Which really means, make background check much more
strict, with an option to revert to the more lenient
behavior.
In this strict version, an app can't have services
started or receive broadcasts at any point when it is
not foreground. Also, it doesn't matter the importance
of a caller trying to start a service, it only depends
on the state of the app whose service is being started.
A new activity shell command allows you to control
whether to use the strict or lenient behavior.
Change-Id: I7f5a50b52881b5c8f9d8b6c8c622d3652a769fd7
Unify all locks to just one lock protecting the entire service.
There is really no need for more complicated locking -- there is
nothing in the code that can take a long time to complete. And
having a single lock will allow various parts of the code to be
much simpler and easier to maintain.
This is just the first step of the change, switching all of the
locking to use one lock. With this done, we can now start
simplifying the code. For example, JobStatus no longer needs
to do any locking (or have atomic variables and such), it can
just rely on its callers holding the global service lock.
Change-Id: I502916ed7f2994b601750c67a59a96b1a4e95c6d
The QSPanel had 0 height when shade was collapsed which messed up
the immediate expand logic.
Bug: 27197894
Change-Id: I623eb5d3b945a83b3410fa5dc52c4f96dd6700cb
This logically reverts a previous CL [1], which added
InputMethodInfo#isEncryptionAware() for File-Based Encryption (FBE)
support, since it turns out that the method in question is unnecessary
to make InputMethodManagerService encryption-aware.
[1]: Icf921fe3661eccf4a589b08b616d05decc561356
69811a98f161a04af8e8ec9978c3a5efe1ea0f29
Bug: 26279466
Change-Id: Ia4884bf5922ad453d4b9e5e3c6f0d17b36dc205d
The shell is often used for developer testing, so we shouldn't
require all of its broadcasts to be protected.
Bug: 27273532
Change-Id: Ic6f262dd11223dee5655614aad30365d51e180c5
In ag/862571 we prevent window states from been removed before the
app is stopped since it can still be rendering to the surface.
The CL also left WindowState.mExiting as true after the exit
transition animation runs. This is okay if the app finishes before
the exit animation is done, but if the exit animation finishes before
the app finishes, then we will always think we need to run an exit
animation and not remove the windows when the app and later activity
manager tries to remove the windows.
mExiting is used to mean exiting animation is running, if it is set to
true then all the code assumes an exit animation is still running and
doesn't remove the window state.
- Always set mExiting when animation is done.
- Renamed mExiting to mAnimatingExit so it is more clear what it is used
for
- Allow window state to be removed is the current surface isn't shown.
This should be save since there won't be any visual effect to the user.
- Rename WindowState.mClientRemoveRequested to WindowState.mWindowRemovalAllowed
and move setting it to true into WMS.removeWindow() so it catches all cases.
- Cleaned-up the code some to be a little clearer.
Bug: 27112965
Change-Id: I6f03d3c75b5b7728e42ceadc8703df40a3b4ae63
- Fix issue when multiple starts cause multiple spurious loads to be sent to
the native layer.
- Fix scrolling issue in the native app.
Bug: 27222043
Change-Id: Iddf36a8b68637f478dc23b001ecbde352db3a5f0
Apparently only the navigation bar is excluded when calculating
Configuration.screenLayout. Make the calculation for non-fullscreen
tasks consistent with fullscreen tasks.
Change-Id: I027e41e49ffe95245116f3d134e0bc93af0ee450
Add a new API to allow a wallpaper bitmap to be obtained for
a specific user.
Without this API, using only getWallpaperFile(..., userId), one
can't get a default wallpaper bitmap if the wallpaper is not set.
Bug: 25185253
Change-Id: Ibe1e9a49d22bee08fd4bed415573c1ee28526aea
When animation happens in a dialog, it is possible for the dialog to
be dismissed and RootRenderNode to be destroyed before we create
animation handle for the staged animators. In that case, we need
to remove the staged animators so they will not run without a
animation handle.
Bug: 26975079
Change-Id: I0c2c6c1b530beaec3984c0b1c410df4fd8f25c95
- Move ClipRectTB/LRAnimation to wm package, because that's the only
place we use it.
- Extend ClipRectTBAnimation to combine it with translation animation
so the clipping gets applied after the translation.
- Fix clip reveal transitions when a window is docked.
- Make the docked divider minimizing animations synchronized with clip
reveal animation.
Bug: 27154882
Bug: 22174716
Change-Id: If5c94c777f3b51c6f53f6f34cc261bf3439cfc88
