a2d92df ash::DisplayController::CreateOrUpdateNonDesktopDisplay()
#15 0x7fcb9a2d937f ash::DisplayController::CreateOrUpdateNonDesktopDisplay()
#16 0x7fcb9a300ea9 ash::(anonymous namespace)::NonDesktopDisplayUpdater::~NonDesktopDisplayUpdater()
#17 0x7fcb9a300df1 ash::DisplayManager::CreateMirrorWindowIfAny()

Original issue's description:
> Force calling PostDisplayConfigurationChange() even if earyling out in UpdateDisplays
>
> When changing from software mirroring mode to sinlge display mode, it
> is possible there is no need to update |displays_| and we early out
> UpdateDisplays(). But we still want to run the PostDisplayConfigurationChange()
> cause there are some clients need to act on this, e.g.
> TouchTransformerController needs to adjust the TouchTransformer when
> switching from dual displays to single display.
>
> BUG=chrome-os-partner:31868
> TEST=tested on Big, after existing software mirroring mode, the touch location
>      transformation is still correct.
>
> Committed: https://crrev.com/4802a8552a40e1f80606ca7171dc2f79930e7fb3
> Cr-Commit-Position: refs/heads/master@{#294481}

TBR=oshima@chromium.org,miletus@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=chrome-os-partner:31868
Committed: https://crrev.com/7f7b757e472769e495204e7fcb84118e7c203e33
Cr-Commit-Position: refs/heads/master@{#294553}

[engedy, 2014-09-12 09:06:54]

Created Revert of Force calling PostDisplayConfigurationChange() even if
earyling out in UpdateDisplays

[commit-bot: I haz the power, 2014-09-12 09:07:13]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patchset/569553002/1

[commit-bot: I haz the power, 2014-09-12 09:09:19]

Committed patchset #1 (id:1) as 40668fb744e39f64a9ea4e35a68c7bb095c4ee06

[commit-bot: I haz the power, 2014-09-12 09:12:27]

Patchset 1 (id:??) landed as
https://crrev.com/7f7b757e472769e495204e7fcb84118e7c203e33
Cr-Commit-Position: refs/heads/master@{#294553}