3 years agoMerge cherrypicks of [3365569, 3365570, 3366860, 3366878, 3365571, 3365572, 3366918... oreo-mr1-release android-8.1.0_r10 android-8.1.0_r11 android-8.1.0_r12 android-8.1.0_r15 android-8.1.0_r16 android-8.1.0_r19 android-wear-8.1.0_r1
android-build-team Robot [Thu, 14 Dec 2017 21:22:39 +0000 (21:22 +0000)]
Merge cherrypicks of [336556933655703366860336687833655713365572336691833655733365589336559033669383366902336557433655753365576336557733669583365824336559133669593366960336696133669623366963336696433669653366919336696633669673366968336696933669703367018336701933655923365593336698533658253366988336698933669903366991336699233669933366994336700433670053367006336700733670083367009336701033670113367012336701333670143367015336701633670173367038336703933670403367041336704233670443367045336704633670493367050336705233670533367054336705533670563366920336692133669223367079] into oc-mr1-release

Change-Id: I41e81da8c61c5c0a7a50749f23e55611a1efd33a

3 years agocas: validate shared buffer size before using
Chong Zhang [Wed, 25 Oct 2017 05:42:30 +0000 (22:42 -0700)]
cas: validate shared buffer size before using

bug: 67962232

cts MediaCasTest
vts VtsHalCasV1_0Target

Change-Id: I20b89d5971db4c89704245bb00c281af8c943697
(cherry picked from commit 16a3cd0b0729bd53265897cf8c790013156e92ce)

3 years agoMerge cherrypicks of [3134552, 3130583, 3131953, 3131954, 3131955, 3131956, 3131957... android-8.1.0_r1 android-8.1.0_r3 android-8.1.0_r4 android-8.1.0_r5 android-8.1.0_r6 android-cts-8.1_r1 android-cts-8.1_r2
android-build-team Robot [Fri, 27 Oct 2017 14:50:01 +0000 (14:50 +0000)]
Merge cherrypicks of [3134552313058331319533131954313195531319563131957313195831319593132062313233631310743133939313102431310253131026313058431308793130880] into oc-mr1-release

Change-Id: I06faeaf1e282dbf73aaa082e2944c5ea563f45fa

3 years agoDon't expect stopOffload to succeed if interface doesn't exist.
Lorenzo Colitti [Mon, 23 Oct 2017 09:46:56 +0000 (18:46 +0900)]
Don't expect stopOffload to succeed if interface doesn't exist.

The AdditionalStopsWithInitReturnFalse test inits offload and
then checks that stopOffload returns true. However, the HAL is
allowed to return false to stopOffload if no upstream was ever
successfully programmed.

To ensure this test doesn't fail in the common case that the
interface is not up, don't expect stopOffload to return true in
this case.

Bug: 67439856
Test: VtsHalTetheroffloadControlV1_0TargetTest passes with and without mobile data connection
Change-Id: I245f9e3e4376a7799572a3b3967e2bf6c52b5e4d
(cherry picked from commit 3189d9de81c937d495482fa3dacc196878cfcda1)

3 years agoSnap for 4411005 from d97f8714f07a762efe5a48747454fb4bdd51be8b to oc-mr1-release
android-build-team Robot [Mon, 23 Oct 2017 02:35:04 +0000 (02:35 +0000)]
Snap for 4411005 from d97f8714f07a762efe5a48747454fb4bdd51be8b to oc-mr1-release

Change-Id: I51e8c1630d4e0f432237b3c6a093754828723122

3 years agoFixup netlink socket creation to make corresponding fix in JNI
Erik Kline [Fri, 20 Oct 2017 08:08:38 +0000 (17:08 +0900)]
Fixup netlink socket creation to make corresponding fix in JNI

Corresponding JNI fix is:


Test: as follows
    - built
    - flashed
    - booted
    - make vts -j30 BUILD_GOOGLE_VTS=true && \
      vts-tradefed run commandAndExit vts --skip-all-system-status-check --primary-abi-only --skip-preconditions --module VtsHalTetherOffloadControlV1_0Target -l DEBUG

      10-20 17:48:28 I/ResultReporter: Invocation finished in 5m 32s. PASSED: 38, FAILED: 0, MODULES: 1 of 1

Bug: 68018148

Change-Id: I5aa019d708abaff61387b83a857e9e8226709842

3 years agoSnap for 4404534 from a555e04f8c37db86c5016218d5b7ab37cccc29c6 to oc-mr1-release
android-build-team Robot [Thu, 19 Oct 2017 06:15:47 +0000 (06:15 +0000)]
Snap for 4404534 from a555e04f8c37db86c5016218d5b7ab37cccc29c6 to oc-mr1-release

Change-Id: Ie01a467265b5abca5b1b4795e79d384adf35321a

3 years agoFix VTS test which turns off radio. am: 536818d17a
Sanket Padawe [Wed, 18 Oct 2017 23:57:54 +0000 (23:57 +0000)]
Fix VTS test which turns off radio. am: 536818d17a
am: 340619c703  -s ours

Change-Id: I4d4c0998ac28fabef207dfb0ce8fae8f08c0afd4

3 years agoFix VTS test which turns off radio.
Sanket Padawe [Wed, 18 Oct 2017 23:56:12 +0000 (23:56 +0000)]
Fix VTS test which turns off radio.
am: 536818d17a

Change-Id: I74ebfa3894e27bdf0e02671cbeccac83b1917424

3 years agoFix VTS test which turns off radio.
Sanket Padawe [Mon, 4 Sep 2017 05:32:43 +0000 (22:32 -0700)]
Fix VTS test which turns off radio.

Since the purpose of most of the test cases in current VTS is to check
if proper errors are returned and there is no crash seen in vendor code,
updating setRadioPower test case to turn on the radio instead of
turning off. We want to avoid test cases which turn off
radio or leads to modem shut down as those test cases affect other tests.

Test: VTS
Change-Id: I4fb9f18884f7ef21162015a0032c4431444f7025
Merged-In: I4fb9f18884f7ef21162015a0032c4431444f7025
Bug: 65230472
(cherry picked from commit 9a721b808743069fc26d6cf226595e2823e347a6)

3 years agoSnap for 4402249 from b70f1bda1f34fd63a9c9f5ad8307c8677a2f1f4d to oc-mr1-release
android-build-team Robot [Wed, 18 Oct 2017 07:22:41 +0000 (07:22 +0000)]
Snap for 4402249 from b70f1bda1f34fd63a9c9f5ad8307c8677a2f1f4d to oc-mr1-release

Change-Id: Ibfb2886719cbdebb901dbf908e9826fcb66a5819

3 years agoMerge "Add REQUEST_NOT_SUPPORTED / SIM_ABSENT for vts tests" into oc-mr1-dev
TreeHugger Robot [Wed, 18 Oct 2017 03:35:42 +0000 (03:35 +0000)]
Merge "Add REQUEST_NOT_SUPPORTED / SIM_ABSENT for vts tests" into oc-mr1-dev

3 years agoAdd REQUEST_NOT_SUPPORTED / SIM_ABSENT for vts tests
sqian [Mon, 16 Oct 2017 21:12:40 +0000 (14:12 -0700)]

Bug: 65230472
Test: run vts
Change-Id: I1fab6f49f04a04a7f1cd3b613fe7049f927171d1

3 years agoMerge "bug fix: during fatal fails bypass freeNode() calls" into oc-mr1-dev
TreeHugger Robot [Tue, 17 Oct 2017 22:37:56 +0000 (22:37 +0000)]
Merge "bug fix: during fatal fails bypass freeNode() calls" into oc-mr1-dev

3 years agobug fix: during fatal fails bypass freeNode() calls
Ram Mohan M [Wed, 11 Oct 2017 10:08:25 +0000 (15:38 +0530)]
bug fix: during fatal fails bypass freeNode() calls

If test application undergoes a fatal failure bypass all the omx
api calls till the application safely returns.

Test: make vts -j99 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 \
&& vts-tradefed run commandAndExit vts \
--skip-all-system-status-check --primary-abi-only \
--skip-preconditions --module VtsHalMediaOmxV1_0Host \

Bug: 63796949

Change-Id: If5a33672a31665db913e53c6f8ba8bea6cfd3ebd

3 years agoSnap for 4399693 from 15e945976b2d8a7047c2a8b87fd639c97423cf51 to oc-mr1-release
android-build-team Robot [Tue, 17 Oct 2017 07:24:00 +0000 (07:24 +0000)]
Snap for 4399693 from 15e945976b2d8a7047c2a8b87fd639c97423cf51 to oc-mr1-release

Change-Id: Icf8e4608c71a03f30cf0a2ae4fd9524bfacc8c73

3 years agotetheroffload VTS: fix stopOffload test conditions
Erik Kline [Sat, 14 Oct 2017 09:29:09 +0000 (18:29 +0900)]
tetheroffload VTS: fix stopOffload test conditions

The tetheroffload HAL is somewhat...and over- and under-specified.
A not unreasonable interpretation is that stopOffload() doesn't have
return success unless offload was actually started (via a call to
setUpstreamParameters()), and that after initOffload() and before
setUpstreamParameters() not actual "offload" has been engaged.

Precision in this matter is not required for the test case:


But for the test case:


we want to ensure the "matching" stopOffload call succeeds. For this
test we add in a call to setUpstreamParameters() for good measure.

Test: as follows
    prompt$ make vts -j30 BUILD_GOOGLE_VTS=true && \
            vts-tradefed run commandAndExit vts \
                --skip-all-system-status-check \
                --primary-abi-only \
                --skip-preconditions \
                --module VtsHalTetherOffloadControlV1_0Target \
                -l DEBUG

    10-16 19:17:17 I/ResultReporter: Invocation finished in 1m 2s. PASSED: 38, FAILED: 0, MODULES: 1 of 1

Bug: 65270149
Bug: 65612227
Bug: 65612332
Change-Id: I924d41f5a20f07707e3d6991cb59d9c6b2b02339

3 years agoSnap for 4393674 from 213c36d71d099b53e9846d320968a36d56b12ef3 to oc-mr1-release
android-build-team Robot [Fri, 13 Oct 2017 07:30:18 +0000 (07:30 +0000)]
Snap for 4393674 from 213c36d71d099b53e9846d320968a36d56b12ef3 to oc-mr1-release

Change-Id: I89cf086fca843d73ea1c4dd2fbc56ac8d5581e7b

3 years agoEnlarge buffer size for direct report testing am: 3c01fef56c
Peng Xu [Thu, 12 Oct 2017 17:05:02 +0000 (17:05 +0000)]
Enlarge buffer size for direct report testing am: 3c01fef56c
am: b0fe06a06b

Change-Id: I06edfcd55eaa52c962f5ff6bda4060d3df9c4efe

3 years agoEnlarge buffer size for direct report testing
Peng Xu [Thu, 12 Oct 2017 17:04:03 +0000 (17:04 +0000)]
Enlarge buffer size for direct report testing
am: 3c01fef56c

Change-Id: I9fce07b88a1ddb12a91554b823d9290c2571eaef

3 years agoSnap for 4391495 from 2528c405f3164c24f6dbf351b904f27265a5e9a8 to oc-mr1-release
android-build-team Robot [Thu, 12 Oct 2017 07:23:07 +0000 (07:23 +0000)]
Snap for 4391495 from 2528c405f3164c24f6dbf351b904f27265a5e9a8 to oc-mr1-release

Change-Id: Ia14cba21c9fc20ab05d5ee78091a1855251d7fda

3 years agoEnlarge buffer size for direct report testing
Peng Xu [Wed, 11 Oct 2017 01:19:18 +0000 (18:19 -0700)]
Enlarge buffer size for direct report testing

This change is part of "Sensor direct report test - gralloc support".

Bug: 67439052
Test: vts test no longer fail on sailfish device
Change-Id: Id67a2a181b2d8847e2b1b9dbe46f5186e97e8563
Merged-In: Ibe1d076c24dc2cfe61dfd19aa5055c9075aa9e14

3 years agoMerge "Added VTS tests for requiresSecureDecoderComponent and release API" into oc...
TreeHugger Robot [Wed, 11 Oct 2017 18:33:25 +0000 (18:33 +0000)]
Merge "Added VTS tests for requiresSecureDecoderComponent and release API" into oc-mr1-dev

3 years agoSnap for 4388906 from 0b9f25d5c9eadb5eda353f5cc01a8bd2d05cf5f2 to oc-mr1-release
android-build-team Robot [Wed, 11 Oct 2017 07:21:12 +0000 (07:21 +0000)]
Snap for 4388906 from 0b9f25d5c9eadb5eda353f5cc01a8bd2d05cf5f2 to oc-mr1-release

Change-Id: I4d9f66c09a4eaa9bb2237ab1cceddc5914e18f76

3 years agoAudio VTS: Allow 4GB buffer prepareToWrite/Read to succeed on 64 bits am: ce607879ba
Mikhail Naganov [Wed, 11 Oct 2017 01:35:02 +0000 (01:35 +0000)]
Audio VTS: Allow 4GB buffer prepareToWrite/Read to succeed on 64 bits am: ce607879ba
am: 5b122a7408  -s ours

Change-Id: Ib60918524329e9ab8c4d07a0c52f5cb2f79c7dd5

3 years agoAudio VTS: Allow 4GB buffer prepareToWrite/Read to succeed on 64 bits
Mikhail Naganov [Wed, 11 Oct 2017 01:33:25 +0000 (01:33 +0000)]
Audio VTS: Allow 4GB buffer prepareToWrite/Read to succeed on 64 bits
am: ce607879ba

Change-Id: I6869d49ab5ceb12e1b0d3e50e0952094c34b3c07

3 years agoMerge "cas-vts: use MemoryDealer to allocate shared memory" into oc-mr1-dev
TreeHugger Robot [Wed, 11 Oct 2017 00:55:43 +0000 (00:55 +0000)]
Merge "cas-vts: use MemoryDealer to allocate shared memory" into oc-mr1-dev

3 years agoMerge "Legacy wrapper: Use arbitrary limit on buffer size" into oc-mr1-dev
TreeHugger Robot [Tue, 10 Oct 2017 23:43:35 +0000 (23:43 +0000)]
Merge "Legacy wrapper: Use arbitrary limit on buffer size" into oc-mr1-dev

3 years agoAudio VTS: Allow 4GB buffer prepareToWrite/Read to succeed on 64 bits
Mikhail Naganov [Tue, 10 Oct 2017 17:31:07 +0000 (10:31 -0700)]
Audio VTS: Allow 4GB buffer prepareToWrite/Read to succeed on 64 bits

Because on oc-dev the runtime image can not be modified and that this is
not a bug but a test and implementation misalignment,
allow prepareForWrite/Read(UINT32_MAX, UINT32_MAX) to succeed on oc-dev.

MR1 will have a better fix where the HAL will always reject > 1GiB
buffers independently of the architecture.

Test: vts-tradefed run commandAndExit vts --module VtsHalAudioV2_0Target
Bug: 67030516
Change-Id: I10a5ee3868c67c45ed5297d9c5c89129d762c7cd
Merged-In: I4cc3efda9bb66e6dae8b4e6785f52d9e51440aee
Signed-off-by: Kevin Rocard <>
3 years agocas-vts: use MemoryDealer to allocate shared memory
Chong Zhang [Tue, 10 Oct 2017 20:03:18 +0000 (13:03 -0700)]
cas-vts: use MemoryDealer to allocate shared memory

cas hal can't (and doesn't need to) use hidl_memory allocated by
hidl allocator. All real clients (MediaDescrambler jni, ACodec)
are using MemoryDealer-allocated ashmem only. Modify the test
to cover these cases.

bug: 67603039
> make vts -j64
> vts-tradefed run vts -m VtsHalCasV1_0Target

Change-Id: Ib9563abc1a14c34781a28d15c7766c6c160d288a

3 years agoVTS: fix VtsHalKeymasterV3_0Target issue am: 926a8e7440
Iris Chang [Tue, 10 Oct 2017 16:27:46 +0000 (16:27 +0000)]
VTS: fix VtsHalKeymasterV3_0Target issue am: 926a8e7440
am: 7b770a8d70

Change-Id: I0bee7c1bdf4e40a2024fadc0b510ca002806c65e

3 years agoVTS: fix VtsHalKeymasterV3_0Target issue
Iris Chang [Tue, 10 Oct 2017 16:22:34 +0000 (16:22 +0000)]
VTS: fix VtsHalKeymasterV3_0Target issue
am: 926a8e7440

Change-Id: I762f63b3f239169d93b33f7164d4eb3f9bd8f3d9

3 years agoLegacy wrapper: Use arbitrary limit on buffer size
Kevin Rocard [Fri, 6 Oct 2017 01:47:02 +0000 (18:47 -0700)]
Legacy wrapper: Use arbitrary limit on buffer size

The legacy HAL wrapper prepareForWrite and prepareForRead used
to return INVALID_ARGUMENTS if the computed buffer size just
under SIZE_MAX.

This meant that the limitation depended on the architecture (32 vs 64
bit size_t). This caused VTS test failure on 64 bits.

Instead of dynamically calculating an arbitrary max size,
choose a fixed one. This max buffer size has been chosen at 1GiB.
It should be enough for the foreseeable future and not too close
from the 4GiB max on 32 bit.

Test: vts-tradefed run commandAndExit vts --module VtsHalAudioV2_0Target
Bug: 67030516
Change-Id: I4cc3efda9bb66e6dae8b4e6785f52d9e51440aee
Signed-off-by: Kevin Rocard <>
3 years agoSnap for 4386596 from e5e339d2a5a5a98fc6fd3643ced6fa938e5bdc44 to oc-mr1-release
android-build-team Robot [Tue, 10 Oct 2017 07:27:15 +0000 (07:27 +0000)]
Snap for 4386596 from e5e339d2a5a5a98fc6fd3643ced6fa938e5bdc44 to oc-mr1-release

Change-Id: Id2551559c381d67cba9357665497f5e58560e592

3 years agoMerge "Refactor test harness for generated VTS tests" into oc-mr1-dev
TreeHugger Robot [Tue, 10 Oct 2017 02:42:13 +0000 (02:42 +0000)]
Merge "Refactor test harness for generated VTS tests" into oc-mr1-dev

3 years agoRefactor test harness for generated VTS tests
I-Jui (Ray) Sung [Fri, 6 Oct 2017 20:22:39 +0000 (13:22 -0700)]
Refactor test harness for generated VTS tests

Address comments in ag/2926388

Test: VtsHalNeuralnetworksV1_0TargetTest on Angler with sample driver
Bug: 67597726
Change-Id: Ie3118dcbfb11c25dbc84d2becdefbbae868ff28f

3 years agoAdded VTS tests for requiresSecureDecoderComponent and release API
Suresh Sivaraman [Wed, 4 Oct 2017 09:07:36 +0000 (14:37 +0530)]
Added VTS tests for requiresSecureDecoderComponent and release API

Bug: 63914034
Bug: 65593293

Build & test vts:

> make vts -j64

> vts-tradefed run vts -m VtsHalCasV1_0Target

Change-Id: I465735768239f2d53427a18e5ad1fa0069f6e35c

3 years agoSnap for 4383937 from 3021f1fafbfb1a4b464effa543aee381f42d8080 to oc-mr1-release
android-build-team Robot [Sun, 8 Oct 2017 07:28:28 +0000 (07:28 +0000)]
Snap for 4383937 from 3021f1fafbfb1a4b464effa543aee381f42d8080 to oc-mr1-release

Change-Id: Ia2f03abe28796187082d7d7a71f79e42a8eb6640

3 years agoMerge "omxvts: exit gracefully during fatal fails" into oc-mr1-dev
TreeHugger Robot [Sat, 7 Oct 2017 02:26:15 +0000 (02:26 +0000)]
Merge "omxvts: exit gracefully during fatal fails" into oc-mr1-dev

3 years agoMerge "Added VTS tests for MediaCas - II" into oc-mr1-dev
TreeHugger Robot [Fri, 6 Oct 2017 16:49:20 +0000 (16:49 +0000)]
Merge "Added VTS tests for MediaCas - II" into oc-mr1-dev

3 years agoSnap for 4380782 from 38cfeee30c4b6ff8d8feed57e721b5d0c0fdf4de to oc-mr1-release
android-build-team Robot [Fri, 6 Oct 2017 07:26:52 +0000 (07:26 +0000)]
Snap for 4380782 from 38cfeee30c4b6ff8d8feed57e721b5d0c0fdf4de to oc-mr1-release

Change-Id: I0daa56402c8dc6483971c440e241bb1490b98792

3 years agoVTS: fix VtsHalKeymasterV3_0Target issue
Iris Chang [Thu, 14 Sep 2017 07:23:18 +0000 (15:23 +0800)]
VTS: fix VtsHalKeymasterV3_0Target issue

Failed cases:

The verify_attestation_record() in Keymaster_hidl_hal_test.cpp calls
parse_attestation_record() to set the value of att_challenge. It fails
to compare att_challenge with challenge by memcmp.
Because setToExternal() method uses buffer pointer to local variable
(record), not use memcpy to copy into itself buffer in
parse_attestation_record(). When it leaves the parse_attestation_record(),
we will get the att_challenge which is null buffer to compare with challenge

Fix: use memcpy to copy the buffer.

Bug: 65039571
Test: build passed. VtsHalKeymasterV3_0Target -> PASSED: 106, FAILED: 0.
Change-Id: I700a9242cc9a5f4cb196b62860823601e4088531
Merged-In: I700a9242cc9a5f4cb196b62860823601e4088531
(cherry picked from 54ca32a130a0450b444848046199d62876b55301)

3 years agoAdd neuralnetworks HAL to the current.txt
Michael Butler [Thu, 5 Oct 2017 20:30:14 +0000 (13:30 -0700)]
Add neuralnetworks HAL to the current.txt

Bug: 63905942
Test: None
Change-Id: I7c247381e8fd97b91d11531d3db595408d8f1b74

3 years agoAdded VTS tests for MediaCas - II
Suresh Sivaraman [Thu, 21 Sep 2017 13:14:24 +0000 (18:44 +0530)]
Added VTS tests for MediaCas - II

Bug: 63914034
Bug: 65593293

Build & test vts:

> make vts -j64

> vts-tradefed run vts -m VtsHalCasV1_0Target

Change-Id: If5a0e7cb3d520843ba45147a085c43b94027e584

3 years agoCorrected performanceInfo documentation
Jean-Luc Brouillet [Thu, 5 Oct 2017 01:24:59 +0000 (18:24 -0700)]
Corrected performanceInfo documentation

We'll be using relative numbers, change the doc to reflect that.

Bug: 63905942
Test: Compiled
Change-Id: Iad67621cfc76490c2dc471865a3f82cebf175440

3 years agoomxvts: exit gracefully during fatal fails
Ram Mohan M [Wed, 27 Sep 2017 11:12:19 +0000 (16:42 +0530)]
omxvts: exit gracefully during fatal fails

Bug: 63796949

Change-Id: I735db42f7de6926adddd328db1a48de35faaa608

3 years agoSnap for 4378450 from 0c7d0a9d0dba6f8d879b8b61cbc1c5b56a5cfd99 to oc-mr1-release
android-build-team Robot [Thu, 5 Oct 2017 07:24:14 +0000 (07:24 +0000)]
Snap for 4378450 from 0c7d0a9d0dba6f8d879b8b61cbc1c5b56a5cfd99 to oc-mr1-release

Change-Id: I7a709f6418fb267d2c4aad0d6cfd14e458181b71

3 years agoMerge "Fix VTS failures on tests with optional inputs" into oc-mr1-dev
TreeHugger Robot [Thu, 5 Oct 2017 06:15:59 +0000 (06:15 +0000)]
Merge "Fix VTS failures on tests with optional inputs" into oc-mr1-dev

3 years agoFix VTS failures on tests with optional inputs
I-Jui (Ray) Sung [Thu, 5 Oct 2017 03:49:57 +0000 (20:49 -0700)]
Fix VTS failures on tests with optional inputs

RequestArgument for optional inputs should have its .hasNoValue set
and other fields left empty. The test harness was incorrectly ignoring
the hasNoValue field and initializing the offset field even for missing
arguments. That inconsistency was caught by the sample driver and caused
an invalid argument error. With this CL the test harness now populates
requests correctly for tests with intentionally missing inputs, and that
fixes lstm* and lsh_projection.

Test: VtsHalNeuralnetworksV1_0TargetTest on Angler
Bug: 63905942
Change-Id: I8729124b720f0d4163f4b9efd452f82b825b0885

3 years agoEarly test termination for unsupported models
Michael Butler [Thu, 5 Oct 2017 01:13:52 +0000 (18:13 -0700)]
Early test termination for unsupported models

Some service drivers won't be able to support all operations. In those
cases, try to prepare the model anyway. If it works, proceed with the
test; if it fails, terminate the test early.

For example, the hvx nn service does not support svdf, so it should
attempt to prepare the model, but the test should not fail if the hvx
driver reports an error.

Bug: 63905942
Test: mm, expected behavior when running on hvx service
Change-Id: I51aa1911145f0d17590906c21314d2bdc24cffea

3 years agoMerge "Only verify output of models a vendor nn service fully supports" into oc-mr1-dev
Michael Butler [Wed, 4 Oct 2017 23:49:51 +0000 (23:49 +0000)]
Merge "Only verify output of models a vendor nn service fully supports" into oc-mr1-dev

3 years agoOnly verify output of models a vendor nn service fully supports
Michael Butler [Wed, 4 Oct 2017 09:38:42 +0000 (02:38 -0700)]
Only verify output of models a vendor nn service fully supports

Some service drivers won't be able to support all operations. In those
cases, run the test as far as it can go, but ignore the output data.

For example, the hvx nn service does not support svdf, so it should
attempt to prepare and execute the model, but the test should not fail
if the hvx driver reports an error. However, hvx indicates it is able to
run MAX_POOL_2D, so the test should fail if that output is incorrect.

Bug: 63905942
Test: mm
Test: run vts binary on hvx service
Change-Id: I3ebc05836c8e070b66daa2e38b107f96637ca812

3 years agoMerge "Fix VtsHalDrmV1_0Target clearkey tests." into oc-mr1-dev
TreeHugger Robot [Wed, 4 Oct 2017 16:38:02 +0000 (16:38 +0000)]
Merge "Fix VtsHalDrmV1_0Target clearkey tests." into oc-mr1-dev

3 years agoMerge "Move android.hardware.tests.libhwbinder@1.0-impl to system.img" into oc-mr1-dev
TreeHugger Robot [Wed, 4 Oct 2017 13:41:00 +0000 (13:41 +0000)]
Merge "Move android.hardware.tests.libhwbinder@1.0-impl to system.img" into oc-mr1-dev

3 years agoSnap for 4376088 from 072811f5fbfcb98b31b9106f69683c04bbbb865b to oc-mr1-release
android-build-team Robot [Wed, 4 Oct 2017 07:21:27 +0000 (07:21 +0000)]
Snap for 4376088 from 072811f5fbfcb98b31b9106f69683c04bbbb865b to oc-mr1-release

Change-Id: I61e48e8ece2ed1c880c8bfd9423627f53651d8ea

3 years agoFix VtsHalDrmV1_0Target clearkey tests.
Edwin Wong [Fri, 29 Sep 2017 22:58:37 +0000 (15:58 -0700)]
Fix VtsHalDrmV1_0Target clearkey tests.

1. Fix the following issues:
  - queryKeyStatus is now implemented, do not return
  - expectedKeyRequest and knownKeyResponse should use
    '-' and '_' instead of '+' and '/' for base64Url encode;
    one uint8_t change from 0x2b to 0x2d for each vector

2. Add tests to create clearkey drm and crypto plugin using
both common pssh box format system ID and clearkey scheme

Test: Vts test on Marlin and Taimen
  adb shell /data/nativetest/VtsHalDrmV1_0TargetTest/

Test: Cts ClearKeyTest on Marlin and Taimen
  ANDROIOP= ./android-cts/tools/cts-tradefed run cts
  -m CtsMediaTestCases

Test: ClearKey plugin unit tests
  adb shell LD_LIBRARY_PATH="/vendor/lib/mediadrm"

Test: netflix and Play Movies & TV for sanity check

bug: 66327050
Change-Id: Icfbe715cf246a7c5253d5f104d0fd446618b18bd

3 years agoMerge "Fix problem passing null optional operands." into oc-mr1-dev
TreeHugger Robot [Wed, 4 Oct 2017 01:40:37 +0000 (01:40 +0000)]
Merge "Fix problem passing null optional operands." into oc-mr1-dev

3 years agoMerge "NNAPI HAL: Change IEvent to explicit callbacks" into oc-mr1-dev
TreeHugger Robot [Wed, 4 Oct 2017 00:07:22 +0000 (00:07 +0000)]
Merge "NNAPI HAL: Change IEvent to explicit callbacks" into oc-mr1-dev

3 years agoMerge "wifi(implementation): Increase stop timeout" into oc-mr1-dev
TreeHugger Robot [Tue, 3 Oct 2017 23:42:36 +0000 (23:42 +0000)]
Merge "wifi(implementation): Increase stop timeout" into oc-mr1-dev

3 years agoFix problem passing null optional operands.
Jean-Luc Brouillet [Thu, 28 Sep 2017 02:22:04 +0000 (19:22 -0700)]
Fix problem passing null optional operands.

Allow optional operands to be specified as having no value in the model.
Make sure that both these and model arguements that have no value can
be passed to through the HAL. Change the validation to ensure that all
required ops are there.

Bug: 63905942
Test: Compiled and ran runtime/test and VTS tests.

Change-Id: Ia20954b9712042f5c0571822967521aedb66150b

3 years agoNNAPI HAL: Change IEvent to explicit callbacks
Michael Butler [Fri, 22 Sep 2017 20:26:12 +0000 (13:26 -0700)]
NNAPI HAL: Change IEvent to explicit callbacks

IEvent was a synchronization primitive which caused some confusion
in the interface. Originally the event object was paired with an
asynchronous task, and the asynchronous task would signal this event
when the corresponding output was ready to be used.

In the case of IDevice::prepareModel, the function call would return an
IPreparedModel object that was not guaranteed to be prepared until the
runtime had returned from waiting on the corresponding event object.
The event object has been changed to two explicit callbacks--
IPreparedModelCallback and IExecutionCallback. Now,
IDevice::prepareModel no longer returns an unfinished IPreparedModel;
instead, it will pass the IPreparedModel object to the runtime through
IPreparedModelCallback::notify. When the runtime retreives the
IPreparedModel object, the asynchronous task has already finished
preparing the model.

The two callbacks are used for different purposes. Each has its own
version of notify to pass the data back to the runtime:
* IPreparedModelCallback::notify(ErrorStatus, IPreparedModel)
* IExecutionCallback::notify(ErrorStatus)

Bug: 63905942
Test: mm, vts, ml/nn/runtime/tests
Change-Id: I0c88cd262ba762e0af15e9da31ebe813a5d150b2

3 years agoCamera: Update VTS for new camera@3.3 interfaces
Eino-Ville Talvala [Fri, 29 Sep 2017 23:37:59 +0000 (16:37 -0700)]
Camera: Update VTS for new camera@3.3 interfaces

- Cross check device version in name with ICameraDeviceSession version
- Switch all calls to configureStreams to configureStreams_3_3 when
  a v3.3 ICameraDeviceSession is available

Test: Test passes on v3.3 devices, doesn't regress on v3.2
Bug: 62358514
Change-Id: I7450e5d5119c5ec1fe9f1360bfc99fc8a47d0936

3 years agoCamera: Enable VTS tests for 3.3 camera devices
Emilian Peev [Thu, 28 Sep 2017 10:10:08 +0000 (11:10 +0100)]
Camera: Enable VTS tests for 3.3 camera devices

A new 3.3 Camera device API is available and devices supporting it
need to be tested as well.

Bug: 67028051
Test: run commandAndExit vts --skip-all-system-status-check
--skip-preconditions --primary-abi-only --module
VtsHalCameraProviderV2_4Target -l INFO

For passthrough devices: run commandAndExit vts
--skip-all-system-status-check --skip-preconditions --abi armeabi-v7a
--module VtsHalCameraProviderV2_4Target -l INFO

Change-Id: I91adb3c1867cdd8f8f810a2bd5b0a8ba2b0c1fcf

3 years agorelease-request-53526352-7b51-4ab1-a661-632ffc55dd7c-for-git_oc-mr1-release-4371241...
android-build-team Robot [Mon, 2 Oct 2017 04:28:14 +0000 (04:28 +0000)]
release-request-53526352-7b51-4ab1-a661-632ffc55dd7c-for-git_oc-mr1-release-4371241 snap-temp-L10900000107789672

Change-Id: Ic92b5738c862081afa132de4abc878619e6a6ab3

3 years agoMerge "omxvts: fix binary for 64bit platforms" into oc-mr1-dev
TreeHugger Robot [Sun, 1 Oct 2017 23:09:21 +0000 (23:09 +0000)]
Merge "omxvts: fix binary for 64bit platforms" into oc-mr1-dev

3 years agoMerge "omxvts: misplaced lock can cause race condition" into oc-mr1-dev
TreeHugger Robot [Sun, 1 Oct 2017 23:09:00 +0000 (23:09 +0000)]
Merge "omxvts: misplaced lock can cause race condition" into oc-mr1-dev

3 years agowifi(implementation): Increase stop timeout
Roshan Pius [Fri, 29 Sep 2017 22:04:33 +0000 (15:04 -0700)]
wifi(implementation): Increase stop timeout

The current timeout of 50ms seems to be insufficient on some devices.
So, changing it to a higher value.

Bug: 66334487
Test:  Ran the VTS tests continously in a loop for 30 minutes
without failure. Previously, failure occurence was about once every
5-10 minutes.
`while adb shell /data/VtsHalWifiSupplicantV1_0TargetTest; do :;
done; adb bugreport`

Change-Id: I884de64d748b46b9710a702f3c72cc30dd51627e

3 years agoomxvts: fix binary for 64bit platforms
Ram Mohan M [Wed, 27 Sep 2017 15:23:22 +0000 (20:53 +0530)]
omxvts: fix binary for 64bit platforms

Bug: 63796949

Change-Id: Ieddae0bd9fc45a88c6e1579bdd588b77b4174a1f

3 years agoomxvts: misplaced lock can cause race condition
Ram Mohan M [Tue, 26 Sep 2017 10:06:16 +0000 (15:36 +0530)]
omxvts: misplaced lock can cause race condition

upon unsuccessful return during timed wait, relock the mutex before
accessing message list.

Bug: 63796949

Change-Id: I84a43e148a6415e629cb39a6b9dfe27259f8f2fd

3 years agoMore tetheroffload VTS fixes for over-specified tests.
Erik Kline [Tue, 26 Sep 2017 09:56:35 +0000 (18:56 +0900)]
More tetheroffload VTS fixes for over-specified tests.

[1] Call addDownstream before removeDownstream in affirmative tests

In order to properly test removeDownstream() where the HAL might
reasonably reject downstreams not previously added we now call
addDownstream() first (for affirmative tests).

[2] Clarify when stopOffload() return values can be safely ignored.

Test: as follows
    - make vts -j30 BUILD_GOOGLE_VTS=true && \
      vts-tradefed run commandAndExit vts \
          --skip-all-system-status-check \
          --primary-abi-only \
          --skip-preconditions \
          --module VtsHalTetherOffloadControlV1_0Target \
          -l DEBUG
      still doesn't pass but it's better than before  :)
Bug: 65270149
Change-Id: I27a574bd2110e3a1626de343f6b57b9efb8cdf83

3 years agoTweaks to tetheroffload VTS: fixes and fd leaks
Erik Kline [Tue, 12 Sep 2017 07:21:55 +0000 (16:21 +0900)]
Tweaks to tetheroffload VTS: fixes and fd leaks

    - always use hidl_handle::setTo(..., true) to force fd ownership
      (and thereby cleanup in hidl_handle destructor)
    - initialize config HAL multiple times to try to trigger fd leaking bugs

Test: as follows
    - make vts -j30 BUILD_GOOGLE_VTS=true && \
      vts-tradefed run commandAndExit vts \
          --skip-all-system-status-check  \
          --primary-abi-only              \
          --skip-preconditions            \
          --module VtsHalTetherOffloadConfigV1_0Target -l INFO

      I/ResultReporter: Invocation finished in 48s. PASSED: 4, FAILED: 0, MODULES: 1 of 1

    - make vts -j30 BUILD_GOOGLE_VTS=true && \
      vts-tradefed run commandAndExit vts \
          --skip-all-system-status-check  \
          --primary-abi-only              \
          --skip-preconditions            \
          --module VtsHalTetherOffloadControlV1_0Target -l INFO
Bug: 29337859
Bug: 32163131
Bug: 64976634
Bug: 65270149
Bug: 65529504
Change-Id: Ib47b35a01092bb2052d241f7bba951bbf7691b66

3 years agorelease-request-4f3e6ea6-241a-4b05-b30f-5977c3f8253d-for-git_oc-mr1-release-4357973...
android-build-team Robot [Mon, 25 Sep 2017 07:22:32 +0000 (07:22 +0000)]
release-request-4f3e6ea6-241a-4b05-b30f-5977c3f8253d-for-git_oc-mr1-release-4357973 snap-temp-L90600000105740988

Change-Id: Idbafe7b031bb75ca50855e77cf4064dbff696242

3 years agoRemove operationTuple.
Jean-Luc Brouillet [Sat, 23 Sep 2017 22:15:58 +0000 (15:15 -0700)]
Remove operationTuple.

Removed operationTuple from Model in the HAL, as the data type was
redundant information.

Removed supportedOperationTuples from Capabilities, as real drivers need
more complex restrictions than this provided.  For the OC-MR1, we'll just
rely on getSupportedNodes.

Also removed the unused cachesCompilation.

Bug: 63905942
Test: Compiled and ran tests.
Change-Id: I15f33d14634f2e1c8d726b1bd01d5b9e123b47ea

3 years agorelease-request-2950741e-0b24-4d70-b748-67bb2b5435b5-for-git_oc-mr1-release-4357583...
android-build-team Robot [Sun, 24 Sep 2017 07:43:47 +0000 (07:43 +0000)]
release-request-2950741e-0b24-4d70-b748-67bb2b5435b5-for-git_oc-mr1-release-4357583 snap-temp-L55500000105655441

Change-Id: Ib6a647f37d4cad0b74b266ec12c979b9a8f6bd51

3 years agoMerge "Removed WIP tag from Neuralnetworks HAL" into oc-mr1-dev
Michael Butler [Sat, 23 Sep 2017 03:38:47 +0000 (03:38 +0000)]
Merge "Removed WIP tag from Neuralnetworks HAL" into oc-mr1-dev

3 years agoMerge "Add negative tests to the Neuralnetworks HAL" into oc-mr1-dev
Michael Butler [Sat, 23 Sep 2017 01:07:43 +0000 (01:07 +0000)]
Merge "Add negative tests to the Neuralnetworks HAL" into oc-mr1-dev

3 years agoMerge "Downgrade generated test harness for NDK libc++" into oc-mr1-dev
Michael Butler [Sat, 23 Sep 2017 00:57:48 +0000 (00:57 +0000)]
Merge "Downgrade generated test harness for NDK libc++" into oc-mr1-dev

3 years agoRemoved WIP tag from Neuralnetworks HAL
Michael Butler [Tue, 19 Sep 2017 23:26:47 +0000 (16:26 -0700)]
Removed WIP tag from Neuralnetworks HAL

Additionally cleaned up some comments and formatting.

Bug: 63905942
Test: not needed
Change-Id: I0306f6d37f3c9ce9b365b5f838c27e81fa001ccd

3 years agoMerge "Added VTS tests for MediaCas" into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 23:26:52 +0000 (23:26 +0000)]
Merge "Added VTS tests for MediaCas" into oc-mr1-dev

3 years agoDowngrade generated test harness for NDK libc++
I-Jui (Ray) Sung [Wed, 20 Sep 2017 20:45:50 +0000 (13:45 -0700)]
Downgrade generated test harness for NDK libc++

NDK libc++ does not yet support full C++14. This CL replaced use of
std::get with type as well as auto lambda arguments with equivalent
C++11 constructs in the VTS test harness for NNAPI.

Test: VtsHalNeuralnetworksV1_0TargetTest
Bug: 63905942
Change-Id: If75e7c088e9221a70bcc47bc647e0dd7b045bfa1

3 years agoAdded VTS tests for MediaCas
Suresh Sivaraman [Fri, 15 Sep 2017 06:21:15 +0000 (11:51 +0530)]
Added VTS tests for MediaCas

Bug: 63914034
Bug: 65593293
Change-Id: I1064ac14513508d2f821a1a27dafcc7389b1a691

3 years agorelease-request-bbc747ae-190b-4d38-8704-432aa6396c16-for-git_oc-mr1-release-4351869...
android-build-team Robot [Thu, 21 Sep 2017 07:32:02 +0000 (07:32 +0000)]
release-request-bbc747ae-190b-4d38-8704-432aa6396c16-for-git_oc-mr1-release-4351869 snap-temp-L93800000104806132

Change-Id: Iae58128454110860d39aa91b2636626938083073

3 years agoMerge "Add flush test" into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 01:01:13 +0000 (01:01 +0000)]
Merge "Add flush test" into oc-mr1-dev

3 years agoMerge "Enable EOSTest_M test" into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 01:00:47 +0000 (01:00 +0000)]
Merge "Enable EOSTest_M test" into oc-mr1-dev

3 years agoMerge "Signal eof flag for each input buffer" into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 01:00:22 +0000 (01:00 +0000)]
Merge "Signal eof flag for each input buffer" into oc-mr1-dev

3 years agoMerge "Add support for preset ANW buffer mode" into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 00:59:59 +0000 (00:59 +0000)]
Merge "Add support for preset ANW buffer mode" into oc-mr1-dev

3 years agoMerge changes I4a4aad3a,I40b94919 into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 00:59:35 +0000 (00:59 +0000)]
Merge changes I4a4aad3a,I40b94919 into oc-mr1-dev

* changes:
  Extend adaptive playback test for graphic buffer modes
  Move graphic buffer allocation to common

3 years agoMerge "Add vts audio dec test for flac decoder" into oc-mr1-dev
TreeHugger Robot [Thu, 21 Sep 2017 00:59:09 +0000 (00:59 +0000)]
Merge "Add vts audio dec test for flac decoder" into oc-mr1-dev

3 years agoAdd negative tests to the Neuralnetworks HAL
Michael Butler [Wed, 20 Sep 2017 02:59:45 +0000 (19:59 -0700)]
Add negative tests to the Neuralnetworks HAL

This CL adds negative tests to:
* IDevice::getSupportedOperations
* IDevice::prepareModel
* IPreparedModel::execute

It does not make sense to have negative tests for the remaining HAL
* IDevice::getStatus (all returned values are valid)
* IDevice::getCapabilities (no input or state)
* IEvent::notify (synchronization callback primitive)

Bug: 63905942
Test: mm, running on sample driver has expected output
Change-Id: Ia3cf5ce05dd07ff3f688395b8d0e581851909028

3 years agoMerge "Add adaptive playback test" into oc-mr1-dev
TreeHugger Robot [Wed, 20 Sep 2017 10:14:12 +0000 (10:14 +0000)]
Merge "Add adaptive playback test" into oc-mr1-dev

3 years agorelease-request-42a2a3ad-8c90-4c84-a0ad-5d067beb8e30-for-git_oc-mr1-release-4349323...
android-build-team Robot [Wed, 20 Sep 2017 07:47:09 +0000 (07:47 +0000)]
release-request-42a2a3ad-8c90-4c84-a0ad-5d067beb8e30-for-git_oc-mr1-release-4349323 snap-temp-L16100000104414353

Change-Id: I5bdbdf8f5c118a593d19cc27aac12505c732adc9

3 years agoMerge "Camera: Add default implementation of camera.device@3.3" into oc-mr1-dev
Eino-Ville Talvala [Tue, 19 Sep 2017 15:53:43 +0000 (15:53 +0000)]
Merge "Camera: Add default implementation of camera.device@3.3" into oc-mr1-dev

3 years agorelease-request-336ddfb0-c762-4bda-8684-20198669642d-for-git_oc-mr1-release-4346879...
android-build-team Robot [Tue, 19 Sep 2017 07:23:46 +0000 (07:23 +0000)]
release-request-336ddfb0-c762-4bda-8684-20198669642d-for-git_oc-mr1-release-4346879 snap-temp-L51800000104038734

Change-Id: I2459bbf08e5500e34c439589d0fc58de13caf78d

3 years agoMerge "Fix typos in IOmxStore.hal" into oc-mr1-dev
TreeHugger Robot [Tue, 19 Sep 2017 01:14:00 +0000 (01:14 +0000)]
Merge "Fix typos in IOmxStore.hal" into oc-mr1-dev

3 years agoMerge "Add capability for sensor hal to use wakelock" into oc-mr1-dev
Peng Xu [Mon, 18 Sep 2017 19:07:09 +0000 (19:07 +0000)]
Merge "Add capability for sensor hal to use wakelock" into oc-mr1-dev

3 years agoAdd vts audio dec test for flac decoder
Ram Mohan M [Fri, 15 Sep 2017 14:18:17 +0000 (19:48 +0530)]
Add vts audio dec test for flac decoder

Test: make vts -j99 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 \
&& vts-tradefed run commandAndExit vts \
--skip-all-system-status-check --primary-abi-only \
--skip-preconditions --module VtsHalMediaOmxV1_0Host \

Bug: 63796949

Change-Id: Ie504de96007c1c0ec6dc9c8b7da78f6c1da45631

3 years agoAdd support for preset ANW buffer mode
Ram Mohan M [Thu, 14 Sep 2017 08:12:12 +0000 (13:42 +0530)]
Add support for preset ANW buffer mode

Test: make vts -j99 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 \
&& vts-tradefed run commandAndExit vts \
--skip-all-system-status-check --primary-abi-only \
--skip-preconditions --module VtsHalMediaOmxV1_0Host \

Bug: 63796949

Change-Id: I17452d9f83cf9749cd654eaf55735563f0c5f82f

3 years agoExtend adaptive playback test for graphic buffer modes
Ram Mohan M [Thu, 14 Sep 2017 06:57:44 +0000 (12:27 +0530)]
Extend adaptive playback test for graphic buffer modes

Test: make vts -j99 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 \
&& vts-tradefed run commandAndExit vts \
--skip-all-system-status-check --primary-abi-only \
--skip-preconditions --module VtsHalMediaOmxV1_0Host \

Bug: 63796949

Change-Id: I4a4aad3a0dd4f528c8e855164cbb2998b33be1cd

3 years agoMove graphic buffer allocation to common
Ram Mohan M [Thu, 14 Sep 2017 06:34:48 +0000 (12:04 +0530)]
Move graphic buffer allocation to common

Test: make vts -j99 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 \
&& vts-tradefed run commandAndExit vts \
--skip-all-system-status-check --primary-abi-only \
--skip-preconditions --module VtsHalMediaOmxV1_0Host \

Bug: 63796949

Change-Id: I40b949199856beab73933a083045ed2e4e018c75

3 years agoEnable EOSTest_M test
Ram Mohan M [Wed, 13 Sep 2017 09:58:46 +0000 (15:28 +0530)]
Enable EOSTest_M test

eostest_m was disabled during the time of o-release. This is
enabled now. Also improved logging

Test: make vts -j99 BUILD_GOOGLE_VTS=true TARGET_PRODUCT=aosp_arm64 \
&& vts-tradefed run commandAndExit vts \
--skip-all-system-status-check --primary-abi-only \
--skip-preconditions --module VtsHalMediaOmxV1_0Host \

Bug: 63796949

Change-Id: I4797fce4f5911d9f37f5c91f179d9ab9263177f0