| 321c5fc0e22706a109aa505735cbb3397ce2c7a6 |
04-Jun-2018 |
Steven Moreland <smoreland@google.com> |
Better error for doc_comment at the end of block.
Previous error:
ERROR: syntax error, unexpected '}' at ...
New error:
ERROR: Doc comments must preceed what they describe at ...
Bug: 78135149
Test: hidl_error_test
Merged-In: Id56e34a4beb33b0323eb86a85190be63ec3b135d
Change-Id: Id56e34a4beb33b0323eb86a85190be63ec3b135d
rror_test/doc_comment_ends_block/1.0/IFoo.hal
rror_test/doc_comment_ends_block/1.0/required_error
|
| 361f943a1108b8facf76a9a8cf7fa0b94b3ac158 |
10-Apr-2018 |
Steven Moreland <smoreland@google.com> |
Fix 'ping' for Java.
ping in java was incorrectly returning an error by
default.
Bug: 77818717
Test: vts_ibase_test
Test: hidl_java_test
Change-Id: I8fa0854ddabdfe48e8c73c385677b8178c14623d
Merged-In: I8fa0854ddabdfe48e8c73c385677b8178c14623d
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 5f93004be6d4c527f261817f0262e2970065913a |
22-Mar-2018 |
Steven Moreland <smoreland@google.com> |
Fix enum toString for inherited enums + test.
This code is duplicated because of b/65200821.
Bug: 76111794
Test: hidl_test
Change-Id: I477318e7909972d21648ab14dc97c257d050e083
Merged-In: I477318e7909972d21648ab14dc97c257d050e083
idl_test/hidl_test_client.cpp
|
| 2c73b4f7d57c89261614b6c71d30e381628e1b89 |
06-Mar-2018 |
Yifan Hong <elsk@google.com> |
Add FqInstance class.
FqInstance is a wrapper around FQName to include instance
name as well.
Bug: 73774955
Fixes: 74361197
Test: libhidl-gen-utils_test
Change-Id: I5c064dec8d0e72ad1c90d3a892986f74d77788ce
Merged-In: I5c064dec8d0e72ad1c90d3a892986f74d77788ce
tils_test/main.cpp
|
| 255c9a11e080847555c6a36ae1f1903a9439c58d |
26-Feb-2018 |
Steven Moreland <smoreland@google.com> |
Clean errors for missing package roots.
Before:
$ hidl-gen -Lcheck a.b.c@1.0
hidl-gen F 02-20 11:09:59 210330 210330 Coordinator.cpp:338] Check failed: ret != mPackageRoots.end() Unable to find package root for a.b.c@1.0
Aborted
After:
$ hidl-gen -Lcheck a.b.c@1.0
ERROR: Package root not specified for a.b.c@1.0
The previous abort error confused several and also incorrectly indicated
that it was an internal error in hidl-gen ('Check failed' 'Aborted') rather
than indicating an error in the command-line arguments.
Fixes: 73649477
Test: hidl's run_all_host_tests.sh, update-makefiles works, + manual
Change-Id: Iedba2cddd2509f90862cbcc4af1462e9ba54f218
ost_test/main.cpp
|
| 71f091347acddfb178dd289a3d1c0af89f1addc7 |
20-Feb-2018 |
Steven Moreland <smoreland@google.com> |
Import fails if parsing fails.
Previously, hidl-gen doesn't distinguish between
failed parse due to parsing errors or due to the
file being missing. This means that any of a number
of times when a file was imported, it could fail
to parse, but the parsing of the file that referenced
it would continue on happily.
A case to hidl_error_test was also added which fails before
this patch and succeeds after.
Change-Id: Ie51a56eff6ef35c7ae5c513252a3a7e6581addb1
Fixes: 64773282
Test: hidl's run_all_host_tests.sh
rror_test/references_broken_package/1.0/IBar.hal
rror_test/references_broken_package/1.0/required_error
rror_test/references_broken_package/bad_package/1.0/IFoo.hal
rror_test/references_broken_package/bad_package/1.0/types.hal
|
| 99231b04b2dd38d3d14ee53d21233ec05fc5050e |
20-Feb-2018 |
Steven Moreland <smoreland@google.com> |
Enable experimental in hidl-gen.
This is for C++17 optional support.
The exact reason is for enabling Coordinator getPackageRoot methods,
etc, to return std::optional so that we can have a nice error
message instead of aborting.
Test: hidl_test
Test: hidl's run_all_host_tests.sh
Change-Id: I0c8a7687032b773a65835e89156de6b72aad08d0
idl_test/hidl_test_client.cpp
|
| 366fe777bc9f1c3a25216020d168f5a8fa990421 |
13-Feb-2018 |
Steven Moreland <smoreland@google.com> |
hidl_test(_java) require dependency.
hidl_test(_java) runs cross 32 + 64-bit tests. Modifying required
statements to require exactly what is needed.
Bug: 7456955
Test: build time errors gone, hidl_test + hidl_test_java
Change-Id: I4e60efbe018fdf0f9a8fc9713c07ce7d97ce0769
idl_test/Android.mk
ava_test/Android.mk
|
| 6d3d3c8b60b56edd309b0366ebd1139ce7f5ae94 |
08-Feb-2018 |
Steven Moreland <smoreland@google.com> |
Coordinator keeps track of output directory.
Previously it was passed around everywhere just to pass it to the
coordinator which was also being passed around everywhere.
This is one stone, here are the birds:
- closer/easier to generate dep files
- code doesn't have to care about passing this around
- we don't worry some code accidentally opens an output file
without being tracked by the coordinator
Bug: 73001417
Test: run_all_host_tests.sh
Change-Id: Idb72e71730a500feecad4a86715115a6f0e2182e
ost_test/main.cpp
|
| d623a9c5a20895973fd7305e959dad59d81aae72 |
02-Feb-2018 |
Steven Moreland <smoreland@google.com> |
Error and test for missing variable name.
Hit this recently.
Bug: N/A
Test: hidl_error_test
Change-Id: Ief7f2a5660de0464cea9381cafb8dd63f54a397a
rror_test/missing_variable_type/1.0/IFoo.hal
rror_test/missing_variable_type/1.0/required_error
|
| 382208356c8b23d9cddee11c0c615524e1b28ab9 |
25-Jan-2018 |
Yuexi Ma <yuexima@google.com> |
config change after VtsPythonVirtualenvPreparer change
VtsPythonVirtualenvPreparer was both a target preparer and multi target preparer.
The new TF suite requires it to be only one of them.
This CL configs vts xmls to use VtsPythonVirtualenvPreparer as multi target preparer
Test: vts-selftest
Bug: 69307585
Change-Id: I3a4de4bbfefcfdf70ffd5aa9609da9e74b38e3c3
Merged-In: I3a4de4bbfefcfdf70ffd5aa9609da9e74b38e3c3
idl_test/AndroidTest.xml
|
| 06a81cf671b91b7626ba702934d043edde76db23 |
17-Jan-2018 |
Steven Moreland <smoreland@google.com> |
Get imported packages based on granular names.
Previously, the names that a hidl package imported where
constructed from a list (mImportedNames) which is constructed
whenever a type is referenced. This list isn't very useful
and removal of it will be explored in the future.
Fixes: 71863483
Test: unreferenced imports successfully registered in generated
Android.bp files.
Test: hardware/interface Android.bp files are unmodified except
where they exhibit this bug and are fixed
Test: hidl's run_all_host_tests.sh
Change-Id: I38858289121dcbdda8a8d4a840fcc290501fe56a
endor/1.1/Android.bp
endor/1.1/IVendor.hal
|
| e56a217d8d1562fef0c3dc3e753d8e72fd9933c2 |
05-Jan-2018 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes Ia29c8ae4,I8ca07335
* changes:
Test for empty generates.
Automatic dependencies for error test.
|
| 1bf7b4f9f26c5d1c4c089899c9abb00f9301be2c |
05-Jan-2018 |
Howard Chen <howardsoc@google.com> |
Merge "Make HIDL java to support a interfaces in structs"
|
| 36f8e6845f90b42f1d2a727d61cefbfc1d3fe538 |
04-Jan-2018 |
Steven Moreland <smoreland@google.com> |
Test for empty generates.
Bug: 71545356
Test: hidl_error_test
Change-Id: Ia29c8ae4ffe9fb2ddf03cf621bddc72b54eba013
rror_test/empty_generates/1.0/IFoo.hal
rror_test/empty_generates/1.0/required_error
|
| 4f20ed6b8bb6ebe7d2a1354924574fc7a618a0e3 |
04-Jan-2018 |
Steven Moreland <smoreland@google.com> |
Automatic dependencies for error test.
Using "**/foo" pattern in order to ensure the test
script is always correct and so we don't have to
keep on updating these things.
Bug: 71545356
Test: hidl_error_test
Change-Id: I8ca073350fd457b5772e7af244de1dc9304ebcc3
rror_test/Android.bp
|
| af084db4490c80b9715901d279ac937d7fe43e94 |
27-Dec-2017 |
Howard Chen <howardsoc@google.com> |
Make HIDL java to support a interfaces in structs
Make hidl-gen to support interfaces in structs for java. Also
use the test method in IBaz to test arguments passing.
Test: hidl_test_java/ internal master
Bug: 70529597
Change-Id: Ife7f77a145690050d0b9191235dd20e73b121994
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 121a75804e5a71da7d09f9ecdf9cde5adcfe8e2c |
25-Dec-2017 |
Hsin-Yi Chen <hsinyichen@google.com> |
Add metadata to modules in system/tools/hidl
Add metadata. key="plan"; value="vts-library".
Test: vts-tradefed run commandAndExit vts-library --include-filter VtsHidlUnitTests
Bug: 70485967
Change-Id: Ifead85e53ca324be6f466727e57ff7907fc24b1c
idl_test/AndroidTest.xml
|
| 3aa0d7114239effdb03abe4dc43c9b3e64564dd4 |
12-Dec-2017 |
Steven Moreland <smoreland@google.com> |
Union in vec in HIDL.
Not sure why this was disallowed in the first place.
Bug: 31973802
Test: hidl_test
Change-Id: I3170560fb9e72277874f7c5861dc0ce4dcf9e40a
idl_test/hidl_test_client.cpp
|
| ebb9b24fde4dd9179a3b92c4988414dc1208743a |
12-Dec-2017 |
Steven Moreland <smoreland@google.com> |
Add missing check to java test.
Test failed to assert that an exception was actually
thrown when a service was missing.
Test: passes
Bug: None
Change-Id: I67536c18d0219369fce6c169c5bf0b9bfca82c86
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 89a9ebb3df3ab0f3694830a53b15565c08477581 |
04-Dec-2017 |
Steven Moreland <smoreland@google.com> |
Module owner support in HIDL.
- "hidl-gen -O OWNER" option supported for -Landroidbp(-impl)?
- hidl_interface owner can be set to propogate ownership to other modules
- update-makefile-helper script takes "-O" option and passes it to
hidl-gen so that you can do this.
Test: all three of the above OWNERS usecases
Test: test/run_all_host_tests.sh
Fixes: 70157937
Change-Id: Iad99a8fea45c0627d6382e42da43b214d22695bd
endor/1.0/Android.bp
endor/1.1/Android.bp
endor/update-makefile.sh
|
| ecfb4517ff68c59294f1ddc99513aef7b6dc2f24 |
21-Nov-2017 |
Howard Chen <howardsoc@google.com> |
HIDL defined type to support a composed interface
This patch supports parcel/un-parcel a struct with a composed
Interface.
Bug: 68295540
Test: hidl_test
Change-Id: Ia1d4493115c9bd2df8a5b2266632d3e88931ee4d
idl_test/Android.bp
idl_test/hidl_test_client.cpp
|
| 6961d3f487ba89ff73d89ec78f148cd130d27fa5 |
17-Nov-2017 |
Steven Moreland <smoreland@google.com> |
Add enum iteration to HIDL.
e.x.:
for (const auto v : hidl_enum_iterator<Foo>) {
...
}
Going with exposing an iterator interface completely
in the header because:
- this should only increase code size if it is used
- we want to be able to change this interface/implementation
to be more efficient or to add new features without having
to update all prebuilts
Bug: 68715899
Test: hidl_test
Change-Id: I613d5c71ebf3f58c5c92338888ad90172896178d
idl_test/hidl_test_client.cpp
|
| 3ba4df6187f5be12b120cc5bcad8e5e9dd1014e5 |
11-Nov-2017 |
Steven Moreland <smoreland@google.com> |
Update makefiles for hidl_interface.
Bug: 35570956
Test: pass
Change-Id: I4c6af124e2ef69efd15855945d0da39386b2b5d4
endor/1.0/Android.bp
endor/1.0/IVendor.hal
endor/1.0/types.hal
endor/1.0/update-base.sh
endor/1.1/Android.bp
endor/1.1/IVendor.hal
endor/Android.bp
endor/Android.mk
endor/update-makefile.sh
|
| d2af9eaac996d55763e446c35a774ad9e81160d2 |
09-Nov-2017 |
Steven Moreland <smoreland@google.com> |
Remove hidl-gen subdirs.
And no longer generate subdirs since they
aren't required.
Test: manual
Change-Id: I37582022424ffb587040559fb955aeeb3992085a
ndroid.bp
|
| ba93480f44bee7601dbbcea0753c94af8468e9a2 |
09-Nov-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix bug in enforceMinorVersionUprevs."
|
| 8775715b80249be323586dbacf01713e51546e38 |
09-Nov-2017 |
Steven Moreland <smoreland@google.com> |
Fix bug in enforceMinorVersionUprevs.
Every minor version uprev should extend the top-level
interfaces from the previous minor version. In hidl-gen,
we check that at least one interface is extended. Here,
there was a problem where if the minor version upgrade
contains an interface which is not in the previous package,
then this was counted as an extended interface.
The exact problem was that this statement:
lastFQName.getPackageAndVersion() == prevPackage.getPackageAndVersion()
considered an interface extended even if !lastFQNameExists. Since
lastFQName was always constructed from prevPackage, this statement
is always true.
Test: problematic interface fails
Test: hidl_version_test
Change-Id: I41983ad5c44eae1903123d49a9106509e4005f7b
ersion_test/Android.bp
ersion_test/bad5/version/2.2/IFoo.hal
ersion_test/bad5/version/2.3/IBaz.hal
|
| cdfe2cf0fae406bd2d7dd9850c7a3347ed9a42b8 |
07-Nov-2017 |
Steven Moreland <smoreland@google.com> |
Move disabled test out.
hidl_test must be static so we can't test preloadPassthroughService
here. Instead, the test has been moved to libhidl_test.
Test: libhidl_test, hidl_test
Change-Id: Ib2c85288c3805449101c7ea6f15370eadb13836f
idl_test/hidl_test_client.cpp
|
| 235485397176f4c800f8b9d7c83ac8413a063b31 |
07-Nov-2017 |
Dan Willemsen <dwillemsen@google.com> |
Install 32-bit hidl_test_java_native too
Fixes: 68941941
Test: m hidl_test_java
Change-Id: I66609407f7e155a937851584a3d28b2edca0ca91
ava_test/Android.mk
|
| f89e1063e162f09b6826bbc4960703649ec757bb |
01-Nov-2017 |
Yifan Hong <elsk@google.com> |
Allow equality checking and hash for HIDL interface proxies.
IFoo.Proxy.equals() -> HidlSupport.equals() ->
IHwInterface.asBinder().equals() -> HwRemoteBinder.equals().
IFoo.Stub.equals() -> default Object.equals()
Notice that IHwInterface.asBinder() returns mRemote(of type
HwRemoteBinder) for proxies and itself (of type HwBinder) for stubs.
If IFoo.Stub.asBinder() had not return "this", its equals()
should also be overridden.
Test: hidl_test_java
Bug: 68727931
Change-Id: I57936b165fbc0460d14792ce58f1f69a909547a1
ava_test/Android.mk
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| a23f1ae530383d3a0d0e5178a0bf988da789e893 |
31-Oct-2017 |
Andreas Huber <andih@google.com> |
Optimize array marshaling to copy linear ranges of memory instead of
doing per-element copies if the array element type corresponds to a primitive
java type.
Bug: 68654583
Test: hidl_test_java
Change-Id: Idb0fee0331de9bd5b3e18252312d86fa5714f633
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 9bf5a09af6faa5ca009314cbf773ae0053a6ab7c |
25-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Reland "HIDL Java getService now the same as C++."
Add getService(retry) in Java for C++ feature parity.
Due to ABI stability reasons, the API can't be made
exactly as the C++ API.
retry no retry
C++ getService tryGetService
Java getService(true) getService()/getService(false)
getService(true) should be preferred. This will get the
HIDL service and return it IFF it is declared in the
VINTF manifest/is served.
Test: hidl_test_java
Bug: 67981006
Change-Id: Ic2e41efafc50afb50126f3e628439fc1d20f4e1c
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| d5ba3da6058b72871b1686fd55e3bb31ceb50d32 |
24-Oct-2017 |
Dan Willemsen <dwillemsen@google.com> |
Remove module names from LOCAL_ADDITIONAL_DEPENDENCIES
This module would always build, since you're depending on phony targets.
It will also trigger the dependencies to be installed even if we're only
building the local module (like during checkbuild).
Bug: 68042065
Test: build/soong/build_test.bash
Test: m hidl_test_java
Change-Id: I6aa4c09b4ec4599fcdc03c0868a17fc01875383b
ava_test/Android.mk
|
| 5ed0a0b376c86d8303fe4d1e3cad4b665a80469d |
23-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Make sure comments in current.txt files are tested.
I had tested these manually. Explicit test here.
Test: hidl_hash_test
Change-Id: I9d15b8e765b1769f68afa65599496490fa4192ee
ash_test/good/current.txt
|
| 9180ecde737a8fffd74a803062957278b814d522 |
14-Oct-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Add fmq_test to hidl's run_all_device_tests script"
|
| 5e33b0cdd74c03fab31fd9ad91359a41d21cd7a9 |
14-Oct-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "hidl_test using static passthrough libs"
|
| 05de21fa8ce3b572332e6d88d5d70742c5489d40 |
13-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Add fmq_test to hidl's run_all_device_tests script
hidl + libhidl affect this test.
Test: run_all_device_tests.sh
Change-Id: I07ac84424ba5b3d40a359ab74784f05c502fa1c3
un_all_device_tests.sh
|
| b4a1a593461a0fca3003404e867e6e08cb6e00a7 |
13-Oct-2017 |
Steven Moreland <smoreland@google.com> |
hidl_test using static passthrough libs
Using static passthrough libs for portability
of the test.
Disabled a test for preloadPassthroughLibrary because
that requires shared libraries.
Bug: 67500459
Test: hidl_test
Change-Id: I29b2d505d557e63b228200db7d049efce03cc1ad
idl_test/Android.bp
idl_test/AndroidTest.xml
idl_test/hidl_test_client.cpp
|
| 87b26d7809cf5f41ccd6c977decafeedaaa51c8c |
11-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Don't enforce minor ver restrictions if types-only
It doesn't make sense to require at least one interface to
extend an interface from the previous package if there
are no previous interfaces.
Test: hidl_version_test
Bug: 67678594
Change-Id: I361f569ebc82154dcdaea98987ad3f09827e0903
ersion_test/Android.bp
ersion_test/good/version/3.0/types.hal
ersion_test/good/version/3.1/ICanExtendTypesOnly.hal
ersion_test/good/version/3.1/types.hal
|
| 887b84c552cfdb074f1bd97ec1cf219d05a97525 |
10-Oct-2017 |
Steven Moreland <smoreland@google.com> |
hidl-gen generates Java with Soong.
Bug: 33420795
Test: boot device + hidl_test_java
Change-Id: I76ab20a60df9f66b28a7f8970e107dee7853903f
endor/1.0/Android.bp
endor/1.0/Android.mk
endor/1.1/Android.bp
endor/1.1/Android.mk
endor/update-makefile.sh
|
| 4eb9df16f11c294961cb4c6d1efaf2f213013824 |
10-Oct-2017 |
Andreas Huber <andih@google.com> |
Fix startup race during execution of hidl_test_java.
Apparently startup/registration of the java service takes just a little
bit longer than it used to - causing the native client to silently fall
back to a local implementation and not talk to the java service at all.
Bug: 67634179
Test: hidl_test_java
Change-Id: I687eaedae43c95010216275fc54f3c45cadce785
ava_test/hidl_test_java
ava_test/hidl_test_java_native.cpp
|
| c4b8b6f49fd1376c8682fbd4a2120b1c40886c66 |
09-Oct-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Configure test timeout for hidlUnitTest
Bug: 67500459
Test: make vts.
Change-Id: I029b1dbfdab437697844ee42da6fe6b0f202be07
idl_test/AndroidTest.xml
|
| 768250faa6806620d97968a3bd6e8e8fc744f029 |
04-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Add hidl-module-defaults to adapter modules.
This default module is for "all hidl-gen-generated modules".
In addition to providing -Wall + -Werror, this also allows
the flags that this default to be used by all hidl-gen
generated code.
Test: update makefiles
Change-Id: I1e207cd46b86f971fe874b6887a65e00da0ec2b9
endor/1.0/Android.bp
endor/1.1/Android.bp
|
| bff4bd2c4fcb038d1ace19a085f9b3c3c8117c3a |
02-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Use wrapPassthrough in getService.
Also added tests for this in hidl_test.
This way in HIDL_FETCH_IParent, you can return an IChild object
and castFrom will work appropriately.
Test: ./test/run_all_device_tests.sh
Bug: 67104214
Change-Id: Ic317d610e733e1ead061a1090c121be30c3c4a9b
idl_test/hidl_test_client.cpp
|
| b8361a2a675455b238b58d49963504da3b05c5a4 |
03-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Remove bad/flaky test from hidl_test.
This test has a couple problems:
- it takes *way* too long to run
- it doesn't test much (of course it will fail if you try to allocate
too much memory).
- it isn't guaranteed to fail (some devices may allocate memory
lazily or have higher capacities).
Test: hidl_test is 3x as fast and passes
Bug: 66222855
Change-Id: If3c05855ddbc7ddbb24b53a58d2e94580cd856fe
idl_test/hidl_test_client.cpp
|
| b953206ee310105d4236f93e695b366a7eb05d25 |
03-Oct-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Use -Werror in system/tools/hidl"
|
| 5880ba2486b9dfca79bb752038b88e13dc5fa081 |
03-Oct-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic "no-L-makefile-prep"
* changes:
Split up generateAndroidBpForPackage.
Coordinator package root locations made private.
Coordinator getFilepath explicit support for dirs.
|
| 489aa29662e6f8c227e82bd7d869556291ed1312 |
03-Oct-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Clarify test cases which differentiate passthrough"
|
| b8e21fd1d774f81dd5ed44c8c0568b1672a9fd95 |
02-Oct-2017 |
Chih-Hung Hsieh <chh@google.com> |
Use -Werror in system/tools/hidl
* Use -Wno-error= for existing warnings, to fix later.
* Some cflags might not be used, but added to silence missing -Werror warnings.
Bug: 66996870
Test: build with WITH_TIDY=1
Change-Id: If558ab43a2c42a2a839fef7a71ad9f06c7c45de5
rror_test/Android.bp
xport_test/Android.bp
ash_test/Android.bp
idl_test/Android.bp
mpl_test/Android.bp
ava_test/Android.bp
ersion_test/Android.bp
|
| 5c70d61b8af15d775d88b27747a0eeece7d01f33 |
03-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Coordinator getFilepath explicit support for dirs.
And tests.
Bug: 33420795
Test: ./test/run_all_host_tests.sh
Change-Id: Id636a342750113033012c375526c96af8723558e
ost_test/main.cpp
|
| b354398d1957635b7870ebbbbc2d2f00fb12a43f |
03-Oct-2017 |
Steven Moreland <smoreland@google.com> |
Clarify test cases which differentiate passthrough
Some tests had TODOs on 31959402 which can't be fixed in some
cases. Instead, reasons why the test won't work are put here.
Test: ./test/run_all_device_tests.sh
Bug: 31959402
Change-Id: Ib97896d2a3e8cef1fa469d0e5afacec5c236c1fb
idl_test/hidl_test_client.cpp
|
| 3c4fea96325737d0dd91823fd4136ee0a972698c |
03-Oct-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Add VTS dependencies to host tests."
|
| f8bd0ec4ef3d8feb100396074067db41020bc4f8 |
20-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Stop generating '_vendor' libs for new makefiles.
Before the VNDK was implemented, two separate library
versions were required: vendor and non-vendor. When
the VNDK is enabled, only a vendor_available variant
is required which acts as both a vendor and a non-vendor
version of the library.
For legacy codebases, this makefile change need not be
applied.
Fixes: 65837082
Test: system makefiles are unchanged
Test: vendor makefiles have '_vendor' variant remvoed
from them.
Change-Id: I25dfd5f0fc33bcb53845a13fbbdcb6cd2ed6d489
endor/1.0/Android.bp
endor/1.1/Android.bp
endor/update-makefile.sh
|
| 3fde0bfe6f76215ee3425c22209a2d67eb95174a |
27-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix typo in update makefiles."
|
| fcdde2e275e9ea0088e054a589224aeb2359f57f |
26-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Fix typo in update makefiles.
Missing newline.
Bug: 37518178
Test: pass
Change-Id: Ibbb0046764b6d4c6053e29ea54554fc36062f4fb
endor/1.0/Android.bp
endor/1.1/Android.bp
|
| c6f3162b885d875943a6d48196b0262d2ac51bf4 |
26-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Additional test for hidl registering.
Make sure other registered interfaces are left untouched
even if they are the same instance.
Test: hidl_test
Change-Id: I2f7a3f1836d270addc7b46ff92c3aeb547e11828
idl_test/hidl_test_client.cpp
|
| 9ee1aa24b14c0f88fe9f4895a426e2ceed05db8a |
26-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add VTS dependencies to host tests.
Added verification of VTS tests.foo profiler + driver.
Although this is much more than just a test of -Lvts, it
can help ensure we don't break VTS as many unrelated changes
can do so.
Test: run_all_host_tests.sh
Change-Id: Ia24758d7624c7a210227a4a728c09a6137283b00
un_all_host_tests.sh
|
| f78c44db648a1f7dd38fed110b8e53cc04f96516 |
26-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Coordinator add getFormatter.
This way Coordinator can keep track of the files
that are used and we can consolidate the code that
builds paths to output files.
Test: all host tests
Bug: 65738957
Change-Id: Iaf1b00e130be892ef1bb620017032c29140ee5b6
ost_test/main.cpp
|
| 3d6e7beee4f180a1cd442775f550059f09d5f7ab |
26-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add test for -Lexport-header.
Statically verify output is valid and that
generation works so that we don't accidentally
break this functionality.
Test: run_all_host_tests.sh
Change-Id: Ia425dfb463a3ac0a932212ce7bdfc64d31ade9c5
ndroid.bp
xport_test/1.0/IFoo.hal
xport_test/1.0/types.hal
xport_test/Android.bp
xport_test/test.cpp
un_all_host_tests.sh
|
| f2c21ebf13941bc2d785c880074a2d99e51fbae7 |
26-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "hidl_test: add expression@1.0 to AndroidTest.xml"
|
| b866d264049c32013863e58169a52a7151a9b779 |
25-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic "hidl-adapter"
* changes:
Add adapters for HIDL interfaces.
Add tests for conflicting interfaces.
|
| 9a6da7a784bea068b1f3fc3758c9cfc785183693 |
16-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add adapters for HIDL interfaces.
An adapter for a HIDL interface takes an x.(y+1) interface
and registers it as an x.y interface.
This adds new autogenerated test libraries to HIDL which
can be used to run tests.
Test: can switch out implementations
~/android/master$ lshal | grep hidl.allocator
Y android.hidl.allocator@1.0::IAllocator/ashmem 0/1 652 471
~/android/master$ lshal | grep hidl.allocator
Y android.hidl.allocator@1.0::IAllocator/ashmem 0/1 7635 471
~/android/master$ lshal | grep hidl.allocator
Y android.hidl.allocator@1.0::IAllocator/ashmem 0/1 652 471
Test: build all adapters in hardware/interfaces
Test: adapt @1.1::IPower to @1.0::IPower
Test: NFC works even after having an adaption layer over it (@1.0 -> @1.0 :/)
Bug: 37518178
Change-Id: Ia009d59352963321f16e085f034f70c462ccceeb
endor/1.0/Android.bp
endor/1.1/Android.bp
|
| 44b86b10a2583e828f2f2d9330271e7c9e671c4b |
25-Sep-2017 |
Yifan Hong <elsk@google.com> |
hidl_test: add expression@1.0 to AndroidTest.xml
Test: vts-tradefed run vts -m VtsHidlUnitTest
Change-Id: Iccb9557b6ea53f2bf9f57d55a666a8b4bca7ddfa
Fixes: 65950515
idl_test/AndroidTest.xml
|
| da4a014304fb2244e15edc8077b6f67317931796 |
25-Sep-2017 |
Steven Moreland <smoreland@google.com> |
More specific host/device test summary.
Since many of the tests being run also have their
own summaries, I'd rather qualify these titles.
Test: output as expected
Change-Id: I86ed3d155df293b5cf2ac5d4e8f50993a62a77ff
un_all_device_tests.sh
un_all_host_tests.sh
|
| 74214204ebc00cdbb7f7f91b1f6be554dd52a84a |
19-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add tests for conflicting interfaces.
For IBar extends IFoo if 1.0::IFoo/default is being registered, remove
IFoo/default. This makes sure the following two things are equivalent
1). IBar::castFrom(IFoo::getService(X))
2). IBar::getService(X)
Test: hidl_test
Bug: 62303973
Change-Id: I3bb106f597b694d20d96fb3aaeed8abe04e0c229
idl_test/hidl_test_client.cpp
|
| 1ef6899717f9571bba7db4b1d6fab9a008d2b8b2 |
22-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add hidl_error_test.sh to test dependencies.
This way, changes to the script cause the test
to be rerun as well.
Test: hidl_error_test
Bug: 66488832
Change-Id: I1fe000be05ad593bcf585d5dd6ca74fdf2e6151f
rror_test/Android.bp
|
| 23f80119e26b1ed52ca129860820cb652a0e21f6 |
22-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Use hidl_error_test.sh from any directory.
Test: hidl_error_test
Test: ./test/error_test/hidl_error_test.sh
Bug: 66488832
Change-Id: I8a9f5749de9a23d31aa27949e173d3be2a5645ef
rror_test/hidl_error_test.sh
|
| 32d29264447a7f638cd5b43beab9ec04e10a12dc |
22-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add required errors for hidl_error_test.
There have been cases in the past when these have failed from checks
rather than from the proper errors. Here we make sure that the paths
are receiving the proper error.
Test: hidl_error_test
Change-Id: I0a3c51edc92269ddee59070f99f76ef7a8200440
Fixes: 66488832
rror_test/Android.bp
rror_test/bad_character/1.0/required_error
rror_test/cpp_forward_decl_restriction/1.0/required_error
rror_test/enum_storage/1.0/required_error
rror_test/enum_unique_field_names/1.0/required_error
rror_test/enum_unique_field_names_extends/1.0/required_error
rror_test/hidl_error_test.sh
rror_test/interface_extends_only_interface/1.0/required_error
rror_test/interface_no_using_reserved_method_names/1.0/required_error
rror_test/interface_semicolon/1.0/required_error
rror_test/interface_unique_method_names/1.0/required_error
rror_test/interface_unique_method_names_inheritance/1.0/required_error
rror_test/method_ends_in_semicolon/1.0/required_error
rror_test/no_cyclic_declaration_1/1.0/required_error
rror_test/no_cyclic_declaration_2/1.0/required_error
rror_test/no_cyclic_declaration_3/1.0/required_error
rror_test/no_cyclic_declaration_4/1.0/required_error
rror_test/no_cyclic_expressions_1/1.0/required_error
rror_test/no_cyclic_expressions_2/1.0/required_error
rror_test/no_cyclic_expressions_3/1.0/required_error
rror_test/no_data_outside_interface/1.0/required_error
rror_test/no_interface_in_types/1.0/required_error
rror_test/no_two_interfaces/1.0/required_error
rror_test/same_name_interface/1.0/required_error
rror_test/same_package_name/1.0/required_error
rror_test/scope_unique_type_names/1.0/required_error
rror_test/struct_unique_field_names/1.0/required_error
|
| 80ccd1aa19b9c0ee0a67391c5f9600718faf11bd |
22-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic "hidl-host-test"
* changes:
Add hidl-gen-host_test.
Move hidl_test into its own folder.
Remove extra '/' from Coordinator paths.
Coordinator checks for package root conflicts.
|
| 745a168a88cdd3ec3d51dc2b75c8c0b5aff54a38 |
22-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix typo in hidl_error_test.sh usage."
|
| 37d261100808c102ce3a2cffba456fb2ca12b384 |
21-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Better error messages for enum storage type.
If enum storage type is not defined, get:
- error: "ERROR: Must explicitly specify enum storage type for A at ..."
- parsing can continue and show other errors
Test: manual, host tests
Fixes: 66233431
Change-Id: I22927ad0b03c365a1dbe7dcd85cd4c303b231eda
rror_test/Android.bp
rror_test/enum_storage/1.0/IFoo.hal
|
| 53b9bfb42a5f1426c5eb21234d2b4d2a4ad0222c |
20-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add hidl-gen-host_test.
Note the primary difference between this and utils_test is that
these tests are host only whereas utils_test is testing the part
of hidl-gen that is available both on host and on device.
This only contains tests for Coordinator and Location because they
have been modified recently. More tests will be added in the future.
Also updated run_all_host_tests.sh to accept additional host tests.
Test: run_all_host_tests
Change-Id: Ie5a0727157ea109f2e6d408c1d5acc84517e9587
ndroid.bp
ost_test/Android.bp
ost_test/main.cpp
un_all_host_tests.sh
|
| 32a9eccae19b96573a21fa5b65e12e3b9c83f8b2 |
20-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Move hidl_test into its own folder.
Rather than mixing it with the other tests.
Test: hidl_test
Change-Id: I964d85429c296129c0c74fa3f1b98610e1811f5c
ndroid.bp
ndroid.mk
ndroidTest.xml
ooCallback.cpp
ooCallback.h
idl_test
idl_test.h
idl_test/Android.bp
idl_test/Android.mk
idl_test/AndroidTest.xml
idl_test/FooCallback.cpp
idl_test/FooCallback.h
idl_test/hidl_test
idl_test/hidl_test.h
idl_test/hidl_test_client.cpp
idl_test/hidl_test_helper
idl_test/hidl_test_servers.cpp
idl_test/static_test.cpp
idl_test_client.cpp
idl_test_helper
idl_test_servers.cpp
tatic_test.cpp
|
| 8e35e65214da0b72fa6425f368bb18c39840c44c |
22-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Fix typo in hidl_error_test.sh usage.
Test: hidl_error_test.sh
Change-Id: If259382fa968131d19d86fd41335df6fe8943343
rror_test/hidl_error_test.sh
|
| 8698469c7f42bde2d73ba09765c8f46f0e21e4e7 |
20-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add run_all_device_tests.sh script.
Convenient script to run all device tests.
Oftentimes, there are regressions in the Java test that
isn't realized until after CL submission. Adding a convenient
script so that all HIDL tests can be run by doing:
$ ./test/run_all_device_tests.sh
$ ./test/run_all_host_tests.sh
Test: run_all_device_tests.sh
Change-Id: I7e4b664dd030256dbdf36de967d702a788828353
un_all_device_tests.sh
|
| 207a9fe818224d9942edb7f680e9c01b7fe7e3fc |
19-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic "hidl-unit-tests"
* changes:
Add script for running all host tests.
Add unit tests for StringHelper.
utils: Add missing header export.
Fix infinte loop in StringHelper trims.
Don't abort on incomplete package.
|
| 275b7b219761982cd0a4a544a3ac7cf86ff62649 |
19-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "hidl_test: add trie@1.0 to AndroidTest.xml"
|
| 960dd87c36b9981e796b696fc8f1ff9be659fe46 |
19-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add script for running all host tests.
Test: run_all_host_tests.sh
Change-Id: I2abf0a5cafefffda16021d56338b08768db6aceb
un_all_host_tests.sh
|
| 69788dc46dd4b831baf0f3d70111940d6f102ddc |
19-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add unit tests for StringHelper.
There were some corner cases. Going to start
hidl-gen unit tests with this.
Test: libhidl-gen-utils_test
Change-Id: I17a6b81cb9910e614684671091eb218636cdc419
ndroid.bp
tils_test/Android.bp
tils_test/main.cpp
|
| fb9a1de22293aa9e77dacf0dbecfeae1ef8432ae |
19-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix error message in hidl_error_test.sh."
|
| 250f0a22df1fa75a78d7749c5ca9289b81793c0d |
18-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Fix error message in hidl_error_test.sh.
- Use "error:" prefix so it's recognized by most tools.
- Specify what failure is.
Test: hidl_error_test
Change-Id: Ie7e77985db75d7e73e1a0388fd6d8e1e611a8339
rror_test/hidl_error_test.sh
|
| 478958888f861868616a5f588d6ba2fb558bc5dc |
18-Sep-2017 |
Yifan Hong <elsk@google.com> |
hidl_test: add trie@1.0 to AndroidTest.xml
This fixes hidl_test in VTS.
Test: vts-tradefed run vts -m VtsHidlUnitTest
Change-Id: Ib76d1305b0d8009f21c7fee802b61395a87a1ef0
ndroidTest.xml
|
| c726d8658431b40630f946490b2e6bb10bbde079 |
16-Sep-2017 |
Yifan Hong <elsk@google.com> |
hidl_test: add expression test
Test: m hidl_test -j
Change-Id: Idce83dd3d3c1fdc8204c377602b83c12f0603ab1
Fixes: 64613737
ndroid.bp
tatic_test.cpp
|
| 461391474f93a6f4d59921bd2da5463d0e000884 |
16-Sep-2017 |
Yifan Hong <elsk@google.com> |
hidl_test: move static test to a separate file.
Test: m hidl_test -j
Bug: 64613737
Change-Id: Ifdc5cf3857eb17fe060a8e172d88ff2cfe227dd6
ndroid.bp
idl_test_client.cpp
tatic_test.cpp
|
| c57402d65b89d25e81db30b66d955728b5ef99ed |
16-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "hidl_test: fix FailedBatchAllocate test."
|
| 61539737f498d1c0d06314714d6f9127173645fc |
16-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Omit HIDL test libraries from the VNDK."
|
| bf3b355ef9c4233a5bbb47e4c4ae9e72121aed35 |
15-Sep-2017 |
Yifan Hong <elsk@google.com> |
hidl_test: fix FailedBatchAllocate test.
The test is not allocating enough hidl_memory 's to fail
the call.
Test: hidl_test
Change-Id: I01bd85acac617baed1d379f1d60844cb9d5a1c78
idl_test_client.cpp
|
| 372832539f5019be5b5382e6cb6c080dc304cff7 |
29-Aug-2017 |
Jiyong Park <jiyong@google.com> |
Omit HIDL test libraries from the VNDK.
Add -t option to hidl-gen to emit omit vndk enabled when generating
Android.bp. This is to filter-out the shared libs of HALs for testing
from the VNDK directory (system/lib/vndk) to save storage.
update-makefile-helper.sh is also updated to set the -t option when
there exists .hidl_for_test file in the directory for an HAL package.
Bug: 64776708
Test: no tests.vendor*.so file in /system/lib[64]/vndk on 2017 pixel
devices
Test: hidl_test
Merged-In: Idb3a14bdecbda43309db8ef862baab59ff8ba647
Change-Id: Idb3a14bdecbda43309db8ef862baab59ff8ba647
(cherry picked from commit 16b2752f484d8d94ec4ef552a503850c750192a0)
leanSpec.mk
endor/1.0/Android.bp
endor/1.1/Android.bp
endor/update-makefile.sh
|
| 87e69dcd459da3a68d6ab464f82334baeb8c2782 |
13-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Fix imports.
Allow imports to reference fully qualified names. Remove
recursive call in import lookup. Add test for references
using fully qualified names.
Bug: 65597200
Test: m hidl_test (type tests done statically)
Change-Id: I912f1201bfa2d00cbaf711a7c34b934271ff37f0
idl_test_client.cpp
|
| a3805bdfe22257a94c6a60570e2015f86c415128 |
14-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Make hidl-gen print full name"
|
| 4df4fa0fa0470da35818e0398663b481b063fb3e |
12-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Make hidl-gen print full name
It was assumed that there was no need to
print full names for types.
hal example:
struct S {};
typedef S A;
struct B {
struct S {
A a;
};
};
output:
struct S final {};
typedef S A;
struct B final {
struct S final {
S a;
}
}
So, at least for this case it is necessary to print
namespace::S instead of S.
Fixes: 65563308
Test: mma
Test: manual check for example above
Change-Id: I612aa3c09a7d543de644b58b45c8dc48be8f6099
idl_test_client.cpp
|
| 458ca365d99ae9f98bd59c0c2ea8b4fa54e2d535 |
13-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Put type definitions in reversed topological order
Puts a warning comment if the order is changed.
Doesn't change the order if that is not necessary.
Fixes: 65636464
Bug: 31827278
Test: mma
Test: ~/android/aosp/out/soong/.intermediates/hardware/interfaces is
changed only for trie
Change-Id: I91ec448859fcdd94228acfaa1f67de7bb54f777c
rror_test/Android.bp
rror_test/cpp_forward_ref_incomplete_restriction/1.0/types.hal
|
| 8cf59e8473af2eba9621627471ca588bc6fdf9ff |
14-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Refactor hidl_error_test build"
|
| bc051af6575bd94254d792157b1554ff5428da9e |
13-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Tests for forward reference restrictions"
|
| bcc8bc7a6252086960d09f192c6f27cf35c12e67 |
09-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Test hidl forward references
Bug: 31827278
Test: hidl_test
Change-Id: I844de5cfb90f4b512d03d8b68a9060343f1be9a4
ndroid.bp
idl_test.h
idl_test_client.cpp
|
| 1becfe9950f1de04a71304ed941f83579838d1f2 |
10-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Refactor hidl_error_test build
bash scripts that assumes that every directory in test/error_test
is a test.
Test: hidl_error_test
Change-Id: Id45d224dba12d96d1fd149d337cf4b22a4f62bf9
rror_test/Android.bp
rror_test/hidl_error_test.sh
|
| b455d3f6ee2a7cdbe2e05c63df3659576a751690 |
09-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Tests for forward reference restrictions
Tests checking acyclic and checking C++ restrictions.
Bug: 31827278
Test: hidl_error_test
Change-Id: Ib41fff0ba7b1d3167dbe99e5d76b392bf033038e
rror_test/Android.bp
rror_test/cpp_forward_decl_restriction/1.0/types.hal
rror_test/cpp_forward_ref_incomplete_restriction/1.0/types.hal
rror_test/no_cyclic_declaration_1/1.0/IFoo.hal
rror_test/no_cyclic_declaration_2/1.0/IFoo.hal
rror_test/no_cyclic_declaration_3/1.0/types.hal
rror_test/no_cyclic_declaration_4/1.0/types.hal
rror_test/no_cyclic_expressions_1/1.0/types.hal
rror_test/no_cyclic_expressions_2/1.0/types.hal
rror_test/no_cyclic_expressions_3/1.0/types.hal
|
| 68a520f1862161fd16f623276aa3c29cad94e375 |
11-Sep-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic "fix-hidl-bad-characters"
* changes:
Make spacing consistent in hidl-gen_l.ll.
Add missing dependencies to error_test.
Fix bad character being disallowed in .hal files.
Make keyword token have void type.
|
| 8ea1d007d59cccc2c86ce773d638d63f2b86c3a0 |
07-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Add missing dependencies to error_test.
Test: error_test
Change-Id: If46f645572116e001a7777eb79b7fb366713d2e6
rror_test/Android.bp
|
| 7a4c662f930d410d4da960ffbc5c7aa87ad128dc |
07-Sep-2017 |
Steven Moreland <smoreland@google.com> |
Fix bad character being disallowed in .hal files.
Instead of ignoring bad characters, we fail.
Note this was taken advantage of in the past by some vendor
hals which use \r\n newlines. Because of this, this CL
also adds explicit support for them.
Test: (added) test that file with emoji fails.
Test: (added) \r\n newlines still work.
Change-Id: I1f221e4db4d0b132de470fc32f8a70b3f345b39a
Fixe: 65422445
rror_test/Android.bp
rror_test/bad_character/1.0/IFoo.hal
endor/1.0/IVendor.hal
|
| bd3096d5397fec2ff4133338098a842b1440080f |
09-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Remove unused imports from hidl_test_servers
These imports became unused after change of hidl_test.h
Test: mma
Change-Id: I2938d758f4b58a03088c71fecf5c340438e87af5
idl_test_servers.cpp
|
| 565b0137b2ad0e8e239d38e92c025f096de62f2d |
07-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Refactor AST::addScopedType.
Makes NamedType receive full name in constructor.
Adds test that defined type names are unique within one scope.
Test: mma
Test: hidl_error_test
Change-Id: If218e1febc2af9f44c5908408f67e772efdda18e
rror_test/Android.bp
rror_test/scope_unique_type_names/1.0/IFoo.hal
|
| 116a5251775e3c34e8241c92f0d3a60cfd530322 |
07-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Test interface method name uniqueness validation
Test: hidl_error_test
Test: manual test run
Change-Id: Ieeea0b4b38b5cf11281a726ac5c7d2b69382ca9b
rror_test/Android.bp
rror_test/interface_no_using_reserved_method_names/1.0/IFoo.hal
rror_test/interface_unique_method_names/1.0/IFoo.hal
rror_test/interface_unique_method_names_inheritance/1.0/IBar.hal
rror_test/interface_unique_method_names_inheritance/1.0/IFoo.hal
|
| 8625e1597b2d7f690f88c1060105269c30d162e1 |
05-Sep-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Remove ambigious reference from test hidl
Necessary for forward references support.
Bug: 31827278
Test: mma
Change-Id: Ib4abe2bede48b0e8fae5a644e7d9e92f5e2796f1
endor/1.1/IVendor.hal
|
| 0737c2a4a10731188ee9a94f864b0c7b462d55db |
31-Aug-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Extending existing arrays is moved from parsing
Moves setting array exact element type and size from
parsing to post parse pass.
Fixed an issue about incorrect output array dimension:
input:
typedef int32_t[2][3] x;
typedef x[4][5] y;
output:
typedef ::android::hardware::hidl_array<int32_t, 4, 2, 3, 5> y;
fixed output:
typedef ::android::hardware::hidl_array<int32_t, 4, 5, 2, 3> y;
Addes test for the described case.
Fixes: 65257916
Test: mma
Test: /hardware/interfaces output is unchanged, however
CL changes the behavior
Change-Id: I91a63e3c5749b092a8dad942c4cf6f294b79eda9
idl_test_client.cpp
|
| 81bcf6819be1ba939b83641a75f0c5097de4db00 |
24-Aug-2017 |
Steven Moreland <smoreland@google.com> |
0 hashes for unreleased interfaces.
This allows us to detect when an interface is being used
which isn't released even when the current.txt file isn't
available at runtime.
Bug: 63388924
Test: hidl_test
Change-Id: I33942c6c51fd74ac8155e7476b3e3fc3d724e365
idl_test_client.cpp
|
| 0344e61434d7af537cbf33c86b47ec165593f7dd |
25-Aug-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Get rid if Reference<Interface>
This change allows to have
std::vector<Reference<Type>*> getReferences()
^
Thus we can make recursive lookup be like:
for ref : getReferences() { ref->lookup() }
The check that Interface expends only interfaces is moved to
Interface::validate.
Bug: 31827278
Test: mma
Test: new error test
Change-Id: I8e81b1bdf5658749d4b33d02e812a49765dc03bb
rror_test/Android.bp
rror_test/interface_extends_only_interface/1.0/IFoo.hal
rror_test/interface_extends_only_interface/1.0/types.hal
|
| ec123f3609c91fb5b1361466aedd8e2591c6bf83 |
22-Aug-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Tests for hidl-gen check for scope field uniqueness
Fixes: 64373764
Test: hidl_error_test
Test: every new test was run separately to make sure that they fail
because of new checks
Change-Id: Iba8ec14c60ce73624459b0ddd2e8e60aa8a113f7
rror_test/Android.bp
rror_test/enum_unique_field_names/1.0/IFoo.hal
rror_test/enum_unique_field_names_extends/1.0/IFoo.hal
rror_test/struct_unique_field_names/1.0/IFoo.hal
|
| 306cc4b6e8f90402a4efea3142628840682998cb |
02-Aug-2017 |
Chih-Hung Hsieh <chh@google.com> |
Fix or suppress misc-macro-parentheses in system/tools/hidl.
* Use NOLINT to suppress wrong warnings around macro parameters
used as type arguments or statements.
* Replace macro EACH_SERVER with template function runOnEachServer,
whose template parmater is a class template.
Bug: 28705665
Test: make with WITH_TIDY=1 WITH_TIDY_CHECKS=-*,misc-macro-* \
WITH_TIDY_FLAGS=-header-filter=system/tools/hidl/.*
Change-Id: Ibf4e8f8c0291662851af615071c58df971844229
idl_test.h
idl_test_client.cpp
idl_test_servers.cpp
|
| 5b527a14e4b4e8213aa0938cea3a92cc1dd9f83d |
12-Aug-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Remove unnecessary folders in error_test: follow up fix
Package check appeares after syntax check, so
I missed to change package name.
Also, forgot to delete one line in build script.
Test: hidl_error_test
Test: interface_semicolon passes if we add semicolon
Change-Id: If7882e2d4dd9c12624119b047674b113c72bb0f7
rror_test/Android.bp
rror_test/interface_semicolon/1.0/IFoo.hal
rror_test/method_ends_in_semicolon/1.0/IFoo.hal
rror_test/no_data_outside_interface/1.0/IFoo.hal
rror_test/no_interface_in_types/1.0/IFoo.hal
rror_test/no_interface_in_types/1.0/types.hal
rror_test/no_two_interfaces/1.0/IFoo.hal
rror_test/same_name_interface/1.0/IFoo.hal
rror_test/same_package_name/1.0/IFoo.hal
|
| f90d06b40e2b1251c63a7d2d578c5c5fbf90e97e |
11-Aug-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Remove unnecessary folders in error_test
There is no need to have an additional folder in the path
Test: hidl_test_error
Test: ran one command manually
Change-Id: Iec0648e89ae832fb53e8ab02b6705d6a31f25abc
rror_test/Android.bp
rror_test/interface_semicolon/1.0/IFoo.hal
rror_test/interface_semicolon/error/1.0/IFoo.hal
rror_test/method_ends_in_semicolon/1.0/IFoo.hal
rror_test/method_ends_in_semicolon/error/1.0/IFoo.hal
rror_test/no_data_outside_interface/1.0/IFoo.hal
rror_test/no_data_outside_interface/error/1.0/IFoo.hal
rror_test/no_interface_in_types/1.0/IFoo.hal
rror_test/no_interface_in_types/1.0/types.hal
rror_test/no_interface_in_types/error/1.0/IFoo.hal
rror_test/no_interface_in_types/error/1.0/types.hal
rror_test/no_two_interfaces/1.0/IFoo.hal
rror_test/no_two_interfaces/error/1.0/IFoo.hal
rror_test/same_name_interface/1.0/IFoo.hal
rror_test/same_name_interface/error/1.0/IFoo.hal
rror_test/same_package_name/1.0/IFoo.hal
rror_test/same_package_name/error/1.0/IFoo.hal
|
| 79e700eebd33a571892c57d2c55424fe9f49a3f7 |
26-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Update makefiles for vndk enabled.
HIDL makefiles now generate "vndk: {enabled: true}"
for interface libraries so that they are included
in the VNDK.
Test: none
Bug: 63866913
Merged-In: I1dbca3933f4bc585b22660fc47d6ba4a5f7fdd51
Change-Id: I1dbca3933f4bc585b22660fc47d6ba4a5f7fdd51
(cherry picked from commit 27d78442df86036dce2d4af82fb49841aeceac88)
endor/1.0/Android.bp
endor/1.1/Android.bp
|
| 3ee43c82a006963d26815ae59605be61a712eeb8 |
04-Aug-2017 |
Steven Moreland <smoreland@google.com> |
Remove Soong variables from HIDL code.
Bug: 64318482
Test: links
Change-Id: I6828d204b4cfa0968fdadc3db372ad0fcddbbf6e
ndroid.bp
|
| bc7e831671ea6679cc57716e15f24ded2d57bd7e |
02-Aug-2017 |
Steven Moreland <smoreland@google.com> |
Remove HIDL static targets.
These are now included in the regular targets.
Bug: 36376126
Test: marlin boots and works w/o static libs
Change-Id: Ic757cfc335a31af034559d29006532177c41e04a
endor/1.0/Android.mk
endor/1.1/Android.mk
|
| 98ca32ffe7f708d56c78fb70f34432587a773ef7 |
31-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Enable HIDL libs to be static."
|
| e404e5ad9ec80376b4f8f5df49d03a32b3b18173 |
29-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Silence impl_test warnings.
This test tests that a boilerplate hidl hal implementation compiles as it is
generated. However, this generated code is expected to leave all function
arguments unused. Rather than modify the output (and make it harder to use), I
have silenced the errors that it emits.
Test: impl_test produces no warnings
Change-Id: I120677c5370a8518516af84eca6da0d8dcf4f53a
mpl_test/Android.bp
|
| a036a0a29082f57597502da779f996501873e236 |
27-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix hidl_test when there is a failure."
|
| 5ea4320aef4b7b1547c70010ade333c71403a0e1 |
26-Jul-2017 |
Keun Soo Yim <yim@google.com> |
mark optional dependency to test/vts/tools/build/Android.host_config.mk
Test: mma
Bug: 64076762
Bug: 63013400
Change-Id: I7371a69bf508c0b679b0bb87fbed83def7f47b10
ndroid.mk
|
| d7995028b401e3263f0736d52efe9a547043ee55 |
26-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Fix hidl_test when there is a failure.
'set -e' was added, but this breaks the test badly when
it fails because the test relies on checking $? and 'set -e'
prevents it from killing the server process.
Test: hidl_test
Bug: 63940457
Change-Id: Ia863f35d899aa025c5576d6617c845f17e9ccb5c
idl_test
|
| 02516de420064d8d26ae21dad479e8f82c21f3b8 |
26-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Enable HIDL libs to be static.
This is required for a couple of reasons:
- enabling HIDL passthrough in recovery
- enabling VTS tests to be static blobs
Bug: 32920003
Bug: 64040096
Test: update-all-google-makefiles.sh
Change-Id: I28eede0844d23f5acf970c5db95860a792625f92
endor/1.0/Android.bp
endor/1.1/Android.bp
|
| 820926c540cc67e9516f1cda69456e59ef1363cf |
26-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic 'hwservicemanager-1-1'
* changes:
Support minor version upgrade of hwservicemanager.
Add test for 1.1 IServiceManager.
|
| f7f2a9acce2e34282c67e71baf4df0bd5ddfe5f5 |
22-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Support minor version upgrade of hwservicemanager.
Also has the side effect of making autogenerated files contain
a summary of everything that is in a specific hal.
- reference manager package instead of specific version
when determining if a hal is the manager.
- fix problem where static methods (specifically
registerForNotifications) are resolved over
instance methods when an instance method is
declared in a parent class and the static method
is declared in a derived class.
Test: can compile android.hidl.manager@1.1
Test: manually inspect output files for changes
Test: hidl_test
Bug: 38128519
Change-Id: Id72cc65cd8cf2286501dfcd40073458313dc32dd
endor/1.0/IVendor.hal
|
| 79d98cd965a25070552c3b5da2124a4dfcce9b13 |
25-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Add test for 1.1 IServiceManager.
This adds a test for IServiceManager::unregisterServiceNotification.
Bug: 63940457
Test: hidl_test
Change-Id: I4212319b687041bdb50706bec76c017b0e4b3018
idl_test
idl_test_client.cpp
|
| b5fc9d57b2af41628559a7627415c586cbe8f26b |
25-Jul-2017 |
Martijn Coenen <maco@google.com> |
Actually build vendor extension.
Test: mma
Change-Id: I2d8453ed9dcda460ee16afdb748f68c17cee15a4
ndroid.bp
|
| 5edf528b901764bb1610e3995d0864d7d8026533 |
20-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Fix _vendor generated makefiles when they are deep
This makes sure the entire closure of modules remains on vendor.
Test: tests/vendor/
Test: make sure system makefiles are unaffected
Bug: 63883276
Change-Id: I8b563207b6d6073e6379d1f2681eb09637140bc2
endor/1.0/update-makefile.sh
endor/1.1/Android.bp
endor/1.1/Android.mk
endor/1.1/IVendor.hal
endor/update-makefile.sh
|
| 57a89366d506a6fb225fe96baacfa9cfa770cfe2 |
21-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Revert "Revert "Allow proxies to be compared and identified.""
This reverts commit a8559fcb28b854ef4f1b1896292a46b9329d2a40.
Now that branches are synced up again.
(originally reverted due to merge conflicts)
Bug: 32172906
Test: hidl_test
Change-Id: I62e53c89cc86cafe63a6ea90309fd6e4bf44b300
idl_test_client.cpp
|
| 057d661196c6bf50b41c238f3d02bbbb4cd3578a |
21-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Fix inconsistency between hidl_test in AOSP.
Extra includes here that aren't in master.
Test: none
Bug: 63899698
Change-Id: I1b62be9662f60b843bc49cd5be2e888cd8a0889e
idl_test_client.cpp
|
| c7230be6f1e7a191eb46e83e8e2d58be4817bdaa |
21-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Merge "Revert "Allow proxies to be compared and identified.""
|
| a8559fcb28b854ef4f1b1896292a46b9329d2a40 |
21-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Revert "Allow proxies to be compared and identified."
This reverts commit 217fc6026b84f2ec3b6818e4def85c28c2a68a59.
Change-Id: Iba38905bed21bc91f1119fe6b56f57228121a752
idl_test_client.cpp
|
| 882fce7885208fd28bf6c49cf7c33491f5c999cc |
21-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic 'fix-bnmap'
* changes:
Allow proxies to be compared and identified.
Restore order from 'Fix infinite recursion'
|
| 0b1ed60b3a20560b45a9fbd01874a8d5bb9bdb1b |
21-Jul-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Merge "Update VtsHidlUnitTest"
|
| 217fc6026b84f2ec3b6818e4def85c28c2a68a59 |
16-May-2017 |
Steven Moreland <smoreland@google.com> |
Allow proxies to be compared and identified.
Note, this is not an ABI breakage because the virtual
destructors being added are already created implicitly.
In this CL, we delete the cached instances of
BnHw* and and then compare BpHw* objects based on
their remote pointer.
Bug: 63864326
Bug: 38348572
Bug: 32172906
Test: boot and use a device
Test: boot a marlin
Test: hidl_test, hidl_test_java
Change-Id: I7cecd9015231bdf9ad4c039df2e4ecb031077dd7
idl_test_client.cpp
|
| 05efc1e4cf9b19f8c9808f543dc61c4c6f8a53db |
21-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Restore order from 'Fix infinite recursion'
Test: pass
Bug: 63899698
Merged-In: Ifc49bcc480b894f2566837865eb350bb8e143df3
Change-Id: I88db5f67b856b121545384e832ead6c0deddf50a
idl_test_client.cpp
|
| 225bc92ef6ab5ea4bc7235d404d44a792e2dae56 |
27-Jun-2017 |
Martijn Coenen <maco@google.com> |
Fix infinite recursion in onTransact().
When sending an unknown message code to a server
object, onTransact() would be called recursively,
which will cause the process to crash.
Bug: 63069272
Bug: 63899698
Test: hidl_test
Merged-In: Ifc49bcc480b894f2566837865eb350bb8e143df3
Change-Id: Ifc49bcc480b894f2566837865eb350bb8e143df3
idl_test_client.cpp
|
| c668a6b11e4f9cde60644df057cb6cdf5bb5108d |
07-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Test preloadPassthroughService.
This will preload the appropriate libraries
in the appropriate namespaces so that getService
will be faster.
Test: hidl_test
Bug: 62376972
Bug: 63899698
Merged-In: I2b46cbe5dabccfbd33e7d1e7bd0c08af643b8dd6
Change-Id: I2b46cbe5dabccfbd33e7d1e7bd0c08af643b8dd6
idl_test_client.cpp
|
| 60aae9fa673ed4e0806c0bae810d024cfcd9c335 |
27-Apr-2017 |
Martijn Coenen <maco@google.com> |
Use String.getBytes().length to determine string length.
String.length() returns the number of UTF-16 code units,
but hidl_string objects are encoded in UTF-8; this will
cause verification failures in strings that are valid UTF-8,
but have a different number of code units in (Java's) UTF-16
encoding than in the native UTF-8 encoding.
Bug: 37722371
Bug: 63899698
Test: hidl_test_java with shiny UTF-8 string (that requires a
different number of code units in UTF-16)
Merged-In: I4fcf12c91b306c54c73013aef11bb049cb09190a
Change-Id: I4fcf12c91b306c54c73013aef11bb049cb09190a
ava_test/hidl_test_java_native.cpp
|
| 2522af64f03afc4225845e2c7699c4055d674281 |
20-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Remove unused using statements Test: links, compiles"
|
| 940ec12a3ce1169ec56454cc3b220a1af121b1b1 |
20-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Support vec<memory> in HIDL."
|
| fd6e6038512eabe1c491db32932ef453ba1e17be |
05-Jul-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Remove unused using statements
Test: links, compiles
Bug: 63899698
Change-Id: If61ea9dd4273616cad183ca401d84c7f0d17eff8
ava_test/hidl_test_java_native.cpp
|
| 397b5e1957cc1a1ad3365e294437324d5c088039 |
08-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Support vec<memory> in HIDL.
Test: hidl interfaces compile with vec<memory>
Test: hidl_test
Bug: 36727951
Bug: 63899698
Merged-In: I68092eda26a6ead11bd09ea6c9bb64655207cb5f
Change-Id: I68092eda26a6ead11bd09ea6c9bb64655207cb5f
idl_test_client.cpp
|
| dc3ada7f5ccba825a7a3ddc522e2b1f27a466fbc |
18-Jul-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Update VtsHidlUnitTest
* Update hidl_test_helper to support additonal option of ld_library_path
* Update AndroidTest.xml to push test files under /data instead of /systm
Bug:63704014
Test: make vts
vts-tradefed run vts -m VtsHidlUnitTests
Change-Id: I6205cf0eba3856ea6a8f927553978b2752a98e9e
ndroidTest.xml
idl_test_helper
|
| f2a0f5b1bac4441c17976ba4c1feda1ac9f226d1 |
19-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Add test for C++ impl.
Let's make sure this continues to compile.
One caveat is that this can't test -Landroidbp-impl.
Also added special hidl-gen targets for generating headers
or implementation files separately.
Fixes: 38175620
Test: hidl_test_impl
Change-Id: Ib19e35ee9579ef14107b8dd9aa4cb48543f17288
ndroid.bp
mpl_test/Android.bp
|
| ad09fb67e4c91f42e589499f39daad908e79c555 |
14-Jul-2017 |
Steven Moreland <smoreland@google.com> |
Move hidl_test to system (as a VTS test).
Test: hidl_test passes
Test: build with BOARD_VNDK_VERSION=current
Bug: 63704014
Change-Id: Ic5005856a6acc028faf11763ac879509a9a0fb69
ndroid.bp
ndroidTest.xml
|
| d300f556e9989d3ce864300c845154f1e612f1cd |
14-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Tests for configureRpcThreadpool"
|
| b8ecafa6b6df978224f11c233b6bdeaef234a5bb |
14-Jul-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix push paths in VtsHidlUnitTests"
|
| 8d9522d09c2bfaa76ec32a781a1875de7c7d2f33 |
06-Jul-2017 |
Timur Iskhakov <iskhakovt@google.com> |
Tests for configureRpcThreadpool
Test: links, compiles
Test: hidl_test
Change-Id: If000c574fa9ba38062cef718af22c9ef9ef22cf4
Merged-In: If000c574fa9ba38062cef718af22c9ef9ef22cf4
ndroid.bp
idl_test.h
idl_test_client.cpp
idl_test_servers.cpp
|
| 34eae6ce5c78e9739be99728c3ca1c557f3732ce |
12-Jul-2017 |
Timur Iskhakov <iskhakovt@google.com> |
configureRpcThreadpool java test
Test: compiles, passes
Change-Id: I0c403a3c5dcbf627f14e8979e745604b6464c272
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 0525101892bcdc280d2cd3048408de3ff27c5ff0 |
12-Jul-2017 |
Ryan Campbell <ryanjcampbell@google.com> |
Fix push paths in VtsHidlUnitTests
Update the paths starting in vendor/ to DATA/vendor/. Tests will still
not pass because of an ld library path issue, but this will no longer
result in a push error.
Test: run vts -m VtsHidlUnitTests
Bug: 63592071
Change-Id: Ib1bcfa34836a7daa92440923fa2ab7162823945f
ndroidTest.xml
|
| 3ecd289d82f68451819b0d23730754dae14ec8d0 |
28-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Add test for NoSuchElementException in java test.
Test: hidl_test_java
Bug: 63087967
Merged-In: I9ed84f3852210f3c6fc5ca312df10f34386fd30b
Change-Id: I9ed84f3852210f3c6fc5ca312df10f34386fd30b
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| f2fd9bae42d0e6af46b8ce305c5008ff19d2fa0b |
26-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Add self-castFrom tests to hidl_test.
This adds additional tests where we make sure that
Ixyz::castFrom(sp<Ixyz>) works for instances of
sp<Ixyz> that are underlayed by Ixyz and subclasses
of Ixyz.
Test: hidl_test
Bug: 62892352
Change-Id: I65bf67fb8a68a71884f6543e113fd18a7e9d5531
idl_test_client.cpp
|
| c59326ec5bcae86051571a0824e39e01fc2c4053 |
21-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Don't check hash while generating hash.
A common usecase of -Lhash is to add a new hash to
a current.txt file. However, doing so requires removing
old hashes which then must be re-added for correctness.
This is because hashes of released interfaces should never be
removed.
Here I've made it so that -Lhash will return a hash for
valid interfaces even if the hash fails. I've also added
this to hidl_hash_test.
Test: hidl_hash_test
Bug: 62858322
Change-Id: I840086ae59656c73c7109b20c142da27b47f80f8
ash_test/Android.bp
|
| 03d022573fdca1992cb15341c86af5cbd9f95899 |
19-Jun-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Explicit error message for all errors."
|
| 610002f6280bb7a873a9b722a7171ca1049d9c08 |
16-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Explicit error message for all errors.
- added for passthrough oneway queue filling up
- added for null sync callback
- added test for null sync callback since it didn't exist
Test: hidl_test
Bug: 62713128
Change-Id: Ib9d3080aaf51161344e67b806babcb6fb45419ec
idl_test_client.cpp
|
| 77cf764b885fffe91244097c2fe5d663c8c39832 |
15-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Explicitly generate _vendor library.
Explicitly create '_vendor' copies of libraries so that
vendor code can link against the extensions. When this is
used, framework code should link against vendor.awesome.foo@1.0
and code on the vendor image should link against
vendor.awesome.foo@1.0_vendor. For libraries with the below extensions,
they will be available even on the generic system image.
Because of this, they should always be referenced without the
'_vendor' name suffix.
Note: vendor_available is still outputted even though
it is for the time being a no-op
Bug: 35813011
Bug: 62616897
Test: system hidl lib definitions are unchanged
Test: vendor hidl lib definitions include vendor definition
Change-Id: I9fcc4b4b6f2093fadec7a832413ae168d1ebcfb9
endor/1.0/Android.bp
|
| e4582d0b184f56624c850794b6f8917c05e71f02 |
07-Jun-2017 |
Steven Moreland <smoreland@google.com> |
Fix hidl_test_java failure.
- set TREBLE_TESTING_OVERRIDE
- print summary of results (so result isn't lost)
- add additional prints to test so you can see why it fails
- mark modules as tests so they don't accidentally get included
in a release.
Test: hidl_test_java
Bug: 62353086
Change-Id: I0a41dd1762aab92bd492ad0a71f6a1758202061e
ava_test/Android.bp
ava_test/Android.mk
ava_test/hidl_test_java
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| fbb5a094d3f323b79c00a4025d2a2566d7fc405d |
26-May-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic 'hidl-better-tests'
* changes:
Negative tests for syntax errors.
hidl_hash_test: declare genrule srcs
Positive/negative tests for versioning.
Remove non-automatic tests from hidl-gen.
|
| 4fdfa73ae9743c4fb941c9cd63051de39276ac28 |
25-May-2017 |
Steven Moreland <smoreland@google.com> |
Negative tests for syntax errors.
Old tests have been broken a couple of times. These
will get checked at build time.
Test: hidl_error_test
Bug: 38242758
Change-Id: Ia32705e2eaf77693880ceac8e19023a1cc0a17a9
ndroid.bp
rror_test/Android.bp
rror_test/interface_semicolon/error/1.0/IFoo.hal
rror_test/method_ends_in_semicolon/error/1.0/IFoo.hal
rror_test/no_data_outside_interface/error/1.0/IFoo.hal
rror_test/no_interface_in_types/error/1.0/IFoo.hal
rror_test/no_interface_in_types/error/1.0/types.hal
rror_test/no_two_interfaces/error/1.0/IFoo.hal
rror_test/same_name_interface/error/1.0/IFoo.hal
rror_test/same_package_name/error/1.0/IFoo.hal
|
| 72a3234b7fb63c78576c6447bc252f911502a16f |
25-May-2017 |
Steven Moreland <smoreland@google.com> |
hidl_hash_test: declare genrule srcs
Test: hidl_hash_test
Change-Id: Ib1d28caf1499a03b8d3fe629c2edeb733592610f
ash_test/Android.bp
|
| d520c9d4f00dca15fa217e87d1e44b55b74e3867 |
25-May-2017 |
Steven Moreland <smoreland@google.com> |
Positive/negative tests for versioning.
Bug: 38242758
Test: hidl_version_test
Change-Id: If96e8b72c943f42e3a01fc040d43ed1fb2d1a334
ndroid.bp
ersion_test/Android.bp
ersion_test/bad1/version/1.0/IFoo.hal
ersion_test/bad1/version/2.2/IBar.hal
ersion_test/bad1/version/2.3/IBar.hal
ersion_test/bad2/version/2.2/IBar.hal
ersion_test/bad2/version/2.3/IBar.hal
ersion_test/bad2/version/2.4/IBar.hal
ersion_test/bad3/version/2.2/IBar.hal
ersion_test/bad3/version/2.2/IFoo.hal
ersion_test/bad3/version/2.3/IBar.hal
ersion_test/bad3/version/2.4/IBar.hal
ersion_test/bad3/version/2.4/IFoo.hal
ersion_test/bad3/version/2.5/IBar.hal
ersion_test/bad3/version/2.5/IFoo.hal
ersion_test/bad4/version/1.0/IFoo.hal
ersion_test/bad4/version/2.2/IBar.hal
ersion_test/bad4/version/2.2/IFoo.hal
ersion_test/bad4/version/2.3/IBaz.hal
ersion_test/good/version/1.0/IFoo.hal
ersion_test/good/version/2.2/IBar.hal
ersion_test/good/version/2.2/IFoo.hal
ersion_test/good/version/2.3/IBar.hal
ersion_test/good/version/2.3/IBaz.hal
ersion_test/good/version/2.4/IBar.hal
ersion_test/good/version/2.4/IFoo.hal
ersion_test/good/version/2.5/IBar.hal
ersion_test/good/version/2.5/IFoo.hal
|
| 83e49ecf0aea62eaeb2b96b95cd5158ca11ddad7 |
25-May-2017 |
Steven Moreland <smoreland@google.com> |
Remove non-automatic tests from hidl-gen.
Test: none
Bug: 38242758
Bug: 38242758
Change-Id: I6fd40989d19c0e9e9b4aeecaae356d53386ffb08
rrors/syntax.output
rrors/syntax/1.0/IEx1.hal
rrors/syntax/1.0/IGood.hal
rrors/syntax/1.0/IGood2.hal
rrors/syntax/1.0/IIdentifier1.hal
rrors/syntax/1.0/README
rrors/versioning/2.0/IFoo.hal
rrors/versioning/2.2/IFoo.hal
rrors/versioning/3.2/IFoo.hal
rrors/versioning/3.3/IBar.hal
rrors/versioning/3.3/IFoo.hal
est.sh
|
| f9cf33b5d923dcd979f394a83bdfc59a54eea334 |
18-May-2017 |
Steven Moreland <smoreland@google.com> |
Require manifest specification for HIDL devices.
Changed vintfEmpty -> vintfLegacy which is based on
__ANDROID_TREBLE__.
This change does two things:
- make sure that race conditions between framework startup and
service startup don't occur unless "tryGetService" is explicitly
used. Since this change makes sure hals are actually in the manifest,
getService will always wait when appropriate.
- prevent framework backwards compatibility breaking for certain
development models (described below).
On Treble devices, allow this safety check to be turned off
during tests (like hidl_test) which register and unregister
services dynamically for testing purposes. On legacy/pre-Treble
devices, behavior is unchanged.
This means that you must set TREBLE_TESTING_OVERRIDE when running a test such
as hidl_test. Ideally these binaries set this value themselves. This allows
test modules to dynamically add and unset services even though they are not
declared in the device manifest. This prevents a problem where framework
changes are accidentally made in a way that is not backwards compatible. For
instance, consider the following situation for two devices developed in the
same tree:
A: serves @1.1::IFoo, declares @1.0::IFoo (incorrect)
B: serves @1.0::IFoo, declares @1.0::IFoo (correct configuration)
If development is done on device A, then framework code like: "V1_1::IFoo::
getService()->doV1_0Api()" will work. However, this will unintentionally break
the feature for devices like device B for which "V1_1::IFoo::getService()
will return nullptr. In order to prevent problems like this, we only allow
fetching an interface if it is declared in a VINTF manifest.
Test: multiple internal devices boot/work, lshal
Test: hidl_test, hidl_test_java
Test: added additional logging to hwservicemanager and made sure no
getService requests anywhere in the system were failing from this.
Bug: 38415912
Change-Id: Ib1d4f37c764470a96fbdfbcf991c8ca244746ea0
idl_test_client.cpp
idl_test_servers.cpp
ava_test/hidl_test_java_native.cpp
|
| 6a3cdff31410d7809ac3ff9af3f687af4ded345e |
19-May-2017 |
Steven Moreland <smoreland@google.com> |
Add 'hidl-module-defaults'
This default library is added to all hals and allows us to
specify special behavior without having to update all hals
everywhere.
Test: boot internal marlin
Bug: 38415912
Change-Id: Iaf71375fabe642d9f28d2790c8a9e7934ca419dc
endor/1.0/Android.bp
|
| 64f1b82a17efc0ddf8a768ca8b2f9fa237e64286 |
12-May-2017 |
Steven Moreland <smoreland@google.com> |
remove unnecessary dependency from hash_test
Test: test passes
Change-Id: I111f0a50e3f238b2566a38ba02090f69f149cbeb
ash_test/Android.bp
|
| 61d3f4b981565ce120b2fdcd36abce2bd4988866 |
29-Apr-2017 |
Steven Moreland <smoreland@google.com> |
Explicitly declare android-base dependency in Bs*
Test: links
Bug: 37791060
Change-Id: I98f6d07c890ea03e9bacb619e859038613aa15ea
endor/1.0/Android.bp
|
| 0e875df3c2b6a2edf96dff550a376c897ca405c0 |
25-Apr-2017 |
Steven Moreland <smoreland@google.com> |
Vendor extension hals on the vendor image.
This is required to make sure
Test: running hidl-gen on vendor image, image built to vendor
Test: update-makefiles is no-op on:
hardware/interfaces
frameworks/hardware/interfaces
system/hardware/interfaces
system/libhidl/transport
Bug: 35813011
(cherry picked from commit ec642438e32309b694a4b3303f8c5a6a524dda20)
Merged-In: Ie3f4aef0f9088c09c108030084e71d5921599c4d
Change-Id: Ie3f4aef0f9088c09c108030084e71d5921599c4d
endor/1.0/Android.bp
|
| f5cf3ce42dd7c76c5deb7d99d0f1b56221f30778 |
22-Apr-2017 |
Steven Moreland <smoreland@google.com> |
hidl-gen: tests for hashing.
Note b/37575883.
Test: hidl_hash_test (also made sure it could fail)
Change-Id: Ifb3dd0aa93237c20895c37bbf155a697284c70de
Fixes: 37513440
ndroid.bp
ash_test/Android.bp
ash_test/bad/current.txt
ash_test/bad/hash/1.0/IHash.hal
ash_test/good/current.txt
ash_test/good/hash/1.0/IHash.hal
|
| e0586b0e450deaf81cc732896bced28c63de714e |
20-Apr-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "hidl_test: 10.986s -> 3.081s"
|
| 7b680ebdc6b150e808dc85ff5fae6344b20a206b |
20-Apr-2017 |
Steven Moreland <smoreland@google.com> |
hidl_test: 10.986s -> 3.081s
- death listeners take ~30ms. Waiting 100ms now instead of 1s.
- don't always sleep 1s waiting for servers to startup. Just add calls
to wait for them to all startup.
Test: `time hidl_test`
Bugs: me
Change-Id: I6c815344e51b8947fbc8ef4a42361d5f9517c1b5
idl_test.h
idl_test_client.cpp
idl_test_helper
idl_test_servers.cpp
|
| 6f7514ba6c90006958a8bac79e322431ecce3c54 |
13-Apr-2017 |
Steven Moreland <smoreland@google.com> |
Mark hidl_test as vendor.
hidl_test requires a lot of 'vendor' modules (-impls) and must be marked
as vendor in order to work on BOARD_VNDK_VERSION := current devices.
Test: hidl_test
Bug: 33241851
(cherry picked from commit b92ff1c6d8a926004f52b272f021268c3a0ff97f)
Merged-In: I745a10af598a20f121ee0f3f1bee7418fa511ace
Change-Id: I745a10af598a20f121ee0f3f1bee7418fa511ace
ndroid.bp
|
| 23d7477765a2b282f7185c90afcd1d5ea551e5f1 |
14-Apr-2017 |
Yuexi Ma <yuexima@google.com> |
fix hidl unit test file path and permission issue.
Fix: 37322148
Test: vts-tradefed
Change-Id: Ib233655494334dfd944e6db6552701832ffeb002
ndroidTest.xml
idl_test
|
| aa573273fba7030405fb3b12d10bd5de85b90aa3 |
13-Apr-2017 |
Andreas Huber <andih@google.com> |
Changed the naming scheme for Treble java libraries (both static and dynamic)
android.hardware.foo@1.0-java => android.hardware.foo-V1.0-java
Bug: 37207894
Test: make
Change-Id: I0f62c49a098400a9798b31d0912f0155bbe75c08
Merged-In: Ia87b63d5c90772e4620aa9fe3ce431b297be117b
ava_test/Android.mk
ava_test/hidl_test_java
endor/1.0/Android.mk
|
| f1a00acefec7045cac2fb55c2e70efcd69131d11 |
07-Apr-2017 |
Steven Moreland <smoreland@google.com> |
Added tests for passthrough service lookup.
Test: hidl_test
Bug: 37107636
Change-Id: I0ef3510d5f272b8a852da1c555a62eaabb801bfb
(cherry picked from commit 731163da922c499d48e867e94c888eddef1af6bf)
idl_test_client.cpp
|
| b9b331ee196ad7afcf9cd2fe5ab87be671f48ce2 |
06-Apr-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Push hidl_test HAL libraries to system/lib and system/lib64"
|
| 9bd61bd327aac4cd38d1211304483ec0d48b9ffd |
06-Apr-2017 |
Andreas Huber <andih@google.com> |
Merge "Fix IHwBinder.asInterface(...) to respect the interfaceChain and return null"
|
| a2abe98a661c20b028bf20c3d8d1cd13d080b66a |
04-Apr-2017 |
Andreas Huber <andih@google.com> |
Fix IHwBinder.asInterface(...) to respect the interfaceChain and return null
given a binder that's unrelated. Also adds public method
castFrom(IHwInterface) that calls through to asInterface(IHwBinder)
without exposing that implementation detail.
Bug: 36749201
Test: hidl_test_java
Change-Id: Ic4bc1f22f05e692dd8e9456382492056b96a809b
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 1cefc3920f649e06365cdaf8ced3759bdb1c508f |
04-Apr-2017 |
Steven Moreland <smoreland@google.com> |
Test: Update for variably sized token.
Bug: 33842662
Test: hidl_test
Change-Id: I49db62f87a2c13d2291858cd8508799a3ce5c91f
(cherry picked from commit cf47e4d11ab8750c848a24f40963783552411ddf)
idl_test_client.cpp
|
| a0e45298d417c03a73a27cbca34b83cb0b0c3552 |
05-Apr-2017 |
Hridya Valsaraju <hridya@google.com> |
Push hidl_test HAL libraries to system/lib and system/lib64
They were incorrectly being pushed to data/lib and data/lib64
earlier.
Bug: 35710876
Test: run VtsHidlUnitTests
Change-Id: If6824ac9350bb2d3f6ab8e1bf6296b6797bd0e3f
ndroidTest.xml
|
| 623ad8e7a28e41c510411dee692ca0c3ae8c97bc |
04-Apr-2017 |
Dan Willemsen <dwillemsen@google.com> |
Use log/log.h instead of android/log.h for ALOG* macros
ALOG* is only defined when including log/log.h, not android/log.h
Test: mmma -j system/tools/hidl
Change-Id: I76550ec7537bbbf20c7028cafa8b724b8ad30187
ooCallback.cpp
|
| 30b5d1ffe964d0b82008bfc8f4b8e61ab3bac86f |
03-Apr-2017 |
Yifan Hong <elsk@google.com> |
Add getHashChain() to IBase.
Each interface has a getHashChain() method that
returns the hash of all .hal files from the rtti
of the interface up to IBase.
Test: hidl_test
Bug: 36602587
Change-Id: I13f2e54bd45593c70064dff05fa934d5cd01996a
ndroid.bp
idl_test_client.cpp
idl_test_servers.cpp
|
| d7899cdce34bbe2ca9f2cfbe9430e498f2550fad |
31-Mar-2017 |
Hridya Valsaraju <hridya@google.com> |
Disable FooCallmeTest and WrapTest cases in hidl_test by default
An command line option has been added to enable them as well.
Bug: 35710876
Test: hidl_test
Change-Id: I395be9a8e3c868da1646722d7ff0ca2fb006df58
idl_test_client.cpp
|
| 673ad8758a8fafa9526e9eba951756d37803ec75 |
30-Mar-2017 |
Hridya Valsaraju <hridya@google.com> |
make VTS run hidl_test
Bug: 35710876
Test: run vts -m VtsHidlUnitTests
Change-Id: I0ce65099e2e9ded11b0f4d23bc6124eaa71fe3d5
ndroid.mk
ndroidTest.xml
|
| 038903b7f99f27179d1e9411f9181b467ccd7ac1 |
30-Mar-2017 |
Steven Moreland <smoreland@google.com> |
Add tryGetService.
Required for cases where a hal is being used but we want to fail in
cases where the remote HAL is down. This is also required when someone
wants to get a handle on a hidl service from the only hwbinder thread in
a pool.
Test: hidl_test
Bug: 36780273
Change-Id: I6544f534ef2dff0773c645cc9953de6d1340d01d
idl_test_client.cpp
|
| cbab402a94995b46560c1226cdc865108d89a3ad |
29-Mar-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic 'hidl_gen_fix'
* changes:
Fix .equals in Java does not handle edge cases.
RemoteException should not be a HIDL keyword.
|
| 45b331b651c3290391518ec1820c65e0dff8515d |
27-Mar-2017 |
Yifan Hong <elsk@google.com> |
Fix .equals in Java does not handle edge cases.
The old .equals code uses Objects.deepEquals,
which does not handle for, e.g. ArrayList<byte[]>.
Create our own recursive HidlSupport.deepEquals
and deepHashCode method to handle these edge cases.
Bug: 36454147
Test: hidl_test_java
Change-Id: I58d86c163e81899da1db266b2589b8652df1e2c1
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 11bd9dd3447fc2a98171866fb84b57bc3d47eb15 |
23-Mar-2017 |
Hridya Valsaraju <hridya@google.com> |
Create hidl_test_helper so that FMQ unit tests can use it as well
Bug: 35214642
Test: hidl_test and fmq_test passes
Change-Id: Ie1983774c9f4a47aa5c113eee62022e33860a2f9
ndroid.mk
idl_test
idl_test_helper
|
| 93a2a0847ff19cb5dc6fc7dd288fbf6957274f9b |
10-Mar-2017 |
Hridya Valsaraju <hridya@google.com> |
Cleanup to_string() usage and use toString() instead
Bug: 35703683
Test: hidl_test_java passes
Change-Id: Ib62828e5fb1c18825974f247c64c1e4dc686b23a
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 2e4d670350c58bb62ef1d86bcba271faf2e12633 |
16-Mar-2017 |
Yifan Hong <elsk@google.com> |
Add test for non-GC'd parcel.
Test: hidl_test_java
Bug: 36088202
Change-Id: I9ac3771b9571ea63a00005d178ca0d82ef60263d
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| fc0601696cc798c6528f74e11f1b2ea11102d0f3 |
11-Mar-2017 |
Steven Moreland <smoreland@google.com> |
Update for read+readRange of memory.
Bug: 35328065
Test: hidl_test
Change-Id: I1df9f7e61135227f8f1c47c0f4c72ce6531990fa
idl_test_client.cpp
|
| 72d948857e957bc7da6be58615beb3e128c7b209 |
14-Mar-2017 |
Steven Moreland <smoreland@google.com> |
Update passthrough test TODOs.
Bug ids are out of sync.
Bug: 32747392
Test: pass
Change-Id: Ie655aa79bea447e1c85bab181601d5dde24d8e00
idl_test_client.cpp
|
| 3ac9f70571828f62152e1c0c84b2d655fd2e83bd |
08-Mar-2017 |
Steven Moreland <smoreland@google.com> |
Remove unused IPCThreadState include from tests.
Test: pass
Change-Id: I11ec3e50e8811e6fc89f32341613e3d6963a3b58
idl_test_client.cpp
|
| 21aa43c93083bab6f1ab5511418fc65bf49de068 |
08-Mar-2017 |
Steven Moreland <smoreland@google.com> |
Better service notification test.
Explicitly test/show that service notifications registered for class A
are sent out when a subclass of A is registered.
Bug: 35988664
Test: hidl_test
Change-Id: Id93ad5c56bf44f537718d26f21ac396116032a98
idl_test_client.cpp
|
| 2ef6b39949680840b2acffa2a46edc0fe8f01d0f |
03-Mar-2017 |
Yifan Hong <elsk@google.com> |
Update system/tools/hidl/test/test.sh for minor version uprev.
Bug: 28825532
Test: system/tools/hidl/test/test.sh -a
Test: system/tools/hidl/test/test.sh -u
Test: system/tools/hidl/test/test.sh
Change-Id: Ie2f23123dee5a5a82e42d70ac0c9506884a36cff
rrors.output
rrors/syntax.output
rrors/versioning/2.0/IFoo.hal
rrors/versioning/2.2/IFoo.hal
rrors/versioning/3.2/IFoo.hal
rrors/versioning/3.3/IBar.hal
rrors/versioning/3.3/IFoo.hal
est.sh
|
| 21171d2de5e3e6b595a6bf1af52819e7c6816daf |
03-Mar-2017 |
Yifan Hong <elsk@google.com> |
Update errors.output.
Test: ./system/tools/hidl/test/test.sh
Bug: 33277700
Change-Id: I644468dcc5cd9dc1e13ed0a907c060bd83b31165
rrors.output
|
| c20e35b887a49a98dfe7b9e52f7aa40002d1a1b5 |
02-Mar-2017 |
Martijn Coenen <maco@google.com> |
Fix ServiceList test.
Test: hidl_test passes again
Change-Id: I0c2587fe01b7039dba513a31348b6cf5216146f9
idl_test_client.cpp
|
| 8b98d8ddadec92c1465d60c25e003dd604517ea7 |
25-Feb-2017 |
Yifan Hong <elsk@google.com> |
IServiceManager/ITokenManager uses "default" service name
The getService() and registerAsService() methods of interface objects
now have default parameters of "default" for the service name. HALs
will not have to use any service name unless they want to register
more than one service.
Test: boots
Test: hidl_test
Bug: 33844934
Change-Id: Ib3227220ed93b2fa574f485284db326529e1f186
idl_test_client.cpp
|
| 8951e55e7af8d9666bd55cff1c6a672b3a94b510 |
27-Feb-2017 |
Hridya Valsaraju <hridya@google.com> |
Test hidl_memory move constructor/assignment operator
Bug: 34093434
Test: hidl_test passes
Change-Id: Ibb25f1fd4fe43fcf48f05df947523baedd783506
idl_test_client.cpp
|
| 179379a911b386023250998f6d257482febb2d86 |
10-Feb-2017 |
Hridya Valsaraju <hridya@google.com> |
Separate servers and client for hidl_test
Make hidl_test configurable to run with 32/64-bit client
and servers. The default setting runs all four combinations.
Test: make hidl_test, adb root && adb remount && adb sync,
adb shell data/nativetest64/hidl_test
Bug: 34986428
Change-Id: Ie06b1d0f33544927fa53583f6fa0e92ed1ef7451
ndroid.bp
ndroid.mk
leanSpec.mk
idl_test
idl_test_client.cpp
idl_test_servers.cpp
ain.cpp
|
| ae588f643275b443df572f28c6ae7c7ffe2d559b |
22-Feb-2017 |
Hridya Valsaraju <hridya@google.com> |
Move IBaz default implementation to hardware/interfaces
Test: adb shell hidl_test_java passes
Change-Id: Ibebdf8b3f14fee4964f5a10728f61d375820465a
ava_test/Android.bp
ava_test/hidl_test_java_native.cpp
|
| 0c87d633d3dd1490465dafad81f4eb6b10f7b54a |
24-Nov-2016 |
Logan Chien <loganchien@google.com> |
Remove unused Android.mk for hidl_gen_test.
This commit removes Android.mk which was originally for hidl_gen_test.
The hidl_gen_test and the related test_output.py were removed in
f31fdd3b. Now, hidl_gen_test won't compile even if ENABLE_HIDL_GEN_TEST
is set to true.
Test: mma -j64
Merged-In: I76659d1380624fd473a32db89e62c5415ca9083e
Change-Id: I76659d1380624fd473a32db89e62c5415ca9083e
ndroid.mk
|
| e45b5303e072043679483a70606f6c00dde17382 |
22-Feb-2017 |
Yifan Hong <elsk@google.com> |
Add toString to Java code
* toString is mainly for debugging purposes.
* For structs and interfaces, add Object.toString.
* For enums, add MyEnum.toString(int) and MyEnum.dumpBitfield(int).
Use them as follows:
* For enums, use the static method E.toString(int).
* For bitfields, use the static method E.dumpBitfield(int).
* For all arrays, use java.utils.Arrays.deepToString(o)
* For everything else, use one of the following:
* o.toString(), if o is not null
* Object.toString(o)
* String.valueOf(o)
* Note that for array / vec of enums / bitfields, the raw integer
value is dumped.
Bug: 33459772
Test: hidl_test_java
Change-Id: Ifb1ed519770b907e0a4e345b2c3109dc322a23b2
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 870d1a7ccd70bd710128993de401278614d1975e |
22-Feb-2017 |
Yifan Hong <elsk@google.com> |
Fix C++ toString for enums doesn't handle for non-bitfields.
Given an enum MyEnum : int32_t { A=1, B=2, C=3 },
toString(MyEnum e):
toString(A) => "A"
toString((MyEnum)3) => "C"
// For invalid values
toString((MyEnum)4) => "0x4"
for bitfields,
toString(int32_t e):
toString(3) => "A | B | C"
// For invalid values, remaining bits are also printed
toString(5) => "A | 0x4"
Test: hidl_test
Bug: 33459772
Change-Id: I5a9ccffca80cc73c127702fc9e677f77446a5366
ain.cpp
|
| fbcdc80e6451c24545cf0a5841d0c40cca58f42c |
23-Feb-2017 |
Yifan Hong <elsk@google.com> |
Use interfaceDescriptor instead of interfaceChain
wherever suits. Sementically, interfaceDescriptor
returns the first element of interfaceChain; use
interfaceDescriptor when the rest of the elements
aren't used.
Bug: 34136228
Test: hidl_test
Change-Id: Ie3332c168a792ab018d9b762f9bf6c15ddcbf998
ain.cpp
|
| a3e8b239a43da534038ae4f4c350d3940713ee4c |
22-Feb-2017 |
Yifan Hong <elsk@google.com> |
Fix Return<T> should not be copied.
Test: pass
Change-Id: I6f40eff96f3b54f53d7eeb037e3744447270e00b
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| ffc4e519aa059811a1e5839c0883a876d3d05aaa |
17-Feb-2017 |
Steven Moreland <smoreland@google.com> |
hidl_test: use separate allocator
Test: hidl_test, device boots with allocator
Bug: 35327976
Merged-In: Id1f2eabd9710210e241c11b2c5d90d0c355a3793
Change-Id: Id1f2eabd9710210e241c11b2c5d90d0c355a3793
ndroid.bp
ain.cpp
|
| 7d3ac806f3f8e9ac3fe5ecc3d17054e8b4e641df |
15-Feb-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Revert "Revert "Move HidlInstrumentor to HidlInternal and namespace details.""
This reverts commit 7f12d6b336f105d4aa8ab3302f43c4d4efa40037.
Change-Id: Icbae613d1ec1a54492f6ac25815be6495264e9ee
ain.cpp
|
| 7f12d6b336f105d4aa8ab3302f43c4d4efa40037 |
15-Feb-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Revert "Move HidlInstrumentor to HidlInternal and namespace details."
This reverts commit 5f1d0313dd627baeae11e6541b0473bf6252266c.
Change-Id: Ib340c08e52b6bc340489d1f90d9fbdace75047a2
ain.cpp
|
| 5f1d0313dd627baeae11e6541b0473bf6252266c |
14-Feb-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Move HidlInstrumentor to HidlInternal and namespace details.
Bug:34644498
Test: run vts -m VibratorHidlTargetProfilingTest
Change-Id: Ic84d1e5bac6cb07b6a4f74e76faaf29bd57afe69
ain.cpp
|
| 424a948265821d514be42e9e2dd9a53cf515d0bb |
14-Feb-2017 |
Steven Moreland <smoreland@google.com> |
IBase: add ping method
Add ping method to IBase. We deleted this from libhwbinder types, so
moving it to hidl types.
Bug: 34501346
Test: hidl_test
Change-Id: I1e13a48f1ca3668450def6434650763483baf34e
ain.cpp
|
| 177a8b1db6432472cdda1311219dd83c2a78d3b5 |
09-Feb-2017 |
Andreas Huber <andih@google.com> |
Updated Java intermediates directory.
Java intermediates are now created in the common rather than the device-
specific generated sources subdirectory.
Bug: 33842609
Test: make
Change-Id: I37dba3a3cb9ba4e65f64150fa74e0d473c5e71e6
endor/1.0/Android.bp
endor/1.0/Android.mk
|
| 264ba8a8ed69accbbafed1287f3b4f3f63ce1911 |
07-Feb-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Verify that strings are properly marshaled from Java to native"
|
| 3c249490571f7d5cd09ed9d2de9c5778df6ef2f8 |
06-Feb-2017 |
Andreas Huber <andih@google.com> |
Verify that strings are properly marshaled from Java to native
in particular verify that strings retain their original size.
Bug: 35038064
Test: hidl_test_java
Change-Id: Id0308ad14fcb0f419ffbc74873efe3a24cfd6662
ava_test/hidl_test_java_native.cpp
|
| 6081863ce80ff231323a54e97072a17227a56820 |
04-Feb-2017 |
Steven Moreland <smoreland@google.com> |
Add -Werror.
Also, fixed several warnings (including one bug in generateCppImpl.cpp).
Test: compiles
Bug: 32559427
Change-Id: I78375b9052c0ad447ebac638258f25bbf562712f
ndroid.bp
ain.cpp
|
| 6e3f9f47eaf9664c605f703e6f48df2648756332 |
30-Jan-2017 |
Steven Moreland <smoreland@google.com> |
hidl_test: make sure string size is correct
Test: hidl_test still passes
Change-Id: I67b792f3d75f8aec5875a2dc8bc371dd7c6959b1
ain.cpp
|
| d10feeaaad681c9f70330838f5d19e54ecfc98b7 |
24-Jan-2017 |
Zhuoyao Zhang <zhuoyao@google.com> |
Update hidl-gen for HidlInstrumentor.
* Generate the code to pass the package and interface name to
HidlInstrumentor
Test: make hidl-gen
mma for hardware/interface
run vts -m VibratorHidlTest
Change-Id: Ic370e41b62dc1b81ddb7794682de53e0515f9193
ain.cpp
|
| 677576012329e4f10a506b1f9623a2345dfbe93a |
17-Jan-2017 |
Yifan Hong <elsk@google.com> |
Merge "Type resolution favor local types."
|
| 87ff823efa7bda4304391987235cc48e6b351b58 |
09-Jan-2017 |
Yifan Hong <elsk@google.com> |
Type resolution favor local types.
Test: mma (see main.cpp; it is statically tested to ensure
types are resolved correctly)
Bug: 31971432
Change-Id: Iaf5c0ac30941269200f4c0b8312437eed61606b4
ain.cpp
|
| 5715fedbd05231bf845e3a120c83b931b7f53aa3 |
16-Jan-2017 |
Steven Moreland <smoreland@google.com> |
hidl-gen: correct package paths in makefiles.
Previously, only IBase and the current package was added to package
paths whereas we want all paths which are required/inherited from.
Also uploadated tests.vendor to inherit something and be java
compatible.
Also fixed package root paths so that they are compared against the
beginning of a package instead of being anywhere in the package.
Bug: 34316257
Test: tests/vendor/1.0 update-base/makefiles, mma, hardware/interfaces
update-makefiles and mma
Change-Id: I9ef32b3867646690363fc9f85fe691ae74796d45
endor/1.0/Android.bp
endor/1.0/Android.mk
endor/1.0/IVendor.hal
endor/1.0/update-base.sh
endor/1.0/update-makefile.sh
endor/Android.mk
|
| 0a344a3e7cb56609ccf510a0e968ba2a5c1a20bc |
16-Jan-2017 |
Martijn Coenen <maco@google.com> |
Merge changes from topic 'java-empty-hidl-vec'
* changes:
Testing struct that has vector of other struct
Add 'nullable' parameter to Java readEmbeddedBuffer calls.
|
| b6b3fcde5ff47b5c1f51bd9b80e190c8dd6782fa |
13-Jan-2017 |
Pavel Maltsev <pavelm@google.com> |
Testing struct that has vector of other struct
Test: hidl_java_test
Bug: b/34255213
Change-Id: I5de1225c25b91af16324019ffb871bfcf0c985c7
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| f7b596bc5f6ab432ed6b664d790b91a0b8bb917a |
13-Jan-2017 |
Martijn Coenen <maco@google.com> |
Fix memory corruption in hidl_test.
MQDescriptor takes ownership of the native_handle_t
that is passed into it; giving that same native_handle
to another MQDescriptor, or freeing it manually, will
cause double-free and memory corruption.
Test: hidl_test doesn't crash
Change-Id: I8dc170aa6279e6fd9a438f0a37fe5b1389bb816e
ain.cpp
|
| afb30cb46a492318261ba49b8c9fc4af67dd5976 |
13-Jan-2017 |
Martijn Coenen <maco@google.com> |
Add test for empty hidl_vec.
Test: mma, hidl_test
Change-Id: Iebf8531206cdb49d8388cd926a6500601f2b4c46
ain.cpp
|
| c75fd47b91f2edea70bb88f5508d9a161b916045 |
11-Jan-2017 |
Yifan Hong <elsk@google.com> |
Add interfaceDescriptor() method to all interfaces
which returns interfaceChain[0].
Bug: 34136228
Test: hidl_test
Test: hidl_test_java
Change-Id: Id0d8139d51f39b1c317d966b305e883e561241b8
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
ain.cpp
|
| f5cc2f74e86504f7904a0a24e7fcc00fa19cd579 |
05-Jan-2017 |
Yifan Hong <elsk@google.com> |
Emit toString functions for all types.
* toString() is mainly for debugging purposes only.
* For HIDL internal types (hidl_string, hidl_vec, etc.)
toString() is found in ::android::hardware::details.
* For a user defined type
android.hardware.foo@1.0::IFoo.Type,
toString() is found in ::android::hardware::foo::V1_0.
* For bitfield<::anroid::hardware::foo::V1_0::T>
that gets translated to the underlying
numeric type of T, it doesn't make sense to override toString().
A templated toString() function for each user-defined HIDL enum \
is introduced into the same namespace; call it with
using namespace ::android::hardware::foo::V1_0;
toString<IFoo::MyEnumType>(value);
Test: hidl_test and look at the output of logcat
Bug: 33459772
Change-Id: I70eee018e31d700bf1376334276dbd343af5615f
ain.cpp
|
| 9d97da2203a07661288709caaec2ed5caaaff0d6 |
12-Jan-2017 |
Martijn Coenen <maco@google.com> |
Allow passing null handles in all cases.
Test: hidl_test
Bug: 30814137
Bug: 34234675
Change-Id: Ia3968632db023723d7a4cba038ef06bcf14d4154
ain.cpp
|
| fcb03f8e1dbb60c2b3b740bd3650418438b05a08 |
12-Jan-2017 |
Andreas Huber <andih@google.com> |
Merge "Test Java functionality with native client/servers in all supported ABIs"
|
| 164f83e5eb4a3b9d39041df3888e59dcf5db3b65 |
06-Jan-2017 |
Andreas Huber <andih@google.com> |
Test Java functionality with native client/servers in all supported ABIs
(32- and 64-bit for now)
Test: hidl_test_java
Change-Id: I338ad0ebd60f39e2b52055f050b44f08aa207322
ndroid.bp
ava_test/Android.bp
ava_test/Android.mk
ava_test/hidl_test_java
|
| f31f165e060c91ac97a1ff139de8a72e824a3308 |
07-Jan-2017 |
Steven Moreland <smoreland@google.com> |
hidl_test: null handle in hidl_memory
Bug: 34134129
Test: hidl-test
Change-Id: I7b69e17340eae264c9d47cca236aee09d3e3b1ea
ain.cpp
|
| 01e7cdef5d4198104a4e772cb742fb88777dcceb |
10-Jan-2017 |
Yifan Hong <elsk@google.com> |
Bp/nFoo -> Bp/nHwFoo
Bug: 33554989
Test: compiles
Test: hidl_test
Change-Id: I41399cd9e8df70df4bf9d2fb1ac13ec4858d3f6c
ain.cpp
endor/1.0/Android.bp
|
| db1b1b638865a2043c9cddd8c865751e9742b181 |
10-Jan-2017 |
Steven Moreland <smoreland@google.com> |
@export: add export_parent
1. export_parent is by default true
2. setting export_parent to false will prevent -Lexport-header from
exporting hal information
Fix: 32539339
Test: verification that output without adding the option is the same,
manual test with export_parent = true/false
Change-Id: I7a6304cdc40a495427d0369fcc054675770df608
endor/1.0/Android.bp
endor/1.0/Android.mk
endor/1.0/types.hal
endor/1.0/update-base.sh
endor/1.0/vendor-base.h
|
| 0b515ab17bcd37b19ba7ec9edddc63f6d9f2858f |
09-Jan-2017 |
Yifan Hong <elsk@google.com> |
Remove unnecessary code in HIDL test.
Bug: 33669138
Test: compiles
Change-Id: I1abcdf7b152a05acb1b88596b0d41eda4ad714b8
ain.cpp
|
| 64af5e8e25039f5d29af1d38d5c4cfa0e1d93996 |
04-Jan-2017 |
Steven Moreland <smoreland@google.com> |
java: get/register checked exceptions.
Bug: 33673120
Test: hidl_test_java
Change-Id: Ie3565279566d80347e2c89830621f5a6a4864e5f
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| fcf94e48639538117e52f81ba7018c2af7b4cc81 |
08-Dec-2016 |
Yifan Hong <elsk@google.com> |
Gives better error message on callback tests.
FooCallback::reportResults will wait for longer time, then
assert if the time left is greater than a threshold. This
gives more information than simply stating "time is up" (by
returning -1.)
Test: hidl_test
Change-Id: I3b72731ee8c772ecb8863f3dfac0faf62182e321
ooCallback.cpp
ain.cpp
|
| 11992a60ff12aa66d13da9731f9578e1165422a0 |
10-Nov-2016 |
Yifan Hong <elsk@google.com> |
move foocallback to hidl_test
Bug: 32778024
Bug: 33820761
Test: hidl_test
Change-Id: I5127a7f34f0fb36fae0fe153ebe700bfeca5d846
ndroid.bp
ooCallback.cpp
ooCallback.h
ain.cpp
|
| b40ef02ca14e507a5a5b878c90eaac52ffe64fd9 |
02-Jan-2017 |
Martijn Coenen <maco@google.com> |
Callback elision for HIDL interfaces.
HIDL interface types are now elided.
Bug: 31380743
Test: mma, hidl_test, hidl_test_java
Change-Id: Ie2bcd67b9f1d8dbd4331a3c29ae1294ba75d0be7
ain.cpp
|
| ec10227f088da9997291ee10029a4577ffd240fc |
21-Dec-2016 |
Yifan Hong <elsk@google.com> |
Emit equals and hashcode for auto-generated Java code.
Test: hidl_test_java
Bug: 32834072
Change-Id: I660852b7d76f79d41b93c9986e4c1188c02ba4a6
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 6906410fd15c48613c901785eec7c0404e94b053 |
04-Jan-2017 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Java: Add checked exceptions to methods."
|
| dad1b30bd4d2aa9641156909acce334b25d3e615 |
21-Dec-2016 |
Steven Moreland <smoreland@google.com> |
Java: Add checked exceptions to methods.
Bug: 33673120
Test: hidl_test_java
Merged-In: Ie46f2895cda02c192b1353c6b9317461061f0d26
Change-Id: Ie46f2895cda02c192b1353c6b9317461061f0d26
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| cd00b9b86431d82ca9b9b60cb97d200af277f220 |
29-Dec-2016 |
Steven Moreland <smoreland@google.com> |
Return<*> getStatus().isOk() -> isOk()
Bug: 31348667
Test: compiles
Change-Id: I3d5dc2ec32f50bc4283924f69ef662bcbd10b92b
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| 1f365395dc3cf47fa31f92204f6a09c6bce2bc9b |
03-Jan-2017 |
Hridya Valsaraju <hridya@google.com> |
Merge "Add FmqType to templatize MQDescriptor"
|
| a32bde87bb8341954d297ff889a62fb6abc6d3a1 |
27-Dec-2016 |
Hridya Valsaraju <hridya@google.com> |
Add FmqType to templatize MQDescriptor
Also removed PredefinedType and renamed MQDescriptorSync to
fmq_sync and MQDescriptorUnsync to fmq_unsync.
Test: FMQ unit tests.
Bug: 33948735
Merged-In: I7f2802b76000a1e1a55d02381cca202530a0cd18
Change-Id: I7f2802b76000a1e1a55d02381cca202530a0cd18
endor/1.0/IVendor.hal
|
| fa55d6ec8496deb5e40865f13f8bf6a971ca622d |
20-Dec-2016 |
Martijn Coenen <maco@google.com> |
Automatically start threadpool.
Generated code should automatically start a threadpool
in one of the following scenarios:
- You pass an asynchronous interface to a remote process
(so you can receive calls on it).
- You register a death recipient on an interface
(so the kernel can notify you on this thread when it dies).
Also, updated tests to use the new configure/joinThreadPool() API.
Bug: 31748996
Test: mma, hidl_test, hidl_test_java
Change-Id: I2262580282c9b25c2e004dfd928b7444f4be78d9
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| 8d12b50cfb20eaef50e0998abe8d458866896003 |
27-Dec-2016 |
Martijn Coenen <maco@google.com> |
Support (un)linkToDeath in Java.
Test: mma, hidl_test_java
Bug: 31632518
Change-Id: Ic324deb5b0b17340328a08e736e07b47a7bea59b
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| c6752dcea3b5dae1e99960b5beb6af394280b393 |
20-Dec-2016 |
Yifan Hong <elsk@google.com> |
Add operator== and != for some structs.
operator== and != are defined for type T where T is
one of the following:
* scalars
* strings
* enums
* U[] where U is in this list
* vec<U> where U is in this list
* structs that contain only types in this list.
If a struct contains a handle, an MQDescriptor,
or a union, for example, then == and != is not generated.
Bug: 32834072
Test: hidl_test
Change-Id: Icac73a14f81274ee6592e2b8fadf26d629151d9f
ain.cpp
|
| 2a75353a696ebc6c4f934e066d32787b3c91c0ce |
15-Dec-2016 |
Steven Moreland <smoreland@google.com> |
hidl_test: Update for binderized IAllocator.
Bug: 32185232
Test: hidl_test
Change-Id: I629b20e78364ce9e26c892cb436db862c6371fef
ndroid.bp
ain.cpp
|
| aa2b83adce439a67ef150e1677a39e5e4154ca34 |
22-Dec-2016 |
Steven Moreland <smoreland@google.com> |
Add tests for android.hidl.token@1.0.
Test: hidl_test
Bug: 33058559
Change-Id: I6ab11620ca56334e79695661618d6225a800977e
ndroid.bp
ain.cpp
|
| 115d428f9c38c078d679a98942c47ce4a17bd599 |
19-Dec-2016 |
Martijn Coenen <maco@google.com> |
Add (un)linkToDeath to generated interfaces.
(un)linkToDeath is now a part of IBase, and can be
called on any generated HIDL interface. The implementation
is a no-op, except in proxy objects, which are by
definition a different process than the interface they point
to.
Since clients are not aware of the transport implementation, we must
wrap the transport-independent callback in a transport-specific
callback object. In case of binder, that object is a
hidl_binder_death_recipient.
The binder proxy object contains a list of registered death
recipients, as well as a mutex to protect access to the list.
The list is required to allow us to map back transport-independent
callbacks to transport-dependent callbacks in unlinkToDeath().
Bug: 31632518
Test: mma, hidl_test
Change-Id: I5083a8789dd706a886a8a09f8c733031a351a36a
ain.cpp
|
| 20e4f9cc51d20add2e1d57231cbb3a9e52b02268 |
20-Dec-2016 |
Yifan Hong <elsk@google.com> |
Add test for Java GC for binder objects.
Test: hidl_test_java
Bug: 33042939
Change-Id: I4c7c5ffe18629798ae7625132b568fcb49551f3b
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 1902088798bc9a68c082111c658ffd3d9e63a910 |
20-Dec-2016 |
Steven Moreland <smoreland@google.com> |
Merge "Use interface for method callbacks."
|
| bcd790346fdc609d214791fd82c6ab1af388e2f2 |
17-Dec-2016 |
Steven Moreland <smoreland@google.com> |
Use interface for method callbacks.
This way people can implement callback methods in java with lambdas.
When you want to call a java method with multiple arguments, right
now you have to say:
foo.method(arg1, arg2, new MethodCallback() {
public void onValues(type1 res1, type2 res2) {
...;
}
}
By changing MethodCallback from an anonymous class to an interface, we
can call the function with a Java 8 lambda instead.
foo.method(arg1, arg2, (res1, res2) -> res1 == OK ? setCallback(res2) : die(res1)));
Test: compiles
Bug: 33744648
Change-Id: I9b3aad019560714e2e6ffff0081c5d590f447cbc
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 5749b2c5842f7ca031349eba3841a243e07d54eb |
28-Nov-2016 |
Yifan Hong <elsk@google.com> |
Test for echoed interfaces.
Test: hidl_test
Bug: 33173166
Change-Id: I11af5eb23fd258af5c99b422841fa06e10d41c0d
ain.cpp
|
| e19fecedecc43976bbf2f73ab4102cee89c6b116 |
17-Dec-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Add test for constructor of HidlInstrumentor."
|
| 3491fc17561f7274f27f363451f25a988ec5ddee |
17-Dec-2016 |
Yifan Hong <elsk@google.com> |
Merge "Add a script to prevent regression in hidl-gen error recovery."
|
| def2cfbad63338aefdaa71dc7a9ae3a0f75c0dd9 |
15-Dec-2016 |
Yifan Hong <elsk@google.com> |
Add test for constructor of HidlInstrumentor.
Bug: 33645959
Test: adb shell ./data/nativetest64/hidl_test/hidl_test64
--gtest_filter="HidlTest.WrapTest" -p
Change-Id: I27d686432a0a0ee1e0f6b1447d1b1b126ec1e57d
ain.cpp
|
| db20ea0d9acae63335ef771deec2797bc4dd8d53 |
15-Dec-2016 |
Martijn Coenen <maco@google.com> |
Merge changes from topic 'removehw'
* changes:
BnFoo class cleanup.
Stop generating IHw classes.
|
| 9d3e403a17e35784d44ddbb0d7b3a23a25d177f5 |
15-Dec-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix hidl_test once again."
|
| a26c1b38846711c888dca20586526f531024b0e2 |
14-Dec-2016 |
Martijn Coenen <maco@google.com> |
Fix hidl_test once again.
Until android.hidl.memory is a service, we need this
to be able to run hidl_test correctly.
Test: mma, hidl_test
Change-Id: I4713f3eb5304d5c7dfd74f694e3f25b75fe9c138
ndroid.bp
|
| 6ec2f0b6fd43eec8e064c392804d3d9aa87b6f2f |
11-Dec-2016 |
Martijn Coenen <maco@google.com> |
BnFoo class cleanup.
BnFoo objects wrap an IFoo implementation, which
is stored in mImpl. But, BnFoo previously also
inherited from IFoo, and the implementation of those
methods just did a call-through on the wrapped
implementation.
What's wrong with this is that BnFoo is a transport
object, that we don't want to expose to clients
directly. This change fixes that by removing the
IFoo inheritance from BnFoo. A consequence of this
is also that, if A passed an IFoo to B, and B passes
this IFoo back to A, A will receive the exact same
IFoo object (its local service implementation).
Additionally, all BnFoo objects now inherit from
BnBase; BnBase is a common base class all BnFoo
objects share, so we can safely cast generic
IBinder objects back to BnFoo objects.
Test: mma, hidl_test, hidl_test_java
Bug: 33173166
Bug: 33492405
Change-Id: I6d6dd0db4810fa47af5a428547a1f7d2f0904d43
ain.cpp
|
| 7763ab3046cb998ac8a32b8d2312981631886813 |
14-Dec-2016 |
Yifan Hong <elsk@google.com> |
Disallow duplicated argument name or result name.
Test: compiles
Test: hidl-gen -Lc++ -r tests:system/tools/hidl/test \
-r android.hidl:system/libhidl/transport -o ~/temp \
tests.errors.syntax@1.0
Bug: 33277700
Change-Id: I67c0f8df1a23b1155664aa448d8422953a8a77c5
rrors/syntax/1.0/IEx1.hal
rrors/syntax/1.0/README
|
| 3d5587cdf1c52e73473299004021f5f422c42e1a |
10-Dec-2016 |
Yifan Hong <elsk@google.com> |
Add a script to prevent regression in hidl-gen error recovery.
Run it like so:
croot
./system/tools/hidl/test/test.sh
Add -h to see help text.
Test: run the script
Change-Id: If9636ca08af88121eb7606c54700fc2e258a40cd
Fix: 32813451
rrors.output
est.sh
|
| a47eef385eff3306cef16553ce98e916beb82a38 |
12-Dec-2016 |
Yifan Hong <elsk@google.com> |
Allow method names, result names and argument names to be the same.
size(int32_t size) generates (int32_t size) is allowed.
result names and argument names themselves may still collide and
cause compilation error; for example, foo(int32_t bar, int32_t bar)
is not detected by hidl-gen.
Bug: 33277700
Test: compiles with mma (withe the new method added to IBar)
Change-Id: Ibfe6db03e2dff40f0cae1429fd94cd0db6ccb3a1
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 8c56cbe4b107934da1bea14475b5cc41280a4f01 |
13-Dec-2016 |
Yifan Hong <elsk@google.com> |
Allow bitfield types to be elidable.
Test: mma
Bug: 33298401
Change-Id: I3066d8f0773c135b181cea497f971d9624a77bd6
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 4b0214d0fdeb1aeb63aba0c8a17ba031fb7d6fcf |
07-Dec-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: fix test names.
Test: compiles
Change-Id: Ib76d6b0fdcb476e50810ba14101de6fb730a6015
ain.cpp
|
| 157953f9a055d969e56ebf0cfa291e869d52e180 |
06-Dec-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Fix hidl_test."
|
| 19cb106fe4e75f7462f014f2fdf53ccddedb73ae |
06-Dec-2016 |
Martijn Coenen <maco@google.com> |
Fix hidl_test.
Test: mma, hidl_test passes again
Change-Id: I346d4253aefb16e60c8bee0980178c950c5b6a31
ndroid.bp
|
| c57c8bb9f368faca636bdb6b39773e72255e8b08 |
01-Dec-2016 |
Yifan Hong <elsk@google.com> |
Add a bitfield type.
It is possible to say
enum MyEnum : int8_t {...};
...
bitfield<MyEnum>
in HIDL. In the above example, bitfield<MyEnum>
simply translates to int8_t.
Change mutating |= and &= to allow only (e.g.)
int8_t s = 0;
s |= MyEnum::VAL1;
Bug: 31702236
Test: hidl_test
Change-Id: I8d79975bb48e1ffd3af4726a52ef3678dac60115
rrors/syntax/1.0/IEx1.hal
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| c89340422f53046bfe24ff3e529161f9194120f8 |
18-Nov-2016 |
Yifan Hong <elsk@google.com> |
Hidlize IBase; use IBase instead of IBinder.
* IBase is hidlized over hwbinder. GenericBinder
is removed and Interface with name
android.hidl.base@1.0::IBase is used instead.
* Removed AST::mImportedNamesForJava, because in Java,
we always refer to a type with the full name. We don't
import anything in the java file.
* IFoo.h now includes HidlSupport.h instead of
HidlTransportSupport.h
Test: hidl_test
Test: hidl_test_java
Fix: 32756130
Fix: 33273293
Change-Id: I435ad7c68dea028a16252148b6f5778e3610b7cd
ava_test/Android.mk
ava_test/hidl_test_java
ain.cpp
endor/1.0/Android.bp
endor/1.0/update-makefile.sh
|
| 99e6bebc96d75d1c9b9b8130cd9b880dacb95b01 |
01-Dec-2016 |
Martijn Coenen <maco@google.com> |
memory -> hidl_memory and pointer -> hidl_pointer
Test: hidl_test
Change-Id: Id71c742867be01e80cd48f689c0c619f6b647aac
ndroid.bp
ain.cpp
|
| 2ae5bcaaa598f2cc1208b4e85a283325b4bdc453 |
01-Dec-2016 |
Steven Moreland <smoreland@google.com> |
Revert "Use get instead of implicit cast of Return<*>."
This reverts commit 90e50cdfe97cb021c3c138ce079d31868d77703d.
Change-Id: I854d8bb0ca7b4c98ca5194533d5815fac380dd45
ain.cpp
|
| ccd782b771e283825ad178f1ab33b8c105e6d55d |
28-Nov-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: Fix server names get out of bounds.
It is easy to forget to update the serverNames array.
The logcat output was clobbered; it is now fixed.
Test: hidl_test, manually viewed logcat output.
Change-Id: Ic87d0f43ceb6d1670ea7ceb4479ed5c5c1f056c0
ain.cpp
|
| c7167ca32679a1127d9504be700dfd6dfbc41a3f |
28-Nov-2016 |
Steven Moreland <smoreland@google.com> |
hidl_test: fail if service registration fails.
Bug: 33109192
Test: hidl_test
Change-Id: Idbbc95052931e21506f0494be89cb34a74400e95
ain.cpp
|
| 90e50cdfe97cb021c3c138ce079d31868d77703d |
22-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Use get instead of implicit cast of Return<*>.
Bug: 31348667
Test: hidl_test
Change-Id: I3b326067c9eb1db20497529e5de7cc852d0fbbff
ain.cpp
|
| 05cd423a858b8911ff073326d138d0efc44ddbc6 |
22-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Proper runtime error handling for _hidl_cb.
If _hidl_cb is called a number of times != 1, we LOG_ALWAYS_FATAL.
Bug: 32943424
Test: hidl_test
Change-Id: Ia56ae7bbe172c7952d9a4327d8db1df646f47543
endor/1.0/Android.bp
|
| 2f69a5b157e9c5c41847e55bf3e7dbb01b7de99d |
18-Nov-2016 |
Martijn Coenen <maco@google.com> |
Map handle to hidl_handle in all cases.
Instead of having inconsistent types for HIDLs 'handle' type
dependent on where the handle is used (eg native_handle_t*
in function parameters, but hidl_pointer<native_handle_t>
in compound structures), make it consistent by always using
hidl_handle. Copy constructors and conversion operators will
still make it easy to use for clients/servers.
Bug: 32089785
Test: builds, hidl_test32/64 passes
Change-Id: I2eb60cb56c4c62ad9479fd57f801349ce0f4b7ef
ain.cpp
|
| 23a0a3a055aac5d3a20ae44a8b4d918844fe45ea |
19-Nov-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Remove test_output.py tests."
|
| f31fdd3b6c0ffad3467d1ae707a84403b3bfafb1 |
18-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Remove test_output.py tests.
These tests are no longer enabled. They allow us to do 'diff' tests
between expected and new output of hidl-gen. However, we're more
concerned about functional tests.
Test: mma -j64
Change-Id: I153851d92a15b6fc75f3004d40ae166fc5ae2e4f
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
ata/android/hardware/tests/expression/1.0/BnExpression.h
ata/android/hardware/tests/expression/1.0/BpExpression.h
ata/android/hardware/tests/expression/1.0/ExpressionAll.cpp
ata/android/hardware/tests/expression/1.0/IExpression.h
ata/android/hardware/tests/expression/1.0/IExpression.java
ata/android/hardware/tests/expression/1.0/IHwExpression.h
est_output.py
|
| a5abe9214a264ad6503c35a97345a855c46a9b7c |
16-Nov-2016 |
Yifan Hong <elsk@google.com> |
Split libhidl into base and transport.
Bug: 32756130
Test: mma
Change-Id: Id5eba7f2acce5b752a7498da566ce1eaf276867a
ndroid.bp
ava_test/Android.mk
endor/1.0/Android.bp
|
| bf21f20426cb99475910cf757495b7db6431193d |
18-Nov-2016 |
Martijn Coenen <maco@google.com> |
Build 32-bit hidl_test.
Test: builds, hidl_test32/64 run.
Change-Id: Ie3087fe83f0c80554ad4d0fd50feb9cce5077651
ndroid.bp
|
| b48a7da5a689a246439dfebf54fdd91ec6949d00 |
11-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Include pure types in hidl_test.
Test: hidl_test
Change-Id: I62f04d34a4e74444e86f6e4906c8f98963d43a1f
ain.cpp
|
| d39133bff0a3df357226f33b424cdf99299cc40a |
11-Nov-2016 |
Steven Moreland <smoreland@google.com> |
IFoo::descriptor String16 -> const char*
This makes them much more useable.
Test: hidl_test/hidl_test_java
Bug: 32745840
Change-Id: Ie9b77965821391a1b5c134c82ec680e8cde1440f
ain.cpp
|
| dcb601e7e5b8e166f028c6d7f5db872ba42584b8 |
11-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Merge "Update for service registration notifications."
|
| 0693f31ad6db66a0ed57e5c52bb61d77337cab3c |
10-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Update for service registration notifications.
- Test to ensure that service notifications are getting called.
- Add required includes.
Bug: 30804608
Test: hidl_test
Change-Id: I3f86c39d733f92d21c07b7b1bf4045fa4adb35ac
ain.cpp
|
| 27e85db4f2a8740d24aa5b200edc9cbd4d7de8a9 |
10-Nov-2016 |
Yifan Hong <elsk@google.com> |
Disallow invalid identifiers.
Bug: 31527497
Test: `hidl-gen -Lc++ -r tests:system/tools/hidl/test -o ~/temp
tests.errors.syntax@1.0::IIdentifier1` correctly reports the error.
Test: mma
Test: on master, cd hardware/interfaces && mma
Change-Id: Ib012fda3ef6d15c4b363bfc7af3370d5b5093370
rrors/syntax/1.0/IIdentifier1.hal
|
| 01bcb774f3b0744774a8a28b5e093a7c1c8dd9ff |
09-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Add tests for listing service names.
Test: hidl_test
Bug: 30804608
Change-Id: I9c9e04a5bf62d5cfabe12a4cf673ed63d8bafe86
ain.cpp
|
| d5419bd2fb8f1a98a46228d67579a28d2c5b048f |
04-Nov-2016 |
Colin Cross <ccross@android.com> |
Support genrules with multiple tools
To allow genrules with more than one tool, rename the tool property to
tools and make it an array, replace $tool with $(location <label>),
and use $() for other variables for consistency.
Bug: 31948427
Test: compare build.ninja
Change-Id: I20a69ff2f30715acf718a08bc195d7a938442a2d
(cherry picked from commit 1b27f616afb31974017978ffc2aecc8e4c7582cc)
endor/1.0/Android.bp
|
| be627b3383497c7cc6d2eeb26052f9d7fd128b6b |
29-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl-gen: some error recovery.
Detect some errors and attempt to recover when a
';' is seen.
Attempt to remind the writer for missing ';'s.
TODO: more error recovery should be done
for hidl/errorhal/error/1.0/IEx1.hal
Bug: 31367057 (partly fixes)
Test: mma
Test: `hidl-gen -Lc++ -r tests.errors:system/tools/hidl/test/errorhal -o
~/temp tests.errors.syntax@1.0`
The output has the correct positions marked for the errors.
Change-Id: I858129e08a10cc12a9978adc7fe0c0b94ffbf83e
rrors/syntax/1.0/IEx1.hal
rrors/syntax/1.0/IGood.hal
rrors/syntax/1.0/IGood2.hal
rrors/syntax/1.0/README
|
| 0752ad5df9ccdda06af7abbe71385420b6913e7c |
02-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Add hidl_test_java dependencies in build system.
Test: fresh make and run hidl_test_java
Change-Id: Ice451a9e0dd5919a4780ec753e7e4fda1c9a9f32
ava_test/Android.mk
|
| c46e9842bd82d9c70e08b37a4507b448b285d6c6 |
02-Nov-2016 |
Steven Moreland <smoreland@google.com> |
Break ::android::hardware namespace assumption.
Test: make test.vendor@1.0
Bug: 32579492
Change-Id: I7afd1d83af5f4c0d1c49b828b9e489c111937dcd
ndroid.bp
endor/1.0/Android.bp
endor/1.0/IVendor.hal
endor/1.0/update-makefile.sh
|
| 19d5c17f7073dd0739fab0cfda948e82880375e8 |
21-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Update for hidlized hwservicemanager.
Test: end to end
Bug: 32313592
Bug: 32579492
Change-Id: Id4803490ae7edb490971962e5b799d97002df233
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| a65bb2cd849eb17274c8477e4959e7bc08555fcb |
27-Oct-2016 |
Yifan Hong <elsk@google.com> |
Add test for null native_handle_t*.
TODO: This currently doesn't work, so the test
still EXPECT_FAIL's. It should be changed to EXPECT_OK
after the kernel support is ready.
Test: hidl_test
Bug: 30814137
Change-Id: I656e43c6d055105a8e2320df46f86abe2806a4d2
ain.cpp
|
| e163823c438558579dfc2018fe136a2808cf5d6c |
26-Oct-2016 |
Martijn Coenen <maco@google.com> |
Allow nullptr interfaces to be passed.
Bug: 32410513
Test: mma, hidl_test
Change-Id: I0ed27dc503f6b4d71b2106e03f70b1994800de52
ain.cpp
|
| 9ec25fc8f22979ced1421f8b9c01b09a22837393 |
31-Oct-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Update hidl-gen to generate full name for vts attributes."
|
| 8c48ad75135ec0394a173b29c2068b3bf84a9dc3 |
14-Oct-2016 |
Yifan Hong <elsk@google.com> |
Test for vec<handle> and embedded handle support.
Bug: 32160974
Test: hidl-test
Change-Id: I57e3c58493bbfbe7d19e17607f324925f98cf438
ain.cpp
|
| 1e81c530eec864805c4f13250538e50a067eafc4 |
19-Oct-2016 |
Yifan Hong <elsk@google.com> |
Add test for binder inheritence.
Test: hidl_test
Test: make hidl_test_java compiles.
Change-Id: I6cee5b3161224b3033b5954725468ceb05775289
ndroid.bp
ava_test/Android.mk
ain.cpp
|
| bf828c8f6e1a1345e352976d5b6b91b2f5c52a2b |
26-Oct-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Update hidl-gen to generate full name for vts attributes.
* This is used to handle namespace issue with vts compiler.
Bug: 32410739
Test: make hidl-gen, update generated vts files for nfc and vibrate then
make vts
Change-Id: I38dc8240afb28c1cfa75723289dfcbcff4766af5
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| 07941dae46082ae539dca6f4480e484dbdabb8a0 |
26-Oct-2016 |
Andreas Huber <andih@google.com> |
Merge "Adds support for specifying the version of a service to register/lookup."
|
| 2bb6e1ed0aba0b264d6f2f0c670afc74153902f8 |
25-Oct-2016 |
Andreas Huber <andih@google.com> |
Adds support for specifying the version of a service to register/lookup.
also autogenerates "IFoo.Stub.registerAsService(String serviceName)"
and "IFoo.getService(String serviceName)" convenience functions that fill in the
proper version from the interface's enclosing package automatically.
Bug: 32405011
Test: hidl_test_java
Change-Id: I59888c8538259a1b0698e597f979c674c7739aa5
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 48562fd28f5bd359ed1fef23fd3366050259c883 |
25-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: disabled some tests that are fixed right away.
Also make the result more clear / easy to read.
Test: hidl_test
Change-Id: Ie69cb736a7b66524b6e949bf1a80a30b3e882025
ain.cpp
|
| 40a472391fcb3771d015c15f33d191ab01d8bec1 |
24-Oct-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "hidl_test: require -impl."
|
| 969d2af84ce758324c7faa0da115cf27fe035e7c |
24-Oct-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Merge "Cleanup hidl-gen to use HidlInstrumentor."
|
| 964f72f0004492590d3d1cd048d180a79a20dbc8 |
21-Oct-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Cleanup hidl-gen to use HidlInstrumentor.
* Create BpFoo/BnFoo to inherit HidlInstrumentor.
* Remove the declaration/initialization of instrumentation related variable from BnFoo/BnFoo
* Rename args to _hidl_args to avoid name confilict.
Merged-In: If1a1c9e29ca60a42300c88df6bf196ba453f9bce
Bug: 32141398
Test: make android.hardware.nfc@1.0
Change-Id: If1a1c9e29ca60a42300c88df6bf196ba453f9bce
ata/android/hardware/tests/expression/1.0/BnExpression.h
ata/android/hardware/tests/expression/1.0/BpExpression.h
ata/android/hardware/tests/expression/1.0/ExpressionAll.cpp
ata/android/hardware/tests/expression/1.0/IExpression.h
ata/android/hardware/tests/expression/1.0/IHwExpression.h
|
| 40d131b0a7f28c613cb5545d945d0a9902372c11 |
24-Oct-2016 |
Steven Moreland <smoreland@google.com> |
hidl_test: require -impl.
Test: hidl_test
Change-Id: I24c7fbabf5ed8672c1edeee0de9886d648ac924a
ndroid.bp
|
| 398e6fb53f387f7d3adfb1083d8381b8e813c9da |
17-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: move duplicated code to shared lib
Test: hidl_test
Change-Id: Ia365049911345d74c1a488541a832a09c0287c49
ndroid.bp
ain.cpp
|
| 24bfddd5bd4a6791167fb1b3c83b9cdce3ba353e |
19-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Removed call to overloaded function in hidl_test.
Bug: 31758541
Test: hidl_test
Change-Id: I50d8421842056128a53c8b3d7950e3a4e075cd10
ain.cpp
|
| 86a112bdfae618d6538307839a6c39555bceb79d |
19-Oct-2016 |
Andreas Huber <andih@google.com> |
Special consideration for passing vec<interface> or vec<IFoo> to and from
methods. These remain unsupported inside structures.
Bug: 30570663
Test: hidl_test
Change-Id: Ic19db58d3c3ae8caa85a773a143fc148fd8affb6
ain.cpp
|
| b054e8c630f27fbc53ea766cf0de2a9f7f4d155a |
19-Oct-2016 |
Andreas Huber <andih@google.com> |
Merge "Vectors are now exposed as ArrayList<T> in the Java backend even for arguments"
|
| 35d6600c0939cef39e68ad5b0470866ea300ae74 |
13-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: kill servers more gracefully.
Test: hidl_test
Change-Id: I52c5fdfb79584c05bff3398b3038a3e190a70096
ain.cpp
|
| d12398dc9796d3b5dd69730c57a5199770e97d68 |
13-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: Add -p/-b option.
* Add an option to let the user
to choose running passthrough tests only or
binderized tests only or both.
Test: The following succeeds.
croot && cd hardware/interfaces
mma
croot && cd system/tools/hidl
mma
adb shell /data/nativetest64/hidl_test/hidl_test
Bug: 31819198
Change-Id: I326b80eb12f9df710e98744ae6b542dfb03d5a75
ain.cpp
|
| 3cccc0f3b74c930ec97759c4fa35d21f308e4bfe |
13-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: Fix validationPointerInterface.
Test: mma passes with HIDL_RUN_POINTER_TESTS.
It cannot be tested because the kernel is not
ready.
Change-Id: Id94f13fa0d15c8d28fbdb68c9c393a9e8e9752ab
ain.cpp
|
| ca89052152606fff3bf634777654584aa4b07efa |
12-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: Move getService to environment setup
so that each test runs faster.
Test: hidl_test
* Note that ForReportResultsTest and FooNullNativeHandleTest
are failing for passthrough mode. It will be fixed in
the upcoming CL.
Change-Id: Idf92b0f2af800851c1816a1ffad42a407f6d721a
ain.cpp
|
| c70f0d8f3c55a2bef3d4d42c804e678c03103aa6 |
10-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: add back binderized mode
Test: hidl_test
Bug: 31819198
Change-Id: I0f0e177c3499a4d5473d8da013e5534c894145b1
ain.cpp
|
| 031a61e70e313cbe532306180fccfb18c818c3d9 |
10-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: passthru; move foo/bar/pointer's impl
-ementation to default implementation of Foo and Bar.
Since they are not imported, tests are now run
in passthrough mode.
Pointer tests would not work because I don't know
how to instantiate validationPointerInterface. This
will be fixed in follow-up CLs in the same topic.
Test: hidl_test
Bug: 31819198
Change-Id: Iaf598beac33645c26839f16f17d9d29922c79611
ain.cpp
|
| 3eac8a3419ce36a4f84e4620474711c138753d3c |
11-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: cleanup
Test: hidl_test
Change-Id: Ifce7f86871b4b2821f8a2a47dd6ffb099f706fcb
ain.cpp
|
| fa4b3c62f07660c92214e9d4846d4d0a1cb1db0f |
18-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Merge "Revert "Removed test for overloaded function.""
|
| 45fa6ebe69252deb9f34009961bd864687ed4c1c |
18-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Revert "Removed test for overloaded function."
This reverts commit 251ed1025b77efc2289301def6a630e5beab931e.
Change-Id: Ib4a7c02f65a0e839df36826ac85a12fd6f01df9f
ain.cpp
|
| c54dab9a7aed0e255c2cb5687d5ded1971c612ab |
18-Oct-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Removed test for overloaded function."
|
| 1b6822b19bf60a98ba8e27b135a5ad66287a5542 |
18-Oct-2016 |
Andreas Huber <andih@google.com> |
Vectors are now exposed as ArrayList<T> in the Java backend even for arguments
and return values to/from methods, they already were inside structures.
In addition vectors of arrays are now properly marshalled when used as arguments
or results to/from methods.
Bug: 32237331, 32180328
Test: hidl_test_java
Change-Id: Ie24f11668f22fc2c31fb7d8e8da534f1c970f58f
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 251ed1025b77efc2289301def6a630e5beab931e |
18-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Removed test for overloaded function.
Bug: 31758541
Test: hidl_test
Change-Id: Ifbba77ee13ab02836c05da094f5412d505282d0b
ain.cpp
|
| 5eac2dee37c9af196743bc273e4dc3db368151eb |
07-Oct-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Update vts files for nfc hidl file style change.
Test: pass
Change-Id: I7f68d29c4973ebff57b8690bdc2110095442e669
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| ef1a9feda3f09f87f0673cc978b36e1d4a8b4223 |
07-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Get rid of call enum.
Bug: 31758541
Test: hidl_test, hardware/interfaces mma, manual inspection of output
files
Change-Id: I16c076dbff9b0b8688e2492b0b9c97e9f6d48943
ain.cpp
|
| b725d679b19a1a5e74e30666450832469bc35cbd |
10-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: add test for null native handle.
Test: hidl_test
Bug: 30814137
Change-Id: I05cf44418c7fcac891d5072b150706f04ada3b31
ain.cpp
|
| 7a827722c5b615585cfd1e664fe434cc47af9dbf |
07-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: abort test gracefully if service not found
instead of SIGABRT without useful output.
Test: hidl_test
Change-Id: I9c37acdc9d17b91e98e416f4c5d7d2ec93fc153a
ain.cpp
|
| c5ea9f589cc7cce0b5e97bd4ac6a8561eb313a02 |
07-Oct-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Update hidl-gen support for vts.
* Support the new Enum type (based on scalar_data)
* Support sub_struct/sub_union defined within compound type.
* Code cleanup: use getVtsType() instead of hard code ones.
Test: make hidl-gen, locally run make hidl_gen_test.
Bug: 30762234
Change-Id: I9a21b5757e0a9fc6cd1bf829ab123565a7990ad5
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| cc59f658497eeb7b58ba566861d22d444ccc1cf2 |
06-Oct-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Merge "Add instrumentation code on proxy side."
|
| 7bc2da67c5ae3ccde4cb1c3f537bedd362be8576 |
06-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Merge "Revert "Revert "Revert "Regenerate test nfc vts files.""""
|
| 89e74b48a04606e17d843d54190bbccec871ef7e |
06-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Revert "Revert "Revert "Regenerate test nfc vts files."""
This reverts commit 803764f9451504994e8b582f69df90e992c6954b.
Change-Id: If737cbf4a72b5577ed3b075052b5f74f20816d32
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| 8f49294db5ac2898bac7c58ca20ff37d4cacdf5e |
28-Sep-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Add instrumentation code on proxy side.
* Only works for C++ backend for now.
Bug: 31266145
Test: make hidl_test
Change-Id: I18a78b274448fee12a8d340a891ba0c4f7a3a738
ata/android/hardware/tests/expression/1.0/BnExpression.h
ata/android/hardware/tests/expression/1.0/BpExpression.h
ata/android/hardware/tests/expression/1.0/ExpressionAll.cpp
ata/android/hardware/tests/expression/1.0/IExpression.h
ata/android/hardware/tests/expression/1.0/IHwExpression.h
|
| be2a3737675b83dd61b52c11f0d7549fa1276739 |
05-Oct-2016 |
Yifan Hong <elsk@google.com> |
hidl-gen: add sanitizedName to emitReaderWriterEmbedded.
* Fix weird variable names in autogenerated code for vec<vec<T>>.
Bug: 31955917
Test: hidl_test
Change-Id: I83363a6d12babae1aba8e118451e1ec80492012c
ain.cpp
|
| 803764f9451504994e8b582f69df90e992c6954b |
04-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Revert "Revert "Regenerate test nfc vts files.""
This reverts commit 86962434a4fa114256ee1eb45af822111e26c7c2.
Change-Id: Ide3f58e8f734c19cba2f863dacfa5dd8132b1728
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| 67f67b4e7a1f0d7d53703daa42d5cc07b375a577 |
29-Sep-2016 |
Steven Moreland <smoreland@google.com> |
Disallow null callbacks.
Bug: 31815793
Test: hidl_test
Change-Id: I3a977df3cb36af46fa5343c1ebbe97df4ff23fca
ain.cpp
|
| 386d6cbb79f6000e48893d3318f702fcc329b272 |
04-Oct-2016 |
Martijn Coenen <maco@google.com> |
Merge "Revert "Regenerate test nfc vts files.""
|
| 7d3f4e35ce839418df72aa8b52c574e7a96613bc |
04-Oct-2016 |
Andreas Huber <andih@google.com> |
Merge "Fix compiler warnings in hidl_test."
|
| 86962434a4fa114256ee1eb45af822111e26c7c2 |
04-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Revert "Regenerate test nfc vts files."
This reverts commit 4046647e5fb2d8d3ecf3f01d86cf880d8ca3fd29.
Change-Id: I2983572b08e17391126a383f94b6b9476746eaf9
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| 1452b4ff5b63e3c093aac5a1fe21c4b20faf5b01 |
04-Oct-2016 |
Steven Moreland <smoreland@google.com> |
Merge "Regenerate test nfc vts files."
|
| e8047bf145564d7dd4479295f1361db28b5109ee |
04-Oct-2016 |
Andreas Huber <andih@google.com> |
Fix compiler warnings in hidl_test.
Test: mma
Change-Id: I44042b940d543c2cb56f1ce747c9f3a0e66d4f85
ain.cpp
|
| 4046647e5fb2d8d3ecf3f01d86cf880d8ca3fd29 |
30-Sep-2016 |
Steven Moreland <smoreland@google.com> |
Regenerate test nfc vts files.
Test: pass
Change-Id: I7fe634f556103a5e4558b641d6e2b982b9c6238e
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| 676abdc8f7f1814c8657f0cb0e8a301e4aa36cb1 |
29-Sep-2016 |
Dan Willemsen <dwillemsen@google.com> |
Generate Android.bp files; Convert to Android.bp
Keep generating Android.mk files for java libraries, but move c++ shared
libraries to Android.bp.
Now that we're building these in Soong, we can build hidl_test in Soong
as well.
Bug: 31742855
Test: hardware/interfaces/update-makefiles.sh
Change-Id: Ifd7b6d423971b04b2120b51d63eacfc6b162e25d
ndroid.bp
ndroid.mk
|
| 84465909d270461251ca0c195640911b4eb6a7fa |
28-Sep-2016 |
Yifan Hong <elsk@google.com> |
Add test for parcel # references cap.
Bug: 31757611 Optimize findBuffer & quickFindBuffer in Parcel for
pointers
Test: `mma`
Test: `make hidl_test && adb sync && adb shell hidl_test`
Change-Id: I54d71065826ef62571fdd0b143f53f06cbee6c6a
ain.cpp
|
| bf459bcedc2ab1a22a9cc290fbcc0a5462e63690 |
24-Aug-2016 |
Yifan Hong <elsk@google.com> |
Pointer support and embedded types in HIDL.
* Pointers work per transaction. Don't work
across transactions.
* ref<T> in HIDL translates to T const* in C++.
* No Java support.
* Embedded types like ref<vec<vec<int32_t>>>
or vec<ref<T>> is supported. Pointers to
pointers like ref<ref<ref<T>>> is supported.
* Array of pointers and pointer to array supported.
* Pointer inside a union is NOT supported.
Test: `mma`
Test: `make hidl_test && adb sync && adb shell hidl_test`
Note that this only works with a kernel patch.
Bug: 31300815
Bug: 31349114
Change-Id: I15b74ca74a801009cc8bdc7132bd53d0185dbcbf
ndroid.mk
ain.cpp
|
| b1c976de3ecfc42d22832b64371969a9c8535d99 |
29-Sep-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Adds a test to verify new hidl_array<T, SIZE1, SIZE...>::size() API(s)"
|
| b15f2483b6ae4aa63b79dd3b8d02c27445e21aae |
29-Sep-2016 |
Dimitry Ivanov <dimitry@google.com> |
Merge "Add missing liblog dependency"
|
| 5d03477ad46fdddc17eabfbc4d909d51f4647798 |
28-Sep-2016 |
Andreas Huber <andih@google.com> |
Adds a test to verify new hidl_array<T, SIZE1, SIZE...>::size() API(s)
Bug: 31805709
Test: hidl_test, hidl_test_java
Change-Id: Ib21275316c7aa69ee69cc8e2fa9d900317a2f65a
ain.cpp
|
| 5e44a291ff57cd25ce6447996b11104ac2938ee9 |
27-Sep-2016 |
Andreas Huber <andih@google.com> |
Avoid name conflicts in the generated C++ proxy code when arguments and results
share the same name, as in a method declaration
sendVec(vec<uint8_t> data) generates (vec<uint8_t> data);
Bug: 30778622
Test: hidl_test, hidl_test_java
Change-Id: I06a12025363b3ca10cb0d548c610c57bdf88af92
ain.cpp
|
| f1abc2a16c3384d5813b69c5ebe8cb27bd256f21 |
28-Sep-2016 |
Andreas Huber <andih@google.com> |
HIDL vectors are now exposed as ArrayLists rather than Vectors in Java.
Bug: 31749239
Test: hidl_test, hidl_test_java
Change-Id: I4dd705ac7962195e3107a67bd62e0c9154339e5b
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| f03332ac955bc6cb22873e236868eacfc3bf78cc |
23-Sep-2016 |
Andreas Huber <andih@google.com> |
Arrays in .hal files are now exposed to C++ as hidl_array<T, SIZE ...>
instead of as native arrays. This allows them to be copied which in turn lets
them exist in vectors. In the Java backend, vectors are limited to
one-dimensional arrays and scalar arrays are properly wrapped,
i.e. vec<uint8_t[]> => Vector<Byte[]>
Change-Id: I47524ec8423dfb41a436df36af8fa05eb8b3c0cc
Bug: 31682327
Test: hidl_test and hidl_test_java
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
ain.cpp
|
| 4724055519f1e4f95da71a3eb5e33032bc707095 |
28-Sep-2016 |
Dimitry Ivanov <dimitry@google.com> |
Add missing liblog dependency
Bug: http://b/31289077
Bug: http://b/27171986
Test: remove liblog from libcutils and recompile
Change-Id: I660031ae94f8ce4176c6cf0d47d67788708bd42d
ndroid.mk
|
| c94ccaf65ec8c7d3d0d2ee9d9ea3479644f83394 |
16-Sep-2016 |
Myles Watson <mylesgw@google.com> |
test: Wait on the correct Pid
Change-Id: I474670179c5a3fe5276dad9358ce7c5cfab24ac6
ain.cpp
|
| 1858dba3eda8d57735060cbcb78c6d3805ec122d |
20-Sep-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Arrays of arrays are now coalesced into a single multi-dimensional array type."
|
| 709b62dbda6184770bb34470ff550e02c1643e67 |
19-Sep-2016 |
Andreas Huber <andih@google.com> |
Arrays of arrays are now coalesced into a single multi-dimensional array type.
This is natively stored as a flat array of the element type.
Change-Id: I1b0eee51b496648a9c8445b835635deff1a90aab
Bug: 31438033
Test: run updated "hidl_test" and "hidl_test_java"
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
ain.cpp
|
| de57800c370f79b52b4b87014ab50b34f8aec7fd |
08-Sep-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Support generating instrumentation code for C++ backend in hidl-gen.
* Register instrumentation callbacks on server side.
* Invoke each callback during the entry/exit of each method.
Bug: 31266145
Test: test pass for hidl_test
Change-Id: I53d8b1b248fc6100e4e29d8a4b31d0994133fb71
ata/android/hardware/tests/expression/1.0/BnExpression.h
ata/android/hardware/tests/expression/1.0/ExpressionAll.cpp
ata/android/hardware/tests/expression/1.0/IExpression.h
|
| ec085bc17cf470f0cda4789431f6f2e3ebcbe3e6 |
16-Sep-2016 |
Steven Moreland <smoreland@google.com> |
Merge "hidl_test: fixed EXPECT_OK and EXPECT_ARRAYEQ"
|
| 4c865b72b320a46f326a335cfd326b66b0e10f67 |
15-Sep-2016 |
Andreas Huber <andih@google.com> |
[Java hidl-gen] Fix multi-dimensional arrays (used within structures, not
as method arguments).
Bug: 31438033
Change-Id: I8e29ae8eca2a9ef8a320e37ca0bb79dcfa8a9b9a
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| e70455b6294fcb1ab98b8cd608000a3af61af2cc |
15-Sep-2016 |
Steven Moreland <smoreland@google.com> |
hidl_test: fixed EXPECT_OK and EXPECT_ARRAYEQ
EXPECT_OK: uses #define instead of function so that the line number is
reported properly in the error message.
EXPECT_ARRAYEQ: renamed to isArrayEqual because it isn't actually
asserting anything.
Change-Id: I03ceb6632f18afdc3802a5cbdfc93296bdbb8fdc
ain.cpp
|
| f9d49f1468bd65595f6f0ae5f4d7300f4e95edf8 |
12-Sep-2016 |
Andreas Huber <andih@google.com> |
[C++ hidl-gen] Fix multi-dimensional arrays, vectors of arrays.
Bug: 31438033
Change-Id: I254e8bfeb8fbf665dd3836825aa271603be6d14c
ain.cpp
|
| 8dd29816611cee8c652a26a16aa03657dd5bb93c |
12-Sep-2016 |
Treehugger Robot <treehugger-gerrit@google.com> |
Merge "Java support for passing vec<STRUCT-TYPE> and STRUCT-TYPE[] to and from methods."
|
| 4bcf97d0862680a60f684dafd623c66cf0910292 |
30-Aug-2016 |
Andreas Huber <andih@google.com> |
MQDescriptor is now a predefined type.
b/31091383 Adding WireMQDescriptor as a hidl type
Change-Id: I8eb7f25dd35bc1e33393b9c563152c3a1551419b
Signed-off-by: Iliyan Malchev <malchev@google.com>
ata/android/hardware/tests/expression/1.0/IExpression.h
|
| f630bc8736003dcf4aac3dfe47167beb6beb6c6a |
09-Sep-2016 |
Andreas Huber <andih@google.com> |
Java support for passing vec<STRUCT-TYPE> and STRUCT-TYPE[] to and from methods.
Bug: 31380705
Change-Id: I7374538d0bc2b3562c997710bfa4ed1f95950589
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| a21f1494a1b75ebd799d3f1156047caecc643270 |
08-Sep-2016 |
Martijn Coenen <maco@google.com> |
Encode hidl_version in each generated interface.
Each generated interface will have a constexpr encoding
the package version. Additionally, there's a client-facing
getInterfaceVersion() method; while for now it always returns
the static version, this is not always correct: in case this
interface has a transport that goes to a remote implementation,
the version needs to be retrieved from that implementation
instead. This depends on other changes and will be added with them.
Bug: 31297066
Change-Id: I15b7f5ccf2dbfa25acfd385c68e8ae1f3e782ade
Signed-off-by: Iliyan Malchev <malchev@google.com>
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| 4ff971cc617b5fbfdd581c50a2c4862a03e20633 |
08-Sep-2016 |
Martijn Coenen <maco@google.com> |
Remove String16 dependency.
Bug: 31297066
Change-Id: I71996af4ca5c0433710cc2df9bf09760818c1043
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| d57066f167eab04f32c8ef7e644e2a036256209b |
08-Sep-2016 |
Iliyan Malchev <malchev@google.com> |
hidl-gen: use Return<void>
C++ HIDL methods now return Return<> objects regadless of elision
status. This hides Status from auto-generated code.
b/31367910
b/31348667
b/31365311
Change-Id: Ib009d2bade022a557f51d3cd4cb3e61ece38bba2
Signed-off-by: Iliyan Malchev <malchev@google.com>
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| 60bce222f72f3ee2d26fc5c8477f3af61e382e08 |
07-Sep-2016 |
Steven Moreland <smoreland@google.com> |
Reduce unnecesary waiting in hidl_test.
Change-Id: Id74e3ff192a99c400620a7f315117276fb13073a
ain.cpp
|
| 34c302c85fb37b523f32d0e4158372361e64f40e |
06-Sep-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Disable hidl_gen_test.
Bug: 31302232
Change-Id: Id9536df1dee691152a4a1c2622d10ae22cb202df
ndroid.mk
|
| 097526fb7a3d4f4bd416ad1883ef4fe8545d5989 |
06-Sep-2016 |
Dan Albert <danalbert@google.com> |
Stop hard coding gtest crap.
The build system handles this.
Test: mm
Bug: None
Change-Id: Ia3eceb53445ee908ac5c0a9564380e88ff17584b
ndroid.mk
|
| 5c78afc26d29e2f54dceab752d73339eb98ca7d0 |
06-Sep-2016 |
Iliyan Malchev <malchev@google.com> |
hidl-gen: update test data
Change-Id: If8fcbb6097d0a87a2eedaa625e075e0ecc2add7b
Signed-off-by: Iliyan Malchev <malchev@google.com>
ata/android/hardware/tests/expression/1.0/BpExpression.h
ata/android/hardware/tests/expression/1.0/IExpression.h
|
| 9391510f0d5ec6340aef2325ce0b92c84c1f56f9 |
01-Sep-2016 |
Martijn Coenen <maco@google.com> |
Move Status to libhidl.
Bug: 31226252
Change-Id: I9615071be3845eb3290fb6c4fa5ae622f56a4d80
ata/android/hardware/tests/expression/1.0/IExpression.h
ata/android/hardware/tests/expression/1.0/IHwExpression.h
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| 0b3b411706f01e094fb0b4bcacf07ca847299c6d |
01-Sep-2016 |
TreeHugger Robot <treehugger-gerrit@google.com> |
Merge changes from topic 'java_makefiles'
* changes:
Emit eval result of constant expressions for Java.
Auto-generate java libraries for compatible packages.
|
| 19ca75ae47df5cd9447b232c31c5df1d110e85d9 |
31-Aug-2016 |
Yifan Hong <elsk@google.com> |
Emit eval result of constant expressions for Java.
b/31223640
Change-Id: I3552e01ecaef8ac570f13c4357cd04c1c3c516bb
Signed-off-by: Iliyan Malchev <malchev@google.com>
ata/android/hardware/tests/expression/1.0/IExpression.java
ava_test/Android.mk
est_output.py
|
| 0fa9e3956e59155c52e40b85b6aef5476ae90912 |
31-Aug-2016 |
Andreas Huber <andih@google.com> |
Auto-generate java libraries for compatible packages.
Change-Id: I753f8bc5ab8ca6b5293c75b548897bbdec16aa65
Bug: 31221188
ava_test/Android.mk
ava_test/hidl_test_java
|
| 03866f5efb7b0e20e8bb4fd416e577a6332f3306 |
30-Aug-2016 |
Andreas Huber <andih@google.com> |
Some minor fixes to hidl_test output.
Change-Id: I14daff48c5b69208dd87c5128246ae5348fda26a
ain.cpp
|
| 93d0ebb5937085cebb4a22c255bd16bd551f3b34 |
31-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
hidl-gen: fix build-time constexpr tests
b/30951879
Change-Id: Icd0eb5db0ff4e6e3c18cda616eef9d651996a0c8
Signed-off-by: Iliyan Malchev <malchev@google.com>
ata/android/hardware/tests/expression/1.0/BnExpression.h
ata/android/hardware/tests/expression/1.0/BpExpression.h
ata/android/hardware/tests/expression/1.0/ExpressionAll.cpp
ata/android/hardware/tests/expression/1.0/IExpression.h
ata/android/hardware/tests/expression/1.0/IHwExpression.h
|
| 5788697381666844eeb23e04e5c6f83ec6ec8b44 |
17-Aug-2016 |
Yifan Hong <elsk@google.com> |
Evaluate constant expressions.
* The rules for evaluating constant expressions are very much
like that in C/C++, in which (u)int32_t are treated as
(unsigned) int and (u)int64_t are treated as (unsigned) long.
* Detailed rules can be found at the end of ConstantExpression.cpp.
* Only bool and integer types are evaluated. float, double
are not evaluated (not even parsed; so a syntax error will
be raised).
* Const-declared identifiers are parsed, but not evaluated.
* Only evaluate expressions for enum types. Array length and
annotation parameter values are not evaluated.
* Original expression is formatted and placed as a trailing comment
(only available in C++ format).
* Tests can be found at Icacba6b6262cc0026a94f64527b80e2c0e35e72e.
Change-Id: Ia33b68a053727747ec87b0e77c361ca19c9f41fd
ata/android/hardware/tests/expression/1.0/BnExpression.h
ata/android/hardware/tests/expression/1.0/BpExpression.h
ata/android/hardware/tests/expression/1.0/ExpressionAll.cpp
ata/android/hardware/tests/expression/1.0/IExpression.h
est_output.py
|
| 4078631353d6d34db21c890d1870e796eca6ea38 |
16-Aug-2016 |
Steven Moreland <smoreland@google.com> |
Split 'IFoo' into 'IFoo' and 'IHwFoo'
The traditional binder "IFoo" interface is coupled
very tightly with Binder; it inherits from Binder
objects and implements several Binder-only methods.
We want to abstract away the RPC mechanism as much as we can; therefore,
we want to create a "clean" IFoo interface, which contains just the
methods defined in the HIDL interface. Separately, there's an IHwFoo
interface, which implements IFoo as well as the necessary Binder
methods.
Changes in hidl-gen to support this:
- Generates IHwFoo.h
- Moved (DECLARE|IMPLEMENT)_HW_BINDER macros from IFoo to IHwFoo
- Added REGISTER_AND_GET_SERVICE macros to IFoo
- Removed all hwbinder/ include paths from IFoo
- When passing an interface through a HIDL method, wrap a
BnInterface object around it
- BnFoo's implementation of IFoo calls through the register
interface implementation
- Updated test code
Tests: hidl_test, hidl_java_test, libhwbinder bench, NFC all work.
Bug: 30588200
Change-Id: Ie7ca4eef905f84aebd06bee971b5f6170e169797
ava_test/hidl_test_java_native.cpp
ain.cpp
|
| d07494ffb43b8aaa49250091296f53cac3d26695 |
30-Aug-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Merge "A few fix of hidl-gen for vts generator."
|
| 199335277986f45bef3aaf8787d57b4c3c6b6b9a |
30-Aug-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
A few fix of hidl-gen for vts generator.
* Remove component_type.
* Use the new VectorDataValueMessage.
Bug: 30762234
Change-Id: I4262240c1b79129bc3801eba3b6037843a484220
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| 401cd16db8a421aaf91338024cea9208549ef67e |
26-Aug-2016 |
Andreas Huber <andih@google.com> |
Adds a test verifying that the Java backend is compatible with the C++ backend.
Change-Id: Ib9cfeeba049281f6bf96c42758864d9e6a800dfa
ndroid.mk
ava_test/Android.mk
ava_test/hidl_test_java
ava_test/hidl_test_java_native.cpp
ava_test/src/com/android/commands/hidl_test_java/HidlTestJava.java
|
| 368491306c932ef23939c3214cd014278e1aa6ca |
26-Aug-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Support typedef for vts generator.
*Since hidl-gen simply treats typedef as alias of the defiend type, no
need to generate type declaration for it in the .vts file.
*Move the make for hidl_gen_test to test/Android.mk.
Bug: 30762234
Change-Id: I509be0dc8e4f32df62448385d58c80a7166c3ff5
ndroid.mk
|
| 0acf4192dc13277bdd4ba8006d3602595ccfd546 |
23-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
hidl_test: add server-timing tests
Test the following:
-- callMe() is a oneway call, so it should return to the
caller immediately;
-- callMe() passes an IFooCallback to Bar, and Bar invokes
three methods on that callback. Of these three methods,
the first one is blocking but with an empty body; the second one is
blocking and the body takes several seconds to execute; the third one
is oneway, but its body takes several seconds to execute
-- hidl_test uses two instrumentation methods on IFooCallback,
reportResults() and youBlockedMeFor(), to collect timing information
on the three methods above. This information is used in a test that
verifies that the timings are as expected, within some predefined
tolerance.
b/30855757 Convert hidl test to gTest
Change-Id: Ie1d8e16ac72a647579b460cbcf3cee8296a9f7c3
Signed-off-by: Iliyan Malchev <malchev@google.com>
ain.cpp
|
| 1dc8793c9aac86fe08dbe3150d96d88435a978df |
19-Aug-2016 |
Yifan Hong <elsk@google.com> |
hidl_test: apply gtest.
b/30855757 Convert hidl test to gTest
Note: only test the values received by the client. Doesn't
test the values received by the server... yet.
Change-Id: I1f50e3fefe9bfc3cc07ff48f764c0f20c90175d5
ndroid.mk
ain.cpp
|
| 864c771ca4ec8a01e31c7c243625b7a5f6316768 |
17-Aug-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Enhanced hidl-gen support for generating vts files.
*Support callflow from mehtod annotations.
*Support inheritance for interfaces and types.
*Support Array and Union type.
b/30762234
Change-Id: I059b780d6362a557bb9cfd70d6c5ec0e73916ce3
ata/android/hardware/nfc/1.0/types.vts
|
| 2b6591b6fc64b21b58f230d7c91cb1367f0bb564 |
19-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
Rename SimpleReturn --> Return
b/30518487 Optimize out lamdas from HIDL C++ auto-generated code in
common cases
Change-Id: I1c5e61d47904b1b70dc03560796654b98d3f3212
Signed-off-by: Iliyan Malchev <malchev@google.com>
ain.cpp
|
| 7473fab0d9902b3befab21670780a5ba5f5ea323 |
19-Aug-2016 |
Martijn Coenen <maco@google.com> |
Generated makefiles link against libhidl.
And updated tests to link to libhidl as well.
Bug: 30839546
Change-Id: I0eff676e114531498859ca548072b0c226fd2ed4
ndroid.mk
ain.cpp
|
| d856b05dd8dd57460271f14b06ec71acdd706a2a |
17-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
hidl_test: update to use elided callbacks
b/30518487 Optimize out lamdas from HIDL C++ auto-generated code in
common cases
Change-Id: I7c7c5d312b5114b034bc16b4dd809362c8a7c7bd
Signed-off-by: Iliyan Malchev <malchev@google.com>
ain.cpp
|
| f505447dc3b3aa0ab6bedc3a9d67e1b34bd22f23 |
17-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
re-enable hidl_test
Change-Id: If2be5c3682823877b6b67582ced1a918b5fc5ceb
Signed-off-by: Iliyan Malchev <malchev@google.com>
ndroid.mk
|
| 82b83bbf40749e747e967175ebe49fc1f7cd180c |
13-Aug-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Add a script to test hidl-gen output
Usage: python testOutput.py hidl-gen output_dir canonical_dir
--output_dir: dir to store generated file, will be cleanup if all tests pass.
--canonical_dir: dir that stores all canonical files used for comparason.
* Currently only test for vts outputs.
b/30762234
Change-Id: I7676585f051a8d2c61c1e7a184e223f27e81ad0f
est_output.py
|
| 9add9fd0cbfda815496e07dacbe966d973c66c51 |
14-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
hidl_test: use booleans
b/30853785 auto-generation for bools is broken
Change-Id: Ia81e4fbde2816a74bbcd49537db674ee6e78d923
Signed-off-by: Iliyan Malchev <malchev@google.com>
ain.cpp
|
| b31e10c04421558569e64c3f275b7577e85f5329 |
14-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
hidl_test: clean up
-- Register FooCallback service so it can respond to requests
-- Add oneway callbacks to FooCallback
-- Make logging output more uniform (CLIENT, SERVER(Foo),
SERVER(FooCallback)
b/30843534: oneway methods seem broken
Change-Id: I0d3572d3d071eb4d627f105c0bd62d96b002f786
Signed-off-by: Iliyan Malchev <malchev@google.com>
ain.cpp
|
| 5158db484e5ab302368f191d75d5b1334c270e52 |
10-Aug-2016 |
Zhuoyao Zhang <zhuoyao@google.com> |
Extend hidl-gen to support generate vts file.
b/30762234
TODO: add regression test.
Change-Id: I1c31fd9a85805cd450ea03cc0ccc750a756d1009
ata/android/hardware/nfc/1.0/Nfc.vts
ata/android/hardware/nfc/1.0/NfcClientCallback.vts
ata/android/hardware/nfc/1.0/types.vts
|
| cd9b4bf6ab94e2f8e373c6076351d71b82807cb2 |
15-Aug-2016 |
Andreas Huber <andih@google.com> |
Merge "fix hidl_test to use proper hidl_vec<> allocation."
|
| 899b0632907d8233f7be46db88e359acd6374411 |
15-Aug-2016 |
Andreas Huber <andih@google.com> |
fix hidl_test to use proper hidl_vec<> allocation.
Change-Id: I5378305c5feee8714357a096e8803077210f1843
ain.cpp
|
| 2831d5145675ead9f2fb767bf5fe4ae56b88349f |
15-Aug-2016 |
Andreas Huber <andih@google.com> |
Initial commit of Java backend to hidl-gen.
Change-Id: I38b62637df74d3e5daf702a8996502d0d5726033
ndroid.mk
|
| 88ca4511354b4bdaed1846486a1b285576788e4b |
11-Aug-2016 |
Steven Moreland <smoreland@google.com> |
Updated hidl_test to reflect changes in hidl-gen.
1). importing binderized HALs as local shared libraries
2). updating functions to correspond to updated versions of
android.hardware.foo/bar@1.0
3). updated code dealing with hidl_vec, who's data members are now
private (no longer
'just' a struct)
4). included test/Android.mk, so you can 'make hidl_test'
Change-Id: I1bf7085f86247414c02417d4a5397c31c878ee9b
ndroid.mk
ain.cpp
|
| b66c3991c5203d6a93444012b9fa8ace22242550 |
08-Aug-2016 |
Iliyan Malchev <malchev@google.com> |
use hidl-gen as the name of the code generator
-- Pass argv[0] as the name of the executable to run in auto-generated
Android.mk files
-- Use hidl-gen in makefiles for the test executable
Change-Id: Ia35eaf9f66608d67730910c73436e4a7b8540752
Signed-off-by: Iliyan Malchev <malchev@google.com>
ndroid.mk
|
| 8a82ff7ce6b5c0288b5239e25e1cbd1f5d8e3d27 |
04-Aug-2016 |
Andreas Huber <andih@google.com> |
hidl::{Parcel,Binder & friends} => hardware::\1
ain.cpp
|
| 6cb08cf9f021a01d9d2b1eaec6729aac6ae70708 |
04-Aug-2016 |
Andreas Huber <andih@google.com> |
Support for inheritance of interfaces.
ndroid.mk
ain.cpp
|
| 9cd48d06c94c25516079e4046be7c8e7e7d2098f |
03-Aug-2016 |
Andreas Huber <andih@google.com> |
Simple test cmdline app.
ndroid.mk
ain.cpp
|