ImportError:/mmcv/_ext.cpython-39-x86_64-linux-gnu.so: undefined symbol: _ZNK3c1010TensorImpl36is_contiguous_nondefault_policy_implENS_12MemoryFormatE

[bearatom]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• I have read the FAQ documentation but cannot get the expected help.
• The bug has not been fixed in the latest version (master) or latest version (1.x).

Task

I'm using the official example scripts/configs for the officially supported tasks/models/datasets.

Branch

master branch https://github.com/open-mmlab/mmediting

Environment

I ran the collect_env.py and got these errors:

Traceback (most recent call last):
File "/data/likang/mmediting/mmedit/utils/collect_env.py", line 17, in
for name, val in collect_env().items():
File "/data/likang/mmediting/mmedit/utils/collect_env.py", line 10, in collect_env
env_info = collect_base_env()
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/utils/env.py", line 110, in collect_env
from mmcv.ops import get_compiler_version, get_compiling_cuda_version
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/ops/init.py", line 2, in
from .active_rotated_filter import active_rotated_filter
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/ops/active_rotated_filter.py", line 8, in
ext_module = ext_loader.load_ext(
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/utils/ext_loader.py", line 13, in load_ext
ext = importlib.import_module('mmcv.' + name)
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/importlib/init.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
ImportError: /home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/_ext.cpython-39-x86_64-linux-gnu.so: undefined symbol: _ZNK3c1010TensorImpl36is_contiguous_nondefault_policy_implENS_12MemoryFormatE

Reproduces the problem - code sample

input:
python demo/inpainting_demo.py configs/inpainting/deepfillv2/deepfillv2_256x256_8x2_places.py checkpoint/deepfillv2_256x256_8x2_places_20200619-10d15793.pth tests/data/image/1.png tests/data/image/1_mask.png tests/data/savepath/1_output.png

Reproduces the problem - command or script

I'm using the official inpainting_demo.py and deepfillv2_256256_82_places.py.

Reproduces the problem - error message

Traceback (most recent call last):
File "/data/likang/mmediting/demo/inpainting_demo.py", line 7, in
from mmedit.apis import init_model, inpainting_inference
File "/data/likang/mmediting/mmedit/apis/init.py", line 4, in
from .matting_inference import init_model, matting_inference
File "/data/likang/mmediting/mmedit/apis/matting_inference.py", line 8, in
from mmedit.models import build_model
File "/data/likang/mmediting/mmedit/models/init.py", line 2, in
from .backbones import * # noqa: F401, F403
File "/data/likang/mmediting/mmedit/models/backbones/init.py", line 14, in
from .sr_backbones import (EDSR, LIIFEDSR, LIIFRDN, RDN, SRCNN, BasicVSRNet,
File "/data/likang/mmediting/mmedit/models/backbones/sr_backbones/init.py", line 3, in
from .basicvsr_pp import BasicVSRPlusPlus
File "/data/likang/mmediting/mmedit/models/backbones/sr_backbones/basicvsr_pp.py", line 6, in
from mmcv.ops import ModulatedDeformConv2d, modulated_deform_conv2d
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/ops/init.py", line 2, in
from .active_rotated_filter import active_rotated_filter
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/ops/active_rotated_filter.py", line 8, in
ext_module = ext_loader.load_ext(
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/utils/ext_loader.py", line 13, in load_ext
ext = importlib.import_module('mmcv.' + name)
File "/home/likang_2/anaconda3/envs/pytorch/lib/python3.9/importlib/init.py", line 127, in import_module
return _bootstrap._gcd_import(name[level:], package, level)
ImportError: /home/likang_2/anaconda3/envs/pytorch/lib/python3.9/site-packages/mmcv/_ext.cpython-39-x86_64-linux-gnu.so: undefined symbol: _ZNK3c1010TensorImpl36is_contiguous_nondefault_policy_implENS_12MemoryFormatE

Additional information

I use cuda11.3 torch1.12.0+cu113 on ubuntu20.04.

[Z-Fran]

same to #1718