OSDN Git Service

android-x86/frameworks-base.git
8 years agoOnly set the task description if the colors are opaque.
Winson [Wed, 17 Feb 2016 22:53:46 +0000 (14:53 -0800)]
Only set the task description if the colors are opaque.

Bug: 27214638
Change-Id: I75e0b8871087f0b8e6acd16dc751bb6374c0f968

8 years agoFix unlock transition when docked stack is visible
Jorim Jaggi [Sat, 13 Feb 2016 02:47:06 +0000 (18:47 -0800)]
Fix unlock transition when docked stack is visible

- Keep surface of docked divider when lockscreen is showing, like
all other apps, so we don't run into synchronization issues when
unlocking. We do this by setting policy visibily to false, but
keep clientVisibility=true
- Update docked stack visibility when starting unlock sequence to
visible.

Bug: 27154882
Change-Id: I6def5c18831e23579fcfc904133749f149eaf376

8 years agoMerge "Protect MotionEvent stream when a11y changes." into nyc-dev
Phil Weaver [Wed, 17 Feb 2016 23:12:35 +0000 (23:12 +0000)]
Merge "Protect MotionEvent stream when a11y changes." into nyc-dev

8 years agoMerge "Notification: Fix tombstones and action alignment" into nyc-dev
Adrian Roos [Wed, 17 Feb 2016 23:10:25 +0000 (23:10 +0000)]
Merge "Notification: Fix tombstones and action alignment" into nyc-dev

8 years agoProtect MotionEvent stream when a11y changes.
Phil Weaver [Fri, 12 Feb 2016 20:05:42 +0000 (12:05 -0800)]
Protect MotionEvent stream when a11y changes.

I found this hole while looking for a solution to a bug when
magnification was enabled when TalkBack was running. I believe
what was happening was that enabling/disabling magnification
resets the AccessibilityInputFilter. The filter batches MotionEvents
for processing with a choreographer, but the disable/enable cycle
when magnification gets enabled could disrupt the event stream.
This change flushes the batched events before objects using them
are disabled.

I've tried a variety of things, but I'm not able to reproduce the
bug on NYC using the same procedure. I suspect that either the
timing of the events is different, or the newer gesture detection
in TalkBack is protecting it somehow. The hole remains, though,
and could create invalid streams of MotionEvents without this fix.

Bug: 26613222
Change-Id: Ie9f3da459929397596e0a80a61d5c7f191001ee8

8 years agoMerge "Fixed typo on Storage Access Framework." into nyc-dev
Felipe Leme [Wed, 17 Feb 2016 22:51:01 +0000 (22:51 +0000)]
Merge "Fixed typo on Storage Access Framework." into nyc-dev

8 years agoMerge "[framework] Don't allow apps on external storage to be active admin" into...
Makoto Onuki [Wed, 17 Feb 2016 22:48:13 +0000 (22:48 +0000)]
Merge "[framework] Don't allow apps on external storage to be active admin" into nyc-dev

8 years agoFixed typo on Storage Access Framework.
Felipe Leme [Wed, 17 Feb 2016 22:45:52 +0000 (14:45 -0800)]
Fixed typo on Storage Access Framework.

BUG: 26742218
Change-Id: I9b695ebe2176f186fc63b20f624e40dd9e257bc3

8 years agoMerge "Removed Intent.OPEN_EXTERNAL_DIRECTORY." into nyc-dev
Felipe Leme [Wed, 17 Feb 2016 22:43:19 +0000 (22:43 +0000)]
Merge "Removed Intent.OPEN_EXTERNAL_DIRECTORY." into nyc-dev

8 years agoMerge "Initial implementation of StorageManager.getVolumesList()." into nyc-dev
Felipe Leme [Wed, 17 Feb 2016 22:42:39 +0000 (22:42 +0000)]
Merge "Initial implementation of StorageManager.getVolumesList()." into nyc-dev

8 years agoMerge "SoundTriggerHelper changes for GenericSoundModels." into nyc-dev
Arunesh Mishra [Wed, 17 Feb 2016 22:42:33 +0000 (22:42 +0000)]
Merge "SoundTriggerHelper changes for GenericSoundModels." into nyc-dev

8 years agoRemoved Intent.OPEN_EXTERNAL_DIRECTORY.
Felipe Leme [Wed, 17 Feb 2016 18:12:04 +0000 (10:12 -0800)]
Removed Intent.OPEN_EXTERNAL_DIRECTORY.

Such intent is now encapsulated by StorageVolume.createAccessIntent().

BUG: 26742218

Change-Id: I2e2bd71126ecd74981f77b0af7d069f51aaece74

8 years ago[framework] Don't allow apps on external storage to be active admin
Makoto Onuki [Wed, 17 Feb 2016 19:17:15 +0000 (11:17 -0800)]
[framework] Don't allow apps on external storage to be active admin

Bug 27149287

Change-Id: I6d959d2e66dc0b19f78e6135fbdcf45ca8551958

8 years agoSoundTriggerHelper changes for GenericSoundModels.
Arunesh Mishra [Tue, 9 Feb 2016 20:15:19 +0000 (12:15 -0800)]
SoundTriggerHelper changes for GenericSoundModels.

 - Refactoring SoundTriggerHelper to handle generic sound models.
     - Ability to store multiple models, callback and state information.
     - Separate out initialization to be done per voice model, per any model
     and per generic model.
 - Minor change to the API exposed -- removing the Handler from the
   createSoundTriggerDetector call.
 - Added callback processing for onRecognitionEvent().
 - Added logic for stopAll().
 - Changes to the SoundTriggerTestApp to start/stop recognition.
     - Multiple models (3).
     - Ability to start/stop/load/unload individual models.

Bug: 22860713
Bug: 27222043
Change-Id: Ie5d811babb956bead653fb560a43f1e549ed11bd

8 years agoMerge "Fix getAcceptedIssuers" into nyc-dev
Chad Brubaker [Wed, 17 Feb 2016 22:20:35 +0000 (22:20 +0000)]
Merge "Fix getAcceptedIssuers" into nyc-dev

8 years agoMerge "Audio record notification: only report SDK recording sources" into nyc-dev
Jean-Michel Trivi [Wed, 17 Feb 2016 22:09:15 +0000 (22:09 +0000)]
Merge "Audio record notification: only report SDK recording sources" into nyc-dev

8 years agoMerge "Clean up App Standby code" into nyc-dev
Amith Yamasani [Wed, 17 Feb 2016 21:50:31 +0000 (21:50 +0000)]
Merge "Clean up App Standby code" into nyc-dev

8 years agoMerge "Fix cross-app drag and drop" into nyc-dev
Vladislav Kaznacheev [Wed, 17 Feb 2016 21:43:30 +0000 (21:43 +0000)]
Merge "Fix cross-app drag and drop" into nyc-dev

8 years agoMerge "Work on QS header redlines and animations" into nyc-dev
Jason Monk [Wed, 17 Feb 2016 21:38:59 +0000 (21:38 +0000)]
Merge "Work on QS header redlines and animations" into nyc-dev

8 years agoClean up App Standby code
Amith Yamasani [Tue, 16 Feb 2016 21:58:46 +0000 (13:58 -0800)]
Clean up App Standby code

Don't continue to check idle states for removed / stopped users.
No need to check for uninstalled packages for all users.

Bug: 27208519
Change-Id: I9a3a3bbec560dd380ce90df9fef102d7f62769ca

8 years agoMerge "Make sure at least one IME is enabled." into nyc-dev
Yohei Yukawa [Wed, 17 Feb 2016 21:36:40 +0000 (21:36 +0000)]
Merge "Make sure at least one IME is enabled." into nyc-dev

8 years agoInitial implementation of StorageManager.getVolumesList().
Felipe Leme [Tue, 9 Feb 2016 00:44:06 +0000 (16:44 -0800)]
Initial implementation of StorageManager.getVolumesList().

This change makes StorageManager.getVolumesList(),
StorageManager.getPrimaryVolume(),  and StorageVolume public and adds a
buildAccessIntent() in the latter to automatically generate the
ACTION_OPEN_EXTERNAL_DIRECTORY intent, but it doesn't change the
ACTION_OPEN_EXTERNAL_DIRECTORY implementation yet (i.e., it still takes an URI with the physical path of the directory, instead of a StorageVolume and
a directorny name).

BUG: 26742218

Change-Id: I36c59c42b6579e125ec7f03c3af141260875a491

8 years agoMerge "[ContactsContract] Add new API PhoneLookup.CONTACT_ID and DATA_ID" into nyc-dev
Tony Mak [Wed, 17 Feb 2016 21:09:37 +0000 (21:09 +0000)]
Merge "[ContactsContract] Add new API PhoneLookup.CONTACT_ID and DATA_ID" into nyc-dev

8 years ago[ContactsContract] Add new API PhoneLookup.CONTACT_ID and DATA_ID
Tony Mak [Wed, 17 Feb 2016 14:11:52 +0000 (14:11 +0000)]
[ContactsContract] Add new API PhoneLookup.CONTACT_ID and DATA_ID

It is a mistake that we put contact id in PhoneLookup._ID.
But for SIP query, the contact id is put in Data.CONTACT_ID instead.
For backward-compability issue, we can't revert the change, but to
save developer from having logic like below, PhoneLookup.CONTACT_ID is
introduced.

if (sip) {
    get contact id in Data.CONTACT_ID column
} else {
    get contact id in PhoneLookup._ID column
}

So for sip query, we put contact id in PhoneLookup.CONTACT_ID.
For other phone lookup query, we put contact id in both PhoneLookup.ID
and PhoneLookup.CONTACT_ID.

BUG:27149017

Change-Id: I4155da29bbacdda0d43164b0ce11c7224a4eee67

8 years agoMerge "Update Region parcelling to use SkRegion::Iter" into nyc-dev
Derek Sollenberger [Wed, 17 Feb 2016 21:07:41 +0000 (21:07 +0000)]
Merge "Update Region parcelling to use SkRegion::Iter" into nyc-dev

8 years agoMerge "Clean-up starting window if resuming activity in another stack" into nyc-dev
Wale Ogunwale [Wed, 17 Feb 2016 21:04:17 +0000 (21:04 +0000)]
Merge "Clean-up starting window if resuming activity in another stack" into nyc-dev

8 years agoRestore DUMP permission check for BluetoothManagerService
Andre Eisenbach [Wed, 17 Feb 2016 19:35:36 +0000 (11:35 -0800)]
Restore DUMP permission check for BluetoothManagerService

Bug: 27150536
Change-Id: I6e07f8f5f7476c6cd6dca79e2fb5e8c8412493ec

8 years agoMake sure we call reportResized exactly once when drag starting
Jorim Jaggi [Sat, 6 Feb 2016 00:54:54 +0000 (16:54 -0800)]
Make sure we call reportResized exactly once when drag starting

If there was another layout happening before the app called
relayoutWindow(), we were issuing multiple reportResized calls,
leading to multiple relayoutWindow() calls, slowing everything
down.

Change-Id: I1f3da04bb7581c655567e1d1a6fe0f8c83c0ffda

8 years agoMerge "allow for slow FrameMetricsListeners" into nyc-dev
Andres Morales [Wed, 17 Feb 2016 20:59:44 +0000 (20:59 +0000)]
Merge "allow for slow FrameMetricsListeners" into nyc-dev

8 years agoMerge "Remove REPORT_EVENT_CONTEXT_HUB gscan option" into mm-wireless-dev
Mitchell Wills [Wed, 17 Feb 2016 20:43:47 +0000 (20:43 +0000)]
Merge "Remove REPORT_EVENT_CONTEXT_HUB gscan option" into mm-wireless-dev
am: 07e27b7541

* commit '07e27b75412de52b3ba5606e35a721e2b31ff094':
  Remove REPORT_EVENT_CONTEXT_HUB gscan option

8 years agoMerge "Guard against NPE in equals checks." into nyc-dev
Steve McKay [Wed, 17 Feb 2016 20:41:28 +0000 (20:41 +0000)]
Merge "Guard against NPE in equals checks." into nyc-dev

8 years agoAudio record notification: only report SDK recording sources
Jean-Michel Trivi [Wed, 17 Feb 2016 20:30:52 +0000 (12:30 -0800)]
Audio record notification: only report SDK recording sources

Bug 22876530

Change-Id: I7575170ee2d7414c8bf9a04a9adc936b66c33e15

8 years agoGuard against NPE in equals checks.
Steve McKay [Tue, 16 Feb 2016 21:08:10 +0000 (13:08 -0800)]
Guard against NPE in equals checks.

Update State test to cover different conditions separately.
Update DirectoryInfo test to best practices.
Update RootInfo to follow DirectoryInfo equals pattern.
Also, allow null cursors from Model...since they happen...I seen it mahself.

Bug: 27118725
Change-Id: Id0a0be053b91c887a745bfca5416e64f0edba995

8 years agoMerge "Remove REPORT_EVENT_CONTEXT_HUB gscan option" into mm-wireless-dev
Mitchell Wills [Wed, 17 Feb 2016 20:30:55 +0000 (20:30 +0000)]
Merge "Remove REPORT_EVENT_CONTEXT_HUB gscan option" into mm-wireless-dev

8 years agoMerge "Remove isPackageInstalled" into nyc-dev
Fyodor Kupolov [Wed, 17 Feb 2016 20:23:54 +0000 (20:23 +0000)]
Merge "Remove isPackageInstalled" into nyc-dev

8 years agoWork on QS header redlines and animations
Jason Monk [Wed, 17 Feb 2016 20:20:29 +0000 (15:20 -0500)]
Work on QS header redlines and animations

Bug: 27201532
Change-Id: I791454dda3a42b588e506f6bcd174a69a0fd29f8

8 years agoMerge "Increase the importance of noisy notifications." into nyc-dev
Julia Reynolds [Wed, 17 Feb 2016 20:19:01 +0000 (20:19 +0000)]
Merge "Increase the importance of noisy notifications." into nyc-dev

8 years agoallow for slow FrameMetricsListeners
Andres Morales [Sat, 13 Feb 2016 02:19:52 +0000 (18:19 -0800)]
allow for slow FrameMetricsListeners

A slow listener could cause a race in the NotifyHandler
where the single reference to the buffer to send would get
updated when it shouldn't have been.

Switch to a queue of available buffers to prevent this race.

Also, stop setting and clearing the observer reference and instead
incStrong/decStrong to mark temporary strong ownership without
colliding with other owners in flight.

Bug: 27097094
Change-Id: Iee647bfae8b80019b6d8290179eed3973230901f

8 years agoMerge "Add a Settings option of "Disable automatic update"." into nyc-dev
Tao Bao [Wed, 17 Feb 2016 20:06:51 +0000 (20:06 +0000)]
Merge "Add a Settings option of "Disable automatic update"." into nyc-dev

8 years agoMerge "[RenderScript] Fix potential leak in RS JNI" into nyc-dev
Miao Wang [Wed, 17 Feb 2016 19:46:57 +0000 (19:46 +0000)]
Merge "[RenderScript] Fix potential leak in RS JNI" into nyc-dev

8 years agoMerge "TIF: Release passthrough session when HardwareSession is released" into nyc-dev
Dongwon Kang [Wed, 17 Feb 2016 19:27:58 +0000 (19:27 +0000)]
Merge "TIF: Release passthrough session when HardwareSession is released" into nyc-dev

8 years agoTIF: Release passthrough session when HardwareSession is released
Dongwon Kang [Tue, 16 Feb 2016 23:50:11 +0000 (15:50 -0800)]
TIF: Release passthrough session when HardwareSession is released

Bug: https://code.google.com/p/android/issues/detail?id=200114
Change-Id: I5d74eb56297e468b95aa0f99b159b2b54f93f5d2

8 years agoNotification: Fix tombstones and action alignment
Adrian Roos [Wed, 17 Feb 2016 19:16:14 +0000 (19:16 +0000)]
Notification: Fix tombstones and action alignment

Bug: 27166437
Change-Id: I8ddacb31b908abcfc238a1559eef22e7b18b2187

8 years agoDon't read raw root from State, use getCurrentRoot
Steve McKay [Wed, 17 Feb 2016 19:00:55 +0000 (11:00 -0800)]
Don't read raw root from State, use getCurrentRoot

Bug: 27213606
Change-Id: I9abf42672d914041ab710bd09886c0c4034b85b6

8 years agoMerge "sysui: refactor in notification panel view" into nyc-dev
Xiaohui Chen [Wed, 17 Feb 2016 19:21:40 +0000 (19:21 +0000)]
Merge "sysui: refactor in notification panel view" into nyc-dev

8 years agoClean-up starting window if resuming activity in another stack
Wale Ogunwale [Wed, 17 Feb 2016 19:16:10 +0000 (11:16 -0800)]
Clean-up starting window if resuming activity in another stack

We currently remove starting windows for initializing activities if we are
resuming another activity in the same stack. However, the starting window
can be orphaned if we are resuming an activity in a different stack and
start another activity in the stack with the starting window in the paused
state at the same time.
We now remove starting windowing for initializing activities that aren't
the top activity in the stack regardless of the stack we are resuming an
activity in.

Bug: 26659857
Change-Id: I6081d52d0c4c0d3556201346cfda7a38e4d02c8e

8 years ago[RenderScript] Fix potential leak in RS JNI
Miao Wang [Thu, 11 Feb 2016 20:32:39 +0000 (12:32 -0800)]
[RenderScript] Fix potential leak in RS JNI

Bug: 27145639
Change-Id: I80cf2971d803a3366a63caa3612a503a88639040
(cherry picked from commit a4ad5f840173ecbfb679153a4f9460d2b06538e7)

8 years agoAdding event log tags for tracking secondary user connections.
Winson [Tue, 16 Feb 2016 19:05:28 +0000 (11:05 -0800)]
Adding event log tags for tracking secondary user connections.

Bug: 27175589
Change-Id: I6025eb6b0cef0057fd006cc044a638c48b481ed7

8 years agoDon't exclude app info from recents.
Winson [Wed, 17 Feb 2016 18:35:47 +0000 (10:35 -0800)]
Don't exclude app info from recents.

Bug: 27216373
Change-Id: Idc419f8922664c618ac6ddc8287b57c89084a2a3

8 years agoFixing issue with double tap sometimes launching the same task.
Winson [Wed, 17 Feb 2016 02:11:35 +0000 (18:11 -0800)]
Fixing issue with double tap sometimes launching the same task.

- When double tapping, the next preload call can come in before activity
  start, which then resets the instance load plan (which is not
  preloaded since Recents is “on top”), thus having no launch task,
  which then falls back to launching the front-most task.  The preload
  call should never create a new plan unless it will preload it.

Change-Id: I001bfc901856c4e6b4e4b9b6a2125085b53c808b

8 years agoDisabling history
Winson [Tue, 16 Feb 2016 23:18:07 +0000 (15:18 -0800)]
Disabling history

- Fixing regression in scrolling from the back of the stack to the front
  where bindVisibleTaskViews() would be called early causing jank and
  task views to be returned to the view pool before the animation was
  complete.

Change-Id: Ib68495a2e3b34f92a4971dd6b32b7bc6c616ac23

8 years agoFix getAcceptedIssuers
Chad Brubaker [Tue, 16 Feb 2016 22:35:41 +0000 (14:35 -0800)]
Fix getAcceptedIssuers

Delegating to the TrustManagerImpl doesn't work correctly with
getAcceptedIssuers, do it in NetworkSecurityTrustManager instead.

Bug: 27124116
Change-Id: Ie527d63aaa115e6137396e07c7d134b1c42bfe87

8 years agoMerge "Fix issue with activity reuse for aliased components." into nyc-dev
Rob Carr [Wed, 17 Feb 2016 18:09:48 +0000 (18:09 +0000)]
Merge "Fix issue with activity reuse for aliased components." into nyc-dev

8 years agoUsing AppWidgetHostView in RemoteViewsAdapter instead of managing the RemoteViews...
Sunny Goyal [Thu, 4 Feb 2016 23:30:48 +0000 (15:30 -0800)]
Using AppWidgetHostView in RemoteViewsAdapter instead of managing the RemoteViews inflation itself

Change-Id: If6dd8a778096a07c58b543efe892bbffbe24098f
(cherry picked from commit 89699a2811967fc0c3ad9e06f716d408d6949c86)

8 years agoChanging hashmaps to sparseArrays and other similar conversions
Sunny Goyal [Thu, 4 Feb 2016 00:29:47 +0000 (16:29 -0800)]
Changing hashmaps to sparseArrays and other similar conversions

Change-Id: I1905120a2dc109e8f383b973e599e2c23e005d0e
(cherry picked from commit c7ccff4b202d4ff8f6ba1f21629af95954d30398)

8 years agoMerge "Expose Intent for launching screen reader tutorial." into nyc-dev
Anna Galusza [Wed, 17 Feb 2016 17:56:32 +0000 (17:56 +0000)]
Merge "Expose Intent for launching screen reader tutorial." into nyc-dev

8 years agoFix cross-app drag and drop
Vladislav Kaznacheev [Wed, 17 Feb 2016 17:51:46 +0000 (09:51 -0800)]
Fix cross-app drag and drop

Bug:27219399
Change-Id: I37199d140ca451326754a321b3a15f82c18157fa

8 years agoMerge "DO NOT MERGE Extend the DownloadManager.addCompletedDownload() method to suppo...
Edward Cunningham [Wed, 17 Feb 2016 17:50:44 +0000 (17:50 +0000)]
Merge "DO NOT MERGE Extend the DownloadManager.addCompletedDownload() method to support origin details." into nyc-dev

8 years agoMerge "Remove "HOME" directory. Use "Documents" instead." into nyc-dev
Steve McKay [Wed, 17 Feb 2016 17:46:23 +0000 (17:46 +0000)]
Merge "Remove "HOME" directory. Use "Documents" instead." into nyc-dev

8 years agoUpdate Region parcelling to use SkRegion::Iter
Derek Sollenberger [Wed, 17 Feb 2016 17:14:50 +0000 (12:14 -0500)]
Update Region parcelling to use SkRegion::Iter

Bug: 26593930
Change-Id: Ie673c776ed7e1c327ca0f3196900fbb229c2a936

8 years agoMerge "Be less aggressive in removing old remote bugreport files" into nyc-dev
Michal Karpinski [Wed, 17 Feb 2016 17:18:18 +0000 (17:18 +0000)]
Merge "Be less aggressive in removing old remote bugreport files" into nyc-dev

8 years agoMerge "Add Copy-On-Write mode to InputMethodSettings." into nyc-dev
Yohei Yukawa [Wed, 17 Feb 2016 17:11:53 +0000 (17:11 +0000)]
Merge "Add Copy-On-Write mode to InputMethodSettings." into nyc-dev

8 years agoMerge "Whitelisting remote bugreport finished broadcast" into nyc-dev
Michal Karpinski [Wed, 17 Feb 2016 16:29:19 +0000 (16:29 +0000)]
Merge "Whitelisting remote bugreport finished broadcast" into nyc-dev

8 years agoMerge "Cancel notifications when user tap on Details or Take Screenshot after service...
Felipe Leme [Wed, 17 Feb 2016 16:11:15 +0000 (16:11 +0000)]
Merge "Cancel notifications when user tap on Details or Take Screenshot after service died." into nyc-dev

8 years agoMake sure at least one IME is enabled.
Yohei Yukawa [Wed, 17 Feb 2016 15:56:46 +0000 (07:56 -0800)]
Make sure at least one IME is enabled.

It turns out that there still exists an edge case where the device can
be fallen into a state where IMMS would pick up no IME at the boot
phase.  Suppose the following case:

  Selecte IMEs  : IME X
  Enabled IMEs  : IME X, IME Y, IME Z
  Available IMEs: IME A, IME B

IMMS has already taken care of the case where the selected IME X is no
longer available at the boot phase for some reasons, which is good, but
what it does in InputMethodManagerService#chooseNewDefaultIMELocked() is
just picking up the best-looking IME only from the "Enabled IMEs"
without considering the case where all the enabled IMEs are unavailable
and we have to enable some IMEs as if it was the first boot.

This issue was found when making IMMS File-Based Encryption (FBE), where
it is relatively easy to see such a situation, e.g. only IME A and IME B
are encryption-aware, but subtle but critical details in storage layer
could cause the same situation.  Bug 6685037 is an example.

As a tentative workaround, this CL introduces a recovery logic in
IMMS#buildInputMethodListLocked() as a workaround.  In future we really
should sort out those initialization code though.

Full FBE support will be added in the next CL [1].

  [1] Ifa2225070bf8223f8964cf063c86889e312c5e9a

Bug: 26279466
Change-Id: Ibd53ef8db7afcfac681b401912876d2a52c743d5

8 years agoAdd Copy-On-Write mode to InputMethodSettings.
Yohei Yukawa [Wed, 17 Feb 2016 15:54:20 +0000 (07:54 -0800)]
Add Copy-On-Write mode to InputMethodSettings.

This is a preparation for File-Based Encryption (FBE) support in IMMS.

In order to support File-Based Encryption (FBE), IMMS needs to reset
its internal state exactly when the device is unlocked by user first
time. This is important because IMMS would recognize only
encryption-aware input method services until the the device is unlocked
by the current user.

Even if we reset the internal state when the device is unlocked by the
current user, there are still two tricky points.

 1. Except for the initial boot, IMMS uses Secure Settings to determine
    what IMEs are enabled and what IME is currently selected.  These
    persistent state may not be suitable for the situation where the
    device is not unlocked yet, because some of IMEs referenced there
    may or may not be encryption-aware.  Depending on the situations,
    we may need to enable at least one encryption-aware IME to ensure
    that the user is able to type password to unlock the device, even if
    such an IME is not Settings.Secure.ENABLED_INPUT_METHODS.  We have
    to be careful when doing this because we don't want non
    pre-installed IMEs to be enabled until the user approves it.
 2. IMMS tends to save its internal state into Secure Settings.
    However, because of the point 1, we may need to automatically enable
    a certain IME to make sure the user is able to type even when the
    device is not unlocked yet.  We don't want such a temporary state
    to be persistent in Secure Settings.

The basic idea of this CL is to implement Copy-On-Write (COW) mode in
InputMethodSettings so that we can later discard any changes made before
the device is unlocked.  As the initial step, we start using this COW
mode until the the ActivityManager becomes ready.  With this change we
can revert a previous commit [1] for Bug 6685037, where forward-locked
encrypted apks need to be taken care of an early boot phase.

  [1] Ifb311f85154beadd4787ec73669bedfdf1f5172d
      4c0e7152e74d091eb78af8baacd38287ba95a1a1

Bug: 26279466
Change-Id: I9c6f9bb3d51174198e5f73588637f87ea0d90e11

8 years agoMerge "Protect against monkey crash" into nyc-dev
Jason Monk [Wed, 17 Feb 2016 15:46:57 +0000 (15:46 +0000)]
Merge "Protect against monkey crash" into nyc-dev

8 years agoMerge "Move detail panel to its own thing" into nyc-dev
Jason Monk [Wed, 17 Feb 2016 15:45:53 +0000 (15:45 +0000)]
Merge "Move detail panel to its own thing" into nyc-dev

8 years agoMove detail panel to its own thing
Jason Monk [Tue, 16 Feb 2016 19:03:21 +0000 (14:03 -0500)]
Move detail panel to its own thing

Move pieces of the detail panel from the header and QSPanel to be
in a single container to make the enter/exit be handled better
and code cleaner.

Change-Id: Id1e79b96830ee50051e340fb09ecde537a64e516

8 years agoProtect against monkey crash
Jason Monk [Wed, 17 Feb 2016 15:39:31 +0000 (10:39 -0500)]
Protect against monkey crash

Bug: 27164002
Change-Id: Ic1b68a55a136106010b6df07766a96533683eece

8 years agoMerge "Import translations. DO NOT MERGE" into nyc-dev
Geoff Mendal [Wed, 17 Feb 2016 15:31:09 +0000 (15:31 +0000)]
Merge "Import translations. DO NOT MERGE" into nyc-dev

8 years agoMerge "Fixes and changes to support settings using UsageView" into nyc-dev
Jason Monk [Wed, 17 Feb 2016 15:30:52 +0000 (15:30 +0000)]
Merge "Fixes and changes to support settings using UsageView" into nyc-dev

8 years agoImport translations. DO NOT MERGE
Geoff Mendal [Wed, 17 Feb 2016 15:30:29 +0000 (07:30 -0800)]
Import translations. DO NOT MERGE

Change-Id: I18d91b33eaa2d759744cdf7c44f41143de0e109f
Auto-generated-cl: translation import

8 years agoMerge "Import translations. DO NOT MERGE" into nyc-dev
Geoff Mendal [Wed, 17 Feb 2016 15:28:48 +0000 (15:28 +0000)]
Merge "Import translations. DO NOT MERGE" into nyc-dev

8 years agoImport translations. DO NOT MERGE
Geoff Mendal [Wed, 17 Feb 2016 15:25:10 +0000 (07:25 -0800)]
Import translations. DO NOT MERGE

Change-Id: I01ab59200468fe5d7a8cb27a9ef16eb46d52f3fe
Auto-generated-cl: translation import

8 years agoMerge "Alarm can end time-based rule." into nyc-dev
Julia Reynolds [Wed, 17 Feb 2016 15:20:09 +0000 (15:20 +0000)]
Merge "Alarm can end time-based rule." into nyc-dev

8 years agoMerge "Suspend packages - one call for multiple packages" into nyc-dev
Andrei Stingaceanu [Wed, 17 Feb 2016 15:17:34 +0000 (15:17 +0000)]
Merge "Suspend packages - one call for multiple packages" into nyc-dev

8 years agoMerge "Import translations. DO NOT MERGE" into nyc-dev
Geoff Mendal [Wed, 17 Feb 2016 15:17:19 +0000 (15:17 +0000)]
Merge "Import translations. DO NOT MERGE" into nyc-dev

8 years agoFixes and changes to support settings using UsageView
Jason Monk [Wed, 17 Feb 2016 13:53:28 +0000 (08:53 -0500)]
Fixes and changes to support settings using UsageView

Bug: 27213136
Change-Id: I1796d710379fc92cf21b196459f34dda4570355f

8 years agoImport translations. DO NOT MERGE
Geoff Mendal [Wed, 17 Feb 2016 15:14:57 +0000 (07:14 -0800)]
Import translations. DO NOT MERGE

Change-Id: If833e6af6b9206411e02db9c081d410c4893d73e
Auto-generated-cl: translation import

8 years agoSuspend packages - one call for multiple packages
Andrei Stingaceanu [Tue, 26 Jan 2016 18:39:55 +0000 (18:39 +0000)]
Suspend packages - one call for multiple packages

Refactor setPackageSuspended into setPackagesSuspended. The rationale
is that the consumers of this API are likely to want to remove
multiple packages at once. Rather than calling the API N times, call
it just once.

The good part is that we already have the broadcast intent for
suspended packages take an array so only one broadcast. Less stress
on the system.

Another good part is that (right now) we only have one consumer of
this API and it will be easy to make changes once this CL goes in.

As a shell command, for consistency only allowed one package at
a time.

Bug: 22776761
Change-Id: Ic8b8cf64d0a288ea3a282bb7b72f9d663b3b0049

8 years agoMerge "Hide java.security.Signatue#getCurrentSpi" into nyc-dev
Narayan Kamath [Wed, 17 Feb 2016 15:13:02 +0000 (15:13 +0000)]
Merge "Hide java.security.Signatue#getCurrentSpi" into nyc-dev

8 years agoMerge "Import translations. DO NOT MERGE" into nyc-dev
Geoff Mendal [Wed, 17 Feb 2016 15:12:27 +0000 (15:12 +0000)]
Merge "Import translations. DO NOT MERGE" into nyc-dev

8 years agoImport translations. DO NOT MERGE
Geoff Mendal [Wed, 17 Feb 2016 15:11:49 +0000 (07:11 -0800)]
Import translations. DO NOT MERGE

Change-Id: Ic8d25da29dbdb3132a51bef32045c166f48b8bc6
Auto-generated-cl: translation import

8 years agoHide java.security.Signatue#getCurrentSpi
Przemyslaw Szczepaniak [Thu, 11 Feb 2016 12:43:30 +0000 (12:43 +0000)]
Hide java.security.Signatue#getCurrentSpi

Internal implementation detail, shouldn't
be public.

Bug: 26689841

(cherry picked from commit 4e559bcbf31a4f60620ed1a7a80e43f51d4dec83)

Change-Id: I86b9a3fdcab8afe580d153fac28ab85606626672

8 years agoIncrease the importance of noisy notifications.
Julia Reynolds [Wed, 17 Feb 2016 14:10:19 +0000 (09:10 -0500)]
Increase the importance of noisy notifications.

For apps with a targetSdkVersion < N.

Bug: 27213267
Change-Id: I4c589d149a28357dbbd26344e7fadb59d8b50672

8 years agoBe less aggressive in removing old remote bugreport files
Michal Karpinski [Wed, 17 Feb 2016 13:47:17 +0000 (13:47 +0000)]
Be less aggressive in removing old remote bugreport files

Always keep all the files of the remote bugreport
operation and keep them for at least a day.

Bug: 27215341
Change-Id: I514956004bf982e868a87b39c705d7c4a4a7b001

8 years agoMerge "Fix BackupTransport#getBackupQuota() javadoc." into nyc-dev
Sergey Poromov [Wed, 17 Feb 2016 13:46:19 +0000 (13:46 +0000)]
Merge "Fix BackupTransport#getBackupQuota() javadoc." into nyc-dev

8 years agoMerge "Synchronize results from runner thread to main full backup thread." into nyc-dev
Sergey Poromov [Wed, 17 Feb 2016 13:46:00 +0000 (13:46 +0000)]
Merge "Synchronize results from runner thread to main full backup thread." into nyc-dev

8 years agoAlarm can end time-based rule.
Julia Reynolds [Thu, 11 Feb 2016 18:49:08 +0000 (13:49 -0500)]
Alarm can end time-based rule.

Bug: 21069008
Change-Id: I857d85a32eb44ca873da4a9f19323c48a922ab9f

8 years agoMerge "Import translations. DO NOT MERGE" into nyc-dev
Geoff Mendal [Wed, 17 Feb 2016 13:30:05 +0000 (13:30 +0000)]
Merge "Import translations. DO NOT MERGE" into nyc-dev

8 years agoImport translations. DO NOT MERGE
Geoff Mendal [Wed, 17 Feb 2016 13:27:36 +0000 (08:27 -0500)]
Import translations. DO NOT MERGE

Change-Id: I01c7b9f2ca1bdff55f474ddf2020051cb09d4b71
Auto-generated-cl: translation import

8 years agoSynchronize results from runner thread to main full backup thread.
Sergey Poromov [Tue, 9 Feb 2016 20:13:45 +0000 (21:13 +0100)]
Synchronize results from runner thread to main full backup thread.

Previously, all results from runner thread - both for preflight
or full backup pass were ignored.
This change adds two synchronized method to get preflight result
and result of full backup pass.
This leads to better coverage of AGENT_ERROR to return in callback
as a result of backup operation.
On the other side we won't start backup pass in main thread
if preflight check hasn't been succeeded.

Change-Id: Id5f9e4c956a1bd5c396d59b7ad2098139a15e69d

8 years agoRemove references to BaseDhcpStateMachine and DhcpStateMachine
Erik Kline [Wed, 17 Feb 2016 13:06:49 +0000 (13:06 +0000)]
Remove references to BaseDhcpStateMachine and DhcpStateMachine
am: 29a215268d

* commit '29a215268dfa18f5023dc3c06b33db68e91ae316':
  Remove references to BaseDhcpStateMachine and DhcpStateMachine

8 years agoRemove references to BaseDhcpStateMachine and DhcpStateMachine
Erik Kline [Wed, 17 Feb 2016 12:44:25 +0000 (21:44 +0900)]
Remove references to BaseDhcpStateMachine and DhcpStateMachine

Bug: 26991160
Change-Id: I3742780a4e5121c163d9d1380951e25edaba19f3

8 years agoAdd IpManager#shutdown() to terminate the state machine.
Erik Kline [Wed, 17 Feb 2016 12:20:19 +0000 (12:20 +0000)]
Add IpManager#shutdown() to terminate the state machine.
am: b1ee39540e

* commit 'b1ee39540ead2e4ee74d4dba9ee1b84211766221':
  Add IpManager#shutdown() to terminate the state machine.

8 years agoSuspend packages - decline suspending the active launcher
Andrei Stingaceanu [Fri, 12 Feb 2016 18:38:19 +0000 (18:38 +0000)]
Suspend packages - decline suspending the active launcher

Bug: 26953019
Bug: 26683041
Bug: 22776761
Change-Id: I7b65abff03aa32f5db8dce68364467a588facdac

8 years agoMerge "Suspend packages - new API for retrieving the suspended status" into nyc-dev
Andrei Stingaceanu [Wed, 17 Feb 2016 11:57:50 +0000 (11:57 +0000)]
Merge "Suspend packages - new API for retrieving the suspended status" into nyc-dev

8 years agoAdd IpManager#shutdown() to terminate the state machine.
Erik Kline [Wed, 17 Feb 2016 09:02:51 +0000 (18:02 +0900)]
Add IpManager#shutdown() to terminate the state machine.

Additionally, remove IpManager.Callback#usingIpReachabilityMonitor()
now that this is now longer used.

Bug: 26991160
Change-Id: I9a17497c82238a9fb37a20d01aeca7bc4913ae2c

8 years agoMerge "Implement getDocumentStreamTypes() in DocumentsProvider." into nyc-dev
Tomasz Mikolajewski [Wed, 17 Feb 2016 11:34:11 +0000 (11:34 +0000)]
Merge "Implement getDocumentStreamTypes() in DocumentsProvider." into nyc-dev