OSDN Git Service

android-x86/frameworks-base.git
9 years agoMerge "Deutsche Telekom VVM free SMS shortcodes DE"
Robert Greenwalt [Tue, 3 Feb 2015 16:44:01 +0000 (16:44 +0000)]
Merge "Deutsche Telekom VVM free SMS shortcodes DE"

9 years agoMerge "Pass in code cache directory when creating a script group."
Yang Ni [Tue, 3 Feb 2015 06:20:19 +0000 (06:20 +0000)]
Merge "Pass in code cache directory when creating a script group."

9 years agoMerge "Fixing up the formatting code which is already merged"
Robert Greenwalt [Mon, 2 Feb 2015 17:00:42 +0000 (17:00 +0000)]
Merge "Fixing up the formatting code which is already merged"

9 years agoFixing up the formatting code which is already merged
Robert Greenwalt [Wed, 21 Jan 2015 19:52:20 +0000 (19:52 +0000)]
Fixing up the formatting code which is already merged

I just apply google coding guide like space.

Uploading the same code about + handling just apply the Google Coding Guide.
The previous code is "https://android-review.googlesource.com/#/c/122770/"

This code is about + plus number converting logic for international NANP.

Change-Id: I23e4709294c603b387b28b3c2a56c932d838c666

9 years agoMerge "Also include the domain when parceling StaticIpConfiguration."
Lorenzo Colitti [Fri, 30 Jan 2015 07:07:48 +0000 (07:07 +0000)]
Merge "Also include the domain when parceling StaticIpConfiguration."

9 years agoMerge "Fix DhcpResults.setDomain() and StaticIpConfiguration.toLinkProperties()"
Lorenzo Colitti [Fri, 30 Jan 2015 07:00:38 +0000 (07:00 +0000)]
Merge "Fix DhcpResults.setDomain() and StaticIpConfiguration.toLinkProperties()"

9 years agoMerge "Fix wrap property creation when truncating."
Christopher Ferris [Thu, 29 Jan 2015 23:41:05 +0000 (23:41 +0000)]
Merge "Fix wrap property creation when truncating."

9 years agoFix wrap property creation when truncating.
Christopher Ferris [Thu, 29 Jan 2015 01:56:32 +0000 (17:56 -0800)]
Fix wrap property creation when truncating.

If a property name gets truncated, make sure it doesn't end in a '.'
since that makes the name illegal.

Bug: 19196358
Bug: https://code.google.com/p/android/issues/detail?id=82947
Change-Id: Icc1a26593237ca19ad0ebd776a60b3d6290bb355

9 years agoPass in code cache directory when creating a script group.
Yang Ni [Fri, 16 Jan 2015 19:06:26 +0000 (11:06 -0800)]
Pass in code cache directory when creating a script group.

Change-Id: Ia73ea917a126a5055ec97f13d90a5feaafd6a2f5

9 years agoAlso include the domain when parceling StaticIpConfiguration.
Lorenzo Colitti [Thu, 29 Jan 2015 08:10:52 +0000 (17:10 +0900)]
Also include the domain when parceling StaticIpConfiguration.

StaticIpConfiguration objects are parceled at least as part of the
IpConfiguration objects that are passed to IEthernetManager when an
application sets static IP configuration on Ethernet.

Change-Id: I49991e2f591cc6cf01b503c18eb343b5929efe29

9 years agoMerge "Support connecting to networks with misconfigured subnet masks."
Lorenzo Colitti [Thu, 29 Jan 2015 08:35:35 +0000 (08:35 +0000)]
Merge "Support connecting to networks with misconfigured subnet masks."

9 years agoFix DhcpResults.setDomain() and StaticIpConfiguration.toLinkProperties()
Paul Jensen [Wed, 5 Nov 2014 14:35:26 +0000 (09:35 -0500)]
Fix DhcpResults.setDomain() and StaticIpConfiguration.toLinkProperties()

setDomain() and toLinkProperties() were not setting the domains.
The setDomain() bug affected Wifi and I believe the toLinkProperties()
bug affected Ethernet and Bluetooth reverse-tethering.

(cherry picked from commit c53113b37f33c7ed19660c8ec5bfd578e8bb5409)

bug:18252947
Change-Id: I6235fcd6b875aee516efbb5f880db1a99380355b

9 years agoSupport connecting to networks with misconfigured subnet masks.
Lorenzo Colitti [Tue, 20 Jan 2015 06:53:02 +0000 (15:53 +0900)]
Support connecting to networks with misconfigured subnet masks.

In K and earlier, we would connect to a network where the gateway
was not covered by the subnet mask of the IP address. This is an
invalid configuration, but it used to work, and other OSes appear
to accept it too, so support it.

Bug: 19067207

(cherry picked from commit 2dfb79a54adeb4bcf1f62332a9db467fce302ced)

Change-Id: I80088f291466dbd5a47f360dcc1620acee5cf57e

9 years agoMerge "Initialize clipping structure"
Jason Sams [Tue, 27 Jan 2015 20:41:42 +0000 (20:41 +0000)]
Merge "Initialize clipping structure"

9 years agoMerge "Fix printf format"
Marco Nelissen [Tue, 27 Jan 2015 20:19:46 +0000 (20:19 +0000)]
Merge "Fix printf format"

9 years agoFix printf format
Marco Nelissen [Tue, 27 Jan 2015 20:17:48 +0000 (12:17 -0800)]
Fix printf format

Change-Id: Ied2f10bb59fe0e13b23bf1f9913ab3b081348a28

9 years agoMerge "Make SoundPool use MediaCodec"
Marco Nelissen [Tue, 27 Jan 2015 17:49:01 +0000 (17:49 +0000)]
Merge "Make SoundPool use MediaCodec"

9 years agoMerge "kill HAVE_PTHREADS."
Yabin Cui [Tue, 27 Jan 2015 06:49:16 +0000 (06:49 +0000)]
Merge "kill HAVE_PTHREADS."

9 years agokill HAVE_PTHREADS.
Yabin Cui [Tue, 27 Jan 2015 03:43:58 +0000 (19:43 -0800)]
kill HAVE_PTHREADS.

Bug: 19083585
Change-Id: Ib466949bb6cd6d1bbc4680e989f0f9fae62ca564

9 years agoInitialize clipping structure
Jason Sams [Tue, 27 Jan 2015 02:14:36 +0000 (18:14 -0800)]
Initialize clipping structure

placeholder until array clipping is added to the API.

Change-Id: Ib20dbd71989c559c0669385a5088ec261596913c

9 years agoMerge "Add missing <string.h> includes."
Elliott Hughes [Sun, 25 Jan 2015 19:17:42 +0000 (19:17 +0000)]
Merge "Add missing <string.h> includes."

9 years agoAdd missing <string.h> includes.
Elliott Hughes [Sun, 25 Jan 2015 16:33:36 +0000 (08:33 -0800)]
Add missing <string.h> includes.

Change-Id: I188bfe1388fa49f6f79d247aa0d0f4fb089de503

9 years agoMake SoundPool use MediaCodec
Marco Nelissen [Thu, 4 Dec 2014 16:59:22 +0000 (08:59 -0800)]
Make SoundPool use MediaCodec

Bug: 18239054
Change-Id: Ia144fc1bbb0d2787638ee972e2224339b4965310

9 years agoMerge "Remove superfluous OpenSSL include paths."
Narayan Kamath [Fri, 23 Jan 2015 10:47:34 +0000 (10:47 +0000)]
Merge "Remove superfluous OpenSSL include paths."

9 years agoRemove superfluous OpenSSL include paths.
Adam Langley [Fri, 23 Jan 2015 01:15:13 +0000 (17:15 -0800)]
Remove superfluous OpenSSL include paths.

The libcrypto and libssl modules (and their respective static and host
versions) use LOCAL_EXPORT_C_INCLUDE_DIRS thus just including the module
is sufficient.

Change-Id: I467f1b1d038b5136fa22b1230da134c394eac54e

9 years agoMerge "Fix default compute thread priority"
Jason Sams [Thu, 22 Jan 2015 21:08:44 +0000 (21:08 +0000)]
Merge "Fix default compute thread priority"

9 years agoMerge "Add FP16 types to the list of RenderScript types."
Jason Sams [Thu, 22 Jan 2015 21:08:03 +0000 (21:08 +0000)]
Merge "Add FP16 types to the list of RenderScript types."

9 years agoMerge "Add support for persist.sys.locale."
Narayan Kamath [Thu, 22 Jan 2015 13:49:43 +0000 (13:49 +0000)]
Merge "Add support for persist.sys.locale."

9 years agoAdd support for persist.sys.locale.
Narayan Kamath [Thu, 15 Jan 2015 14:48:15 +0000 (14:48 +0000)]
Add support for persist.sys.locale.

AndroidRuntime has been changed to read "ro.product.locale" and
"persist.sys.locale" instead of "ro.product.locale.language" etc.
This is passed down as "-Duser.locale" to the runtime.

The system_server has been changed to write out persist.sys.locale
on locale changes.

bug: 17691569

Change-Id: I93360c8795c9620a133656dc491d13d7b6ed162e

9 years agoFix default compute thread priority
Jason Sams [Wed, 21 Jan 2015 20:55:14 +0000 (12:55 -0800)]
Fix default compute thread priority

bug 16651474

Compute inherited graphics default thread priority of Display.  This
was not intended.

Change-Id: I0dd9a230ce8ceba64e971b024cbe518927cd2550

9 years agoMerge "Add + code handling logic for SMS Global Roaming"
Robert Greenwalt [Wed, 21 Jan 2015 19:52:20 +0000 (19:52 +0000)]
Merge "Add + code handling logic for SMS Global Roaming"

9 years agoMerge "[ActivityManager] No home activity resumed and cause key dispatch ANR"
Craig Mautner [Wed, 21 Jan 2015 17:01:01 +0000 (17:01 +0000)]
Merge "[ActivityManager] No home activity resumed and cause key dispatch ANR"

9 years agoMerge "Choose the right executable for wrapped applications."
Narayan Kamath [Wed, 21 Jan 2015 14:01:55 +0000 (14:01 +0000)]
Merge "Choose the right executable for wrapped applications."

9 years ago[ActivityManager] No home activity resumed and cause
louis_chang [Wed, 21 Jan 2015 11:01:52 +0000 (19:01 +0800)]
[ActivityManager] No home activity resumed and cause
key dispatch ANR

Symptom:
In some scenario, there might have two home tasks
in the home stack. If the top home task was finishing
and expected to return to home stack, no activity
will resume.

Root Cause:
Unable to perform resumeHomeStackTask() because the
resumeTopActivityLocked() has been prevented to invoke
recursively.

Solution:
Just resume the next activity if already in home stack

Change-Id: I24ecbcac6cee4ddbd90cdd27c4f4915cffb9b354

9 years agoMerge "New trick to install bad dex file."
Ying Wang [Tue, 20 Jan 2015 21:43:51 +0000 (21:43 +0000)]
Merge "New trick to install bad dex file."

9 years agoAdd FP16 types to the list of RenderScript types.
Jason Sams [Wed, 5 Nov 2014 23:16:26 +0000 (15:16 -0800)]
Add FP16 types to the list of RenderScript types.

Change-Id: Ibf9a6d391fb4d6a5c4b4e05ab73fb58bd74dc586

9 years agoNew trick to install bad dex file.
Ying Wang [Thu, 15 Jan 2015 00:18:51 +0000 (16:18 -0800)]
New trick to install bad dex file.

Previously we use LOCAL_JAVA_RESOURCE_FILES to override the normally
generated classes.dex with a bad dex file.
Now as we have moved Java resource processing before running jarjar, dx
would fail for the bad classes.dex inside the classes.jar.
Instead we override the target specific variable PRIVATE_DEX_FILE to
inject the bad dex file.

(cherry-picked from commit f122d90c290c789e0d0085127beebdcf1935c127)

Bug: 18837479
Change-Id: I73f82c8a15e1d8935d9c2f290ba9456184a56ec7

9 years agoMerge "Fix for broken docs build."
Neil Fuller [Tue, 20 Jan 2015 15:18:59 +0000 (15:18 +0000)]
Merge "Fix for broken docs build."

9 years agoFix for broken docs build.
Neil Fuller [Tue, 20 Jan 2015 15:08:20 +0000 (15:08 +0000)]
Fix for broken docs build.

Broken by commit 50a01d89606b40dd151ad407d05c3b2867bbdbad

Change-Id: Iab460aa6254d320349f67e747f763bbf06d3e9e7

9 years agoMerge "Changes associated with an OkHttp upgrade"
Neil Fuller [Tue, 20 Jan 2015 12:50:47 +0000 (12:50 +0000)]
Merge "Changes associated with an OkHttp upgrade"

9 years agoChanges associated with an OkHttp upgrade
Neil Fuller [Mon, 12 Jan 2015 16:49:06 +0000 (16:49 +0000)]
Changes associated with an OkHttp upgrade

Change-Id: I2a4db602aa7ffdef886e0f1a955715a2551a87a5

9 years agoChoose the right executable for wrapped applications.
Narayan Kamath [Mon, 19 Jan 2015 16:05:24 +0000 (16:05 +0000)]
Choose the right executable for wrapped applications.

64 bit zygotes should exec app_process64, whereas 32 bit
zygotes should exec app_process32.

Change-Id: I63b56400e0478cd2bb206c98da6858be558f0525

9 years agoMerge "Remove support for --classpath and make --runtime-init the default."
Narayan Kamath [Mon, 19 Jan 2015 15:09:16 +0000 (15:09 +0000)]
Merge "Remove support for --classpath and make --runtime-init the default."

9 years agoRemove support for --classpath and make --runtime-init the default.
Narayan Kamath [Thu, 8 Jan 2015 12:45:37 +0000 (12:45 +0000)]
Remove support for --classpath and make --runtime-init the default.

--classpath hasn't been properly supported or used for as long as
git history exists, and the non --runtime-init classpath is useless
without it.

Also, forking and exec' a bare dalvikvm instance from the zygote seems
somewhat pointless since we might as well just execute it directly.

Change-Id: Ib07d822d448e7b431d7564381f9cd11735fd1113

9 years agoMerge "Fix handling of wrapped processes [part 1]."
Narayan Kamath [Mon, 19 Jan 2015 15:00:00 +0000 (15:00 +0000)]
Merge "Fix handling of wrapped processes [part 1]."

9 years agoFix handling of wrapped processes [part 1].
Narayan Kamath [Fri, 16 Jan 2015 16:26:54 +0000 (16:26 +0000)]
Fix handling of wrapped processes [part 1].

Change 3fe595108f47fffd changed our pipe creation to CLOEXEC,
which doesn't work so well for wrapper processes that call exec.

Sending down the PID pre-exec won't work so well because wrappers
are arbitrary and might fork themselves before finally starting
the child application process.

Change-Id: I065f0f2150dc1ff04d2b2e8b25f06b4c754809b9

9 years agoMerge "Fix handling of wrapped processes [part 2]"
Narayan Kamath [Mon, 19 Jan 2015 09:52:20 +0000 (09:52 +0000)]
Merge "Fix handling of wrapped processes [part 2]"

9 years agoMerge "Fast loadInverse() implementation for the common case"
Chris Craik [Fri, 16 Jan 2015 17:55:39 +0000 (17:55 +0000)]
Merge "Fast loadInverse() implementation for the common case"

9 years agoFix handling of wrapped processes [part 2]
Narayan Kamath [Fri, 16 Jan 2015 17:22:41 +0000 (17:22 +0000)]
Fix handling of wrapped processes [part 2]

Drop privileges only if we're root. Zygote.preloadClasses can
be called from a non-root uid / gid in the case of wrapped classes,
so we shouldn't die from the resulting EPERM.

Change-Id: Ie3fbe50a17ec49bcbcb875c3a2ee1e1a3a62e88d

9 years agoMerge "Prevent AudioService dead lock issue."
Jean-Michel Trivi [Fri, 16 Jan 2015 17:23:36 +0000 (17:23 +0000)]
Merge "Prevent AudioService dead lock issue."

9 years agoMerge "Remove DSA support from Android KeyStore and KeyChain."
Alex Klyubin [Fri, 16 Jan 2015 16:44:21 +0000 (16:44 +0000)]
Merge "Remove DSA support from Android KeyStore and KeyChain."

9 years agoMerge "Add KeyPairGenerator.EC backed by Android KeyStore."
Alex Klyubin [Fri, 16 Jan 2015 16:44:06 +0000 (16:44 +0000)]
Merge "Add KeyPairGenerator.EC backed by Android KeyStore."

9 years agoMerge "Fix a property name that's longer than 31 chars."
Narayan Kamath [Fri, 16 Jan 2015 14:06:39 +0000 (14:06 +0000)]
Merge "Fix a property name that's longer than 31 chars."

9 years agoFix a property name that's longer than 31 chars.
Narayan Kamath [Fri, 16 Jan 2015 14:04:56 +0000 (14:04 +0000)]
Fix a property name that's longer than 31 chars.

The name wasn't chosen without much thought and is intended
to be a temporary fix to keep devices booting while a better
name is decided upon.

bug: 18335678
Change-Id: Ic3451255d870f3395cfbcaa08a7764d93defc350

9 years agoMerge "Removing HttpResponseCacheTest"
Neil Fuller [Fri, 16 Jan 2015 10:19:26 +0000 (10:19 +0000)]
Merge "Removing HttpResponseCacheTest"

9 years agoMerge "Offer to detect non-SSL/TLS network traffic."
Jeff Sharkey [Fri, 16 Jan 2015 01:11:26 +0000 (01:11 +0000)]
Merge "Offer to detect non-SSL/TLS network traffic."

9 years agoFast loadInverse() implementation for the common case
Romain Guy [Thu, 15 Jan 2015 19:53:44 +0000 (11:53 -0800)]
Fast loadInverse() implementation for the common case

Most matrices used by the UI toolkit are translation matrices, whose
inverses can be quickly computed by using the negated translation
vector.

Change-Id: I54a28a634a586085779bfc26f3a4160cd5ab2b22

9 years agoOffer to detect non-SSL/TLS network traffic.
Jeff Sharkey [Tue, 4 Nov 2014 21:34:06 +0000 (13:34 -0800)]
Offer to detect non-SSL/TLS network traffic.

Introduces new module that provides network-related features for
the StrictMode developer API.  The first feature offers to detect
sockets sending data not wrapped inside a layer of SSL/TLS
encryption.

When a developer enables, we ask netd to watch all outgoing traffic
from our UID, and penalize us accordingly if cleartext sockets are
detected.  When enabled, netd captures the offending packet and
passes it back to the owning process to aid investigations.  When
death penalty is requested, all future traffic on the socket is
blocked, which usually results in a useful stacktrace before the
app is actually killed.

Bug: 18335678
Change-Id: I3adbc974efd8d3766b4b1a23257563bb82d53c29

9 years agoRemoving HttpResponseCacheTest
Neil Fuller [Thu, 15 Jan 2015 16:00:42 +0000 (16:00 +0000)]
Removing HttpResponseCacheTest

Moving HttpResponseCacheTest.java from
frameworks/base/core/tests/coretests/src/android/net/http
to
cts/tests/tests/net/src/android/net/http/cts

Change-Id: I61a07c53f8d0ef55d35b928f244e4680725b76ef

9 years agoMerge "Fix unexpected language changes on SIM card insertion."
Narayan Kamath [Thu, 15 Jan 2015 11:53:31 +0000 (11:53 +0000)]
Merge "Fix unexpected language changes on SIM card insertion."

9 years agoFix unexpected language changes on SIM card insertion.
Narayan Kamath [Tue, 13 Jan 2015 15:51:57 +0000 (15:51 +0000)]
Fix unexpected language changes on SIM card insertion.

There are two separate issues here that need to be fixed, both
boil down to the fact that adding an imperative (userSetLocale)
to the Configuration is a bad idea. Because of this:

- We'd never persist the first user set configuration if it was en_US,
  because of an erroneous call to Configuration.setLocale.

- ActivityManager.getConfiguration would sometimes return a
  Configuration with userSetLocale == true, which means callers with
  the right permissions would inadvertently persist a locale they didn't
  want to persist.

bug: 18879010

(cherry picked from commit c63a16b8e46f2f3ecce6c548663cc3185f011fd6)

Change-Id: I4d714eb1f75afc0a5bf4d53a374df4ed16e328bf

9 years agoRemove DSA support from Android KeyStore and KeyChain.
Alex Klyubin [Thu, 15 Jan 2015 00:45:51 +0000 (16:45 -0800)]
Remove DSA support from Android KeyStore and KeyChain.

We're switching from OpenSSL to BoringSSL which does not support DSA.

Bug: 17409664
Change-Id: Id9b52666ba9ef234076105c925610b5b312988a5

9 years agoAdd KeyPairGenerator.EC backed by Android KeyStore.
Alex Klyubin [Wed, 14 Jan 2015 21:35:32 +0000 (13:35 -0800)]
Add KeyPairGenerator.EC backed by Android KeyStore.

The Android KeyStore backed KeyPairGenerator can generate EC key
pairs, but it cannot be instantiated via the standard JCA approach of
KeyPairGenerator.getInstance("EC", "AndroidKeyStore"). Instead, the
user must invoke KeyPairGenerator.getInstance("RSA",
"AndroidKeyStore") and then tell it to generate an EC key pair.

This CL fixes this weirdness.

The fix requires the introduction of late resolution of key algorithm
and default key size. Prior to this CL, these parameters were resolved
prior to KeyPairGenerator initialization, inside KeyPairGeneratorSpec.
In this CL, these parameters are resolved during KeyPairGenerator
initialization. This is fine because KeyPairGeneratorSpec should be as
dumb as possible and all the logic should reside in KeyPairGenerator
and lower layers.

Bug: 19018089
Change-Id: I114502356e6c9691518cf05b6d9eb0920b4fe0b2

9 years agoMerge "Generate IKeystoreService using aidl"
Chad Brubaker [Wed, 14 Jan 2015 20:31:29 +0000 (20:31 +0000)]
Merge "Generate IKeystoreService using aidl"

9 years agoGenerate IKeystoreService using aidl
Chad Brubaker [Wed, 14 Jan 2015 01:41:22 +0000 (17:41 -0800)]
Generate IKeystoreService using aidl

This replaces IKeystoreService.java with IKeystoreService.aidl and
changes the methods that passed down a byte[][] to instead pass down a
KeystoreArguments which is currently a thin parcelable wrapper around a byte[][].

Change-Id: I6367bcf57562f41a27aab14f1903b74995cb65c2

9 years agoMerge "Find wall clock RTC through sysfs"
Greg Hackmann [Tue, 13 Jan 2015 20:43:35 +0000 (20:43 +0000)]
Merge "Find wall clock RTC through sysfs"

9 years agoFind wall clock RTC through sysfs
Greg Hackmann [Wed, 26 Feb 2014 20:22:13 +0000 (12:22 -0800)]
Find wall clock RTC through sysfs

Devices may have multiple RTCs.  By default the kernel uses rtc0 to
store the system time, but devices may override this (or even specify
that none of them should be used for system time).

Userspace can indirectly find the designated RTC through sysfs.  During
AlarmManagerService initialization, enumerate through all rtc class
devices to locate the device with attribute hctosys=1.

This is only done on devices without /dev/alarm, which has its own
in-kernel mechanism to pick the RTC.

Change-Id: Ife2b342c3590133ed316ddaf1799cbc1bfa6e6d9
Signed-off-by: Greg Hackmann <ghackmann@google.com>
9 years agoMerge "Prevent system server dump stuck by pipe buffer full."
Craig Mautner [Tue, 13 Jan 2015 04:08:03 +0000 (04:08 +0000)]
Merge "Prevent system server dump stuck by pipe buffer full."

9 years agoMerge "Use more intention-revealing tests for S_ISLNK and S_ISSOCK."
Elliott Hughes [Tue, 13 Jan 2015 01:12:59 +0000 (01:12 +0000)]
Merge "Use more intention-revealing tests for S_ISLNK and S_ISSOCK."

9 years agoPrevent AudioService dead lock issue.
seunghwan.hong [Wed, 29 Oct 2014 08:43:20 +0000 (17:43 +0900)]
Prevent AudioService dead lock issue.

Dead lock sequence:
 1. called onSetA2dpConnectionState()
  1-1. synchronized(mConnectedDevices)
 2. called onServiceDisconnected()
  2-1. synchronized(mA2DPAvrcpLock)
 3. waiting to lock(mConnectedDevices) in onServiceDisconnected()
 4. waiting to lock(mA2DPAvrcpLock) in onSetA2dpConnectionState()

========================================================================================
- watchdog issue
----- pid 3306 at 2014-10-14 16:15:12 -----
Cmd line: system_server

"main" prio=5 tid=1 MONITOR
  - waiting to lock <0x4343e9b0> (a java.util.HashMap) held by tid=46 (AudioService)
      at android.bluetooth.BluetoothA2dp$2.onServiceDisconnected(BluetoothA2dp.java:529)

"AudioService" prio=5 tid=46 MONITOR
      - waiting to lock <0x42ab7a58> (a java.lang.Object) held by tid=1 (main)
========================================================================================

Signed-off-by: Seunghwan Hong <seunghwan.hong@lge.com>
Change-Id: I99e061c07be01aabcd26786ef2ebb71f46717b93

9 years agoUse more intention-revealing tests for S_ISLNK and S_ISSOCK.
Elliott Hughes [Mon, 12 Jan 2015 22:33:04 +0000 (14:33 -0800)]
Use more intention-revealing tests for S_ISLNK and S_ISSOCK.

Change-Id: Ia7f13eb7d5a36cf2b4c0a253b2b4d7f7af396951

9 years agoMerge "New Script Group API."
Yang Ni [Mon, 12 Jan 2015 21:25:12 +0000 (21:25 +0000)]
Merge "New Script Group API."

9 years agoMerge "Fix HttpResponseCacheTest in anticipation of an OkHttp upgrade."
Neil Fuller [Mon, 12 Jan 2015 14:50:33 +0000 (14:50 +0000)]
Merge "Fix HttpResponseCacheTest in anticipation of an OkHttp upgrade."

9 years agoMerge "[ActivityManager] Distinguish FG or BG receiver finished"
Olawale Ogunwale [Mon, 12 Jan 2015 13:45:10 +0000 (13:45 +0000)]
Merge "[ActivityManager] Distinguish FG or BG receiver finished"

9 years agoMerge "bindService can't start up service process"
Olawale Ogunwale [Mon, 12 Jan 2015 13:44:33 +0000 (13:44 +0000)]
Merge "bindService can't start up service process"

9 years agoMerge "[ActivityManager] Fix process record leakage."
Olawale Ogunwale [Mon, 12 Jan 2015 13:44:15 +0000 (13:44 +0000)]
Merge "[ActivityManager] Fix process record leakage."

9 years agoAdd + code handling logic for SMS Global Roaming
sohryun.shin [Mon, 12 Jan 2015 04:42:59 +0000 (13:42 +0900)]
Add + code handling logic for SMS Global Roaming

When sending SMS from international CDMA NANP country, outgoing SMS is failed due to + code converting.
(Send a SMS text message: to an international phone number using the plus code (+)).

In kitkat, when user send a sms with plus code, the "+" converted to 011(NANP_IDP_STRING). And it works well.
However, in lollipop, the "+" converted to current IDP, not NANP_IDP_STRING. And it failed to send the SMS.

I think, we need to seperate SMS and Call logic like cdmaCheckAndProcessPlusCodeForSms and cdmaCheckAndProcessPlusCode.
sohryun.shin@lge.com

Change-Id: Ie9d402cbeb4de4ac30299d5820b33909d0e40320

9 years agoMerge "MediaCodec/Image: handle null cropRect correctly"
Lajos Molnar [Sat, 10 Jan 2015 01:12:31 +0000 (01:12 +0000)]
Merge "MediaCodec/Image: handle null cropRect correctly"

9 years agoMerge "media: add support for 160-byte image info in MediaCodec"
Lajos Molnar [Sat, 10 Jan 2015 01:10:23 +0000 (01:10 +0000)]
Merge "media: add support for 160-byte image info in MediaCodec"

9 years agoFix HttpResponseCacheTest in anticipation of an OkHttp upgrade.
Neil Fuller [Fri, 9 Jan 2015 11:43:42 +0000 (11:43 +0000)]
Fix HttpResponseCacheTest in anticipation of an OkHttp upgrade.

OkHttp recently changed the behavior of their caching with
commit e74e3f3bf744ef7f4d8ee724a7cf2347e486cfab - it is now
neccessary to close the inputstream (or disconnect the
HttpURLConnection) for a response to be cached.

This change is (effectively) a no-op prior to the upgrade.

The behavior is undefined as to whether closing the
input stream is required for caching. OkHttp's new behavior
is consistent with other HttpURLConnection implementations
tried.

Change-Id: Iaf57371651296ac84850971ef60a9338cead57c0

9 years agoMerge "Call VMRuntime.clampGrowthLimit for small heap apps"
Mathieu Chartier [Fri, 9 Jan 2015 00:17:15 +0000 (00:17 +0000)]
Merge "Call VMRuntime.clampGrowthLimit for small heap apps"

9 years agoCall VMRuntime.clampGrowthLimit for small heap apps
Mathieu Chartier [Thu, 8 Jan 2015 22:42:20 +0000 (14:42 -0800)]
Call VMRuntime.clampGrowthLimit for small heap apps

If the manifest doesn't specify large heaps, we now call
VMRuntime.clampGrowthLimit to release heap virtual address space
which won't ever get used.

Bug: 18387825
Bug: 17131630
Change-Id: I61fdcd70c70234256637eeebefe3abb22b91095d

9 years agoMerge "Fixup aapt usage message"
Adam Lesinski [Thu, 8 Jan 2015 20:15:38 +0000 (20:15 +0000)]
Merge "Fixup aapt usage message"

9 years agobindService can't start up service process
chad [Sun, 21 Dec 2014 07:43:01 +0000 (15:43 +0800)]
bindService can't start up service process

Issue 85758 https://code.google.com/p/android/issues/detail?id=85758

Change-Id: I28645445ee5c46b2ab4cf78189f143ea97df63dd
Signed-off-by: chad <guochongscut@gmail.com>
9 years agoMerge "check ntpserver as empty string instead of null"
Jeff Sharkey [Thu, 8 Jan 2015 00:57:42 +0000 (00:57 +0000)]
Merge "check ntpserver as empty string instead of null"

9 years agocheck ntpserver as empty string instead of null
Young-Ho Cha [Wed, 7 Jan 2015 02:24:35 +0000 (11:24 +0900)]
check ntpserver as empty string instead of null

mServer cannot set null, because string from resource always returns
non-null charsequence

Change-Id: I8d6a6fdbc34267ee361e7bd20719887268161870
Signed-off-by: Young-Ho Cha <ganadist@gmail.com>
9 years agoNew Script Group API.
Yang Ni [Fri, 24 Oct 2014 15:52:24 +0000 (08:52 -0700)]
New Script Group API.

Change-Id: I73d3572276a38dac775025b472ba229ad8770c84

9 years agoMerge "Core/jni: Disallow relocation for decryption"
Andreas Gampe [Wed, 7 Jan 2015 22:05:11 +0000 (22:05 +0000)]
Merge "Core/jni: Disallow relocation for decryption"

9 years agoCore/jni: Disallow relocation for decryption
neo.chae [Wed, 7 Jan 2015 06:32:15 +0000 (15:32 +0900)]
Core/jni: Disallow relocation for decryption

When decrypting a device, a tmpfs is temporarily mounted as /data,
the size of which is usually small. When the zygote, system server
and necessary apps are brought up, they will be compiled into the
tmpfs.

If the system image contains prebuilts, they will be relocated instead
of compiled. This is unnecessary. In this special situation it is
acceptable to run out of the prebuilt oat files without relocation,
which can save space in the tmpfs.

This patch ensures that the boot image is not being relocated.

Change-Id: I42bfb7e3039574b7e4f2772e0d395f093d59ed1b
Signed-off-by: Hyangseok Chae <neo.chae@lge.com>
9 years agoMerge "camera2: Clarify the documentation on the ImageReader usecase"
Narayan Kamath [Wed, 7 Jan 2015 18:31:55 +0000 (18:31 +0000)]
Merge "camera2: Clarify the documentation on the ImageReader usecase"

9 years agocamera2: Clarify the documentation on the ImageReader usecase
Martin Storsjo [Wed, 19 Nov 2014 09:58:03 +0000 (11:58 +0200)]
camera2: Clarify the documentation on the ImageReader usecase

Uncompressed JPEG data is an oxymoron. Also mention the other
usecases for ImageReader.

Change-Id: I75261baf162a58c415705db99afa474c3826f900

9 years agoMerge "Check for OOM in BitmapFactory's getMimeTypeString()."
Vladimir Marko [Wed, 7 Jan 2015 10:20:48 +0000 (10:20 +0000)]
Merge "Check for OOM in BitmapFactory's getMimeTypeString()."

9 years agoFixup aapt usage message
Johan Redestig [Wed, 7 Jan 2015 08:24:39 +0000 (09:24 +0100)]
Fixup aapt usage message

The --preferred-configurations option was renamed to
--preferred-density in fab5087 but only part of the
usage message was changed.

Change-Id: I89d270990023beca19605901d956d29d0b0b848b

9 years agoMerge "NPE in SearchManager.triggerSearch"
Amith Yamasani [Wed, 7 Jan 2015 00:05:10 +0000 (00:05 +0000)]
Merge "NPE in SearchManager.triggerSearch"

9 years agoMerge "Fix crash caused by toHex returning exception"
Paul Lawrence [Tue, 6 Jan 2015 23:34:06 +0000 (23:34 +0000)]
Merge "Fix crash caused by toHex returning exception"

9 years agoFix crash caused by toHex returning exception
Paul Lawrence [Tue, 6 Jan 2015 21:11:23 +0000 (13:11 -0800)]
Fix crash caused by toHex returning exception

toHex was changed to throw an exception in
I4986a8e806d9066129f696ab9f2e80655424e723, but its caller was not adjusted
accordingly, causing a crash whenever an unencrypted device was booted.

Bug: 18886749
Change-Id: If0505f617001cf5e0d99cf14c8b09e6a6a377167

9 years agoCheck for OOM in BitmapFactory's getMimeTypeString().
Vladimir Marko [Tue, 6 Jan 2015 18:17:52 +0000 (18:17 +0000)]
Check for OOM in BitmapFactory's getMimeTypeString().

Bug: 18909596
Change-Id: If65fe45f5fd569a99a0cae1d31117af262bca430

9 years agoMerge "Do not allow the navigation bar to move if width == height"
Olawale Ogunwale [Tue, 6 Jan 2015 16:43:54 +0000 (16:43 +0000)]
Merge "Do not allow the navigation bar to move if width == height"

9 years agoMerge "Use NULL instead of 0 for pointer arguments of CallStaticVoidMethod."
Narayan Kamath [Tue, 6 Jan 2015 10:06:05 +0000 (10:06 +0000)]
Merge "Use NULL instead of 0 for pointer arguments of CallStaticVoidMethod."

9 years agoUse NULL instead of 0 for pointer arguments of CallStaticVoidMethod.
Haitao Feng [Wed, 31 Dec 2014 05:48:08 +0000 (13:48 +0800)]
Use NULL instead of 0 for pointer arguments of CallStaticVoidMethod.

CallStaticVoidMethod is varargs function, and calling it with
a literal 0 like CallStaticVoidMethod(..., 0) will treat the
argument as a 4 byte int in both 32 and 64 bit processes.

This is incorrect for pointer arguments where NULL should be
used instead.

Reviewed-by: Liao, Bruce <bruce.liao@intel.com>
Signed-off-by: Yong Yao <yong.yao@intel.com>
Change-Id: I9d700d3790a80dbee6826f64baf9ef5d81ca390f