The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/dbfea6ff41a33e73312113c90f055ce7ac35ba1d

commit dbfea6ff41a33e73312113c90f055ce7ac35ba1d
Author: sky <sky@chromium.org>
Date: Mon Apr 17 22:34:47 2017

Disables more tests for ash_unittests --mus

And disables creation of a couple of objects in ash that make use of a
class (DeviceDataManager) that is not created.

BUG= 712290 , 705715 
TEST=none
R=msw@chromium.org

Review-Url: https://codereview.chromium.org/2820103003
Cr-Commit-Position: refs/heads/master@{#465045}

[modify] https://crrev.com/dbfea6ff41a33e73312113c90f055ce7ac35ba1d/ash/shell.cc
[modify] https://crrev.com/dbfea6ff41a33e73312113c90f055ce7ac35ba1d/testing/buildbot/filters/ash_mus_unittests.filter

I'm assuming this is related to touchscreens and setting the transforms/device mappings?

Mus clients talk to InputDeviceManager instead of DeviceDataManager. There is an InputDeviceManager implementation InputDeviceClient that uses mojo to talk to DeviceDataManager in mus-ws.

This works properly in mash, or at least did. In mushrome if we are moving the display management code to mus+chrome process then we need expand the InputDeviceManager interface and associated mojom with the touchscreen functions.

This change is pretty simple, although there should probably be two mojo interfaces? First, the unprivileged interface that lets client find out about input devices that already exists. Second, the privileged interface that lets a client set touchscreen device mappings and touchscreen transforms.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46

commit 83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46
Author: sky <sky@chromium.org>
Date: Mon Apr 24 20:29:28 2017

Wires up InputDeviceClient for mash/mushrome

Classic ash gets InputDeviceClient by way of creating
OzonePlatform. Mushrome, when run from chrome --mus gets
InputDeviceClient by way of Chrome explicitly creating it. This means
in tests run in MUS mode the test harness needs to create
InputDeviceClient. WindowManager when run in MASH mode needs to create
InputDeviceClient too.

BUG= 712290 
TEST=covered by tests
R=kylechar@chromium.org

Review-Url: https://codereview.chromium.org/2837893002
Cr-Commit-Position: refs/heads/master@{#466743}

[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/mus/BUILD.gn
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/mus/test/wm_test_helper.cc
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/mus/test/wm_test_helper.h
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/mus/window_manager.cc
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/mus/window_manager.h
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/shell.cc
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/test/BUILD.gn
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/test/ash_test_helper.cc
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/ash/test/ash_test_helper.h
[modify] https://crrev.com/83b1bb154f0fcf2f3ecb5b28f1d76994d0d13b46/testing/buildbot/filters/ash_mus_unittests.filter

Both mushrome and mash now have a DeviceDataManager.