OSDN Git Service
Robert Carr [Wed, 21 Jun 2017 17:25:05 +0000 (10:25 -0700)]
ListPopupWindow: Wrap new bounds checking in targetSdk check.
While a negative height is pretty silly, crashing apps on
the new version of android makes them sad.
Test: Existing CTS passes.
Bug:
62434804
Change-Id: I5fc3fc50fb6ccfa9e96f38ded4fb8e338f263f09
Philip P. Moltmann [Tue, 20 Jun 2017 15:42:13 +0000 (15:42 +0000)]
Merge "Correct lastId on onCreate with unexpected state" into oc-dev
TreeHugger Robot [Tue, 20 Jun 2017 04:24:08 +0000 (04:24 +0000)]
Merge "Hack around b/
36066697." into oc-dev
TreeHugger Robot [Tue, 20 Jun 2017 01:38:36 +0000 (01:38 +0000)]
Merge "Remove legacy advertiser from map after it's no longer used" into oc-dev
Adam Lesinski [Tue, 20 Jun 2017 01:07:06 +0000 (01:07 +0000)]
Merge "Fix potential race in Theme creation" into oc-dev
TreeHugger Robot [Tue, 20 Jun 2017 00:27:32 +0000 (00:27 +0000)]
Merge "Fixed updateAutofillValue() when mText is not set." into oc-dev
Philip P. Moltmann [Mon, 19 Jun 2017 19:57:13 +0000 (12:57 -0700)]
Correct lastId on onCreate with unexpected state
onCreate might be called with a state not produced by
onSaveInstanceState. In this case the last autofill ID is not set, so we
incorrectly restore it to NO_ID instead of LAST_APP_AUTOFILL_ID
Change-Id: Id40c06bf223c0e3c6235b2d50779e3f4532898b5
Fixes:
62296699
Test: cts-tradefed run cts-dev -m CtsAutoFillServiceTestCases
(cherry picked from commit
1266d08be53c6ee2494d18363092c68825ef449e)
TreeHugger Robot [Mon, 19 Jun 2017 23:50:23 +0000 (23:50 +0000)]
Merge "Fix issue #
62787070: restart due to NPE in JobServiceContext..." into oc-dev
Siarhei Vishniakou [Mon, 19 Jun 2017 23:47:48 +0000 (23:47 +0000)]
Merge "Revive DeadZone" into oc-dev
Jakub Pawlowski [Wed, 14 Jun 2017 01:59:07 +0000 (18:59 -0700)]
Remove legacy advertiser from map after it's no longer used
Test: manual
Bug:
62597369
Change-Id: Ief1e1b054f29b59b1ed72514b138d4f209f9847a
TreeHugger Robot [Mon, 19 Jun 2017 23:29:04 +0000 (23:29 +0000)]
Merge "Make --no-window-animation flag apply to animators, too" into oc-dev
TreeHugger Robot [Mon, 19 Jun 2017 23:16:14 +0000 (23:16 +0000)]
Merge "Added back removed ctor from DisplayEventReceiver" into oc-dev
Adam Lesinski [Mon, 12 Jun 2017 19:22:10 +0000 (12:22 -0700)]
Fix potential race in Theme creation
Bug:
38190555
Test: none
Change-Id: Id627bd6088dc469baffb1abb8310cd9e992996aa
TreeHugger Robot [Mon, 19 Jun 2017 21:52:14 +0000 (21:52 +0000)]
Merge "[Companion] Fix NPE in associate" into oc-dev
Dianne Hackborn [Mon, 19 Jun 2017 21:24:51 +0000 (14:24 -0700)]
Fix issue #
62787070: restart due to NPE in JobServiceContext...
...handleOpTimeoutLocked
Don't try to print the running job if it is null.
Also, the timeout message should be scheduled with the actually
running job it is for, so we can ignore it if we happen to
dispatch one after the job is over.
Test: bit CtsJobSchedulerTestCases:*
Change-Id: I7bc55f55da645a9e116d3f0ee02f2ee115383ea9
Josh Gao [Mon, 19 Jun 2017 20:38:20 +0000 (13:38 -0700)]
Hack around b/
36066697.
Avoid calling libbinder's static destructors by calling _exit instead
of returning in main.
Bug: http://b/
36066697
Test: treehugger
Change-Id: I81dd50453f1ea64d97ec97d207fff732b3c62512
Siarhei Vishniakou [Sat, 3 Jun 2017 00:20:34 +0000 (17:20 -0700)]
Revive DeadZone
Override dispatchTouchEvent for the root FrameLayout
of NavigationBar to process ACTION_OUTSIDE MotionEvents
and dispatch directly to DeadZone to keep track of the
most recent outside tap.
Clarified documentation of ACTION_OUTSIDE behaviour.
Bug:
37552674
Test: open IME, tap any key, then quickly tap on top half
of the home button. The home button tap is ignored
and device does not go to homescreen.
Change-Id: Icb5cf6c76959f3514b8b94c09e38cc5434f31b23
TreeHugger Robot [Mon, 19 Jun 2017 21:01:42 +0000 (21:01 +0000)]
Merge "Do not update currently open tooltip." into oc-dev
Wale Ogunwale [Mon, 19 Jun 2017 20:53:32 +0000 (13:53 -0700)]
Added back removed ctor from DisplayEventReceiver
...because some app can't help but access hidden APIs...
Test: builds
Change-Id: Ia1d4a71b46a53b44e3ebe5844e446009785a90ac
Fixes:
62719884
TreeHugger Robot [Mon, 19 Jun 2017 20:32:13 +0000 (20:32 +0000)]
Merge "Reduce no-op notification log messages." into oc-dev
Felipe Leme [Mon, 19 Jun 2017 16:45:48 +0000 (09:45 -0700)]
Fixed updateAutofillValue() when mText is not set.
updateAutofillValue() was crashing some apps when the mText was not set at the
time it was called. One solution would be to not set mText at all - since the
Autofill Service should rely only on getAutofillValue() - but that could break
existing services.
Hence, a safer solution is to set that field if it's null.
Test: existing CtsAutoFillServiceTestCases tests pass
Test: manual verification using Fly Delta app
Fixes:
62751039
Change-Id: I91a8e0ed5db4148f5eb5729b8e254aa3531f15e4
Dianne Hackborn [Mon, 19 Jun 2017 17:04:04 +0000 (17:04 +0000)]
Merge "Maybe fix issue #
62199092: Alarm didn't ring, upcoming alarm..." into oc-dev
Siyamed Sinir [Mon, 19 Jun 2017 16:53:11 +0000 (16:53 +0000)]
Merge "Fix crash during TextView.onCreateContextMenu" into oc-dev
TreeHugger Robot [Mon, 19 Jun 2017 15:26:11 +0000 (15:26 +0000)]
Merge "Fix flash of ambient on fp unlock" into oc-dev
Todd Kennedy [Sun, 18 Jun 2017 17:06:17 +0000 (17:06 +0000)]
Merge "Track both framework and app overlays per package" into oc-dev
TreeHugger Robot [Sat, 17 Jun 2017 05:57:54 +0000 (05:57 +0000)]
Merge "Import translations. DO NOT MERGE" into oc-dev
TreeHugger Robot [Sat, 17 Jun 2017 02:49:55 +0000 (02:49 +0000)]
Merge "Only preload selected pages" into oc-dev
TreeHugger Robot [Sat, 17 Jun 2017 02:15:02 +0000 (02:15 +0000)]
Merge "APs from old scans should not have RSSI values set." into oc-dev
Bill Yi [Sat, 17 Jun 2017 01:39:16 +0000 (18:39 -0700)]
Import translations. DO NOT MERGE
Change-Id: Id2e2ab20c78bfcf7ab82807c33466fed3116236b
Auto-generated-cl: translation import
Dianne Hackborn [Sat, 17 Jun 2017 00:59:35 +0000 (17:59 -0700)]
Maybe fix issue #
62199092: Alarm didn't ring, upcoming alarm...
...notification was stuck
Don't deliver to a ProcessRecord that has been killed, instead
go through creating a new process.
Test: manual
Change-Id: I74eb0843200b5b99d6496e4b3f6eef5ed38c926e
TreeHugger Robot [Sat, 17 Jun 2017 00:52:57 +0000 (00:52 +0000)]
Merge "Revert "Trim text from autofillvalue when checking for empty."" into oc-dev
Phil Weaver [Sat, 17 Jun 2017 00:45:59 +0000 (00:45 +0000)]
Merge "Separate a11y and autofill view ids" into oc-dev
Casey Burkhardt [Sat, 17 Jun 2017 00:40:57 +0000 (00:40 +0000)]
Merge "Resolve 3 inconsistencies in accessibility button API - Unify logic for detecting availability of the accessibility button - Ensure the initial visibility state is propagated to A11yMS - Ensure services only receive availability callbacks for changes" into oc-dev
Eugene Susla [Sat, 17 Jun 2017 00:20:41 +0000 (17:20 -0700)]
[Companion] Fix NPE in associate
The root cause of the exception was that the activity destroy listener was
reacting to any activity being destroyed instead of just the one used with
the CompanionDeviceManager
Fixes:
62549525
Test: Ensure the attached bug no longer reproduces
Change-Id: I2f977e9ac9176247f5be9d08d19b3875f2b4a703
Bryce Lee [Fri, 16 Jun 2017 23:15:03 +0000 (23:15 +0000)]
Merge "Create native Surface object on updateWindow for legacy apps." into oc-dev
Felipe Leme [Fri, 16 Jun 2017 22:57:58 +0000 (22:57 +0000)]
Revert "Trim text from autofillvalue when checking for empty."
This reverts commit
c5d4e068f2a0c47dc69b72200f8a9fc4bd5a2c28.
Change-Id: I4027cf0f91ebb9032c21efad7025118891188bc6
TreeHugger Robot [Fri, 16 Jun 2017 22:50:42 +0000 (22:50 +0000)]
Merge "Trim text from autofillvalue when checking for empty." into oc-dev
Tyler Freeman [Fri, 16 Jun 2017 22:43:12 +0000 (22:43 +0000)]
DO NOT MERGE ANYWHERE Vibrator: Allow priority vibrations in low-power mode for accessibility. am:
79613777df -s ours
am:
cf82b09998 -s ours
Change-Id: I128c5a5b07fc2aef5ada9118e3cc491a87970758
Tyler Freeman [Fri, 16 Jun 2017 22:36:26 +0000 (22:36 +0000)]
DO NOT MERGE ANYWHERE Vibrator: Allow priority vibrations in low-power mode for accessibility.
am:
79613777df -s ours
Change-Id: I965bac24a8d62b44b08679bda58b8ba37dbed5b4
TreeHugger Robot [Fri, 16 Jun 2017 21:57:20 +0000 (21:57 +0000)]
Merge "Log last home activity and result we tried to start" into oc-dev
TreeHugger Robot [Fri, 16 Jun 2017 21:55:28 +0000 (21:55 +0000)]
Merge "Disable notification sounds on TV" into oc-dev
TreeHugger Robot [Fri, 16 Jun 2017 21:45:39 +0000 (21:45 +0000)]
Merge "Use an optimized findViewNodeByAutofillId(id) to get just one node." into oc-dev
Phil Weaver [Thu, 15 Jun 2017 17:10:06 +0000 (10:10 -0700)]
Separate a11y and autofill view ids
Autofill seems to need IDs to be preserved across things
like configuration changes, while accessibility will not
function without views with unique ids. Separating out the
two types of IDs. We can re-combine them once it's clear
that both demands can be satisfied.
Bug:
62301218
Test: Run a11y and autofill CTS, and verify that the play
store app functions with TalkBack after a screen rotation.
Change-Id: I17a99de2874768fc0ade3aa354130dd1f6b4cd7e
Winson Chung [Fri, 16 Jun 2017 21:32:39 +0000 (21:32 +0000)]
Merge changes I20bdab1b,Id82f9081 into oc-dev
* changes:
Fix backwards flag causing crash.
Fixing issue with PiP aspect ratio being clobbered.
Bryce Lee [Fri, 16 Jun 2017 14:20:34 +0000 (07:20 -0700)]
Create native Surface object on updateWindow for legacy apps.
There are some apps that use the Surface object itself to indicate
changes. As a result, recycling the existing Surface object for
updates can lead to such apps ignoring events such as size changes.
This changelist restores the original behavior for legacy apps, where
the underlying native Surface object is recreated during updates.
Fixes:
62108743
Test: go/wm-smoke
Test: Open affected application, observe expansion to fullscreen when
nav bar disappears. Rotate to other orientation and observe
expansion to fullscreen.
Change-Id: I874602b6b8686c6ecb05cf7b1a04ec4b700ad3f9
Todd Kennedy [Fri, 16 Jun 2017 20:55:13 +0000 (13:55 -0700)]
Track both framework and app overlays per package
Always bundle framework and app overlays. The old implementation, where
framework and app overlays were tracked independently, lead to an error
in the following scenario:
1. Enable app overlay -> change reflected in app
2. Enable framework overlay -> error: no change reflected in app
3. Disable app overlay -> change reflected in app, including framework
overlay
This change also leads to better architecture since the package manager
no longer needs to know that an app's overlays consist of both framework
and app specific overlays. Instead, that knowledge is handled by the
overlay manager.
Also, correct indentation in "cmd package dump packages" output and
remove obsolete constant DUMP_ENABLED_OVERLAYS.
Test: Manual
Change-Id: I707fc00052a15b22fb8c17e6155732520e6b2e52
Bug:
62680061
Felipe Leme [Fri, 16 Jun 2017 20:24:32 +0000 (13:24 -0700)]
Trim text from autofillvalue when checking for empty.
Test: VirtualContainerActivityTest#testSaveDialogNotShownWhenBackIsPressed
Fixes:
62667931
Change-Id: I7e673eedd39fff7af89c694d0bfc5a495f14ea26
Tyler Freeman [Fri, 5 May 2017 00:23:35 +0000 (17:23 -0700)]
DO NOT MERGE ANYWHERE Vibrator: Allow priority vibrations in low-power mode for accessibility.
This is mainly for Wear devices that don't have speakers.
Bug:
37543478
Bug:
37257715
Merged-In: Ia6dffa7e528886e7c84390dbb33b44f298c7afa4
Change-Id: Ia6dffa7e528886e7c84390dbb33b44f298c7afa4
Rob Carr [Fri, 16 Jun 2017 19:55:07 +0000 (19:55 +0000)]
Merge "ViewRootImpl: More null checks for performTraversals." into oc-dev
Wale Ogunwale [Fri, 16 Jun 2017 19:42:51 +0000 (12:42 -0700)]
Log last home activity and result we tried to start
To help debug ANR.
Bug:
38121026
Test: manual
Change-Id: I484301810d07dcf7330dfdbab7310eac14638b90
Fyodor Kupolov [Fri, 16 Jun 2017 19:51:00 +0000 (19:51 +0000)]
Merge "Always log new user state to event log" into oc-dev
Tony Mantler [Fri, 16 Jun 2017 17:50:00 +0000 (10:50 -0700)]
Disable notification sounds on TV
Bug:
38495875
Test: BuzzBeepBlinkTest
Change-Id: Iffebe2063de3d73c1538a3bbbe9edd2233005ba3
Siyamed Sinir [Fri, 16 Jun 2017 01:22:31 +0000 (18:22 -0700)]
Fix crash during TextView.onCreateContextMenu
Test: bit CtsWidgetTestCases:TextViewTest
Bug:
62450393
Change-Id: I162c28d253ca197eba9c4c148c76179c2880daa5
TreeHugger Robot [Fri, 16 Jun 2017 19:23:37 +0000 (19:23 +0000)]
Merge "Throw IAE when service add null AutofillIds on SaveInfo methods." into oc-dev
Felipe Leme [Fri, 16 Jun 2017 18:49:48 +0000 (11:49 -0700)]
Use an optimized findViewNodeByAutofillId(id) to get just one node.
Session was using findViewNodesByAutofillIds(ids) before, which not only was
not optimal, but error prone (for example, it could return a non-empty array
with an empty value).
Test: CtsAutoFillServiceTestCases pass
Fixes:
62532979
Change-Id: If984f1263cc3f2aac1d1e098687fe02d73c55211
Philip P. Moltmann [Fri, 16 Jun 2017 18:30:52 +0000 (11:30 -0700)]
Only preload selected pages
Otherwise if only some pages are selected all (non-selected) pages in
between are preloaded.
Test: Selected only some pages of a large document and observed which
pages got preloaded
cts-tradefed run cts-dev -m Print
Fixes:
62296301
Change-Id: I3a97cfa6991e5a95ff73628c4bb540c629160c0c
Vladislav Kaznacheev [Fri, 16 Jun 2017 18:31:23 +0000 (11:31 -0700)]
Do not update currently open tooltip.
Currently if View.setTooltipText is called while
the tooltip is being shown for that view, it will
update the displayed text. The tooltip then will
resize to wrap around the new text, but not change
its position. This looks confusing if the new text
is significantly shorter or longer.
Removing this functionality until proper
re-positioning is implemented.
Bug:
38491655
Test: android.view.cts.TooltipTest passes
Change-Id: I79689288185888854b992b89e19fe381d3ac50e4
TreeHugger Robot [Fri, 16 Jun 2017 18:07:46 +0000 (18:07 +0000)]
Merge "enable/disable dexopt for instant apps remotely" into oc-dev
Winson Chung [Fri, 16 Jun 2017 18:07:29 +0000 (11:07 -0700)]
Fix backwards flag causing crash.
Bug:
62692677
Test: Use an activity options that requires the bitmap copy, ensure
that it does not crash.
Change-Id: I20bdab1b91dfe47f7fe134fd17fe104eb4b27ec1
Fyodor Kupolov [Fri, 16 Jun 2017 17:45:51 +0000 (10:45 -0700)]
Always log new user state to event log
Previously it wasn't logged for stopping users...
Test: manual
Bug:
38121026
Change-Id: I0fc37224ce1043745dfffcc78639f003c35b402e
Winson Chung [Fri, 16 Jun 2017 17:52:40 +0000 (10:52 -0700)]
Fixing issue with PiP aspect ratio being clobbered.
- We should only initialize the initial aspect ratio to the default
aspect ratio once, instead of when the resources are reloaded.
Otherwise, they will be clobbered when the configuration changes.
This is a regression from ag/
2295004.
Bug:
62702553
Test: Set PiP aspect ratio from landscape-locked activity, ensure that
it stays in that aspect ratio after the pip menu dismisses.
Change-Id: Id82f9081c25c7d6482cf7bcf0268fedb9068eabe
Marco Nelissen [Fri, 16 Jun 2017 17:38:19 +0000 (17:38 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours am:
ce4c7df6a0 -s ours am:
035e2a5c37 -s ours am:
4e5ed845c7 -s ours am:
3cd244e872 -s ours am:
a9bf75f50c -s ours am:
21caf4aee4 -s ours
am:
009b3c6aec -s ours
Change-Id: Ied441cf591abb8933a5601c6d36066e0288dfbd7
Marco Nelissen [Fri, 16 Jun 2017 17:24:37 +0000 (17:24 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours am:
ce4c7df6a0 -s ours am:
035e2a5c37 -s ours am:
4e5ed845c7 -s ours am:
3cd244e872 -s ours am:
a9bf75f50c -s ours
am:
21caf4aee4 -s ours
Change-Id: I25f1f4394b6a5a6b3708df13be41299865e76d52
Marco Nelissen [Fri, 16 Jun 2017 17:12:09 +0000 (17:12 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours am:
ce4c7df6a0 -s ours am:
035e2a5c37 -s ours am:
4e5ed845c7 -s ours am:
3cd244e872 -s ours
am:
a9bf75f50c -s ours
Change-Id: Ib75388665c3586c09e963917bc54b020aae22dce
Marco Nelissen [Fri, 16 Jun 2017 16:59:57 +0000 (16:59 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours am:
ce4c7df6a0 -s ours am:
035e2a5c37 -s ours am:
4e5ed845c7 -s ours
am:
3cd244e872 -s ours
Change-Id: I256e1413df73fd577708d5dcb84074e3a8ed61ab
Marco Nelissen [Fri, 16 Jun 2017 16:51:14 +0000 (16:51 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours am:
ce4c7df6a0 -s ours am:
035e2a5c37 -s ours
am:
4e5ed845c7 -s ours
Change-Id: Iae9f79e4b91c4ef5e844259709e3055607a781b8
Marco Nelissen [Fri, 16 Jun 2017 16:41:20 +0000 (16:41 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours am:
ce4c7df6a0 -s ours
am:
035e2a5c37 -s ours
Change-Id: Ib6f84a832e27329e0eabf000b89e93145c43a31a
Marco Nelissen [Fri, 16 Jun 2017 16:26:30 +0000 (16:26 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours am:
45cd809caa -s ours
am:
ce4c7df6a0 -s ours
Change-Id: I1fb798481ca56b8c81386792b5642a544aae1c7e
Marco Nelissen [Fri, 16 Jun 2017 16:19:02 +0000 (16:19 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours am:
23118a9b77 -s ours
am:
45cd809caa -s ours
Change-Id: Idd7b2d2536d1c25072794ba228f33fc2459f2aaa
Marco Nelissen [Fri, 16 Jun 2017 16:11:01 +0000 (16:11 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours am:
49fc959be3 -s ours
am:
23118a9b77 -s ours
Change-Id: Id314511cf3e6af08e0f00e84e8945ea404b85c6a
Marco Nelissen [Fri, 16 Jun 2017 16:02:19 +0000 (16:02 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours am:
398d50feeb -s ours
am:
49fc959be3 -s ours
Change-Id: I0eb248911ce1a582a4939e2bd5419659c586128f
Marco Nelissen [Fri, 16 Jun 2017 15:51:48 +0000 (15:51 +0000)]
libmedia_jni.so doesn't need libjhead.so am:
9a4a34afd8 -s ours
am:
398d50feeb -s ours
Change-Id: Ide31821909dde91443cdf650216a6ada7277ce9a
Marco Nelissen [Fri, 16 Jun 2017 15:43:46 +0000 (15:43 +0000)]
libmedia_jni.so doesn't need libjhead.so
am:
9a4a34afd8 -s ours
Change-Id: I557d7e96401b742d916e2eef5aa03949018382c7
TreeHugger Robot [Fri, 16 Jun 2017 15:12:30 +0000 (15:12 +0000)]
Merge "Import translations. DO NOT MERGE" into oc-dev
Julia Reynolds [Fri, 16 Jun 2017 13:30:29 +0000 (13:30 +0000)]
Merge "Fix two IndexOutOfBoundsException crashes." into oc-dev
Bill Yi [Fri, 16 Jun 2017 12:16:09 +0000 (05:16 -0700)]
Import translations. DO NOT MERGE
Change-Id: Icb9547e4d5fb47bfab83f8012358047189787fda
Auto-generated-cl: translation import
TreeHugger Robot [Fri, 16 Jun 2017 01:47:07 +0000 (01:47 +0000)]
Merge "Make RankingHelper$Record.groups thread-safe" into oc-dev
Casey Burkhardt [Wed, 14 Jun 2017 03:12:42 +0000 (20:12 -0700)]
Resolve 3 inconsistencies in accessibility button API
- Unify logic for detecting availability of the accessibility button
- Ensure the initial visibility state is propagated to A11yMS
- Ensure services only receive availability callbacks for changes
Test: Manual, created test accessibility services
targeting specific breakages
Bug:
38345417
Change-Id: I2250b32830cdfc2ecdc1dff7b7130dced2c1db29
TreeHugger Robot [Fri, 16 Jun 2017 01:04:09 +0000 (01:04 +0000)]
Merge "Don't even think about changing keyguard transit" into oc-dev
Sundeep Ghuman [Fri, 16 Jun 2017 00:55:21 +0000 (17:55 -0700)]
APs from old scans should not have RSSI values set.
This change creates parity between 'initWithScanResult' and
'update(ScanResult)' so that RSSIs are set similarly. This avoids an
edge case problem when starting WifiSettings and only having very old
scan results available (eg, coming from a screen off state), in which
the old APs would have been shown and then instantly removed due to a
sequential WifiTracker call to update the APs based on sticky
broadcasts.
Otherwise worded, creating an AP with a scan result and then instantly
updating it with that same scan result would drastically change the
state of the AccessPoint object.
Bug: b/
38212080
Test: With old scan results in the scanner (screen off more than 15
seconds), start wifi settings. Old APs should not be shown and then
instantly removed.
Change-Id: Ida13713dc115ba04882a7651eab4ebfd5cccbe7c
TreeHugger Robot [Fri, 16 Jun 2017 00:46:45 +0000 (00:46 +0000)]
Merge "Fix issue #
62524613: Background limitation is not working as expected..." into oc-dev
Jason Monk [Thu, 15 Jun 2017 18:21:34 +0000 (14:21 -0400)]
Fix flash of ambient on fp unlock
If we are in fp mode WAKE_AND_UNLOCK assume always set dozing
to false because we may not have received the keyguard callback
yet but we still should be hiding doze content.
Test: Add delay to exit animation callback and use fp unlock
Change-Id: I3d9256482cd5a35f6fa8864577b799432e2ab189
Fixes:
34916735
TreeHugger Robot [Thu, 15 Jun 2017 23:52:01 +0000 (23:52 +0000)]
Merge "When updating a split app, copy compiled files from base.apk only." into oc-dev
Felipe Leme [Thu, 15 Jun 2017 22:29:23 +0000 (15:29 -0700)]
Throw IAE when service add null AutofillIds on SaveInfo methods.
Test: existing CtsAutoFillServiceTestCases pass
Test: manual verification using Contacts app
Test: manual verification adding a CTS test case that crashes the app, but such
test cannot be commit because once the issue is fixed, it crashes the
service (the right way to test this fix is through unit tests against
exceptional conditions, but we don't support those on Autofill yet).
Fixes:
62649290
Change-Id: I8fc01fa929270219cd40035ff02eaf0dda5ecbfa
TreeHugger Robot [Thu, 15 Jun 2017 22:05:43 +0000 (22:05 +0000)]
Merge "don't block uninstall of unknown packages" into oc-dev
Dianne Hackborn [Thu, 15 Jun 2017 21:33:16 +0000 (14:33 -0700)]
Fix issue #
62524613: Background limitation is not working as expected...
...for most background services.
Whoops, was only stopping the first service it found!
Test: bit CtsAppTestCases:ActivityManagerProcessStateTest
Change-Id: I42af2872afed8b49fab0f53eff36b62e55d23409
Julia Reynolds [Thu, 15 Jun 2017 19:24:01 +0000 (15:24 -0400)]
Fix two IndexOutOfBoundsException crashes.
Do not remove (sometimes multiple) items from lists
you are iterating over.
Test: runtest systemui-notification
Change-Id: I130cc63ae2f5721e7b434006f4306e0b1eaef77d
Fixes:
62622503
Winson Chung [Thu, 15 Jun 2017 21:05:44 +0000 (21:05 +0000)]
Merge "Update TV to show custom actions." into oc-dev
TreeHugger Robot [Thu, 15 Jun 2017 20:57:16 +0000 (20:57 +0000)]
Merge "Fix companion uses-permission enforcement" into oc-dev
TreeHugger Robot [Thu, 15 Jun 2017 20:56:38 +0000 (20:56 +0000)]
Merge "ImageReader: Set BufferItem mGraphicBuffer to null when returning it" into oc-dev
Jorim Jaggi [Thu, 15 Jun 2017 19:10:38 +0000 (15:10 -0400)]
Don't even think about changing keyguard transit
When camera was launched with a lockscreen wallpaper set, the
wallpaper target was launcher in that case, which was also in
mClosingApps because it was first getting shown by keyguard
exit but then immediately hidden by starting the camera,
before the transition started.
Now since lockscreen wasn't the wallpaper target, launcher was
already for some reason, and we changed the transit to
WALLPAPER_CLOSE as a window with the wallpaper target was in
mClosingApps.
Fix this by never ever changing away from keyguard transits.
Test: go/wm-smoke
Test: ActivityManagerTransitionSelectionTests
Test: Set lockscreen wallpaper, set animation duration scale to
0.5, insert a random sleep statement in SystemUI, launch
camera from screen off while in trusted state and camera
wasn't running before.
Fixes:
37677242
Change-Id: I984b66d7f117034f3d55591284dd822b5ec76cbd
Robert Carr [Thu, 15 Jun 2017 19:39:34 +0000 (12:39 -0700)]
ViewRootImpl: More null checks for performTraversals.
Let's first understand how mView could become null. Notice
at the beginning of performTraversals, there is a check that
mView != null, and so it is nulled while we are in the function.
mView is package private and there are only two places which assign
to it, ViewRootImpl#setView and ViewRootImpl#doDie. setView is guarded
by mView == null. But mView was not null (per the check at the beginning
of performTraversals) and so mView is being nulled by doDie().
doDie() only has 3 callpoints:
1. ViewRootImpl#die(). Here though, calling it is guarded by
!mIsInTraversal. !mIsInTraversal is unconditionally set at the
beginning of performTraversals, and so this isn't our caller.
2. ViewRootHandler, handling MSG_DIE. This runs on the same thread
as performTraversal, and so it can't be our nuller.
3. WindowManagerGlobal#addView. This must be our nuller.
We see WindowManagerGlobal#addView will call doDie in the case that
we attempt to add a view which we had previously set to be removed
but deferred removal of. Now we can construct a reasonable sequence
for getting here:
1. requestLayout(). Perform traversals ends up on handler.
2. removeView(). MSG_DIE ends up on handler, View ends up in mDyingViews
3. performTraversals is executed by the handler
4. From a callback initiated by performTraversals (e.g. measure)
the client calls WindowManagerGlobal#addView on the view which
was just removed.
5. We are still in performTraversals so MSG_DIE hasn't been processed
yet. This means that WindowManagerGlobal will perform the doDie
immediately nulling mView.
6. We return to performTraversals and crash.
We can see shortly after the offending call to doDie, a new ViewRoot will be
constructed and so whatever traversal we are doing on the old one doesn't
seem particularly important. It doesn't seem that we can do any better
than letting it fall through without crashing.
Bug:
38421184
Test: go/wm-smoke. Feed to the monkeys.
Change-Id: I55f310a3533175c9df4a82878be5a60fd01b80c1
Sudheer Shanka [Thu, 15 Jun 2017 19:25:02 +0000 (19:25 +0000)]
Merge "Use correct timestamps of events for logging." into oc-dev
Jeff Hao [Tue, 13 Jun 2017 18:09:10 +0000 (11:09 -0700)]
When updating a split app, copy compiled files from base.apk only.
This fixes issues with some splits failing to be recompiled if they
haven't changed, but the dex files they depend on have.
The real fix will be for frameworks to generate the new expected
classpath and check in DexFile.getDexOptNeeded. Then we can undo
this change and copy over all the compiled split files again.
Bug:
62269291
Test: cts-tradefed run singleCommand cts -d --module
CtsAppSecurityHostTestCases -t android.appsecurity.cts.SplitTests
(cherry-picked from commit
d1235f54d4943a0a3a920013a5875b5193bd0490)
Change-Id: I6e640a966ef3b43054d163326878adebe2329693
Dianne Hackborn [Thu, 15 Jun 2017 18:45:28 +0000 (18:45 +0000)]
Merge "Work on issue #
36891897: Need to ensure foreground services..." into oc-dev
Eugene Susla [Thu, 15 Jun 2017 18:22:57 +0000 (11:22 -0700)]
Fix companion uses-permission enforcement
Fixes: b/
62662686
Test: Call CompanionDeviceManager#associate from app without uses-permission
declaration
Ensure exception is thrown
Call the method from app with the declaration
Ensure no exception is thrown
In debug mode catch the #isCallerSystem call from NotificationManagerService
(which runs in the system process) and ensure it passes the check
Change-Id: I26e5d2a7f5e63a346d4ab50b9ded1ec7dbc246a5
Todd Kennedy [Thu, 15 Jun 2017 16:44:34 +0000 (09:44 -0700)]
enable/disable dexopt for instant apps remotely
Change-Id: I997b8ae515f7bf2570edca4ed7ab4b46198148a5
Fixes:
62591398
Test: Manual; install instant app and see that it doesn't dexopt
Test: Manua; update gservices flag, install instant app and see that it does dexopt
TreeHugger Robot [Thu, 15 Jun 2017 16:20:07 +0000 (16:20 +0000)]
Merge "add a log for peek to shade expansion" into oc-dev
Dan Sandler [Wed, 14 Jun 2017 18:03:18 +0000 (14:03 -0400)]
Reduce no-op notification log messages.
Only log when channels or groups actually change.
Change-Id: I43e5a4f946962fc2d200f0d40d1744a938c83ca3
Fixes:
62487398
Test: runtest systemui-notification
Todd Kennedy [Thu, 15 Jun 2017 13:38:09 +0000 (06:38 -0700)]
don't block uninstall of unknown packages
If a package can't be found because it's not installed or it's
been filtered due to lack of visibility, don't claim the uninstall
should be blocked.
Change-Id: I932b77cea5749e5964ca626558bef51cdfdfd17b
Fixes:
62450984
Test: Manual