Initial implementation, tracking use of the vibrator, GPS,
and location reports.
Also includes an update to battery stats to also keep track of
vibrator usage (since I had to be in the vibrator code anyway
to instrument it).
The service itself is only half-done. Currently no API to
retrieve the data (which once there will allow us to show you
which apps are currently causing the GPS to run and who has
recently accessed your location), it doesn't persist its data
like it should, and no way to tell it to reject app requests
for various operations.
But hey, it's a start!
Change-Id: I05b8d76cc4a4f7f37bc758c1701f51f9e0550e15
For finishDrawingWindow queue the performLayoutAndPlaceSurfaces call
and return immediately.
For setTransparentRegionHint call the WindowStateAnimator method
immediately, removing the previous queueing of it.
Fixes bug 7174665.
Change-Id: Ia52f9a6685842220e4ffca6e214ee366470ff666
Replace two classes that did similar things in a complicated manner
with one class that does it more simply.
Bug 7064755 fixed.
Change-Id: I8c415671f60d1d2ece9da5916421f4d24aed2d65
Make app transition states easier to understand.
Remove unnecessary dependence on ActivityOptions.
Change-Id: If3942133e919a4121340f8ef5ca1c50df22f370d
systemReady() was returning before the LocationManagerService was
actually ready. Applications making LocationManager transactions
during their startup could possibly hit a race condition with the
yet-uninitialised LocationManagerService.
To guarantee that LocationManagerService is actually ready before
returning from systemReady(), we simply do the startup work on the
thread that called systemReady(), rather than spin up a secondary
thread to do the work asynchronously.
LocationWorkerHandler still needs a thread to do its work on, so
rather than have it run on the secondary thread that was
previously used for systemReady()'s work, we create a HandlerThread
for it.
Additionally, LocationManagerService.init() really needed to grab
lock for some of the things it was doing. I moved all of the code
that could benefit from mutex protection to a single section of
systemReady() and wrapped it up with a lock while I was at it.
Bug: 7723944
Change-Id: I51d480e2781622c3a14769c3a2019a2407dcfd8a
This keeps Recents from taking two identical screenshots, one for
the Recents thumbnail and one for the pause activity thumbnail.
Fixes bug 7351766.
Change-Id: Ia4d12802151666ec36e4d9b395cf10e1e02dc37f
Several problems were causing animations to jump to the end when
launching an app while a previous app was animating away.
- Keep the two app token lists in tighter synch. These were separated
when we hoped to completely separate layout form animation. This is
not as critical a goal any more.
- Use new test criteria for starting and stopping animations.
Bug 7885350 fixed.
Change-Id: Ib679117f627d0957cda17cc6ffca2bc2cdd6ecdd
- Reduce the gpu load by fading the recents thumbnail to an alpha of
0.0 before the remaining animations are completed. When alpha hits
0 the gpu treats the layer as hidden and can render the remaining
layers faster.
- Refactoring of animations to:
o Remove unused setInterpolator() calls on AnimationSet constituents.
o Remove unnecessary setFillBefore() calls.
o Consolidate setDuration() calls into AnimationSet.
o Create Interpolators once.
o Group animation set calls with their Animations.
o Use same animation timing and Interpolator for all animations.
This is a partial fix for 7729214.
Change-Id: Ic3c47bcf7c84944128effb699efcdd1f89200fc4
Bug: 7872918
This is a serious issue which the disabled system auxilialy IME is unexpectedly re-enabled by re-building internal IMI cache.
Change-Id: I0727cc973dfaea9823194021ce94af8665b98373
WifiInfo changed to return SSIDs with quotes included, so relax our
identity matching to accept matches regardless of quotedness.
Bug: 7695807
Change-Id: Ib5d06666035e06b2d4b2f7cb498767d63b1b1828
startUsingNetworkFeature will ignore errors from
reconnect - causing ConnectivityService to send
faulty information back to requester.
Change-Id: I4e0fcc1addd84da409cdc1eed1a95d25d925e020
Reduce the gpu load by fading the recents thumbnail to an alpha of
0.0 before the remaining animations are completed. When alpha hits
0 the gpu treats the layer as hidden and can merge the remaining
layers in time.
This is a partial fix for 7729214.
Change-Id: I9761bbd0554db6454c7eec0485be798b11672ff5
1. This patch takes care of the case where a magnified window is covering an unmagnigied
one. One example is a dialog that covers the IME window.
bug:7634430
2. Ensuring that the UI automator tool can connect and correctly dump the screen.
bug:7694696
3. Removed the partial implementation for multi display magnification. It adds
unnecessary complexity since it cannot be implemented without support for
input from multiple screens. We will revisit when necessary.
4. Moved the magnified border window as a surface in the window manager.
5. Moved the mediator APIs on the window manager and the policy methods on the
WindowManagerPolicy.
6. Implemented batch event processing for the accessibility input filter.
Change-Id: I4ebf68b94fb07201e124794f69611ece388ec116
Pass targetSdkVersion to installd so it knows the appropriate
permissions to apply to the app's home directory.
Bug: 7208882
Change-Id: Ia62ed36b32ee5af01077fb10a586024411be8ed4
Change the power manager to use two different kernel wakelocks
to distinguish between an application or the display keeping the
CPU alive. This may help make the output of "dumpsys batteryinfo"
easier to interpret.
Bug: 7726759
Change-Id: Iaff96ad74030d00200617b459679ea16390a8da5
