OSDN Git Service

android-x86/frameworks-base.git
11 years agoMerge "Use the newer ZoneInfoDB API."
Elliott Hughes [Fri, 26 Apr 2013 18:41:47 +0000 (18:41 +0000)]
Merge "Use the newer ZoneInfoDB API."

11 years agoUse the newer ZoneInfoDB API.
Elliott Hughes [Fri, 26 Apr 2013 18:30:54 +0000 (11:30 -0700)]
Use the newer ZoneInfoDB API.

Change-Id: I846c720ae6c262afb4fd7f8a749008c837e76cc4

11 years agoMerge "Can not mark the same content of browser page after copy"
Victoria Lease [Fri, 26 Apr 2013 17:17:39 +0000 (17:17 +0000)]
Merge "Can not mark the same content of browser page after copy"

11 years agoMerge "Human-readable POSIX capabilities for SystemServer."
Alex Klyubin [Wed, 24 Apr 2013 05:54:46 +0000 (05:54 +0000)]
Merge "Human-readable POSIX capabilities for SystemServer."

11 years agoMerge "Don't set ADDR_COMPAT_LAYOUT on the emulator"
Nick Kralevich [Tue, 23 Apr 2013 23:33:52 +0000 (23:33 +0000)]
Merge "Don't set ADDR_COMPAT_LAYOUT on the emulator"

11 years agoDon't set ADDR_COMPAT_LAYOUT on the emulator
Nick Kralevich [Tue, 23 Apr 2013 22:30:45 +0000 (15:30 -0700)]
Don't set ADDR_COMPAT_LAYOUT on the emulator

Don't work around third party app bugs on the emulator.

Change-Id: I4440fcd56b72e3a74881ddec2a469f037a060525

11 years agoMerge "Adding matching @deprecated."
Ian Rogers [Tue, 23 Apr 2013 15:13:04 +0000 (15:13 +0000)]
Merge "Adding matching @deprecated."

11 years agoHuman-readable POSIX capabilities for SystemServer.
Alex Klyubin [Fri, 19 Apr 2013 17:01:42 +0000 (10:01 -0700)]
Human-readable POSIX capabilities for SystemServer.

Change-Id: I8fecdc55b2cf0041ceed643eb48df53d22ecd600

11 years agoMerge "KeyEvent: update comment"
Michael Wright [Sun, 21 Apr 2013 05:34:28 +0000 (05:34 +0000)]
Merge "KeyEvent: update comment"

11 years agoMerge "Correct a small doc error"
Christopher Tate [Fri, 19 Apr 2013 20:37:31 +0000 (20:37 +0000)]
Merge "Correct a small doc error"

11 years agoCorrect a small doc error
Scott Kennedy [Fri, 19 Apr 2013 20:35:16 +0000 (13:35 -0700)]
Correct a small doc error

Change-Id: I2adcc169193452549e55ac10958f0e28216aeb15

11 years agoAdding matching @deprecated.
Ian Rogers [Fri, 19 Apr 2013 18:31:48 +0000 (11:31 -0700)]
Adding matching @deprecated.

Bug: 8657266
Change-Id: Idbd22dc647f11441d15da12eb23723539768cf4c

11 years agoMerge "Fixed the problem ime invisible status despite being the icon that appears...
Satoshi Kataoka [Thu, 18 Apr 2013 17:43:37 +0000 (17:43 +0000)]
Merge "Fixed the problem ime invisible status despite being the icon that appears in the statusbar."

11 years agoMerge "AndroidKeyStore: Add encrypted flag"
Kenny Root [Mon, 15 Apr 2013 22:44:41 +0000 (22:44 +0000)]
Merge "AndroidKeyStore: Add encrypted flag"

11 years agoAndroidKeyStore: Add encrypted flag
Kenny Root [Wed, 10 Apr 2013 18:30:58 +0000 (11:30 -0700)]
AndroidKeyStore: Add encrypted flag

Add the encrypted flag for the KeyPairGenerator and the KeyStore so that
applications can choose to allow entries when there is no lockscreen.

(partial cherry pick from commit 2eeda7286f3c7cb79f7eb71ae6464cad213d12a3)

Bug: 8122243
Change-Id: I5ecd9251ec79ec53a3b68c0fff8dfba10873e36e

11 years agoMerge "keystore: Add flag for blobs to be unencrypted"
Kenny Root [Mon, 15 Apr 2013 18:29:44 +0000 (18:29 +0000)]
Merge "keystore: Add flag for blobs to be unencrypted"

11 years agokeystore: Add flag for blobs to be unencrypted
Kenny Root [Wed, 10 Apr 2013 17:37:55 +0000 (10:37 -0700)]
keystore: Add flag for blobs to be unencrypted

In order to let apps use keystore more productively, make the blob
encryption optional. As more hardware-assisted keystores (i.e., hardware
that has a Keymaster HAL) come around, encrypting blobs start to make
less sense since the thing it's encrypting is usually a token and not
any raw key material.

(cherry picked from commit a3788b00bb221e20abdd42f747d2af419e0a088c)

Bug: 8122243
Change-Id: Ifc1c64743651b23a4eace208ade0176af47ea989

11 years agoMerge "Remove keystore entries when app data cleared"
Kenny Root [Mon, 15 Apr 2013 18:03:08 +0000 (18:03 +0000)]
Merge "Remove keystore entries when app data cleared"

11 years agoRemove keystore entries when app data cleared
Kenny Root [Fri, 12 Apr 2013 17:06:51 +0000 (10:06 -0700)]
Remove keystore entries when app data cleared

When an application's user data is cleared, the keystore entries need to
be cleared as well. Previously we were only clearing entries when the
application was uninstalled for all users. Now we cover the case of
multiuser as well.

(cherry picked from commit 6fd1c85d7f909580582e5ebc357564331ecef283)

Bug: 8566369
Change-Id: Id4df5e50661b676b8f6507b915764400982c01a0

11 years agoMerge "Check outInfo is not null"
Kenny Root [Mon, 15 Apr 2013 17:44:59 +0000 (17:44 +0000)]
Merge "Check outInfo is not null"

11 years agoCheck outInfo is not null
Kenny Root [Mon, 8 Apr 2013 23:05:00 +0000 (16:05 -0700)]
Check outInfo is not null

Only call keystore if outInfo is not null

(cherry picked from commit 3b9933d0c2f2b9a2e973dca296ee45a69087e31c)

Bug: 8568820
Change-Id: I764980fd396dbc637732ad290eea2bd11036ec28

11 years agoKeyEvent: update comment
Chirayu Desai [Mon, 15 Apr 2013 14:41:37 +0000 (20:11 +0530)]
KeyEvent: update comment

Change-Id: I1ea491e7e23670bf0ea83392710010b557803178

11 years agoMerge "Remove /data/system location for mac_permissions.xml."
Geremy Condra [Fri, 12 Apr 2013 01:18:51 +0000 (01:18 +0000)]
Merge "Remove /data/system location for mac_permissions.xml."

11 years agoCan not mark the same content of browser page after copy
Peter Nilsson [Thu, 7 Feb 2013 13:31:25 +0000 (14:31 +0100)]
Can not mark the same content of browser page after copy

Marking a text on the web page and then press copy works,
but trying to mark the same text again does not work.
The reason for this is that the selection never gets
cleared in webkit.
The fix, calling cleaSelection in the onDestroyActionMode.
Also added clearSelection when getting an
onConfigurationChange.

Change-Id: I59b384cb5441b6a3a05007ea7e77f9699889a87c

11 years agoMerge "Add /data/security path for mac_permissions.xml."
Geremy Condra [Fri, 5 Apr 2013 23:32:31 +0000 (23:32 +0000)]
Merge "Add /data/security path for mac_permissions.xml."

11 years agoRemove /data/system location for mac_permissions.xml.
Stephen Smalley [Fri, 5 Apr 2013 17:38:02 +0000 (13:38 -0400)]
Remove /data/system location for mac_permissions.xml.

/data/system is no longer the right location for mac_permissions.xml.
Remove it.

Change-Id: Iabefb734b40d6785013bd6f6427931939fd0648b
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
11 years agoAdd /data/security path for mac_permissions.xml.
Stephen Smalley [Fri, 5 Apr 2013 15:27:19 +0000 (11:27 -0400)]
Add /data/security path for mac_permissions.xml.

/data/security is now the preferred location for updated policy files.
Add it for mac_permissions.xml.
Separate change will remove the old /data/system location.

Change-Id: I2774661e84c858c9b698ec60e224bbbcdac732f3
Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov>
11 years agoMerge "Add new physical keyboard layouts."
Jeff Brown [Fri, 5 Apr 2013 09:39:49 +0000 (09:39 +0000)]
Merge "Add new physical keyboard layouts."

11 years agoMerge "Update list of all audio files."
Jean-Baptiste Queru [Wed, 3 Apr 2013 22:15:10 +0000 (22:15 +0000)]
Merge "Update list of all audio files."

11 years agoUpdate list of all audio files.
Jean-Baptiste Queru [Wed, 3 Apr 2013 21:42:47 +0000 (14:42 -0700)]
Update list of all audio files.

This adds a script that can generate the list from the other
makefiles.

This also regenerates the list.

Bug: 6336895
Change-Id: If1546c4a3f5b1f0d7c3ca79d29383e9843b2a89d

11 years agoMerge "Adjust the HttpResponseCache for OkHttp."
Brian Carlstrom [Wed, 3 Apr 2013 20:17:25 +0000 (20:17 +0000)]
Merge "Adjust the HttpResponseCache for OkHttp."

11 years agoAdjust the HttpResponseCache for OkHttp.
jwilson [Sat, 30 Mar 2013 13:06:48 +0000 (09:06 -0400)]
Adjust the HttpResponseCache for OkHttp.

Change-Id: I2a80f3c3f30a3d15673d797fa3d1211286f75030

11 years agoMerge "Remove keystore entries when package removed"
Kenny Root [Wed, 3 Apr 2013 18:29:24 +0000 (18:29 +0000)]
Merge "Remove keystore entries when package removed"

11 years agoMerge "MediaCodec: docs: Clarify that audio/mp4a-latm is plain AAC, not in LATM"
Andreas Huber [Wed, 3 Apr 2013 16:09:15 +0000 (16:09 +0000)]
Merge "MediaCodec: docs: Clarify that audio/mp4a-latm is plain AAC, not in LATM"

11 years agoRemove keystore entries when package removed
Kenny Root [Mon, 1 Apr 2013 22:59:59 +0000 (15:59 -0700)]
Remove keystore entries when package removed

Add a hook into PackageManagerService so that when app IDs are
completely removed, we erase all entries from keystore for those UIDs
that have gone away.

Bug: 3020069
Change-Id: Id4b1d51a5fa4c418865055635a84bebcf5b65ec8

11 years agoMerge "Fix potential deadlock between LockScreen and WindowManagerService"
Craig Mautner [Tue, 2 Apr 2013 14:59:17 +0000 (14:59 +0000)]
Merge "Fix potential deadlock between LockScreen and WindowManagerService"

11 years agoFix potential deadlock between LockScreen and WindowManagerService
Sangkyu Lee [Wed, 16 Jan 2013 05:53:17 +0000 (14:53 +0900)]
Fix potential deadlock between LockScreen and WindowManagerService

If LockScreen is enhanced using SurfaceView/GLSurfaceView,
deadlock problem between LockScreen and WindowManagerService
can occur because of IWindow.resized() callback.
And it must lead to watchdog and reset.

IWindow.resized() callback is one-way function so calling resized()
callback of a remote IWindow object is never blocked.
However, calling resized() callback of a local IWindow object
(LockScreen is running on the same system_server process)
is always blocked until resized() callback returns.
Because resized() callback of SurfaceView/GLSurfaceView can lead to
WindowManagerService.relayoutWindow() call, deadlock can occur
between relayoutWindow() and performLayoutAndPlaceSurfacesLockedInner().
(Both functions need locking mWindowMap)

So this patch simulate one-way call when calling resized() callback
of a local IWindow object.

Change-Id: I2a6a5c74ed22d8e6b7a3bea3424ff2879d227105
Signed-off-by: Sangkyu Lee <sk82.lee@lge.com>
11 years agoMerge "It helps to catch the right exception."
Jean-Baptiste Queru [Mon, 1 Apr 2013 17:26:19 +0000 (17:26 +0000)]
Merge "It helps to catch the right exception."

11 years agoAdd new physical keyboard layouts.
Olivier Gay [Wed, 20 Mar 2013 15:45:43 +0000 (16:45 +0100)]
Add new physical keyboard layouts.

Add support for the following physical keyboard layouts:
English US international and Brazilian layout.

Change-Id: Iae928713a94611de18a92f6959adf7fa3500b2b2
Signed-off-by: Olivier Gay <ogay@logitech.com>
Signed-off-by: Myriam Galaret <mgalaret@logitech.com>
11 years agoMerge "Clarify the documentation of SparseArray.indexOfValue."
Elliott Hughes [Fri, 29 Mar 2013 23:25:07 +0000 (23:25 +0000)]
Merge "Clarify the documentation of SparseArray.indexOfValue."

11 years agoMerge "KeyStore: add API to query storage type"
Kenny Root [Fri, 29 Mar 2013 21:09:24 +0000 (21:09 +0000)]
Merge "KeyStore: add API to query storage type"

11 years agoKeyStore: add API to query storage type
Kenny Root [Fri, 29 Mar 2013 18:14:17 +0000 (11:14 -0700)]
KeyStore: add API to query storage type

Add an API to keystore daemon to query what kind of storage is currently
in use.

Change-Id: I5a83ae92250ca63b691dcf1beb8b3e1703797745

11 years agoMerge "AndroidKeyStore: add Builder for param spec"
Kenny Root [Thu, 28 Mar 2013 23:13:28 +0000 (23:13 +0000)]
Merge "AndroidKeyStore: add Builder for param spec"

11 years agoAndroidKeyStore: add Builder for param spec
Kenny Root [Thu, 28 Mar 2013 22:05:03 +0000 (15:05 -0700)]
AndroidKeyStore: add Builder for param spec

Change-Id: I13403197e1ac7ac607efa10979eb73bde0135a2a

11 years agoMerge "Save off the seinfo value with packages.list."
Geremy Condra [Thu, 28 Mar 2013 21:45:22 +0000 (21:45 +0000)]
Merge "Save off the seinfo value with packages.list."

11 years agoMerge "Add ability to install credentials as other UID"
Kenny Root [Thu, 28 Mar 2013 20:56:58 +0000 (20:56 +0000)]
Merge "Add ability to install credentials as other UID"

11 years agoMerge "Correct executable bit for source files [Take 2]"
Elliott Hughes [Thu, 28 Mar 2013 20:50:34 +0000 (20:50 +0000)]
Merge "Correct executable bit for source files [Take 2]"

11 years agoMerge "Add data validation on seinfo labels."
Geremy Condra [Thu, 28 Mar 2013 20:30:36 +0000 (20:30 +0000)]
Merge "Add data validation on seinfo labels."

11 years agoAdd ability to install credentials as other UID
Kenny Root [Thu, 28 Mar 2013 16:25:51 +0000 (09:25 -0700)]
Add ability to install credentials as other UID

We need the ability to install from the system UID to wifi UID
to explicitly bind WiFi credentials to the WiFi profile. This adds the
ability for Wifi Settings to invoke installation of a PKCS12 file for
the wifi UID.

Bug: 8183258
Change-Id: I652b7e6fa93deda6d6d310be33f224e5a356c787

11 years agoAdd data validation on seinfo labels.
Robert Craig [Thu, 28 Mar 2013 10:22:12 +0000 (06:22 -0400)]
Add data validation on seinfo labels.

Ensure that policy contains a clean seinfo
string. Where clean means no whitespace characters.

Change-Id: I814411cbc8d16eaed99a1389f5487529e36e617b
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
11 years agoMerge "Add interrupter library to simulate EINTR"
Kenny Root [Thu, 28 Mar 2013 00:16:30 +0000 (00:16 +0000)]
Merge "Add interrupter library to simulate EINTR"

11 years agoCorrect executable bit for source files [Take 2]
Chirayu Desai [Wed, 27 Mar 2013 11:22:35 +0000 (16:52 +0530)]
Correct executable bit for source files [Take 2]

Change Ieb51bafb46c895a21d2e83696f5a901ba752b2c5 left out some
files, this fixes them.

Change-Id: Ia949a8581668836ea0251602e048f09c852f5169

11 years agoMerge "Two REMOVE_PROVIDER messages caused by race condition."
Dianne Hackborn [Wed, 27 Mar 2013 00:17:41 +0000 (00:17 +0000)]
Merge "Two REMOVE_PROVIDER messages caused by race condition."

11 years agoTwo REMOVE_PROVIDER messages caused by race condition.
Guobin Zhang [Thu, 21 Mar 2013 05:57:11 +0000 (13:57 +0800)]
Two REMOVE_PROVIDER messages caused by race condition.

Fix a bug in unstable ContentProvider.
IllegalStateException: ref counts can't go to zero here: stable=0 unstable=0
IllegalStateException: unstable count < 0: -1

There is a race between main thread and background database thread. Main thread
is responsible for handling the REMOVE_PROVIDER message. Database thread starts
insert or query request again and again. acquireProvider in db thread will often
snatch provider from the jaws of death, sometime it fails to remove REMOVE_PROVIDER
which is already fired out from MessageQueue. But completeRemoveProvider in main
thread gets suspended when trying to execute the critical section. If db thread
released the provider before main thread resumes the execution, then two
REMOVE_PROVIDER messages will be executed.

Change-Id: I8588aa1d1a8bc444dcd2adf6f8bc3f055cebbdc4
Signed-off-by: Guobin Zhang <guobin.zhang@intel.com>
11 years agoClarify the documentation of SparseArray.indexOfValue.
Elliott Hughes [Tue, 26 Mar 2013 23:34:30 +0000 (16:34 -0700)]
Clarify the documentation of SparseArray.indexOfValue.

Bug: https://code.google.com/p/android/issues/detail?id=53297
Change-Id: I28725a2d03d4c11948bfe4fb90d730c86bf41481

11 years agoMerge "Remove dead DateUtils code that was never in an API level."
Elliott Hughes [Tue, 26 Mar 2013 22:25:22 +0000 (22:25 +0000)]
Merge "Remove dead DateUtils code that was never in an API level."

11 years agoRemove dead DateUtils code that was never in an API level.
Elliott Hughes [Tue, 26 Mar 2013 22:11:28 +0000 (15:11 -0700)]
Remove dead DateUtils code that was never in an API level.

Also rewrite @deprecated comments to use an initial capital so they
look right in the generated documentation.

Also improve formatIpAddress documentation.

Change-Id: I317ebe411ef76e16a4535318ce73e00b63af38da

11 years agoMerge "Remove the fixed-length buffer in DdmHandleNativeHeap.getLeakInfo."
Elliott Hughes [Tue, 26 Mar 2013 21:39:51 +0000 (21:39 +0000)]
Merge "Remove the fixed-length buffer in DdmHandleNativeHeap.getLeakInfo."

11 years agoRemove the fixed-length buffer in DdmHandleNativeHeap.getLeakInfo.
Elliott Hughes [Tue, 26 Mar 2013 20:18:52 +0000 (13:18 -0700)]
Remove the fixed-length buffer in DdmHandleNativeHeap.getLeakInfo.

It was possible to overrun this buffer, and even if you did fill
the buffer in a single read and exit the loop, the "string" still
wouldn't be terminated.

Bug: 8468088
Change-Id: Ia19f4d26dfd79e7b63807a7ec9540b932163d246

11 years agoIt helps to catch the right exception.
Jeff Sharkey [Tue, 26 Mar 2013 17:29:01 +0000 (10:29 -0700)]
It helps to catch the right exception.

Bug: 7238149
Change-Id: I7b60ec44d1d1d5ece8a871cb089f811b611f4a4a

11 years agoSave off the seinfo value with packages.list.
Robert Craig [Tue, 26 Mar 2013 12:21:37 +0000 (08:21 -0400)]
Save off the seinfo value with packages.list.

Patch adds the seinfo label per package to the file.
This is of particular interest to the run-as program
which uses the seinfo tag to correctly label the
app security context before running the shell.

Change-Id: I9d7ea47c920b1bc09a19008345ed7fd0aa426e87
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
11 years agoMerge "BackupManagerService files need new security labeling."
Geremy Condra [Tue, 26 Mar 2013 05:10:37 +0000 (05:10 +0000)]
Merge "BackupManagerService files need new security labeling."

11 years agoMerge "Proper security labeling of multi-user data directories."
Geremy Condra [Mon, 25 Mar 2013 21:17:46 +0000 (21:17 +0000)]
Merge "Proper security labeling of multi-user data directories."

11 years agoMerge "Add seinfo parsing to PackageManagerService."
Geremy Condra [Mon, 25 Mar 2013 21:07:37 +0000 (21:07 +0000)]
Merge "Add seinfo parsing to PackageManagerService."

11 years agoProper security labeling of multi-user data directories.
rpcraig [Wed, 28 Nov 2012 13:37:06 +0000 (08:37 -0500)]
Proper security labeling of multi-user data directories.

This patch covers 2 cases. When an app is installed
and the resulting data directory is created for all
existing users. And when a new user is created and
all existing app data directories are created for
the new user.

Change-Id: Iacaba6d9d18d5337e65713960d14efe32006b330
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
11 years agoAdd seinfo parsing to PackageManagerService.
Robert Craig [Mon, 25 Mar 2013 10:33:03 +0000 (06:33 -0400)]
Add seinfo parsing to PackageManagerService.

This patch set allows the PMS to parse the
mac_permissions.xml file which contains the
seinfo values. Each package that is installed
on the device will be assigned an seinfo value
based on policy. This seinfo value will help label
the app process and data directory.  Modifications
include adjustments to ApplicationInfo.java
to store the seinfo tag per package as well as
adjustments to installd to communicate the seinfo
tag to libselinux.

Change-Id: I61ad1ea12fb6a9a6d0b108ec163bc4bf4c954b58
Signed-off-by: rpcraig <rpcraig@tycho.ncsc.mil>
11 years agoMerge "KeyStore: change migrate to duplicate"
Kenny Root [Thu, 21 Mar 2013 21:56:53 +0000 (21:56 +0000)]
Merge "KeyStore: change migrate to duplicate"

11 years agoKeyStore: change migrate to duplicate
Kenny Root [Thu, 21 Mar 2013 21:21:50 +0000 (14:21 -0700)]
KeyStore: change migrate to duplicate

After discussion, it was determined that duplicate would be less
disruptive and it still fit in the current HAL model.

Change-Id: I2f9cae48d38ec7146511e876450fa39fc92cda55

11 years agoMerge "Fix Watchdog HeartbeatHandler to run on correct thread"
Wink Saville [Thu, 21 Mar 2013 04:49:28 +0000 (04:49 +0000)]
Merge "Fix Watchdog HeartbeatHandler to run on correct thread"

11 years agoMerge "Update SELinux JNI to use helpers"
Kenny Root [Wed, 20 Mar 2013 22:26:03 +0000 (22:26 +0000)]
Merge "Update SELinux JNI to use helpers"

11 years agoUpdate SELinux JNI to use helpers
Kenny Root [Fri, 19 Oct 2012 21:15:26 +0000 (14:15 -0700)]
Update SELinux JNI to use helpers

Change-Id: Ifc707eafd1ecbba64c93fa2250c4334d0ce40005

11 years agoMerge changes I65f7a915,I32098dc0
Kenny Root [Wed, 20 Mar 2013 22:00:10 +0000 (22:00 +0000)]
Merge changes I65f7a915,I32098dc0

* changes:
  KeyStore: add "migrate" command
  KeyStore: add API to uid versions

11 years agoKeyStore: add "migrate" command
Kenny Root [Wed, 20 Mar 2013 18:36:50 +0000 (11:36 -0700)]
KeyStore: add "migrate" command

To support the WiFi service, we need to support migration from the
system UID to the wifi UID. This adds a command to achieve the
migration.

Bug: 8122243
Change-Id: I65f7a91504c1d2a2aac22b9c3051adffd28d66c1

11 years agoKeyStore: add API to uid versions
Kenny Root [Thu, 14 Feb 2013 01:02:57 +0000 (17:02 -0800)]
KeyStore: add API to uid versions

In previous commits, we added the ability to specify which UID we want to
target on certain operations. This commit adds the ability to reach those
binder calls from the KeyStore class.

Also fix a problem where saw() was not reading all the values returned via
the Binder call. This changes the semantics to return a null instead of
failing silently when it's not possible to search.

Change-Id: I32098dc0eb42e09ace89f6b7455766842a72e9f4

11 years agoFix Watchdog HeartbeatHandler to run on correct thread
John Michelau [Mon, 18 Mar 2013 23:28:23 +0000 (18:28 -0500)]
Fix Watchdog HeartbeatHandler to run on correct thread

The HeartbeatHandler for the System Server Watchdog has been running
on the wrong thread due to a race condition in initialization.  It's
designed to run on ServerThread, so that it can catch lockups in the
main looper of the System Server.  It has been running on
ActivityManagerThread instead, so it does not detect lockups on the
ServerThread as it should.

ActivityManagerService is calling Watchdog.getInstance() before
ServerThread calls Watchdog.getInstance().init(), so the handler is
being bound to the ActivityManagerThread instead of the ServerThread.

Explicitly bind HeartbeatHandler to ServerThread, so that the Watchdog
catches lockups on this critical thread.

Change-Id: Iccb184ac3adb817feb86ed4ee0e50e443bf74636

11 years agoMerge "Remove dead code from DateUtils."
Elliott Hughes [Tue, 19 Mar 2013 18:45:54 +0000 (18:45 +0000)]
Merge "Remove dead code from DateUtils."

11 years agoRemove dead code from DateUtils.
Elliott Hughes [Tue, 19 Mar 2013 18:36:12 +0000 (11:36 -0700)]
Remove dead code from DateUtils.

This stuff is @hide, @deprecated, unused, and was never in any API level.

Change-Id: I87d132d77d7296a2de8527e05c2def362cab1c37

11 years agoMerge "Fix times in debugging output to use H rather than k."
Elliott Hughes [Tue, 19 Mar 2013 16:50:38 +0000 (16:50 +0000)]
Merge "Fix times in debugging output to use H rather than k."

11 years agoMerge "Fix build."
Elliott Hughes [Tue, 19 Mar 2013 01:11:02 +0000 (01:11 +0000)]
Merge "Fix build."

11 years agoFix build.
Elliott Hughes [Tue, 19 Mar 2013 01:10:24 +0000 (18:10 -0700)]
Fix build.

Change-Id: Icb7fb60f00251ab01c8f9844572bb349e45c6c41

11 years agoFix times in debugging output to use H rather than k.
Elliott Hughes [Tue, 19 Mar 2013 00:47:22 +0000 (17:47 -0700)]
Fix times in debugging output to use H rather than k.

Bug: 8359981
Change-Id: I4fccf3ed7e469b77f6e755dacbcd025005390505

11 years agoMerge "Clean up (some of) our many clocks to use H instead of k."
Elliott Hughes [Tue, 19 Mar 2013 00:28:41 +0000 (00:28 +0000)]
Merge "Clean up (some of) our many clocks to use H instead of k."

11 years agoClean up (some of) our many clocks to use H instead of k.
Elliott Hughes [Tue, 19 Mar 2013 00:21:33 +0000 (17:21 -0700)]
Clean up (some of) our many clocks to use H instead of k.

Bug: 8359981
Change-Id: I3c0c38c2bf5adc4b5bdc11c3d10cf41dd8fb6462

11 years agoMerge "Deliberately break DateFormat.format's formatting of 'k' for compatibility."
Elliott Hughes [Mon, 18 Mar 2013 22:07:22 +0000 (22:07 +0000)]
Merge "Deliberately break DateFormat.format's formatting of 'k' for compatibility."

11 years agoDeliberately break DateFormat.format's formatting of 'k' for compatibility.
Elliott Hughes [Mon, 18 Mar 2013 21:59:59 +0000 (14:59 -0700)]
Deliberately break DateFormat.format's formatting of 'k' for compatibility.

Bug: 8359981
Change-Id: I44ed816c644f94783c77952d544c5444b966868b

11 years agoMerge "Correct the target app token for input method window"
Craig Mautner [Mon, 18 Mar 2013 15:56:48 +0000 (15:56 +0000)]
Merge "Correct the target app token for input method window"

11 years agoCorrect the target app token for input method window
Carrie Xu [Tue, 30 Oct 2012 09:28:39 +0000 (17:28 +0800)]
Correct the target app token for input method window

The root cause is:
There is a defect in window manager service: When a new
activity that can be ime target is added into window manager
but the Z order of input method window don't need to be
changed, then the target app token of input method window
would not be updated to new one. This defect may cause that
the layer of input method window is calculated incorrectly.

The solution:
Correct the target app token for input method window.

Change-Id: I008311e3c9b1cf5fc320b614d8675c183c506d50

11 years agoMerge "update APN version for MVNO"
Robert Greenwalt [Sat, 16 Mar 2013 00:18:52 +0000 (00:18 +0000)]
Merge "update APN version for MVNO"

11 years agoMerge "app_process: don't pollute the environment"
Nick Kralevich [Fri, 15 Mar 2013 20:55:53 +0000 (20:55 +0000)]
Merge "app_process: don't pollute the environment"

11 years agoapp_process: don't pollute the environment
Nick Kralevich [Fri, 15 Mar 2013 20:12:28 +0000 (13:12 -0700)]
app_process: don't pollute the environment

Give a hoot — don't pollute!

Change-Id: I30d6c05cecab483f0e9d8063907f7bc0ef2bb104

11 years agoMerge "Never call app_main more than once"
Nick Kralevich [Fri, 15 Mar 2013 19:08:50 +0000 (19:08 +0000)]
Merge "Never call app_main more than once"

11 years agoNever call app_main more than once
Nick Kralevich [Fri, 15 Mar 2013 18:38:29 +0000 (11:38 -0700)]
Never call app_main more than once

Different kernels seem to handle ADDR_COMPAT_LAYOUT differently,
sometimes passing it to its children, sometimes not. If it's not
passed to its child successfully, we can end up in a restart loop.

Instead of testing for the presence of ADDR_COMPAT_LAYOUT, use an
environment variable instead, which is handled more predictably.

Bug: 8392487
Change-Id: Ia531dd2abb4e1cd46f3430d844e644f53581f530

11 years agoMerge "Don't use ADDR_COMPAT_LAYOUT on the emulator"
Nick Kralevich [Thu, 14 Mar 2013 23:46:20 +0000 (23:46 +0000)]
Merge "Don't use ADDR_COMPAT_LAYOUT on the emulator"

11 years agoDon't use ADDR_COMPAT_LAYOUT on the emulator
Nick Kralevich [Thu, 14 Mar 2013 23:31:34 +0000 (16:31 -0700)]
Don't use ADDR_COMPAT_LAYOUT on the emulator

For the emulator, we want people to see memory as it
actually is, not how we're hacking around buggy apps. Don't
set ADDR_COMPAT_LAYOUT on the emulator.

For reasons that I don't understand, personality(ADDR_COMPAT_LAYOUT)
does not persist across an exec on the emulator. app_main gets
into a tight loop restarting itself because of this. This change
also works around that bug.

Change-Id: Ia73a7d2d623c25cf39d248145d97307945d554da

11 years agoMerge "Respawn app_main if ADDR_COMPAT_LAYOUT is not set"
Nick Kralevich [Thu, 14 Mar 2013 22:28:01 +0000 (22:28 +0000)]
Merge "Respawn app_main if ADDR_COMPAT_LAYOUT is not set"

11 years agoRespawn app_main if ADDR_COMPAT_LAYOUT is not set
Nick Kralevich [Thu, 14 Mar 2013 20:23:52 +0000 (13:23 -0700)]
Respawn app_main if ADDR_COMPAT_LAYOUT is not set

Bug: 8358800
Change-Id: Id81aa26d74dfd16a49f0ee5926ea2cdb3c2106d8

11 years agoMerge "Fix NPE when pasting non-text from Clipboard"
John Reck [Tue, 12 Mar 2013 20:24:13 +0000 (20:24 +0000)]
Merge "Fix NPE when pasting non-text from Clipboard"

11 years agoMerge "Switch TLS Channel ID API from ECPrivateKey to PrivateKey."
Kenny Root [Tue, 12 Mar 2013 20:05:09 +0000 (20:05 +0000)]
Merge "Switch TLS Channel ID API from ECPrivateKey to PrivateKey."

11 years agoMerge ""[+>" more icon was never shown in PhoneStatusBar"
Daniel Sandler [Tue, 12 Mar 2013 19:51:32 +0000 (19:51 +0000)]
Merge ""[+>" more icon was never shown in PhoneStatusBar"

11 years ago"[+>" more icon was never shown in PhoneStatusBar
Mats H Carlsson [Wed, 10 Oct 2012 11:37:38 +0000 (13:37 +0200)]
"[+>" more icon was never shown in PhoneStatusBar

"[+>" more icon was never show in status bar because
the member variable for this icon was not initialized
from resources. This fix enables "[+>" icon to appear
in status bar when the number of indications in status
bar becomes large.

Bug: 8368569
Change-Id: Ieb3412eed831052d69c0cf63c9b4230c38171e4a