onnxruntime
488544b7 - [MIGraphX EP] Fix CopyTensorAsync and add guards for stream sync CopyTensors (#16787)

Commit
2 years ago
[MIGraphX EP] Fix CopyTensorAsync and add guards for stream sync CopyTensors (#16787) Add compile guards to gate functionality based on MIGRAPHX_STREAM_SYNC for adding the following - remove excess hipStreamSyncronize to nullstream on CopyTensor calls - Add proper call for stream synchronized CopyTensorAsync for DeviceToHost case Without this change subsequent CopyTensorAsync() calls will fail for cards that don't use pinned memory thus causing hipMemcpy() calls to occur before certain kernel operations occur. ![image](https://github.com/microsoft/onnxruntime/assets/107195283/4915c18a-fb2d-40c9-a50e-a7c6613c324b) becomes ![image](https://github.com/microsoft/onnxruntime/assets/107195283/f661acf4-e2af-4c9a-b26a-30fca339cf1d) --------- Co-authored-by: Ted Themistokleous <tthemist@amd.com>
Parents
Loading