OSDN Git Service

android-x86/frameworks-base.git
9 years agoMerge "[View] Clear Hardware-Layers before system destroys HardwareRenderer​."
John Reck [Thu, 31 Jul 2014 18:28:07 +0000 (18:28 +0000)]
Merge "[View] Clear Hardware-Layers before system destroys HardwareRenderer​."

9 years agoMerge "Fix a bug that the RecentTask list of other’s may show up to non-primary users."
Craig Mautner [Thu, 31 Jul 2014 16:04:18 +0000 (16:04 +0000)]
Merge "Fix a bug that the RecentTask list of other’s may show up to non-primary users."

9 years agoFix a bug that the RecentTask list of other’s may show up to non-primary users.
Eunae Kim [Fri, 25 Jul 2014 00:16:13 +0000 (09:16 +0900)]
Fix a bug that the RecentTask list of other’s may show up to non-primary users.

When switching to a newly created user, the user may face this bug when he opens the RecentTask screen.
A possible bug scenario is described as follows:
A user id of a removed user may be recycled when created a new user.
However, mRecentTasks is not correctly controlled so that old information may still remain and be possibly mapped to wrong user.
This patch prevents this bug by explicitly removing old information in mRecentTasks when removing existing user.

Change-Id: I1874dbd604598a5d740ae1e034981e21214c15c6
Signed-off-by: Eunae Kim <eunae.kim@lge.com>
9 years agoMerge "[ActivityManager] Prevent coredump file may not complete for testing."
Craig Mautner [Thu, 31 Jul 2014 03:53:15 +0000 (03:53 +0000)]
Merge "[ActivityManager] Prevent coredump file may not complete for testing."

9 years ago[ActivityManager] Prevent coredump file may not complete for testing.
riddle_hsu [Wed, 30 Jul 2014 16:15:05 +0000 (00:15 +0800)]
[ActivityManager] Prevent coredump file may not complete for testing.

Sympton:
During testing, skip kill native crash process manually because it will continue to die by default.

Root Cause:
Large process may take some time to do coredump.In auto test, crash process will be killed immediately that results incomplete coredump file.

Solution:
If the tester (IActivityController) will handle app crash event,
Do not kill native crashed process if the rom is debuggable.

Change-Id: Ia360af147d694125d440e5ba2f958c4759a50494

9 years agoMerge "[ActivityManager] Avoid mistaking visibility by finishing task."
Craig Mautner [Thu, 31 Jul 2014 03:52:14 +0000 (03:52 +0000)]
Merge "[ActivityManager] Avoid mistaking visibility by finishing task."

9 years ago[ActivityManager] Avoid mistaking visibility by finishing task.
riddle_hsu [Wed, 30 Jul 2014 16:26:51 +0000 (00:26 +0800)]
[ActivityManager] Avoid mistaking visibility by finishing task.

Sympton:
Next activity only adds to history but does not launch/resume then results ANR.

Root Cause:
In a rare timing, some windows are switched at the same time,
it will cause some finishing records on the top temporarily,
then set startIt to false that skip to resume the real top activty.

Solution:
If all activities in a task are finishing, do not use it to check.
The behavior/checking is the same concept as in JellyBean:
            // If starting in an existing task, find where that is...
            boolean startIt = true;
            for (int i = NH-1; i >= 0; i--) {
                ActivityRecord p = mHistory.get(i);
                if (p.finishing) { // <--
                    continue;
                }

Change-Id: I9d81a7b5182400c52e173da23eee61c74692beee

9 years agoMerge "Don't assume languages are 2 letter codes."
Narayan Kamath [Thu, 31 Jul 2014 12:13:23 +0000 (12:13 +0000)]
Merge "Don't assume languages are 2 letter codes."

9 years agoDon't assume languages are 2 letter codes.
Narayan Kamath [Wed, 30 Jul 2014 14:42:25 +0000 (15:42 +0100)]
Don't assume languages are 2 letter codes.

Also, note that this method never worked. Locale
settings were stored with underscores (like Locale.toString)
but matched against AssetManager.getLocales() which
returned language-tag like output.

bug: 10090157
Change-Id: I36ffea6e39ff7d1907b2787f8fd4545c24d326a8

9 years agoMerge "Switch TextUtils over to new ICU API."
Narayan Kamath [Thu, 31 Jul 2014 11:29:56 +0000 (11:29 +0000)]
Merge "Switch TextUtils over to new ICU API."

9 years agoSwitch TextUtils over to new ICU API.
Narayan Kamath [Wed, 30 Jul 2014 15:51:27 +0000 (16:51 +0100)]
Switch TextUtils over to new ICU API.

.. and use Locale.getScript() instead of ICU.getScript.

bug: 15876704
Change-Id: Idf9462d8ef568dbc88c95e65a971184952e97872

9 years ago[View] Clear Hardware-Layers before system destroys HardwareRenderer​.
henry.uh_chen [Fri, 18 Jul 2014 07:11:18 +0000 (15:11 +0800)]
[View] Clear Hardware-Layers before system destroys HardwareRenderer​.

Symptom: Sometimes HardwareLayer display is corrupted.

Root Cause: In some cases, ViewRootImpl destroys HardwareRenderer​ without
clearing HardwareLayers in the view hirarchy. If the system trims memory fully
and kill the glContext at the same time, HardwareLayer may corrupt.

Solution: Clear Hardware-Layers before system destroys HardwareRenderer​.

Reproduce steps: Modify the cache's trim memory mechanism to make it easier to
trim fully, HardwareLayers implemented in app will easily be corrupted.
Modification of the cache is not required but make it easier to reproduce.

Change-Id: If57c802cdacf8dc37cbbcc199e74f9482834f981

9 years agoMerge "Update makefile for clang host build."
Tim Murray [Thu, 24 Jul 2014 20:59:33 +0000 (20:59 +0000)]
Merge "Update makefile for clang host build."

9 years agoMerge "Restore untrusted glStencilMask value on resume after functor"
Chris Craik [Wed, 23 Jul 2014 00:14:48 +0000 (00:14 +0000)]
Merge "Restore untrusted glStencilMask value on resume after functor"

9 years agoRestore untrusted glStencilMask value on resume after functor
Chris Craik [Tue, 22 Jul 2014 19:30:39 +0000 (12:30 -0700)]
Restore untrusted glStencilMask value on resume after functor

Fixes an issue where the stencil buffer is corrupted
(used in overdraw debugging, path clipping) when used in
conjunction with a HW accelerated WebView.

bug:15023700
bug:16465551

Change-Id: If201b503931380f7107a34e9a9a2f877a50f72b1

10 years agoMerge "[Bitmap] Add null pointer protection in Bitmap_sameAs()"
Chris Craik [Tue, 22 Jul 2014 19:08:23 +0000 (19:08 +0000)]
Merge "[Bitmap] Add null pointer protection in Bitmap_sameAs()"

10 years agoMerge "[HWUI] Fix invisible views are still shown in popup window"
Chris Craik [Tue, 22 Jul 2014 19:07:19 +0000 (19:07 +0000)]
Merge "[HWUI] Fix invisible views are still shown in popup window"

10 years agoUpdate makefile for clang host build.
Tim Murray [Thu, 17 Jul 2014 23:37:42 +0000 (16:37 -0700)]
Update makefile for clang host build.

bug 16172793

Change-Id: If7484c5dbcccce7d925bec97bff0a3e4c30e9434

10 years agoMerge "Rewriting android.text.format.Time without the native _tz functions"
Neil Fuller [Tue, 22 Jul 2014 15:26:08 +0000 (15:26 +0000)]
Merge "Rewriting android.text.format.Time without the native _tz functions"

10 years agoRewriting android.text.format.Time without the native _tz functions
Neil Fuller [Wed, 25 Jun 2014 10:13:25 +0000 (11:13 +0100)]
Rewriting android.text.format.Time without the native _tz functions

Bug: 15765976
Change-Id: I666b72ecf9da8a9dcfb97cc503006b415909a558

10 years agoMerge "Skip broadcasting to a receiver if the receiver seems to be dead"
Craig Mautner [Mon, 21 Jul 2014 15:26:58 +0000 (15:26 +0000)]
Merge "Skip broadcasting to a receiver if the receiver seems to be dead"

10 years agoSkip broadcasting to a receiver if the receiver seems to be dead
Koji Fukui [Fri, 21 Feb 2014 12:31:14 +0000 (21:31 +0900)]
Skip broadcasting to a receiver if the receiver seems to be dead

If an application that has a broadcast receiver is killed
during broadcasting, thread variable of ProcessRecord becomes null
so that IIntentReceiver#performReceive() is called
in BroadcastQueue#performReceiveLocked(). But if binder driver has not
noticed the death of the application yet, it can't throw
DeadObjectException. After that, binder driver notices. But it can't
notify DeadObjectException to the caller because performReceive() is
async call. So broadcasting keeps on waiting for finishing
performReceive() until timeout.

This change checks the death of the application before calling
performReceive() and skips broadcasting to the receiver
if the application has already died.

Change-Id: Ifa02b8b1a7e7b6fd314de90fedff5b7a5326825d

10 years agoMerge "Add support.multidex tests app for an update bug."
Yohann Roussel [Fri, 18 Jul 2014 10:54:42 +0000 (10:54 +0000)]
Merge "Add support.multidex tests app for an update bug."

10 years agoMerge "Remove copy of Locale.adjustLanguageCode."
Narayan Kamath [Tue, 22 Jul 2014 11:15:42 +0000 (11:15 +0000)]
Merge "Remove copy of Locale.adjustLanguageCode."

10 years agoRemove copy of Locale.adjustLanguageCode.
Narayan Kamath [Fri, 18 Jul 2014 15:09:36 +0000 (16:09 +0100)]
Remove copy of Locale.adjustLanguageCode.

Change-Id: Iba8c67890f95d50fadcd1245dc083d646fe6bcc1

10 years agoMerge "[HWUI]: fix residual line on FrameBuffer"
Chris Craik [Thu, 17 Jul 2014 19:28:32 +0000 (19:28 +0000)]
Merge "[HWUI]: fix residual line on FrameBuffer"

10 years agoMerge "[ActivityManager]: Update home process when home activity resumed"
Craig Mautner [Wed, 16 Jul 2014 17:17:46 +0000 (17:17 +0000)]
Merge "[ActivityManager]: Update home process when home activity resumed"

10 years ago[ActivityManager]: Update home process when home activity resumed
louis_chang [Mon, 28 Apr 2014 08:59:22 +0000 (16:59 +0800)]
[ActivityManager]: Update home process when home activity resumed

Symptom:
  When switch between two home activities, the home process might not be updated to the correct one. In that case, the home activity might be killed easily.

Root Cause:
  The home process is updated only when a home activity is newly created or being restarted. ActivityManager did not update the home process when simply resume a home activity.

Solution:
  Update home process when home activity resumed

Reproduce Steps:
  1. Install a launcher application, such as Apex Launcher
  2. Press home key to change to Apex Launcher by "Just once" option
  3. Press home key to switch back to original home activity by "Just once" option.
     (The home process is still the Apex Launcher's process because ActivityManager does not update the home process when resuming the original home activity.)

Change-Id: I046279ca7ba851a283ee67ea19202890f7b3f343

10 years agoAdd support.multidex tests app for an update bug.
Yohann Roussel [Tue, 15 Oct 2013 08:10:47 +0000 (10:10 +0200)]
Add support.multidex tests app for an update bug.

The failing scenario is
- install v1
- run
- update to v2
- update to v3
- run
last run was failing with initial multidex library versions because
it's still running on v1 extracted secondary dex files.

(cherry picked from commit 05e2a94c8b510131f43a686f5188d4c0f2a5eebd)

Change-Id: Ibb5d16642b127ee4c0baddc4f6ba461c11d25f90

10 years agoMerge "[ActivityManager] Do not finish root activity when reset task"
Craig Mautner [Wed, 16 Jul 2014 15:46:30 +0000 (15:46 +0000)]
Merge "[ActivityManager] Do not finish root activity when reset task"

10 years ago[ActivityManager] Do not finish root activity when reset task
louis_chang [Wed, 16 Jul 2014 10:15:35 +0000 (18:15 +0800)]
[ActivityManager] Do not finish root activity when reset task

Symptom:
The root activity not always located at index 0 of the task.
For example, the index 1 activity will become the new root
when the original root activity (at index 0) finished.
The new root activity might be finished unexpectedly before
the original root activity actually destroyed.

Solution:
Check frontOfTask to avoid finishing the root activity unexpectedly

Change-Id: I623ab97e9c95c83b3cfe7c9dfc151a291a391ea4

10 years agoMerge "[ActivityManager] Move top task to top in window manager"
Craig Mautner [Wed, 16 Jul 2014 15:42:50 +0000 (15:42 +0000)]
Merge "[ActivityManager] Move top task to top in window manager"

10 years ago[ActivityManager] Move top task to top in window manager
louis_chang [Wed, 16 Jul 2014 09:42:04 +0000 (17:42 +0800)]
[ActivityManager] Move top task to top in window manager

Symptom:
In certain situations, application starts activity while
it is in background. When an existing background activity
starts a new activity which results to be kept in the same
background task, the new activity won't be resumed.
In that case, the background task (sourceTask) should not be
moved to top in window manager.

Solution:
Move top task of the target stack to top in window manager

Change-Id: Id7a37ea67ce1f80e0c2b5399865c51fd7113deb8

10 years agoMerge "Removed useless test"
Narayan Kamath [Wed, 16 Jul 2014 13:31:58 +0000 (13:31 +0000)]
Merge "Removed useless test"

10 years agoMerge "Add an internal API to get all asset locales."
Narayan Kamath [Wed, 16 Jul 2014 12:51:30 +0000 (12:51 +0000)]
Merge "Add an internal API to get all asset locales."

10 years agoMerge "Optimize Blend composites."
Deepanshu Gupta [Tue, 15 Jul 2014 20:07:22 +0000 (20:07 +0000)]
Merge "Optimize Blend composites."

10 years agoMerge "Add BlendComposite.java"
Deepanshu Gupta [Tue, 15 Jul 2014 20:07:08 +0000 (20:07 +0000)]
Merge "Add BlendComposite.java"

10 years agoAdd an internal API to get all asset locales.
Narayan Kamath [Mon, 14 Jul 2014 15:28:24 +0000 (16:28 +0100)]
Add an internal API to get all asset locales.

This will allow us to not copy paste this code verbatim into
bundled apps.

bug: 10090157
Change-Id: I008dc683ecbef2ad8b7a26968cb3cbda7e5a8388

10 years agoMerge "Allow 3 letter language codes in InputMethodUtils."
Narayan Kamath [Mon, 14 Jul 2014 08:49:13 +0000 (08:49 +0000)]
Merge "Allow 3 letter language codes in InputMethodUtils."

10 years agoAllow 3 letter language codes in InputMethodUtils.
Narayan Kamath [Fri, 11 Jul 2014 10:50:24 +0000 (11:50 +0100)]
Allow 3 letter language codes in InputMethodUtils.

Replace locale.substring(0, 2) with a function that always
returns the first component of the locale (assumed to be the
language).

bug: 10090157

(cherry picked from commit cefc79c6b18bb2e824c299e7b9e212071e5ebc43)

Change-Id: I56b02dd0c6dc222daa3ceade841a30879449bb4a

10 years agoMerge "Logging in LocalTransport on as default"
Christopher Tate [Thu, 10 Jul 2014 18:51:58 +0000 (18:51 +0000)]
Merge "Logging in LocalTransport on as default"

10 years agoLogging in LocalTransport on as default
Henrik Baard [Thu, 10 Jul 2014 06:46:39 +0000 (08:46 +0200)]
Logging in LocalTransport on as default

The logging is LocalTransport.java is enabled by
default.

Change default to false to remove logs, as this class
states that it is only for debug.

Change-Id: Iae63c7b62edbd503c606f0aee671a9579e78843d

10 years agoMerge "GpsLocationProvider: look for config file in /etc/gps.<ro.hardware.gps>.conf"
Colin Cross [Thu, 10 Jul 2014 17:49:35 +0000 (17:49 +0000)]
Merge "GpsLocationProvider: look for config file in /etc/gps.<ro.hardware.gps>.conf"

10 years agoMerge "Update copyFrom(BaseObj[]) for large objects."
Stephen Hines [Thu, 10 Jul 2014 00:10:05 +0000 (00:10 +0000)]
Merge "Update copyFrom(BaseObj[]) for large objects."

10 years agoUpdate copyFrom(BaseObj[]) for large objects.
Tim Murray [Tue, 1 Jul 2014 23:56:18 +0000 (16:56 -0700)]
Update copyFrom(BaseObj[]) for large objects.

Change-Id: I99cadbd1ad925cad0dd357c9abbd1a49c07785c9

10 years agoMerge "Switch frameworks/base/libs/androidfw to the new icu."
Elliott Hughes [Wed, 9 Jul 2014 23:59:58 +0000 (23:59 +0000)]
Merge "Switch frameworks/base/libs/androidfw to the new icu."

10 years agoSwitch frameworks/base/libs/androidfw to the new icu.
Elliott Hughes [Wed, 9 Jul 2014 23:58:28 +0000 (16:58 -0700)]
Switch frameworks/base/libs/androidfw to the new icu.

Change-Id: Id82830750dd7e83bbc66811a1133b403cc697c97

10 years agoSwitch frameworks/base/core/jni to the new icu.
Elliott Hughes [Wed, 9 Jul 2014 23:56:11 +0000 (16:56 -0700)]
Switch frameworks/base/core/jni to the new icu.

Change-Id: Id12ce02da377ce78f318e10633c47f500237d9d9

10 years agoMerge "Switch frameworks/base/core/jni to the new icu."
Elliott Hughes [Wed, 9 Jul 2014 23:59:14 +0000 (23:59 +0000)]
Merge "Switch frameworks/base/core/jni to the new icu."

10 years agoOptimize Blend composites.
Deepanshu Gupta [Mon, 7 Jul 2014 21:22:19 +0000 (14:22 -0700)]
Optimize Blend composites.

Removed redundant array allocations to improve performance for various
blending modes.

Change-Id: Iaba1d6ff3ad03eebdc859c599b610cc593370438

10 years agoAdd BlendComposite.java
Deepanshu Gupta [Mon, 7 Jul 2014 17:35:15 +0000 (10:35 -0700)]
Add BlendComposite.java

The class is adapted from a demo tool for Blending Modes written by
Romain Guy (romainguy@android.com). The tool is available at
http://www.curious-creature.org/2006/09/20/new-blendings-modes-for-java2d/

Change-Id: I8f7c7ca08d3078106056764a4e2f1ce95d990137

10 years agoMerge "[ActivityManager] Prevent potential deadlock from system error dialog and...
Craig Mautner [Mon, 7 Jul 2014 21:14:33 +0000 (21:14 +0000)]
Merge "[ActivityManager] Prevent potential deadlock from system error dialog and InputMethodManager."

10 years ago[ActivityManager] Prevent potential deadlock from system error dialog and InputMethod...
riddle_hsu [Mon, 7 Jul 2014 20:30:19 +0000 (04:30 +0800)]
[ActivityManager] Prevent potential deadlock from system error dialog and InputMethodManager.

Sympton:
When application crash in a special timing and system server's InputMethodManager is binding, deadlock may occur.

Root Cause:
Thread(1): When using InputMethodManager in system server, it will lock mH(handler) of InputMethodManager,
and sometimes it will call to InputMethodManagerService::showCurrentInputLocked and will call bindService at some condition,
then it will also lock ActivityManagerService.
Thread(2): When an application crashed, it will lock ActivityManager when showing crash dialog, inside the dialog,
it will call setEnabled of Button and lock mH of InputMethodManager.setEnable of TextView will lock IMM's handler.
So the deadlock happened as the flow: (2) lock AMS -> (1)lock mH -> (2)wait mH -> (1) wait AMS

Solution:
Reduce nested lock of error dialog: post message to let (2) lock mH after release AMS lock.

Change-Id: Id85c29406236db3b5fca9655fde1fcaf0afd1337

10 years agoMerge "Remove"
Brian Carlstrom [Mon, 7 Jul 2014 20:48:48 +0000 (20:48 +0000)]
Merge "Remove"

10 years agoRemove
Brian Carlstrom [Tue, 17 Jun 2014 20:57:58 +0000 (13:57 -0700)]
Remove

(cherry picked from commit 9881203920c8445e9a4bdd9e4a90ba0014f34c5e)

Change-Id: Ib40823b7c03af4c71da902b0e5c504e074cf0dd9

10 years agoMerge "Fix issue when converting fil->tl."
Narayan Kamath [Mon, 7 Jul 2014 08:52:10 +0000 (08:52 +0000)]
Merge "Fix issue when converting fil->tl."

10 years agoFix issue when converting fil->tl.
Narayan Kamath [Sat, 5 Jul 2014 14:33:28 +0000 (15:33 +0100)]
Fix issue when converting fil->tl.

We should call ResourceTable::getLocales directly,
and not AssetManager::getLocales. The latter will convert
"tl" to "fil" so we'll end up thinking we have resources
for "fil" when we don't really have any.

bug: 15873165

Change-Id: I9753e4608aaecede328a40ee1f3ee6b016d0dedc

10 years ago[Bitmap] Add null pointer protection in Bitmap_sameAs()
henry.uh_chen [Thu, 3 Jul 2014 12:40:22 +0000 (20:40 +0800)]
[Bitmap] Add null pointer protection in Bitmap_sameAs()

Symptom: SkBitmap::getAddr(int, int) may return NULL due to unrecognized config
(ex: kRLE_Index8_Config). This will cause memcmp method to crash. Since bm0 and
bm1 both have pixel data() (have passed NULL == getPixels() check), those 2
bitmaps should be valid (only unrecognized), we return JNI_FALSE to warn user
those 2 unrecognized config bitmaps may be different.

Change-Id: I6970c27de412110a3035d0a783112c4cd3ebc35b

10 years ago[HWUI] Fix invisible views are still shown in popup window
henry.uh_chen [Thu, 3 Jul 2014 10:01:37 +0000 (18:01 +0800)]
[HWUI] Fix invisible views are still shown in popup window

Symptom: Invisible views are still shown in popup window

Root Cause: glClear isn't called if there is no draw command

Solution: Even if there is no draw command, it still calls glClear
          by startFrame().

Reproduce steps: App set one view in a popup window to invisible,
                 it will be still visible.

Change-Id: I36bc7b752434ad388fbb80ea63341ac778f48d35

10 years ago[HWUI]: fix residual line on FrameBuffer
henry.uh_chen [Wed, 2 Jul 2014 11:36:56 +0000 (19:36 +0800)]
[HWUI]: fix residual line on FrameBuffer

Symptom: If app applies animation to enlarge a bitmap, there will be an residual
         line on the screen

Root Cause: On platform which has Tile Rendering implementation (ex. Qualcomm
            CPU), startTiling() call will restrict the framebuffer region which
            GPU can affect. So the expansion of clear region by 1 will not take
            effect if startTiling region is not expanded.

Solution: Expand the startTiling region by 1, too.

Reproduce steps: Apply animation to enlarge (and then shrink) a bitmap icon.

Change-Id: I7b4a59e180daa29dbe909d9e11f4093ae1d7396f

10 years agoMerge "fix AudioTrack and AudioRecord JNI"
Narayan Kamath [Fri, 4 Jul 2014 08:41:17 +0000 (08:41 +0000)]
Merge "fix AudioTrack and AudioRecord JNI"

10 years agofix AudioTrack and AudioRecord JNI
Eric Laurent [Wed, 30 Apr 2014 18:59:21 +0000 (11:59 -0700)]
fix AudioTrack and AudioRecord JNI

AudioTrack and AudioRecord JNI should not cast
jshortArray to jbyteArray. This appeared to work with Dalvik
but causes data corruption with ART.

(cherry picked from commit 9d02848e902d04417df616354db3a18c03e639b7)

Change-Id: Ie36624d3ea06042373c64edced4b5e30b7b1ee86

10 years agoMerge "Fix LocalePicker locale filtering."
Narayan Kamath [Fri, 4 Jul 2014 08:23:08 +0000 (08:23 +0000)]
Merge "Fix LocalePicker locale filtering."

10 years agoFix LocalePicker locale filtering.
Narayan Kamath [Thu, 3 Jul 2014 12:26:07 +0000 (13:26 +0100)]
Fix LocalePicker locale filtering.

This was broken by commit b27c137087ee92674 which
removed a badly phrased check. Prior to that commit,
there was a check that excluded locales whose toString()
length was not 5.

This change reinstates that check in a less roundabout
way, by excluding BCP-47 tags that don't have a country
subtag.

bug: 16038949

Change-Id: Ic24b1df87b2e2ce47d9ae04c7759088721f27b4f

10 years agoMerge "Typo in method name: Clone"
Nick Kralevich [Sat, 5 Jul 2014 17:11:25 +0000 (17:11 +0000)]
Merge "Typo in method name: Clone"

10 years agoMerge "Fix typo in the exception message."
Nick Kralevich [Sat, 5 Jul 2014 17:07:51 +0000 (17:07 +0000)]
Merge "Fix typo in the exception message."

10 years agoMerge "Fix typo in comment of FragmentTransaction"
Nick Kralevich [Sat, 5 Jul 2014 17:05:08 +0000 (17:05 +0000)]
Merge "Fix typo in comment of FragmentTransaction"

10 years agoMerge "Added missing brace."
Nick Kralevich [Sat, 5 Jul 2014 17:00:02 +0000 (17:00 +0000)]
Merge "Added missing brace."

10 years agoMerge "Correct typo in doc comment"
Nick Kralevich [Sat, 5 Jul 2014 16:57:45 +0000 (16:57 +0000)]
Merge "Correct typo in doc comment"

10 years agoMerge "Fix typo - issue 56671"
Nick Kralevich [Sat, 5 Jul 2014 16:54:56 +0000 (16:54 +0000)]
Merge "Fix typo - issue 56671"

10 years agoMerge "fix trivial typo: s/meansure/measure"
Nick Kralevich [Sat, 5 Jul 2014 16:53:17 +0000 (16:53 +0000)]
Merge "fix trivial typo: s/meansure/measure"

10 years agoMerge "Fix a copy/paste typo in ViewPropertyAnimator"
Nick Kralevich [Sat, 5 Jul 2014 16:47:22 +0000 (16:47 +0000)]
Merge "Fix a copy/paste typo in ViewPropertyAnimator"

10 years agoMerge "Fix Issue 59989"
Nick Kralevich [Sat, 5 Jul 2014 16:43:58 +0000 (16:43 +0000)]
Merge "Fix Issue 59989"

10 years agoMerge "Fix a typo in a javadoc comment"
Nick Kralevich [Sat, 5 Jul 2014 16:41:54 +0000 (16:41 +0000)]
Merge "Fix a typo in a javadoc comment"

10 years agoMerge "Fix typo in DevicePolicyManager documentation"
Nick Kralevich [Sat, 5 Jul 2014 16:40:50 +0000 (16:40 +0000)]
Merge "Fix typo in DevicePolicyManager documentation"

10 years agoMerge "Fix a typo in NumberPicker javadoc."
Brian Duff [Thu, 3 Jul 2014 06:05:07 +0000 (06:05 +0000)]
Merge "Fix a typo in NumberPicker javadoc."

10 years agoMerge "Fix javadoc typo in Message.java: "arg1" -> "arg2""
Nick Kralevich [Sat, 5 Jul 2014 16:30:08 +0000 (16:30 +0000)]
Merge "Fix javadoc typo in Message.java: "arg1" -> "arg2""

10 years agoMerge "Typo fix in permission string"
Nick Kralevich [Sat, 5 Jul 2014 16:18:31 +0000 (16:18 +0000)]
Merge "Typo fix in permission string"

10 years agoMerge "Make misc config directory during user creation"
Robin Lee [Wed, 9 Jul 2014 17:36:21 +0000 (17:36 +0000)]
Merge "Make misc config directory during user creation"

10 years agoMerge "Make dex2oat heap size product configurable [frameworks/base]"
Brian Carlstrom [Wed, 9 Jul 2014 04:07:56 +0000 (04:07 +0000)]
Merge "Make dex2oat heap size product configurable [frameworks/base]"

10 years agoMake dex2oat heap size product configurable [frameworks/base]
Brian Carlstrom [Tue, 8 Jul 2014 17:40:59 +0000 (10:40 -0700)]
Make dex2oat heap size product configurable [frameworks/base]

Bug: 15919420
Change-Id: I9b7b4f60826fc9b0cc6bb3765ceaa36542425006

10 years agoMerge "Adds support for multi-input kernels to Frameworks/Base/RS."
Stephen Hines [Tue, 8 Jul 2014 21:18:17 +0000 (21:18 +0000)]
Merge "Adds support for multi-input kernels to Frameworks/Base/RS."

10 years agoMerge "Add Telefonica/Movistar specific APN details for dun"
Robert Greenwalt [Tue, 8 Jul 2014 16:53:49 +0000 (16:53 +0000)]
Merge "Add Telefonica/Movistar specific APN details for dun"

10 years agoAdd Telefonica/Movistar specific APN details for dun
Albert [Tue, 8 Jul 2014 04:56:08 +0000 (06:56 +0200)]
Add Telefonica/Movistar specific APN details for dun

Movistar Spain needs to add a specific APN to offer tethering
in their devices.

Change-Id: I5554ff98a5ac4c5f33b9af0083e1fac8305c9e13
Signed-off-by: Albert <alberto.crespell@gmail.com>
10 years agoAdds support for multi-input kernels to Frameworks/Base/RS.
Chris Wailes [Wed, 11 Jun 2014 19:01:28 +0000 (12:01 -0700)]
Adds support for multi-input kernels to Frameworks/Base/RS.

* Added a new JNI call to pass arrays of Allocations to the RS runtime.
* Added a new version of ForEach that takes an array of Allocations.
* Added some casts to disambiguate existing calls to forEach.

Change-Id: I46d2834c37075b2a2407fd8b010546818a4540d1

10 years agoMerge "[ActivityManager] Avoid keeping restarting home when only home activity exists."
Craig Mautner [Mon, 7 Jul 2014 16:34:06 +0000 (16:34 +0000)]
Merge "[ActivityManager] Avoid keeping restarting home when only home activity exists."

10 years agoMerge "[ActivityManager] Prevent kill a restarted process again."
Craig Mautner [Mon, 7 Jul 2014 15:29:50 +0000 (15:29 +0000)]
Merge "[ActivityManager] Prevent kill a restarted process again."

10 years ago[ActivityManager] Avoid keeping restarting home when only home activity exists.
riddle_hsu [Fri, 4 Jul 2014 09:01:56 +0000 (17:01 +0800)]
[ActivityManager] Avoid keeping restarting home when only home activity exists.

Root Cause:
When there is only home activity existed,
updating home apk will call forceStopPackageLocked to finish the activity.
Then activity history becomes empty, then home will be launched,
but its package is still target to close that results a loop.

Solution:
If home activity has been force-stopped, do not stop the same home activity again.

Change-Id: Icff12028d407873c2e6f50a06bcad231b908ccbd

10 years ago[ActivityManager] Prevent kill a restarted process again.
riddle_hsu [Fri, 4 Jul 2014 04:19:41 +0000 (12:19 +0800)]
[ActivityManager] Prevent kill a restarted process again.

Root Cause:
Use removeTask with flag ActivityManager.REMOVE_TASK_KILL_PROCESS will set waitingToKill = "remove task"
to the target process when its setSchedGroup is not BG_NONINTERACTIVE.
Later the target process may be killed when applying oom-adj due to setSchedGroup has changed to BG_NONINTERACTIVE.
If the process is needed to restart, the process record will be resued.
Then the restarted process may be killed again because its waitingToKill is not null.

Solution:
Clean waitingToKill when process is dead.

Change-Id: I5ffb5388127f4221da4c700d3f1c224f7ca6e7b2

10 years agoMerge "[ActivityManager] Ensure alive process is not killedByAm."
Craig Mautner [Mon, 7 Jul 2014 14:56:11 +0000 (14:56 +0000)]
Merge "[ActivityManager] Ensure alive process is not killedByAm."

10 years ago[ActivityManager] Ensure alive process is not killedByAm.
riddle_hsu [Fri, 4 Jul 2014 04:37:45 +0000 (12:37 +0800)]
[ActivityManager] Ensure alive process is not killedByAm.

Symptom: Unable to kill by Am again and will be skipped updating its oom adj.

Root Cause:
A restarted process will reuse original process record.
The flag killedByAm will keep previous state.

Solution:
Reset the flag killedByAm to false when the process is started.

Note: Found another similiar patch If95137d91939cc44882ad2813131bcde0edd0c1b

Change-Id: I59a86648ca8d0aed4c489d92751af120aae5ef90

10 years agoMerge "Fix broken assumptions in LocalePicker."
Narayan Kamath [Wed, 2 Jul 2014 08:09:01 +0000 (08:09 +0000)]
Merge "Fix broken assumptions in LocalePicker."

10 years agoMerge "Remove bogus 2 char limit for locales and countries."
Narayan Kamath [Wed, 2 Jul 2014 08:08:36 +0000 (08:08 +0000)]
Merge "Remove bogus 2 char limit for locales and countries."

10 years agoMerge "Fix native crashes when APKs can't be opened."
Narayan Kamath [Wed, 2 Jul 2014 12:07:57 +0000 (12:07 +0000)]
Merge "Fix native crashes when APKs can't be opened."

10 years agoMerge "Fall back to "tl" if "fil" is absent."
Narayan Kamath [Wed, 2 Jul 2014 09:54:25 +0000 (09:54 +0000)]
Merge "Fall back to "tl" if "fil" is absent."

10 years agoFall back to "tl" if "fil" is absent.
Narayan Kamath [Thu, 26 Jun 2014 15:01:28 +0000 (16:01 +0100)]
Fall back to "tl" if "fil" is absent.

For JB-MR1, there was a hack that used "tl" where
we really meant "fil" because ICU didn't have localizations
for "fil". This has now been fixed, and we now support 3 letter
language codes for AAPT so we can use "fil" where required.

For the benefit of apps that need to target older platforms,
we fall back to "tl" if the app has assets for "tl" and the
resource locale is "fil".

See bugs 72913557207176 and 8049507 for more context.

Change-Id: I1ac8502525f99b40f9091d5efd2df33518d47a41

10 years agoMerge "Add profiler type options"
Brian Carlstrom [Tue, 1 Jul 2014 22:25:58 +0000 (22:25 +0000)]
Merge "Add profiler type options"

10 years agoMerge "Allow different SELinux policies for third party apps."
Nick Kralevich [Tue, 1 Jul 2014 21:09:28 +0000 (21:09 +0000)]
Merge "Allow different SELinux policies for third party apps."

10 years agoAllow different SELinux policies for third party apps.
Robert Craig [Tue, 1 Jul 2014 20:53:11 +0000 (13:53 -0700)]
Allow different SELinux policies for third party apps.

Prior support forced all third party apps
to be resolved against the default stanza
of the mac_permissions.xml file when assigning
seinfo labels. This meant that all third party
apps, in effect, were untrusted regardless of
cert and therefore received the same selinux domain.
This also had the unfortunate side effect of forcing
certain third party apps into the wrong domains
because of shared userid requests among apps.

This patch removes that restriction and instead
allows all apps, regardless of location, to be
matched against the full mac_permissions.xml
policy file. This then allows all apps signed
with known good certs to receive the same selinux
domains of other apps with whom they share trust.

Change-Id: Iba569c046135c0e81140faf6296c5da26a243037
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
10 years agoMerge "Make sure we clean up"
Marco Nelissen [Tue, 1 Jul 2014 18:10:37 +0000 (18:10 +0000)]
Merge "Make sure we clean up"