cupy: WSL cuda 11.2: CUDARuntimeError: cudaErrorInvalidValue: invalid argument

I installed cupy on WSL, and get a runtime error when trying to use cupy.

1. Kernel

» uname -a
Linux DESKTOP-1OG5J92 4.19.128-microsoft-standard #1 SMP Tue Jun 23 12:58:10 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux
» cat /etc/os-release
NAME="Ubuntu"
VERSION="20.04.1 LTS (Focal Fossa)"

1. Cupy

» python -c 'import cupy; cupy.show_config()'
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "/home/plate/.local/lib/python3.8/site-packages/cupy/__init__.py", line 866, in show_config
    _sys.stdout.write(str(_cupyx.get_runtime_info()))
  File "/home/plate/.local/lib/python3.8/site-packages/cupyx/_runtime.py", line 240, in get_runtime_info
    return _RuntimeInfo()
  File "/home/plate/.local/lib/python3.8/site-packages/cupyx/_runtime.py", line 138, in __init__
    self.cublas_version = _eval_or_error(
  File "/home/plate/.local/lib/python3.8/site-packages/cupyx/_runtime.py", line 49, in _eval_or_error
    return func()
  File "/home/plate/.local/lib/python3.8/site-packages/cupyx/_runtime.py", line 140, in <lambda>
    cupy.cuda.device.get_cublas_handle()),
  File "cupy/cuda/device.pyx", line 46, in cupy.cuda.device.get_cublas_handle
  File "cupy/cuda/device.pyx", line 47, in cupy.cuda.device.get_cublas_handle
  File "cupy/cuda/device.pyx", line 29, in cupy.cuda.device._get_device
  File "cupy_backends/cuda/api/runtime.pyx", line 275, in cupy_backends.cuda.api.runtime.getDevice
  File "cupy_backends/cuda/api/runtime.pyx", line 247, in cupy_backends.cuda.api.runtime.check_status
cupy_backends.cuda.api.runtime.CUDARuntimeError: cudaErrorInvalidValue: invalid argument

1. CUDA

» nvidia-smi.exe
Fri Apr  9 13:13:07 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 465.12       Driver Version: 465.12       CUDA Version: 11.2     |
|-------------------------------+----------------------+----------------------+
| GPU  Name            TCC/WDDM | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  GeForce GTX 165... WDDM  | 00000000:02:00.0 Off |                  N/A |
| N/A   40C    P8     3W /  N/A |    132MiB /  4096MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+


+-----------------------------------------------------------------------------+
| Processes:
 |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found
 |
 
 » apt list --installed | grep cuda

WARNING: apt does not have a stable CLI interface. Use with caution in scripts.

cuda-command-line-tools-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-compiler-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-cudart-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-cudart-dev-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-cuobjdump-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-cupti-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-cupti-dev-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-cuxxfilt-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-documentation-11-2/unknown,unknown,now 11.2.154-1 amd64 [installed,automatic]
cuda-driver-dev-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-gdb-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-libraries-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-libraries-dev-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-memcheck-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nsight-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nsight-compute-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-nsight-systems-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-nvcc-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvdisasm-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvml-dev-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvprof-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvprune-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvrtc-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvrtc-dev-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvtx-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-nvvp-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-samples-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-sanitizer-11-2/unknown,unknown,now 11.2.152-1 amd64 [installed,automatic]
cuda-toolkit-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed]
cuda-tools-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]
cuda-visual-tools-11-2/unknown,unknown,now 11.2.2-1 amd64 [installed,automatic]

I also verified my environment using pytorch, and it looks like it’s working, torch.cuda.is_available() is True