Before there was only one key type supported, so we didn't need to query
a key type. Now there is DSA, EC, and RSA, so there needs to be another
argument.
(cherry picked from commit b91773bce1126d28a93f73fbef18f3a79245f24e)
Bug: 10600582
Change-Id: I5d2be3c514ff7404a90df76dc263484c5f226e6f
* Make sure the recycler is cleared when the adapter is changed.
Bug 10689596
* Don't add views to measure the baseline.
Bug 10690353
Change-Id: I841dc05652eef3319721631f96332f7900d1a631
The ChangeBounds transition causes its target view parents to suppress
layout for the duration of the transition. This is done to avoid artifacts
caused by running layout while layout bounds are being animated between
different scene values.
In order to react correctly to new transition/scene information (such as
running a new transition while another is already running), the transition
system calls pause() on all running transitions to allow layout (among other
things) to work correctly, then resume() after this is finished. This works
in general, but pause/resume do not take account of child transitions that
have already played and ended. ChangeBounds, uses pause/resume to
restore normal layout and re-enable layout suppression. This means that
when resume() is called on a ChangeBounds transition that has already
ended, that transition will leave the parent in a suppressed-layout
state, basically forever. This can cause artifacts like we're seeing in
recent ActionBar changes where the account spinner is never repopulated because
layout is never allowed to run.
The fix is to note when a transition ends and noop future calls to pause/resume.
Issue #10648936 Transitions don't work with new ActionBar functionality
Change-Id: Id9d41c0352b2270d46cfd5ad4dba130767bbf303
Use the new content provider API to canonicalize Uris.
If the provider doesn't support it, don't save the value,
unless it's a silent ringtone.
Bug: 10130785
Change-Id: Id08bb2812b9b2a7036a25801d1997661b0017629
Instead of local instance of the default HostnameVerifier, use it
directly from HttpsURLConnection. This avoids class preloading creating
an instance of it before it's necessary.
(cherry picked from commit 928ee1e48fa89302d02fdf8a8a2c7315d7195e7c)
Bug: 9984058
Change-Id: I9017256b3b5d137aeb5a6f455868af7bdfec7885
Added a hidden API under ConnectivityManager to toggle airplane mode.
This may be a temp solution for b/10653570.
bug:10653570
Change-Id: I0b2b42230073289eb8dc6891317d62b84e26c133
This API could be used for camera recording when MediaMuxer is used to write
output media file.
Bug: 10594784
Change-Id: Ide2d6e1d87b246100a5def49bfb8646dc984a512
Makes it clear that the intent and service APIs are likely to stream
audio over the network and notes that impact this has on bandwidth and
battery life.
Also clarifies that the service API should not be used for continuous
recognition as a result. The intent API only runs when the activity is
in the foreground so I didn't mention continuous recognition for it.
Bug: 10674392
Change-Id: Ib5e0c16b3124f4b7d892ef5bd4167f23662d2ca3
In cases where the client is waiting for an activity to launch
(startActivityMayWait()) it is a bad idea to clear
ActivityRecord.displayStartTime when going into the pause state. If
displayStartTime is cleared before the activity is displayed,
the client will never be released.
This fix keeps pause from clearing displayStartTime if any client
is waiting for the activity to be displayed.
Fixes bug 10095558. But not a permanent fix, startActivityMayWait()
should not be called by any production code.
Change-Id: I7cbdcb04256f4a26233867c52aedd3bc4151adc3
- Deprecates SurfaceTexture.OutOfResourcesException, it wasn't used
- Make all JNI code throw only Surface.OutOfResourcesException
- Get rid of redundant SurfaceControl.OutOfResourcesException
Bug: 10566539
Change-Id: I58126260771b9ccff6a69c672ce7719b9f98138d
Keep any previous value in the stack frame and restore when the
current call is finished.
Bug: 10659409
Change-Id: I02b760ae9ca06a4b3602725e02f649f1ada460a0
Bug: 10667740
A3D creates Allocations before knowing the underlying Type (and thus size).
Moving the mSize calculation to be conditional for concrete Allocations
eliminates the bug. In the future, we could let A3D dynamically update the
Allocation size if it is necessary to improve heap behavior.
Change-Id: I520246806b6ead0387a1a41372dade1a6e7c2271
