diff options
| author | Nathalie Le Clair <lcnathalie@google.com> | 2021-05-26 10:10:07 +0200 |
|---|---|---|
| committer | Nathalie Le Clair <lcnathalie@google.com> | 2021-05-26 12:39:20 +0000 |
| commit | d9be028d41b3ca81026016c4ddd2312ec3298fdb (patch) | |
| tree | a2d06352a3b7ff5b8cdf59ad3df97afdf2c5efa7 /rs/java/android/renderscript/ProgramFragmentFixedFunction.java | |
| parent | d261ad7bfe4ed784755cc27bb1cb7bceac2dc3c5 (diff) | |
Fix OneTouchPlayActionTest failure on playback devices
Playback devices have HdmiProperties.cec_device_types() set to 'playback' and as a consequence, initializing HdmiControlService will automatically initialize a local playback device. This local playback device interferes with the playback device that is added in tests. Concretely, in the failing test the <Report Power Status> messages was being fed to the local playback device, instead of the test one.
To resolve this, HdmiControlService is initialized with empty local
devices, practically ignoring HdmiProperties.cec_device_types().
This will also avoid interference with other local devices while running these unit tests on f.e. a TV panel device.
Also, isActiveSource() doesn't guarantee that oneTouchPlay() was executed, since the active source might have been set in a different method. We should additionally check if the correct CEC messages were sent.
Bug: 187909454
Test: atest OneTouchPlayActionTest on cuttlefish and ADT3
Change-Id: I4599a240fa8a32f47a23c1f183d41e0d280152c3
Diffstat (limited to 'rs/java/android/renderscript/ProgramFragmentFixedFunction.java')
0 files changed, 0 insertions, 0 deletions