The agent instance wasn't properly being conveyed from the generic
restore engine implementation to the state machine handling the
lifecycles. On top of that, the lifecycle wasn't advancing to the
restore-finished callback phase properly in the full-data restore
case.
Bug 22194736
Change-Id: Ic649d6a196adbd21a4a0f3083c7eed2fff383e52
Grant storage access to the Storage Access Framework provider that
offers to delegate access to storage devices. Also grant to the
default music app.
Bug: 22104872
Change-Id: I8cf7dceec8b4c1e31f8ddd40b094ccd4ecb1caf8
Just forgot to add the request to the active set.
Also eradicate a bunch of old cruft that has been replaced
by the final APIs, and improve voice interaction test to
sit fully on top of the final APIs and have a test for
command request.
Change-Id: Ieff7a6165ebf2a4c5fb80c1ebd020511a2ae63ee
Expanding was already set to false when we decided which QS
overscroll animation to run. We can't change the order of expanding
finished/tracking finished because of another bug. Instead, let the
stack scroller know about tracking state and combine both of them.
Bug: 21705230
Change-Id: I041ff1fabc6c939c233528f9d08d804ead64bbab
This is a manual reland of I3680256d41793f62def42fda00e26db1dcc9,
which was certainly merged into lmp-mr1-dev then auto-merged into
master but silently lost there for unknown reasons when
I8c42a1e6091b0fe1253e90265ac248087e was auto-merged into master.
Changes in WindowAnimator.java was already covered by
I2933eaf0ab55fe31cb382c46c411033e33a756e0 so this CL does not
include that change.
Bug: 18021493
Bug: 22158649
Change-Id: Ib1bf9f5bef44d0400230afc32231f7d1f3a1c98b
Before mounting the requested storage into place, fully unmount all
existing mounts inherited from the root namespace.
Bug: 22192518
Change-Id: I6f7f12c1219d25ee487a031d37c60fb4051d24ad
A method call exists to find all inner classes. Use it instead of
manually enumerating the possibilities and checking.
Also fix a typo.
Change-Id: I2af678e92ea0c11b53617b247d8cda324391c185
This bug caused views which used to be quick-rejected to continue to not
be drawn because the parent's DisplayList no longer contained them. The
fix is to notice when offsetting a view puts it back on screen and invalidates
the parent appropriately.
Issue #21413554 offsetTopAndBottom requires invalidating parent when view becomes visible again
Change-Id: I0f5d2cc48e3da912a41635712d9c37fb30e70c86
The beginning divider should draw before the first child that is not
set to GONE. Incidentally, there needs to be a check for such a child,
which involves checking the visibility of all the children with lower
indices. This CL takes into account the scenarios where children with
lower indices could be null, in which case we'll treat them as if they
were GONE.
Bug: 22199594
Change-Id: Iefc73ef87beec219898c967b9a3ccebd1e3b5b4c
Provides consistency with setImageDrawable(). Also updates docs and
annotations for setImageZzz() methods that can take null.
Bug: 21664964
Change-Id: Ia9b7f8ee9edb23eeea457520bea4dd261decc794
bug:19062769
Glyphs were stored to the output buffer with the assumption of
non-overlap, which is incorrect for certain fonts/strings. Instead,
blend src into dst, so new glyphs don't clobber existing content.
Change-Id: I6e22037500e67d5348ee2a43d939416c23c4d1ea
