Simulated trackball should not generate KeyEvent if dispatchGenericMotionEvent
returns true
b/8022205
Change-Id: I1857e25407c508c98ef4db85fe146b1e25a0803e
Bug #7146141
When non-rectangular clipping occurs in a layer the render buffer
used as the stencil buffer is not cached. If this happens on a
View's hardware layer the render buffer will live for as long
as the layer is bound to the view. When a stencil buffer is
required because of a call to Canvas.saveLayer() it will be allocated
on every frame. A future change will address this problem.
If "show GPU overdraw" is enabled, non-rectangular clips are not
supported anymore and we fall back to rectangular clips instead.
This is a limitation imposed by OpenGL ES that cannot be worked
around at this time.
This change also improves the Matrix4 implementation to easily
detect when a rect remains a rect after transform.
Change-Id: I0e69fb901792d38bc0c4ca1bf9fdb02d7db415b9
Hierarchy Viewer currently interfaces to the host via a socket
opened by ViewServer which resides in the WindowManagerService.
Since this has access to all windows, it is enabled only on
debug builds.
This CL adds necessary support to DDM to handle all the commands
required for Hierarchy Viewer. It only misses two commands that
are sent to the Window Manager (which we don't have access to
from the applications).
A future CL will remove the ViewServer functionality.
Change-Id: I1dae316a00737b0cae4e640ccc97bf9bb1d05973
The efficiency of key, trackball and generic motion event
dispatch is greatly influenced by the IME dispatch cycle.
This change simplifies the dispatch of focused input events
and avoids causing event processing to be requeued on the
handler and delayed unnecessarily.
Bug: 7984576
Change-Id: Id82624a3f32c05efe6ee5c322bd55bf2ab21525d
The efficiency of key, trackball and generic motion event
dispatch is greatly influenced by the IME dispatch cycle.
This change simplifies the dispatch of focused input events
and avoids causing event processing to be requeued on the
handler and delayed unnecessarily.
Bug: 7984576
Change-Id: Id82624a3f32c05efe6ee5c322bd55bf2ab21525d
Region clipping, using Canvas.clipPath or Canvas.clipRegion, requires
a stencil buffer to be always present. In addition, extra wiring is
required in JNI and display lists.
This change only adds the necessary JNI/C++ APIs and some extra
plumbing to start the real work on properly supporting region
clipping.
A new debug define called DEBUG_CLIP_REGIONS can be used to draw
the current clip region. It is off by default, as is region
clipping.
The default implementation of clipPath() and clipRegion(), now
in native, mimics the previous Dalvik implementation to prevent
regressions.
Change-Id: I7903e7cfd7412b9b9b622566d4dbfce7bdcec00c
- add mirrored version of the icons
- make VolumePanel respond to layout direction changes
- make AudioService propagate layout direction changes to the VolumePanel
Change-Id: Ibb884ab81641c319a9b7bea1381066f3f19581f0
1. ViewRootImpl was keeping reference to the old focused view so it can
call back the global on focus change listener when another view gets
focus. The stashed reference, however was not cleared which caused a
memory leak if the last focused view was removed from the view tree.
In general keeping additional state for the last focus in ViewRootImpl
is not a good idea since this add complexity due to additional book
keeping work that is required. The view tree already keeps track of
where the focus is and it should be the only place that holds this
data. Since focus does not change that frequently it is fine to look
up the focus since this operation is O(m) where m is the depth of the
view tree. This change removes the duplicate book keeping from
ViewRootImpl and the focus is looked up as needed.
2. ViewRootImpl was calling the global focus change callbacks when focus
is gained, lost, or transferred to another view. This was done in
*ChildFocus methods. In the case of a child losing focus, i.e. in
clearChildFocus, there was a check whether focus searh yields a view
to take focus and if so it did not call back the global focus listener
assuming the the found view will take focus (the view tree gives focus
to the first focusable when a view looses focus). This is not a correct
assumption since some views override methods called as a result of
View.requestFocus that determine what the next focused view should
be. For example, HorizontalScrollView overrides onRequestFocusInDescendants
and changes the direction of the search. In fact focus search does not
take into accound ViewGroup descendant focusability. Hence, the view found
by calling the focus search from the root is not necessarily the one
that will get focus after calling requestFocus. Actually, it is
possible that the focus search will find a view but no view will
take focus. Now the View class is responsible for calling the
global focus listeners which avoids the above problem. Also this
saves book keeping in ViewRootImpl.
bug:7962363
Change-Id: Ic95a18b364e997021f3f6bb46943559aac07d95a
This change merges debug.hwui.profiling and profiling_visualize.
Now only the first property exists and accepts the values true, false
and visual. This enables setting the property from Settings.
Change-Id: I16a11fe23ac1a6bb907a1a826e418d6b3994b56c
When profiling is enabled with debug.hwui.profile set to true,
setting debug.hwui.profile_visualizer to true will display the
profiling data directly on screen.
Change-Id: I3d5fe3f0347090815087b1cbfce66b8e76d9347b
