The GLES driver appears to be left in a bad state without destroying the
render engine instance explicitly after every test case.
Bug: 149043811
Test: build, adb push out/target/product/<platform>/testcases/VtsHalGraphicsComposerV2_2TargetTest/<abi>/VtsHalGraphicsComposerV2_2TargetTest /data/local/tmp && adb shell /data/local/tmp/VtsHalGraphicsComposerV2_2TargetTest
Change-Id: I8e3c4e30eaa2ef3b9831da7b335b9a830be4b4ed
Add a TestCommandReader to composer 2.4 VTS to parse correctly the
new commands introduced in 2.4.
Test: adb shell /data/nativetest64/VtsHalGraphicsComposerV2_4TargetTest/VtsHalGraphicsComposerV2_4TargetTest
Bug: 149124892
Change-Id: Ide4f98fa3bfc771e4095cc672f59b87e1aa03427
All grahpics composer HAL tests have been converted to parameterized
gtest.
Bug: 147894326
Test: atest VtsHalGraphicsComposerV2_1TargetTest \
VtsHalGraphicsComposerV2_2TargetTest \
VtsHalGraphicsComposerV2_3TargetTest \
VtsHalGraphicsComposerV2_4TargetTest
Change-Id: I84515a5698035d7ceb299a5cc140ecd7bb28fe84
Prework for the client composition cache logic which extends
LayerSettings. Passing in a set of pointers will reduce the
amount of copying needed when passing the LayerSettings to
RenderEngine.
Bug: 136561771
Test: atest librenderengine_test libcompositionengine_test
Test: atest VtsHalGraphicsComposerV2_2TargetTest
Test: go/wm-smoke
Change-Id: I0581be7762bf56cd7a3aab0511adf95cd6e2df76
getClientTargetProperty will give hardware composer the ability to request some
properties of the client target that hardware composer wants. Prior to this
API, the client will does its best to produce the client target of which the
properties are pretty much fixed.
BUG: b/145968912
Test: mmma -j32 hardware/interfaces/graphics/composer/2.4/
Change-Id: I055f46b1eeba1d3e20d6a92a9d50a83e0f1ee694
When calling to setActiveConfigWithConstraints, the implementation may
need the client to send a refresh frame before the active config can be
changed. In addition, testing with a device with composer 2.4 revealed
few bugs which are fixed by this change.
Fix: 143775556
Test: adb shell data/nativetest64/VtsHalGraphicsComposerV2_4TargetTest/VtsHalGraphicsComposerV2_4TargetTest
Change-Id: Iafa1e85de60d99190d5d813f1d42924a62d94cc5
Command engine always uses V2.1 CommandWriterBase, change the creation of
command writer inside command engine so that we can use the right version of
command writer.
BUG: b/147365206
Test: build and boot.
Change-Id: I61fbbce19a44244725621e9a1cef750bd4c78b4d
All required interfaces must be able to initialized, make sure passthrough
fails if the requirement is not met.
BUG: b/147365206
Test: boot
Change-Id: I148655bc5392b1c0bc45f33ea3d2020824bb6f4b
Add a callback to notify the client it should retry a
setActiveConfigWithConstraints call in case of SEAMLESS_NOT_POSSIBLE error.
Test: rev up composer to 2.4
Bug: 141329414
Change-Id: I6176638fde937e3916f58cc577d50cb755997c61
The original test case was only tested with bad display. The patch adds the
test case with normal displays.
Test: atest VtsHalGraphicsComposerV2_4TargetTest
Bug: 145566569
Change-Id: Id6a25ebc1f63092194ca1d8837cd450e5bc19def
Also extending the passthrough interface with setLowLatencyMode
and setContentType.
Bug:132731049
Test: make -j$(nproc)
Test: vts-tradefed run vts-hal --skip-device-info -m
VtsHalGraphicsComposerV2_4Target
Change-Id: I14dd61a3d9c60fdb06f9028ccd6f49797267e9f4
Based on Forrest run with earlier base build 6069142, the build with this submission failed apct/bluetooth/instrumentation_test
Reason for revert: Break tests
Change-Id: I4f59a7e3d5c5f401a260f3c8ab3457cf4e2fe386
Also extending the passthrough interface with setLowLatencyMode
and setContentType.
Bug:132731049
Test: make -j$(nproc)
Test: vts-tradefed run vts-hal --skip-device-info -m
VtsHalGraphicsComposerV2_4Target
Change-Id: I0ec5a5be796700074bdd8cac75cd67f96cd5df58
Composer 2.4 extends vsync callback by exposing a new vsync_2_4
callback. Both should not be registered at the same time.
Test: rev up composer to 2.4 and test refresh rate switching
Bug: 141329414
Change-Id: If41f45cab21ed16a7b2f36b3b95fb3591b04c900
IMapper 4.0 does support lockYCbCr functionality through lock and
BufferMetadata getters. However, we will wait to add the support in
one central gralloc library. For now just stub out the call so
there aren't any compiler errors.
Bug: 141631415
Test: Compiles
Change-Id: I9d2f70c87412f8ac2114db85eb6dc01539876e2b
- getDisplayConfigs is required to remain backward compatible with
composer@2.3 and older versions. To keep it that way a new config
attribute was added to group similar configs.
- getSupportedDisplayVsyncPeriods was removed as the vsync period
can be obtained by getDisplayAttribute for each individual config.
- renamed setActiveConfigAndVsyncPeriod -> setActiveConfigWithConstraints
to better describe what this function does.
- added VsyncPeriodChangeTimeline and onVsyncPeriodTimingChanged
With these changes composer is expected to return all possible
configurations when getDisplayConfigs (similar to older composer versions).
The client knows whether a seamless vsync period is possible or not by
the config groups.
Test: rev up composer to 2.4 and test refresh rate switching
Bug: 141329414
Change-Id: I1b4525f7e00b62cdaf260274abe4a6a5962ed1ab
Return BAD_DISPLAY or BAD_PARAMETER as expected.
Bug: 140158240
Test: run vts -m VtsHalGraphicsComposerV2_2Target
Change-Id: I0f5a7103bc134ed87984bd13248aee7597a95d68
SET_LAYER_PER_FRAME_METADATA_BLOBS command length
was calculated wrong. With wrong length
it's impossible to read current and any next commands right.
Test: manual
Bug: 144367999
Change-Id: I09a23bbf507996972ce7de0634f8866bbb13152a
Signed-off-by: o.v.yevtushenko <o.v.yevtushenko@globallogic.com>
IEEE 754 single precision cannot precisely represent the value of large
32b integrals. Accept the imprecision from implicit casts by making the
casts explicit.
One case is comparing the value before and after converting a float to
an int32_t and back, the other is used when printing a value.
Bug: 139945549
Test: mm
Change-Id: I23375d73cd16be0741defc6395046bd3b22d1335
Merged-in: Id30edce2cd29c0f9c24cd52ba5fb33f7c56a3100
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
IEEE 754 single precision cannot precisely represent the value of large
32b integrals. Accept the imprecision from implicit casts by making the
casts explicit.
One case is comparing the value before and after converting a float to
an int32_t and back, the other is used when printing a value.
Bug: 139945549
Test: mm
Change-Id: Id30edce2cd29c0f9c24cd52ba5fb33f7c56a3100
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
