llvm
llvm copied to clipboard
Published
20 hours ago •
intel
intel
whole_update_double_buffer.cpp failing in CI on CUDA
Describe the bug
FAIL: SYCL :: Graph/Update/Explicit/whole_update_double_buffer.cpp (1108 of 2032)
******************** TEST 'SYCL :: Graph/Update/Explicit/whole_update_double_buffer.cpp' FAILED ********************
Exit Code: -6
Command Output (stdout):
--
# RUN: at line 1
/__w/llvm/llvm/toolchain/bin//clang++ -fsycl -fsycl-targets=nvptx64-nvidia-cuda /__w/llvm/llvm/llvm/sycl/test-e2e/Graph/Update/Explicit/whole_update_double_buffer.cpp -o /__w/llvm/llvm/build-e2e/Graph/Update/Explicit/Output/whole_update_double_buffer.cpp.tmp.out
# executed command: /__w/llvm/llvm/toolchain/bin//clang++ -fsycl -fsycl-targets=nvptx64-nvidia-cuda /__w/llvm/llvm/llvm/sycl/test-e2e/Graph/Update/Explicit/whole_update_double_buffer.cpp -o /__w/llvm/llvm/build-e2e/Graph/Update/Explicit/Output/whole_update_double_buffer.cpp.tmp.out
# note: command had no output on stdout or stderr
# RUN: at line 2
env SYCL_PI_CUDA_ENABLE_IMAGE_SUPPORT=1 ONEAPI_DEVICE_SELECTOR=cuda:gpu /__w/llvm/llvm/build-e2e/Graph/Update/Explicit/Output/whole_update_double_buffer.cpp.tmp.out
# executed command: env SYCL_PI_CUDA_ENABLE_IMAGE_SUPPORT=1 ONEAPI_DEVICE_SELECTOR=cuda:gpu /__w/llvm/llvm/build-e2e/Graph/Update/Explicit/Output/whole_update_double_buffer.cpp.tmp.out
# .---command stdout------------
# | Unexpected value at index 256 for DataB2: 1305 (got) vs 1574 (expected)
# `-----------------------------
# .---command stderr------------
# | whole_update_double_buffer.cpp.tmp.out: /__w/llvm/llvm/llvm/sycl/test-e2e/Graph/Update/Explicit/../../Inputs/whole_update_double_buffer.cpp:86: int main(): Assertion `check_value(i, ReferenceB2[i], HostDataB2[i], "DataB2")' failed.
# `-----------------------------
# error: command failed with exit status: -6
--
********************
Slowest Tests:
To reproduce
No response
Environment
No response
Additional context
No response
This came up in https://github.com/intel/llvm/pull/13719#issuecomment-2103009985 too, and has also been reported internally as failing on Windows runs. Will disable the test on CUDA while @intel/sycl-graphs-reviewers investigate.
