Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[cuSOLVER] *trd* tests are failed with latest oneMKL and LLVM #231

Closed
yhmtsai opened this issue Sep 26, 2022 · 0 comments · Fixed by #498
Closed

[cuSOLVER] *trd* tests are failed with latest oneMKL and LLVM #231

yhmtsai opened this issue Sep 26, 2022 · 0 comments · Fixed by #498
Assignees

Comments

@yhmtsai
Copy link
Contributor

yhmtsai commented Sep 26, 2022

Summary

With our docker image ginkgohub/oneapi:cuda11.6, 16 *trd* test in cuSOLVER are failed on NVIDIA devices (I try TitanX and A100)

Environment

  • HW you use: A100 and TitanX
  • Backend library version: cuda 11.6.2
  • oneMKL version: 9abb8cf
  • Compiler version: e1794b668cd2e539672f500fcb0e6bcbc9766077

Steps to reproduce

use nvidia docker runner to run the image from the above summary.
cd /var/tmp/oneMKL
mkdir build_cusolver && cd build_cusolver

cmake -DCMAKE_CXX_COMPILER=clang++ -DENABLE_MKLCPU_BACKEND=OFF -DENABLE_MKLGPU_BACKEND=OFF \
-DENABLE_NETLIB_BACKEND=OFF -DENABLE_CUBLAS_BACKEND=OFF -DENABLE_CUSOLVER_BACKEND=ON \
-DENABLE_CURAND_BACKEND=OFF -DBUILD_FUNCTIONAL_TESTS=ON \
-DREF_LAPACK_ROOT=/var/tmp/lapack-release/build ..

make -j24
ctest -R trd --output-on-failure

Observed behavior

16 tests are failure

The following tests FAILED:
	177 - LAPACK/RT/Hetrd/HetrdAccuracyUsm.ComplexSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	178 - LAPACK/RT/Hetrd/HetrdAccuracyUsm.ComplexDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	179 - LAPACK/RT/Hetrd/HetrdAccuracyBuffer.ComplexSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	180 - LAPACK/RT/Hetrd/HetrdAccuracyBuffer.ComplexDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	323 - LAPACK/RT/Sytrd/SytrdAccuracyUsm.RealSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	324 - LAPACK/RT/Sytrd/SytrdAccuracyUsm.RealDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	325 - LAPACK/RT/Sytrd/SytrdAccuracyBuffer.RealSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	326 - LAPACK/RT/Sytrd/SytrdAccuracyBuffer.RealDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	575 - LAPACK/CT/Hetrd/HetrdAccuracyUsm.ComplexSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	576 - LAPACK/CT/Hetrd/HetrdAccuracyUsm.ComplexDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	577 - LAPACK/CT/Hetrd/HetrdAccuracyBuffer.ComplexSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	578 - LAPACK/CT/Hetrd/HetrdAccuracyBuffer.ComplexDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	721 - LAPACK/CT/Sytrd/SytrdAccuracyUsm.RealSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	722 - LAPACK/CT/Sytrd/SytrdAccuracyUsm.RealDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	723 - LAPACK/CT/Sytrd/SytrdAccuracyBuffer.RealSinglePrecision/NVIDIA_A100_SXM4_40GB (Failed)
	724 - LAPACK/CT/Sytrd/SytrdAccuracyBuffer.RealDoublePrecision/NVIDIA_A100_SXM4_40GB (Failed)

the corresponding testing output:
oneMKL_cuSOLVER_A100_error.txt
oneMKL_cuSOLVER_TitanX_error.txt

Expected behavior

all tests should be passed

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants