OSDN Git Service

android-x86/frameworks-base.git
7 years agoExpand seccomp whitelist
Paul Lawrence [Tue, 7 Feb 2017 16:18:03 +0000 (08:18 -0800)]
Expand seccomp whitelist

Bug: 34979910
Test: System boots, app runs
Change-Id: I8b60a24ac855679251e73edcdec57eb4af4a5610

7 years agoMerge "Fixes Treble marshaling of strings embedded in structures from Java"
Treehugger Robot [Mon, 6 Feb 2017 23:41:27 +0000 (23:41 +0000)]
Merge "Fixes Treble marshaling of strings embedded in structures from Java"

7 years agoMerge "Add a mechanism to configure the default A2DP codec priorities"
Treehugger Robot [Mon, 6 Feb 2017 22:59:50 +0000 (22:59 +0000)]
Merge "Add a mechanism to configure the default A2DP codec priorities"

7 years agoMerge "Hotspot Tile: longclick should open Tethering settings"
Jason Monk [Mon, 6 Feb 2017 22:53:58 +0000 (22:53 +0000)]
Merge "Hotspot Tile: longclick should open Tethering settings"

7 years agoMerge "Remove uncessary dependency of libhwui"
Miao Wang [Mon, 6 Feb 2017 22:47:15 +0000 (22:47 +0000)]
Merge "Remove uncessary dependency of libhwui"

7 years agoMerge "Renamed voiceNetworkStateChanged and dataNetworkStateChanged"
Treehugger Robot [Mon, 6 Feb 2017 22:41:10 +0000 (22:41 +0000)]
Merge "Renamed voiceNetworkStateChanged and dataNetworkStateChanged"

7 years agoMerge "Self managed CS implementation."
Tyler Gunn [Mon, 6 Feb 2017 22:32:13 +0000 (22:32 +0000)]
Merge "Self managed CS implementation."

7 years agoRenamed voiceNetworkStateChanged and dataNetworkStateChanged
Jack Yu [Thu, 26 Jan 2017 20:45:57 +0000 (12:45 -0800)]
Renamed voiceNetworkStateChanged and dataNetworkStateChanged

The misleading voiceNetworkStateChanged should be renamed
to networkStateChanged to reflect either voice or data
network changes will trigger this indication.
And dataNetworkStateChanged should be renamed to
dataCallListChanged as well.

Test: Build
Merged-In: Icf20ceb3718794f7aa62650816714ac8679d101a
Change-Id: Icf20ceb3718794f7aa62650816714ac8679d101a

7 years agoMerge "Framework fixes to support VoLTE conf calls via RemoteConnectionServices."
Brad Ebinger [Mon, 6 Feb 2017 22:03:29 +0000 (22:03 +0000)]
Merge "Framework fixes to support VoLTE conf calls via RemoteConnectionServices."

7 years agoAdd a mechanism to configure the default A2DP codec priorities
Pavlin Radoslavov [Sun, 5 Feb 2017 23:45:06 +0000 (15:45 -0800)]
Add a mechanism to configure the default A2DP codec priorities

Previously, the relative codec priorities were hard-codec internally.
The new mechanism uses the following configurable resources in
packages/apps/Bluetooth/res/values/config.xml to re-assign the
default codec priorities per device, or to explicitly disable a codec.

 - a2dp_source_codec_priority_sbc
 - a2dp_source_codec_priority_aac
 - a2dp_source_codec_priority_aptx
 - a2dp_source_codec_priority_aptx_hd
 - a2dp_source_codec_priority_ldac

Those values are assigned on startup.
Also, they can be changed per device by using an overlay:
device/<VENDOR>/<DEVICE>/overlay/packages/apps/Bluetooth/res/values/config.xml

Test: Manually streaming to a headset
Change-Id: Ic4da3a51ac73f00cbae731156cb7878c8fadee06

7 years agoMerge "Expand whitelist"
Treehugger Robot [Mon, 6 Feb 2017 20:43:01 +0000 (20:43 +0000)]
Merge "Expand whitelist"

7 years agoFramework fixes to support VoLTE conf calls via RemoteConnectionServices.
Tyler Gunn [Mon, 17 Oct 2016 22:48:19 +0000 (15:48 -0700)]
Framework fixes to support VoLTE conf calls via RemoteConnectionServices.

Fixing some issues with the addExistingConnection and addConference APIs
on ConnectionService.  When a connection manager relays the addition of
an existing connection or a conference to Telecom, it will assign a new
ID to the new connection/conference.  Due to how RemoteCSes work, the
Connection/Conf will be added directly via TelephonyConnectionService and
also via the connection manager's connection service.  Because the ID
changes, we ended up adding these twice. Conferences weren't a problem in
the GSM conference case because the TElephonyConnectionService's
ConnectionServiceWrapper didn't know of the IDs for the children of the
conference.  However, due to how the existing connections work its not the
case for VoLTE conferences.  To mitigate this, I'm passing the original
connection/conference ID to the connection manager via extras (ugh) and
using this to ensure that when the new existing connection/conference is
added to telecom that the same ID is used.  This ensures that we can
properly de-dupe the requests from TelephonyConnectionService and the
connection manager.

Also, there was some missing code in RemoteConnectionService which would
cause it to not properly track existing connections.

Bug: 31464792
Change-Id: I436f4438fd000ea48ebea7ceb75105bd3f456e46

7 years agoRemove uncessary dependency of libhwui
Miao Wang [Wed, 1 Feb 2017 01:26:52 +0000 (17:26 -0800)]
Remove uncessary dependency of libhwui

  - libRS.so is not directly used. It is accessed by libRScpp.so.

Test: mm, and boot-up test on sailfish
Change-Id: I5742d2ee3b34ef09bcb997524408affce1ace8fd
(cherry picked from commit 629fd3d48ba1dbedc6e84e91a26f72fb1fba0e22)

7 years agoFixes Treble marshaling of strings embedded in structures from Java
Andreas Huber [Mon, 6 Feb 2017 19:49:26 +0000 (11:49 -0800)]
Fixes Treble marshaling of strings embedded in structures from Java

Bug: 35038064
Test: hidl_test_java
Change-Id: Ibfdeae06fdf3982d226a7fa87d4c23feeb8e766e

7 years agoMerge changes from topic 'debuggerd_ambient'
Josh Gao [Mon, 6 Feb 2017 18:37:54 +0000 (18:37 +0000)]
Merge changes from topic 'debuggerd_ambient'

* changes:
  zygote: make all capabilities inheritable.
  Revert "zygote: don't drop CAP_SYS_PTRACE from the bounding set."

7 years agoExpand whitelist
Paul Lawrence [Mon, 6 Feb 2017 18:16:38 +0000 (10:16 -0800)]
Expand whitelist

Bug: 35034743
Test: Builds, boots
Change-Id: I5ab77554d23a8498cf32953d447534742683deb8

7 years agoSelf managed CS implementation.
Tyler Gunn [Tue, 31 Jan 2017 18:49:05 +0000 (10:49 -0800)]
Self managed CS implementation.

Add required AIDL changes to pass along connection fails.

Test: Manual
Bug: 34159263
Merged-In: I0a24ff786b2867aa4e19ac79959df54a04983251
Change-Id: I0a24ff786b2867aa4e19ac79959df54a04983251

7 years agoMerge "Add tile information to ACTION_QS_TILE_PREFERENCES intent"
Jason Monk [Mon, 6 Feb 2017 15:34:37 +0000 (15:34 +0000)]
Merge "Add tile information to ACTION_QS_TILE_PREFERENCES intent"

7 years agoMerge "MethodHandles: Expose additional API."
Narayan Kamath [Mon, 6 Feb 2017 14:09:45 +0000 (14:09 +0000)]
Merge "MethodHandles: Expose additional API."

7 years agoMerge "Add bytecodes for invoke-polymorphic"
Orion Hodson [Mon, 6 Feb 2017 10:21:53 +0000 (10:21 +0000)]
Merge "Add bytecodes for invoke-polymorphic"

7 years agoAdd tile information to ACTION_QS_TILE_PREFERENCES intent
Akira Oshimi [Tue, 24 Jan 2017 07:50:53 +0000 (16:50 +0900)]
Add tile information to ACTION_QS_TILE_PREFERENCES intent

One application can provide two or more custom Quick Settings tiles.
But there is no way to know which tile is long-pressed from application
side because ACTION_QS_TILE_PREFERENCES intent that is sent when
long-pressed doesn't have any additional information. So the component
name and state of the tile should be added to the intent.

Bug: 34832801
Test: manual - long press a custom tile

Change-Id: Iaa884cd944f19a2f007cbde645e8f8b1198bffb7

7 years agoHotspot Tile: longclick should open Tethering settings
Jake Whatley [Fri, 3 Feb 2017 01:25:12 +0000 (20:25 -0500)]
Hotspot Tile: longclick should open Tethering settings

Change-Id: I96729a1d74b3552a3b17a31f04043f75d3ace41d
Signed-off-by: Jake Whatley <jacob.whatley91@gmail.com>
7 years agoMerge "Add ShadowShaderAnimation benchmark DO NOT MERGE"
Chris Craik [Fri, 3 Feb 2017 21:44:34 +0000 (21:44 +0000)]
Merge "Add ShadowShaderAnimation benchmark DO NOT MERGE"

7 years agoMerge "Expand whitelist"
Paul Lawrence [Fri, 3 Feb 2017 21:16:59 +0000 (21:16 +0000)]
Merge "Expand whitelist"

7 years agoMerge "Add "wifi lost" disconnect cause."
Tyler Gunn [Fri, 3 Feb 2017 21:03:22 +0000 (21:03 +0000)]
Merge "Add "wifi lost" disconnect cause."

7 years agoMerge "Optimize seccomp"
Paul Lawrence [Fri, 3 Feb 2017 20:14:14 +0000 (20:14 +0000)]
Merge "Optimize seccomp"

7 years agoAdd ShadowShaderAnimation benchmark DO NOT MERGE
Chris Craik [Thu, 2 Feb 2017 23:41:58 +0000 (15:41 -0800)]
Add ShadowShaderAnimation benchmark DO NOT MERGE

Bug: 34809371
Test: manual

New benchmark uses several levels of overdraw from just shadows to benchmark
shadow shader performance.

Tessellation workload is minimal, due to simple rect caster shape.

Change-Id: I572476bd54e04566851f39027ada7254597b37f8

7 years agoAdd "wifi lost" disconnect cause.
Tyler Gunn [Mon, 30 Jan 2017 17:52:55 +0000 (09:52 -0800)]
Add "wifi lost" disconnect cause.

New disconnect cause used when wifi call is dropped due to loss of wifi
connection.

Test: Manual, unit.
Merged-In: Iedaa9f54ff6dbfe90a6ce01fc9eb0aad126d05f0
Bug: 32571209
Change-Id: Iedaa9f54ff6dbfe90a6ce01fc9eb0aad126d05f0

7 years agoExpand whitelist
Paul Lawrence [Fri, 3 Feb 2017 18:10:32 +0000 (10:10 -0800)]
Expand whitelist

Bug: 34908783
Test: Device boots. Unable to test app due to Fake OEM error
Change-Id: I2a0f5646fa39f95875ecb661599d76f4b28a43ac

7 years agoOptimize seccomp
Paul Lawrence [Thu, 2 Feb 2017 20:42:02 +0000 (12:42 -0800)]
Optimize seccomp

Bug: 34946764
Test: Make sure boots, seccomp still blocks, and is faster
Change-Id: I2b4da512f8a9eb8a32f4435561285d42e4b0395f

7 years agoMerge "wifi: WifiConfiguration: add isHomeProviderNetwork field"
Treehugger Robot [Thu, 2 Feb 2017 19:29:51 +0000 (19:29 +0000)]
Merge "wifi: WifiConfiguration: add isHomeProviderNetwork field"

7 years agoMerge "Install seccomp into zygote not init"
Paul Lawrence [Thu, 2 Feb 2017 18:09:17 +0000 (18:09 +0000)]
Merge "Install seccomp into zygote not init"

7 years agoMerge "Compile native network functionis into libandroid_net."
Treehugger Robot [Thu, 2 Feb 2017 17:59:52 +0000 (17:59 +0000)]
Merge "Compile native network functionis into libandroid_net."

7 years agoInstall seccomp into zygote not init
Paul Lawrence [Tue, 31 Jan 2017 17:54:31 +0000 (09:54 -0800)]
Install seccomp into zygote not init

Bug: 34710876
Test: Boots, correct size filter installed, blocks unallowed calls

Change-Id: I3e66e8724ace37f7d30451b98a5e574766252ea0

7 years agoMerge "Add ConnectivityManager.networkCapabilitiesForType()"
Treehugger Robot [Thu, 2 Feb 2017 07:15:03 +0000 (07:15 +0000)]
Merge "Add ConnectivityManager.networkCapabilitiesForType()"

7 years agoMerge "Define Network notification proto constants."
Hugo Benichi [Thu, 2 Feb 2017 06:36:21 +0000 (06:36 +0000)]
Merge "Define Network notification proto constants."

7 years agoMerge "Add test for repeated calls to registerMobileUpstreamRequest()"
Treehugger Robot [Thu, 2 Feb 2017 05:07:22 +0000 (05:07 +0000)]
Merge "Add test for repeated calls to registerMobileUpstreamRequest()"

7 years agoAdd ConnectivityManager.networkCapabilitiesForType()
Erik Kline [Thu, 26 Jan 2017 09:08:28 +0000 (18:08 +0900)]
Add ConnectivityManager.networkCapabilitiesForType()

This static method returns a NetworkCapabilities instance with
transports and capabilities set according to the given legacy type.

Also:
    - add NetworkRequest.Builder.setCapabilities(), to be able to use
      the NetworkCapabilities instances returned from the above
    - update UpstreamNetworkMonitor to make immediate use of this

Test: as follows
    - build (bullhead)
    - flashed
    - booted
    - runtest frameworks-net passes
    - WiFi to DUN upstream tethering works
Bug: 32163131

Change-Id: Idfe1ddd2815c355cbf27cf29eb0e3de177de84e9

7 years agozygote: make all capabilities inheritable.
Josh Gao [Wed, 1 Feb 2017 22:56:09 +0000 (14:56 -0800)]
zygote: make all capabilities inheritable.

Set CapInh to the full set of permitted capabilities, so that the crash
handler can raise ambient capabilities to allow crash_dump to ptrace
zygote-descended children without needing CAP_SYS_PTRACE. CapInh is
restricted by filesystem capabilities if the ambient capability set
hasn't been raised, so this should only have an effect when ambient
capabilities are modified.

Bug: http://b/34853272
Test: debuggerd -b `pidof system_server`
Change-Id: I27d2aa4ab4fc47a183ce2b254be178de4d58840c

7 years agoRevert "zygote: don't drop CAP_SYS_PTRACE from the bounding set."
Josh Gao [Wed, 1 Feb 2017 21:19:56 +0000 (13:19 -0800)]
Revert "zygote: don't drop CAP_SYS_PTRACE from the bounding set."

This reverts commit 59972218f8c0d397fc50400a2b8b9be92ceeaf93.

Bug: http://b/34853272
Test: debuggerd -b `pidof system_server`
Change-Id: I440932210b6e84647c2848bab2e082a58338fb1e

7 years agoMerge "Cleanup hwbinder getService code."
Treehugger Robot [Wed, 1 Feb 2017 20:46:45 +0000 (20:46 +0000)]
Merge "Cleanup hwbinder getService code."

7 years agoMerge "Turn off Dynamic IMS switching for now"
Brad Ebinger [Wed, 1 Feb 2017 19:36:47 +0000 (19:36 +0000)]
Merge "Turn off Dynamic IMS switching for now"

7 years agoMerge "Adds @hide permission granting API for ImsServices"
Treehugger Robot [Wed, 1 Feb 2017 19:24:10 +0000 (19:24 +0000)]
Merge "Adds @hide permission granting API for ImsServices"

7 years agoCleanup hwbinder getService code.
Steven Moreland [Wed, 1 Feb 2017 17:58:00 +0000 (09:58 -0800)]
Cleanup hwbinder getService code.

- Remove memory leaks.
- Make sure proper errors are getting thrown.
- Avoid SIGABRT

Test: hidl_test_java
Change-Id: Ia5680fd0ebf83d0f174bf0d8109bf60327be5f9d

7 years agoAdd bytecodes for invoke-polymorphic
Orion Hodson [Wed, 1 Feb 2017 12:02:16 +0000 (12:02 +0000)]
Add bytecodes for invoke-polymorphic

Bug: 30550796
Test: make
Change-Id: Iacdf0f6553556f04723fd31f76642cabd74a3553

7 years agoAdd test for repeated calls to registerMobileUpstreamRequest()
Erik Kline [Mon, 23 Jan 2017 10:05:28 +0000 (19:05 +0900)]
Add test for repeated calls to registerMobileUpstreamRequest()

Test: as follows
    - built
    - flashed
    - booted
    - runtest framework-net passes
Bug: 32163131

Change-Id: I6d2519601b7a9dcfda2917118214a7ceb9dbbb7f

7 years agoMerge "Fix a use-of-uninitialized-value warning"
Treehugger Robot [Wed, 1 Feb 2017 07:09:50 +0000 (07:09 +0000)]
Merge "Fix a use-of-uninitialized-value warning"

7 years agoMerge "hotspot2: implement hashCode() for classes with equals() function"
Treehugger Robot [Wed, 1 Feb 2017 03:12:12 +0000 (03:12 +0000)]
Merge "hotspot2: implement hashCode() for classes with equals() function"

7 years agoMerge "Print the package name and creator UID in dump()."
Treehugger Robot [Wed, 1 Feb 2017 02:40:59 +0000 (02:40 +0000)]
Merge "Print the package name and creator UID in dump()."

7 years agoMerge "hotspot2: update PasspointConfiguration APIs"
Treehugger Robot [Wed, 1 Feb 2017 02:35:19 +0000 (02:35 +0000)]
Merge "hotspot2: update PasspointConfiguration APIs"

7 years agoMerge changes If420bd66,Icbb161ca,Ida190e5a,Iae75781f
Lorenzo Colitti [Wed, 1 Feb 2017 02:25:42 +0000 (02:25 +0000)]
Merge changes If420bd66,Icbb161ca,Ida190e5a,Iae75781f

* changes:
  ConnectivityServiceTest: remove remaining poll loops.
  Remove assertEventuallyTrue as it's unnecessary.
  Rename AvoidBadWifiTracker to MultinetworkPolicyTracker
  Attempt to make waitForIdleHandler reliable.

7 years agoMerge "Post runnables instead of sending messages."
Treehugger Robot [Wed, 1 Feb 2017 01:26:04 +0000 (01:26 +0000)]
Merge "Post runnables instead of sending messages."

7 years agoFix a use-of-uninitialized-value warning
George Burgess IV [Tue, 24 Jan 2017 23:59:19 +0000 (15:59 -0800)]
Fix a use-of-uninitialized-value warning

This is the result of running ./gen after applying
I90bea0c94cf8d7f92a0d9a4db4e64e7e563bfb09.

Bug: None
Test: Still builds; warning is now gone.
Change-Id: Ibcd5e8dcec96406b529c5790ba0a2822dace25c2

7 years agoTurn off Dynamic IMS switching for now
Brad Ebinger [Tue, 31 Jan 2017 23:36:53 +0000 (15:36 -0800)]
Turn off Dynamic IMS switching for now

The new ImsService implementations must be in place before
ImsResolver/ImsManager can use the new Dynamic ImsService switching
system.

Test: Manual
Change-Id: Ie6454aec08a3f1ebd92271f1c0554c16dc57a868

7 years agoPrint the package name and creator UID in dump().
Jeremy Joslin [Tue, 31 Jan 2017 21:48:21 +0000 (13:48 -0800)]
Print the package name and creator UID in dump().

Add a little more info to help differentiate the registered score
caches during a dump.

Test: adb shell dumpsys network_score
Change-Id: I59cb877b3237aaa85cb564795fb8a41c3d95e315

7 years agohotspot2: implement hashCode() for classes with equals() function
Peter Qiu [Wed, 25 Jan 2017 20:27:08 +0000 (12:27 -0800)]
hotspot2: implement hashCode() for classes with equals() function

According to API guideline, classes that override equals()
function should override hashCode() as well.  Make it so.

Bug: 34198926
Test: frameworks/base/wifi/tests/runtests.sh
Change-Id: I4c3afed78a2fc06cdc26dd04d076f579b3472139

7 years agohotspot2: update PasspointConfiguration APIs
Peter Qiu [Wed, 25 Jan 2017 00:01:43 +0000 (16:01 -0800)]
hotspot2: update PasspointConfiguration APIs

Based on the API guideline, use of public variables are discouraged.
So update PasspointConfiguration and its associated classes to use
private variables with public accessor methods.

While there, cleanup unit tests to reduce code duplications.

Bug: 34627062
Test: frameworks/base/wifi/tests/runtests.sh
Change-Id: I6ea45bbcf03aec01c187425a66094fad6098d75d

7 years agoPost runnables instead of sending messages.
Jeremy Joslin [Tue, 31 Jan 2017 19:51:25 +0000 (11:51 -0800)]
Post runnables instead of sending messages.

Gaining access to a Handler's Looper can be problematic in custom
Handler implementations. Instead of creating a new internal Handler
to dispatch the method calls on we simply post Runnables to the
given Handler.

Test: adb shell am instrument -e class android.net.NetworkRecommendationProviderTest -w com.android.frameworks.coretests/android.support.test.runner.AndroidJUnitRunner
Bug: 34845471
Change-Id: I6fe380a92aebd630781abe5853f94737bcfeaa5f

7 years agowifi: WifiConfiguration: add isHomeProviderNetwork field
Peter Qiu [Fri, 20 Jan 2017 19:26:05 +0000 (11:26 -0800)]
wifi: WifiConfiguration: add isHomeProviderNetwork field

This is a boolean field indicating if the given WiFi network is provided
by a Passpont home provider. This is useful for the Passpoint provisioning
app to determine if the current WiFi connection is allowed to perform
Passpoint subscription or policy updates when it needs to.

This field will be set by PasspointNetworkEvaluator when generating
a WifiConfiguration for a Passpoint network.

Bug: 34467212
Test: make update-api; make -j32
Change-Id: Ia8ec2508de9658a9ee520087359ff2cd68091b63
Merged-In: I47310cda078cbf1f2d1b0a9bf79f487230fb668f

7 years agoMerge "Two digit call support for Claro"
Treehugger Robot [Tue, 31 Jan 2017 19:30:23 +0000 (19:30 +0000)]
Merge "Two digit call support for Claro"

7 years agoMerge "Update the A2DP Codec Config API"
Pavlin Radoslavov [Tue, 31 Jan 2017 19:02:35 +0000 (19:02 +0000)]
Merge "Update the A2DP Codec Config API"

7 years agoUpdate the A2DP Codec Config API
Pavlin Radoslavov [Thu, 26 Jan 2017 00:54:07 +0000 (16:54 -0800)]
Update the A2DP Codec Config API

Previously, the JNI upcall would contain only the current codec config.
In the new API, the upcall contains:
 1. The current codec config
 2. The list of codecs containing the local codecs capabilities
 3. The list of codecs containing the selectable codecs capabilities.
    This list is the intersection of the local codecs capabilities
    and the capabilities of the paired device.

Also, refactored the Java internals to accomodate the extra information:
 * Added new class BluetoothCodecStatus that contains the extra info:
   current codec config, local codecs capabilities and selectable
   codecs capabilities
 * Renamed method getCodecConfig() to getCodecStatus() and return the
   corresponding BluetoothCodecStatus object.
 * Updates to class BluetoothCodecConfig:
   new methods isValid(), getCodecName(), and updated toString()
   so it is more user friendly
 * Removed BluetoothCodecConfig.EXTRA_CODEC_CONFIG and
   EXTRA_PREVIOUS_CODEC_CONFIG.
   The former is superseded by BluetoothCodecStatus.EXTRA_CODEC_STATUS;
   the latter is not really used.

Test: A2DP streaming with headsets and switching the codecs
Change-Id: Ia1af2c22e521e863e28a360610aca49f7e62d31b

7 years agoMerge "hotspot2: enclose conditional statement in parenthesis"
Treehugger Robot [Tue, 31 Jan 2017 18:12:56 +0000 (18:12 +0000)]
Merge "hotspot2: enclose conditional statement in parenthesis"

7 years agoMethodHandles: Expose additional API.
Narayan Kamath [Tue, 31 Jan 2017 14:07:35 +0000 (14:07 +0000)]
MethodHandles: Expose additional API.

Tracks libcore change f4ee5c6dccff8f7e037f95b26767e552ea93d41b.

Test: make checkapi docs
Change-Id: Iac4d4741b9729f3409247b6a3a2b1428f3aa66f0

7 years agoConnectivityServiceTest: remove remaining poll loops.
Lorenzo Colitti [Mon, 30 Jan 2017 08:45:49 +0000 (17:45 +0900)]
ConnectivityServiceTest: remove remaining poll loops.

All the tests are now asynchronous. The only remaining uses of
Thread.sleep() are in the tests that check for NetworkRequest
timeouts.

Test: ConnectivityServiceTest passes 100 times in a row
Bug: 32561414
Change-Id: If420bd66c692a90d5031ee06a888a8cc3b4398a8

7 years agoRemove assertEventuallyTrue as it's unnecessary.
Lorenzo Colitti [Fri, 27 Jan 2017 09:46:03 +0000 (18:46 +0900)]
Remove assertEventuallyTrue as it's unnecessary.

The calls in testAvoidBadWifi are no longer necessary now that
waitForIdle is reliable, and the calls in testPacketKeepalives
are replaced with a wait for the NetworkAgent to disconnect.

Test: ConnectivityServiceTest passes 100 times in a row.
Bug: 32561414
Change-Id: Icbb161ca6e343bd14764a1c9ccfdd14b6cd6803f

7 years agoRename AvoidBadWifiTracker to MultinetworkPolicyTracker
Lorenzo Colitti [Tue, 24 Jan 2017 00:41:36 +0000 (09:41 +0900)]
Rename AvoidBadWifiTracker to MultinetworkPolicyTracker

Rename the class and ensure that all code related to bad wifi
avoidance has names that reflect its purpose.  This will allow us
to reuse most of the the code for other purposes in future CLs.

Test: runtest frameworks-net passes
Bug: 34630278
Change-Id: Ida190e5a79e7a12f4eec07e1a3e4d9d138e83bcd

7 years agoAttempt to make waitForIdleHandler reliable.
Lorenzo Colitti [Wed, 25 Jan 2017 14:09:07 +0000 (23:09 +0900)]
Attempt to make waitForIdleHandler reliable.

The current implementation of IdleableHandlerThread is based on
the assumption that MessageQueue#isIdle will return true iff the
message loop has finished processing its messages.

Unfortunately, this is incorrect: isIdle returns true iff are no
more messages waiting in the queue; thus, it will also return
true while it is processing the last message before going idle.

Instead of using idle handlers, take the simpler approach of
posting a runnable to the message loop and waiting for it to
be processed.

Test: see other CLs in this patch series
Bug: 31479480
Change-Id: Iae75781f067b762c8653a488a5e4d5ee0c789e01

7 years agoMerge "ApfTest: fix flaky testApfFilterRa."
Hugo Benichi [Tue, 31 Jan 2017 05:53:51 +0000 (05:53 +0000)]
Merge "ApfTest: fix flaky testApfFilterRa."

7 years agoApfTest: fix flaky testApfFilterRa.
Hugo Benichi [Mon, 19 Dec 2016 05:50:52 +0000 (14:50 +0900)]
ApfTest: fix flaky testApfFilterRa.

testApfFilterRa is failing with probabiliy 1/10 ~ 1/15 on the following
assert: assertDrop(program, packet, lifetime/6), for lifetime values
that are multiple of 6, where 6 is the hardcoded fraction of RA lifetime
to filter in ApfFilter.java.

When the lifetime is not a multiple of 6, the remainder of 1 to 5
seconds gives enough margin so that when the APF program is simulated
the faked lifetime of the program is less than lifetime/6 away and the
packet is dropped.

However for lifetimes which are exact multiples of 6, this margin is
always 0s and that result in nondeterminism in the result. This is
consistent with the obervation that the only failed assert was for a
lifetime of 300s, the only multiple of 6.

This can be observed by detecting the age limit at which the filter
stops dropping packet oscillating between lifetime/6 and lifetime/6 + 1
for lifetimes which are multiple of 6.

This patch fixes the flakyness by freezing the flow of time in tests so
that the expected filter age threshold is consistent and stable.

Test: no failure observed in 1000 runs.
Bug: 32561414
Change-Id: I5251d047039f34b82ce8a5d20ae46563e1e0cce8

7 years agohotspot2: enclose conditional statement in parenthesis
Peter Qiu [Tue, 31 Jan 2017 01:47:53 +0000 (17:47 -0800)]
hotspot2: enclose conditional statement in parenthesis

Couple unit tests are failing due to conditional statements
not enclosed in the parenthesis, which causes unexpected
returned value for the equals function.  So fix it.

Bug: 34764955
Test: frameworks/base/wifi/tests/runtests.sh
Test: frameworks/opt/net/wifi/tests/wifitests/runtests.sh
Change-Id: Iaf0151803cb62eb8af308e5de40558617f84cd2f

7 years agoMerge "[AWARE] Minor hardening of manager."
Etan Cohen [Tue, 31 Jan 2017 01:19:48 +0000 (01:19 +0000)]
Merge "[AWARE] Minor hardening of manager."

7 years agoMerge changes I8eca5f24,I9d79c67c
Jerry Zhang [Tue, 31 Jan 2017 00:41:55 +0000 (00:41 +0000)]
Merge changes I8eca5f24,I9d79c67c

* changes:
  Check and send usb state broadcast when devices boots
  Only send usb disconnect intents after boot.

7 years agoCheck and send usb state broadcast when devices boots
Badhri Jagan Sridharan [Tue, 24 Jan 2017 00:28:38 +0000 (16:28 -0800)]
Check and send usb state broadcast when devices boots

Broadcasts are not sent until BOOT_COMPLETED is received.
When the device boots with usb plugged, the UPDATE_STATE
for connted and configured would be received way before
BOOT_COMPLETED is received. Since, devices can continue
to stay in same usb configuration(set through persistent
property trigger), check and send a usb state broadcast
if needed.

BUG: 34451258
Test: Manually checked to verify the broadcasts sent.
Change-Id: I8eca5f241ad758f5c4f1afa6c9807868dfd6195a

7 years agoOnly send usb disconnect intents after boot.
Jerry Zhang [Mon, 23 Jan 2017 22:14:09 +0000 (14:14 -0800)]
Only send usb disconnect intents after boot.

This prevents cases where the device starts up and attempts
to change the config, but intents cannot yet be sent.

A patch to resend this skipped intent is necessary to
finish fixing the bug.

Bug: 34451258
Test: Check that usb state is properly handled on boot
Change-Id: I9d79c67ce8982fd5b43fa40781b66bb9ae820694

7 years agoAdds @hide permission granting API for ImsServices
Brad Ebinger [Tue, 24 Jan 2017 00:44:23 +0000 (16:44 -0800)]
Adds @hide permission granting API for ImsServices

Adds the @hide API to grant ImsServices default permissions
using the PackageManager and DefaultPermissionGrantPolicy APIs.

Test: Manual
Bug: 32460405
Change-Id: Ia143ede42623b4c2203261fc6e83624d73c478f1

7 years ago[AWARE] Minor hardening of manager.
Etan Cohen [Mon, 30 Jan 2017 21:01:50 +0000 (13:01 -0800)]
[AWARE] Minor hardening of manager.

Callbacks are possible even if no listener was defined.
Perform a check and print error (as opposed to crashing).

Bug: 32315992
Test: unit-tests + integration tests passing
Change-Id: Ia5eb81448ace6b88abcd448e4f378da6530ac630

7 years agoMerge "SFR sims should treat Spain as roaming"
Treehugger Robot [Mon, 30 Jan 2017 20:03:11 +0000 (20:03 +0000)]
Merge "SFR sims should treat Spain as roaming"

7 years agoMerge "Bluetooth: add getDiscoveryEndMillis() call"
Marie Janssen [Mon, 30 Jan 2017 17:46:17 +0000 (17:46 +0000)]
Merge "Bluetooth: add getDiscoveryEndMillis() call"

7 years agoMerge "Introduce DEBUG_JAVA_DEBUGGABLE."
Nicolas Geoffray [Mon, 30 Jan 2017 15:09:44 +0000 (15:09 +0000)]
Merge "Introduce DEBUG_JAVA_DEBUGGABLE."

7 years agoDefine Network notification proto constants.
Hugo Benichi [Thu, 8 Dec 2016 00:36:52 +0000 (09:36 +0900)]
Define Network notification proto constants.

This patch defines 4 new constants in metrics_constants.proto
corresponding to the 4 network notifications shown by
NetworkNotificationManager.

Estimates for the average number of notifications:
  - SIGN_IN:        3/day/device
  - NO_INTERNET:    5/day/device
  - LOST_INTERNET:  5/day/device
  - NETWORK_SWITCH: 3/day/device (turned off right now)

Test: no functional change
Bug: 32198726
Bug: 33030620

(cherry picked from commit e1cbf15f1ad7f2f7e1cc1e96e6a82c3fa2f70bdf)

Change-Id: I6020c283329f752170cef451725b4e5287509bc5

7 years agoMerge "KeySetManagerService: prevent NPE"
Treehugger Robot [Sun, 29 Jan 2017 17:14:36 +0000 (17:14 +0000)]
Merge "KeySetManagerService: prevent NPE"

7 years agoBluetooth: add getDiscoveryEndMillis() call
Marie Janssen [Tue, 24 Jan 2017 22:09:59 +0000 (14:09 -0800)]
Bluetooth: add getDiscoveryEndMillis() call

Method to tell when the adapter finished (or will finish) being in
discovery mode.

Test: compiles and still can scan
Bug: 34395439
Change-Id: I41b48c2b934c0a1d5e1727cec08f3f762e3cb309

7 years agoMerge changes from topic 'enterprise-client-chain'
Treehugger Robot [Sat, 28 Jan 2017 01:38:30 +0000 (01:38 +0000)]
Merge changes from topic 'enterprise-client-chain'

* changes:
  Test passing null cert/keys to WifiEnterpriseConfig
  Account for null client certificate
  Add a client chain to WifiEnterpriseConfig

7 years agoMerge "Adds @hide ImsResolver API and device overlays (2/3)"
Treehugger Robot [Fri, 27 Jan 2017 22:55:15 +0000 (22:55 +0000)]
Merge "Adds @hide ImsResolver API and device overlays (2/3)"

7 years agoMerge "Fix a bunch of repeated reads of a ro.* property"
Treehugger Robot [Fri, 27 Jan 2017 22:49:03 +0000 (22:49 +0000)]
Merge "Fix a bunch of repeated reads of a ro.* property"

7 years agoMerge "Handles #9 as Dial number"
Treehugger Robot [Fri, 27 Jan 2017 21:30:42 +0000 (21:30 +0000)]
Merge "Handles #9 as Dial number"

7 years agoAdds @hide ImsResolver API and device overlays (2/3)
Brad Ebinger [Tue, 24 Jan 2017 00:20:33 +0000 (16:20 -0800)]
Adds @hide ImsResolver API and device overlays (2/3)

- Adds config_ims_package device overlay, which specifies the default
package name to bind to.
- Adds config_dynamic_bind_ims device overlay, which determines
whether or not the new dynamic binding system is used.
- Adds @hide TelephonyManager API to interface with the ImsResolver

Test: Manual
Merged-In: I82a41da00e6da34629a40db431f13b968dfafe2e
Change-Id: I82a41da00e6da34629a40db431f13b968dfafe2e

7 years agoTest passing null cert/keys to WifiEnterpriseConfig
Paul Stewart [Fri, 27 Jan 2017 20:03:47 +0000 (12:03 -0800)]
Test passing null cert/keys to WifiEnterpriseConfig

Ensure that null certificates and keys don't crash.

Bug: 34765004
Test: This is a test
Change-Id: I439b4f985c1b88ad4a9b58ee6f4eb4f90bd81246

7 years agoAccount for null client certificate
Paul Stewart [Fri, 27 Jan 2017 17:37:17 +0000 (09:37 -0800)]
Account for null client certificate

If a null certificate is passed to setClientKeyEntry() we should
not pass a non-null array with a single null element to the
setClientKeyEntryWithCertificateChain helper method.  Instead we
should pass a null array.

Cherry-pick of 410a3498ac28dccf69212d94a533040893c7ce0c

Bug: 34765004
Test: cts-tradefed run cts -d --module CtsNetTestCases --test android.net.wifi.cts.WifiEnterpriseConfigTest
Change-Id: I02793b4b29bc7325f98833c58bf652ba68353827

7 years agoMerge "Use the new radio_metadata wrapper for safer memory management."
Tomasz Wasilczyk [Fri, 27 Jan 2017 20:26:06 +0000 (20:26 +0000)]
Merge "Use the new radio_metadata wrapper for safer memory management."

7 years agoIntroduce DEBUG_JAVA_DEBUGGABLE.
Nicolas Geoffray [Tue, 20 Dec 2016 14:05:05 +0000 (14:05 +0000)]
Introduce DEBUG_JAVA_DEBUGGABLE.

For notifying ART it will run a debuggable app.

Also rename ENABLE_DEBUGGER to ENABLE_JDWP.

Test: builds and runs
bug: 28769520
Change-Id: Ic096a176edfd5bf0bbe92b8367fbaa687a07d284

7 years agoMerge "Switch to listening for all network changes."
Treehugger Robot [Fri, 27 Jan 2017 19:07:06 +0000 (19:07 +0000)]
Merge "Switch to listening for all network changes."

7 years agoFix a bunch of repeated reads of a ro.* property
John Reck [Tue, 20 Sep 2016 21:24:21 +0000 (14:24 -0700)]
Fix a bunch of repeated reads of a ro.* property

SystemProperties.get() is not particularly fast,
especially if a string is returned. Since ro.* values
are unable to be changed, there's no need to
continously re-query them. Cache the value at
static init time to trivially fix this.

Test: refactoring CL.
Change-Id: Iccb021d3cb2ba3a4a1d0048ddec6811bb7409eec
(cherry picked from commit aa67f684ff43c81e3280c846245ec6ebe907787e)

7 years agoMerge "[Frameworks] Add an 'am' cmd option to enable streaming in profiling."
Treehugger Robot [Fri, 27 Jan 2017 18:49:43 +0000 (18:49 +0000)]
Merge "[Frameworks] Add an 'am' cmd option to enable streaming in profiling."

7 years agoAdd a client chain to WifiEnterpriseConfig
Paul Stewart [Wed, 25 Jan 2017 03:36:05 +0000 (19:36 -0800)]
Add a client chain to WifiEnterpriseConfig

Add a list of supporting certificates to be presented in the process
of presenting client credentials.

Cherry-pick of e3511767169357a1409119b5666c62d50e005583

Bug: 34688653
Test: Compile, unit tests
Change-Id: I6afd8baf67312e8ddaaeefd26f30dacc51aa33bb

7 years agoUse the new radio_metadata wrapper for safer memory management.
Tomasz Wasilczyk [Mon, 23 Jan 2017 22:36:47 +0000 (14:36 -0800)]
Use the new radio_metadata wrapper for safer memory management.

Bug: b/34054813
Test: VTS, manual - both done in internal branch
Change-Id: I5f4962f86b7732e14c2584a562e9737c92310697

7 years agoMerge "ConnectivityServiceTest: fix flaky tests"
Lorenzo Colitti [Fri, 27 Jan 2017 09:40:46 +0000 (09:40 +0000)]
Merge "ConnectivityServiceTest: fix flaky tests"

7 years agoMerge "ConnectivityServiceTest: remove flaky waitForIdle test."
Lorenzo Colitti [Fri, 27 Jan 2017 09:38:51 +0000 (09:38 +0000)]
Merge "ConnectivityServiceTest: remove flaky waitForIdle test."

7 years agoHandles #9 as Dial number
koji.x.shigehara [Fri, 20 Nov 2015 10:42:53 +0000 (19:42 +0900)]
Handles #9 as Dial number

Adding #9 as a country-specific exception of MMI codes so that
we can treat it as Dial number. This config change is necessary
to fulfill carrier's requirement.

Test: manual - check dialing #9
Bug: 34755732
Change-Id: I485e48cf30610ff012f84d43935454e693b0d94d