ROCm / rpp

AMD ROCm Performance Primitives (RPP) library is a comprehensive high-performance computer vision library for AMD processors with HIP/OpenCL/CPU back-ends.
https://rocm.docs.amd.com/projects/rpp/en/latest/
MIT License
55 stars 40 forks source link

Develop - Merge to master #308

Closed kiritigowda closed 8 months ago

kiritigowda commented 8 months ago

@r-abishek can you see why this is failing

The following tests FAILED:
      4 - rpp_qa_tests_tensor_hip_all (Timeout)
Errors while running CTest
Output from these tests are in: /var/jenkins_home/workspace/main_precheckin_rpp_develop/LNL4mFrrF/rpp/build/release/Testing/Temporary/LastTest.log
Use "--rerun-failed --output-on-failure" to re-run the failed cases verbosely.
make: *** [Makefile:91: test] Error 8
script returned exit code 2
kiritigowda commented 8 months ago

@r-abishek this PR is failing all tests for Tensor_voxel_hip

 ---------------------------------- Results of QA Test - Tensor_voxel_hip ----------------------------------
5: 
5: fused_multiply_add_scalar_Tensor_PKD3: FAILED
5: fused_multiply_add_scalar_Tensor_PLN3: FAILED
5: fused_multiply_add_scalar_Tensor_PLN1: FAILED
5: slice_Tensor_PKD3: FAILED
5: slice_Tensor_PLN3: FAILED
5: slice_Tensor_PLN1: FAILED
5: add_scalar_Tensor_PKD3: FAILED
5: add_scalar_Tensor_PLN3: FAILED
5: add_scalar_Tensor_PLN1: FAILED
5: subtract_scalar_Tensor_PKD3: FAILED
5: subtract_scalar_Tensor_PLN3: FAILED
5: subtract_scalar_Tensor_PLN1: FAILED
5: multiply_scalar_Tensor_PKD3: FAILED
5: multiply_scalar_Tensor_PLN3: FAILED
5: multiply_scalar_Tensor_PLN1: FAILED
5: 
5: -------------------------------------------------------------------
5: 
5: Final Results of Tests:
5:     - Total test cases including all subvariants REQUESTED = 15
5:     - Total test cases including all subvariants PASSED = 0
5: 
5: General information on Tensor voxel test suite availability:
5:     - Total augmentations supported in Tensor test suite = 5
5:     - Total augmentations with golden output QA test support = 5
5:     - Total augmentations without golden ouput QA test support (due to randomization involved) = 0
5: 
5: -------------------------------------------------------------------
r-abishek commented 8 months ago

@kiritigowda This issue might to be discussed on a call. This seems to be a RHEL-alone issue for all HIP - including Tensor voxel-image and Tensor pixel-image

All tests are passing on Ubuntu and SLES. Ubuntu screenshot from today's run below: image