Also tidy up the bookkeeping for a few settings that were earlier
moved to Global without the redirect tables being fixed up.
Change-Id: I69275db3b2636cd6ba9c8c51b88e97d8ba4b7b7d
Bug: 7073422
This change makes @JavascriptInterface public and it requires
using this annotation to allow javascript access to public java
methods for API level JELLY_BEAN_MR1 and above. The behavior
does not change for API levels JELLY_BEAN and below.
Change-Id: I4108b17cf71b9ac273d7b61b1c8f7f5581e922ee
When building external storage paths, always include user in path
to enable cross-user paths and aid debugging.
Each Zygote process continues to only have access to the appropriate
user-specific emulated storage through bind mounts. A second set of
mounts continue supporting legacy /sdcard-style paths. For example,
a process running as owner has these mount points:
/storage/emulated_legacy
/storage/emulated_legacy/Android/obb
/storage/emulated/0
/storage/emulated/obb
Since Environment is created before Zygote forks, we need to update
its internal paths after each process launches.
Bug: 7131382
Change-Id: I6f8c6971f2a8edfb415c14cb4ed05ff97e587a21
1. The way for computing the magnified region was simplistic and
incorrect. It was ignoring window layering resulting in broken
behavior. For example, if the IME is up, then the everything else
is magnifed and the IME not. Now the keyguard appears and covers
the IME but the magnified region does not expand while it would
since the keyguard completely covers the not magnified IME window.
bug:7138937
Change-Id: I21414635aefab700ce75d40f3e913c1472cba202
Also correct some now-misleading terminology in a permission-check
log message, and fix a bug in which a system component trying to
write to a secondary user's settings would wind up writing the
owner's settings instead.
Bug 7132405
Change-Id: I5b8fafc35720390a01652e386ab5b7c0ad751abe
Indicate link up state based on flags/interface up, and not on IP address.
This is for ethernet interfaces that already exists.
Change-Id: Ib342d519c483bbb2dfa08cfac2c0c1a288cee7c0
Signed-off-by: Vishal Mahaveer <vishalm@ti.com>
-> Persisting certain appwidget options
-> Adding ability to specify appwidget options on bind
so as to avoid AppWidgetProvider update call with no
options.
Change-Id: I5631039f19f1822b8e123b559b6353c880c0192e
Dream settings are stored per-user, so dream manager operations
must act according to the calling or current user.
Bug:7041514
Change-Id: I4a0bbbd76886e6440b1afd89c61af5f4569b0e18
1. The implementation was not taking into account the transformation
matrices if the views.
2. The rectangle that was passed as an argument to
ViewParent.requestChildRectangleOnScreen was modified by
some implementations - now care is taken to prevent it.
3. The scroll of child was used when a rectangle of its coordinate
system was mapped to one in the parent system. However, the
scroll shows how much a parent has scrolled its descendants, so
the scroll of the parent has to be used not the child.
bug:7139556
Change-Id: I5b09eb7f105047e95282f74308968d5465831c84
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
We now support scaling the logical display to fit the
physical display, whatever size it is. So we can allow
adb shell am display-size to use more or less arbitrary sizes
although we do enforce an upper and lower bound to
protect the user.
Change-Id: I5fe6ba32ad1f9e4fbcd6915f7d36850b987bbcc0
(1) It's okay to write literal null as a settings element value
(2) Properly convey the user handle in the put-for-user variant
Bug 7137201
Bug 7139826
Change-Id: I0ed77d65e8377f0e0580a2668f10b7167ad34928
Bug: 7073422
Create the plumbing to use an annotation to allow access to
inherited methods of jsinterface objects. The default webview
behavior has not changed yet. However internally an a flag is
introduced to restrict javascript access to methods that have an annotation.
Change-Id: I41927248e6bc4b09d17c0707c60fe5e6ab681e66
Bug: 6961494
Don't call clearHelpers in destroy(). It is already called
in onDetach, which is where it needs to be. It can't be
called by the finalizer
Switch onCancel listener to onDismiss listener. Fixes an issue
where destroying can live the webcore thread hung because onCancel
won't be called, but onDismiss is
Change-Id: I669c389c8c9a81d92c27f88d83bcb711c6c0a5ee
Now we default to the current user instead of user 0 for most commands
(except where we can do the command for all users).
Many more commands take a user argument: force-stop, kill, profile,
dumpheap.
Improved help text.
Change-Id: I719a13b4d31b668f57ca21e51d7043ac3e0d4e1b
This CL extends the HELO and APNM packets to include the user id
of the application. This allows ddms users to differentiate between
the same app running for multiple users.
Bug 7110696.
Change-Id: I490d0e3781b8fb9db65cf81188677a0955650511
Trackball and generic motion events now pass through the IME in case
it would like to handle them before passing them on to the view
hierarchy.
While I was at it, I also...
...fixed the documentation on InputMethodService.onKeyUp()
...added documentation to InputMethodService.onTrackballEvent()
...added trackball and generic motion events to the "input" command
...fixed input consistency verification involving ACTION_OUTSIDE
Bug: 7050005
Change-Id: I40ab68df4a9542af6df25de6ec2ec500e4c02902
