There is this stupid fudge factor applied to window transformations
when doing a screen rotation animation. We need this when rotating,
but when not rotating it causes very visible artifacts. Historically
the non-rotation case only happened due to configuration changes, so
wasn't that big a deal. Now however that we use this when switching
users, it is more annoying. So get rid of it for such cases.
Change-Id: I6b343866c1bad9b16984b4a629917c2f1bb37b9e
The window manager now has a facility to provide a full-screen
animation, which the activity manager uses every time a user
switch happens.
The current animation is just a simple dumb slide until we get
a design from UX.
Also some cleanup: moved the portrait task animations to the
default config so we always have an animation for them, and finally
got the java symbol stuff out of public.xml.
Change-Id: I726f77422b2ef5f2d98f961f8da003e045f0ebe8
Added more complete support for logical displays with
support for mirroring, rotation and scaling.
Improved the overlay display adapter's touch interactions.
A big change here is that the display manager no longer relies
on a single-threaded model to maintain its synchronization
invariants. Unfortunately we had to change this so as to play
nice with the fact that the window manager wants to own
the surface flinger transaction around display and surface
manipulations. As a result, the display manager has to be able
to update displays from the context of any thread.
It would be nice to make this process more cooperative.
There are already several components competing to perform
surface flinger transactions including the window manager,
display manager, electron beam, overlay display window,
and mouse pointer. They are not manipulating the same surfaces
but they can collide with one another when they make global
changes to the displays.
Change-Id: I04f448594241f2004f6f3d1a81ccd12c566bf296
Cleaned up the implementation of Surface and SurfaceSession
to use more consistent naming and structure.
Added JNI for all of the new surface flinger display API calls.
Enforced the requirement that all Surfaces created by
the window manager be named.
Updated the display manager service to use the new methods.
Change-Id: I2a658f1bfd0437e1c6f9d22df8d4ffcce7284ca2
Make better use of Display object by saving it in DisplayContent.
Only use layerStack when referring to Surfaces. Get displayId from
default Display or default DisplayContent. Remove warnings.
Fixes bug 7038151.
Change-Id: Ie493f0f5e755dc9b91ee969ff561c2a098283ead
Split WindowManagerImpl into two parts, the WindowManager
interface implementation remains where it is but the global
communications with the window manager are now handled by
the WindowManagerGlobal class. This change greatly simplifies
the challenge of having separate WindowManager instances
for each Context.
Removed WindowManagerImpl.getDefault(). This represents the
bulk of this change. Most of the usages of this method were
either to perform global functions (now handled by WindowManagerGlobal)
or to obtain the default display (now handled by DisplayManager).
Explicitly associate each new window with a display and make
the Display object available to the View hierarchy.
Add stubs for some new display manager API features.
Start to split apart the concepts of display id and layer stack.
since they operate at different layers of abstraction.
While it's true that each logical display uniquely corresponds to a
surface flinger layer stack, it is not necessarily the case that
they must use the same ids. Added Display.getLayerStack()
and started using it in places where it was relatively easy to do.
Change-Id: I29ed909114dec86807c4d3a5059c3fa0358bea61
BlackSurface transparency was tracking animation transparency causing
background images to peek around the corners.
Fixes bug 4998851.
Change-Id: I48ac7bf5d0cc560b655c9f12faccda411985cbad
In the old code orientationChangeComplete was set to true on each pass
through perfomLayout. If any window was rotating the variable was set
to false on the way through the performLayout. Since we can now make
passes through performLayout before any animation step occurs we were
seeing mOrientationChangeComplete true prior to rotation completing.
This change sets mOrientationChangeComplete false at the start of a
rotation and sets it to true if we ever get through an animation step
without encountering any rotating windows.
Change-Id: I37690cf20868dfbaac94a81640bc4d9cb9fb8f00
Animation steps are now executed on a Thread launched from the
Choreographer rather than being called at the end of the WindowManager
layout process. Animations and layout are still tightly coupled in
that they share considerable state information and neither can be
executed without holding a lock on WindowServiceManager.mWindowMap.
Change-Id: Ie17d693706971507b50aa473da1b7258e9e67764
Further work to isolate layout from animation and surface operations.
Remove cruft and minor refactoring.
Change-Id: I6f910ed72c7c614996641c353870c2b2ab5e8bb4
(Dianne) pulled the animation steps out of the layout. Changes to
exposed layers cause repeated calls to layout code.
Combined animation steps into start and finish animation code.
Change-Id: I3602d1d6249d20987d102a54e3a67a7a39361b55
The black frame is now a separate animation element. This allows us
to have it move through the aspect ratio change, without the old
and new screen elements doing weird aspect ratio scalings. This
makes the animation generally look different (more like a reveal),
not sure if it is better or worse, but definitely has less artifacts
in things like the gallery.
Also tweaked the animation definitions a bit.
Change-Id: I94c274ba04bbc7a675946dc6c1778e19386e4a66
Modify test so that rotation animation is entered one final time after rotation completes. This last time causes mUpdateRotation to be set true forcing a test for orientation change.
Fixes bug 6109189.
Change-Id: I25475f336a3439a56b1c540205d09ee48c1d2b7c
If we went through the update loop multiple times, and had finished
animating the first time, we would blow away that information the
second time and never kill the animation.
Also moved killing the animation back up to the animation step --
this involves destroying surfaces and such, and so really should
be done as part of the surface transaction. We can also consider
the screen rotation animation object to be owned by the animation,
so it can destroy it when done.
Change-Id: If24356c509c66d046f2ddfd9ad5bfe12504d7716
We now have an animation that can start as soon as we detect to
rotate the screen, and have a different phase once the new UI
is ready to complete the transition into it.
Change-Id: I05b9c12a699232acbddc36b4a994a4452db71476
The ScreenShot layer is now created hidden. The screenshot itself
is aquired during the transaction when the layer is made visible.
This guarantees the screenshot and the layer happen atomically
with respect to screen updates.
Bug: 5534521
Change-Id: Ida23e1f13d5716ec83b78a15712e0646d6cf8729
A LayerScreenshot is a special type of layer that contains a screenshot of
the screen acquired when its created. It works just like LayerDim.
Make sure to call compositionComplete() after rendering into a FBO.
Bug: 5446982, 5467587, 5466259
Change-Id: I5d8a1b4c327f9973d950cd4f4c0bca7f62825cd4
A LayerScreenshot is a special type of layer that contains a screenshot of
the screen acquired when its created. It works just like LayerDim.
Bug: 5446982
Change-Id: I7814aff2380e7e146937f2b641907be2a30c76cc
The window manager now uses the app screen dimensions to compute
the various configuration properties, as it should.
This means that prime is official a "not long" device. Poor prime.
It probably feels inadequate now.
Because it is.
Oh and all that other stuff? Debugging logs. Turned off. And
why the heck not, debugging logs are great.
Change-Id: Iaaf8ef270d986d34fd046d699ef4c0ecea1981fc
The input reader needs this information so that it knows how to
interpolate touches on an external touch screen.
Changed Display so that it asks the WindowManager what the real
display size is (as opposed to the raw display size). This means
it now takes into the forced display size set by
adb shell am display-size.
Replaced all calls to getRealWidth() / getRealHeight() /
getRealMetrics() in the WindowManager and replaced them with direct
usages of the mCurDisplayWidth / mCurDisplayHeight so that the WM
doesn't end up making a reentrant Binder call into itself.
Fixed the table status bar HeightReceiver so that it updates the
height on all configuration changes since it is possible that the
display size changed independently of an external HDMI display
being plugged / unplugged.
Improved the Display class documentation to make the distinctions
betweeen the various sizes clearer.
Change-Id: I3f75de559d3ebffed532ab46c4ae52c5e7f1da2b
...for Market App iRunner
There were a lot of serious issues with how we updated (or often didn't update)
the display and resource state when switching compatibility mode in conjunction
with restarting and updating application components. This addresses everything
I could find.
Unfortunately it does *not* fix this particular app. I am starting to think this
is just an issue in the app. This change does fix a number of other problems
I could repro, such as switching the compatibility mode of an IME.
Also a few changes here and there to get rid of $#*&^!! debug logs.
Change-Id: Ib15572eac9ec93b4b9966ddcbbc830ce9dec1317
There is now an API, which is used for task switching.
Also improved how we handle rotation animation when we can't take a
screen shot, to cleanly revert to the old freeze behavior. This removes
the need to special case the emulator.
Change-Id: I7227432a2309370437ec6ac78db02c6f1e7eedd5
