Files
clang-p2996/mlir/lib/ExecutionEngine/RocmRuntimeWrappers.cpp
max 8f7c8a6ea7 Add gpu::HostUnregisterOp
Without explicitly unregistering you will get

```
'cuMemHostRegister(ptr, sizeBytes, 0)' failed with 'CUDA_ERROR_HOST_MEMORY_ALREADY_REGISTERED'
```

in CUDA (for example) after repeated runs (e.g., during benchmarking the same kernel).

Reviewed By: ftynse

Differential Revision: https://reviews.llvm.org/D147277
2023-04-06 15:07:12 -05:00

7.3 KiB