[Bug] NPU环境:RuntimeError: xxx_impl: implementation for device xla:1 not found.

[GenerallyCovetous]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• The bug has not been fixed in the latest version(https://github.com/open-mmlab/mmcv).

Environment

OrderedDict([('sys.platform', 'linux'), ('Python', '3.7.10 | packaged by conda-forge | (default, Oct 13 2021, 22:05:51) [GCC 9.4.0]'), ('CUDA available', False), ('MUSA available', False), ('numpy_random_seed', 2147483648), ('GCC', 'gcc (GCC) 7.3.0'), ('PyTorch', '1.11.0a0+gitbc2c6ed'), ('PyTorch compiling details', 'PyTorch built with:\n - GCC 7.5\n - C++ Version: 201402\n - OpenMP 201511 (a.k.a. OpenMP 4.5)\n - LAPACK is enabled (usually provided by MKL)\n - NNPACK is enabled\n - CPU capability usage: NO AVX\n - Build settings: BLAS_INFO=generic, BUILD_TYPE=Release, CXX_COMPILER=/opt/buildtools/gcc-7.5.0/bin/c++, CXX_FLAGS= -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOCUPTI -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-sign-compare -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-unused-local-typedefs -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -DMISSING_ARM_VST1 -DMISSING_ARM_VLD1 -Wno-stringop-overflow, LAPACK_INFO=generic, TORCH_VERSION=1.11.0, USE_CUDA=OFF, USE_CUDNN=OFF, USE_EIGEN_FOR_BLAS=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=OFF, USE_MKLDNN=OFF, USE_MPI=OFF, USE_NCCL=OFF, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, \n'), ('TorchVision', '0.12.0'), ('OpenCV', '4.3.0'), ('MMEngine', '0.10.3'), ('MMCV', '2.0.1'), ('MMCV Compiler', 'GCC 7.3'), ('MMCV CUDA Compiler', 'not available')])

Reproduces the problem - code sample

python tools/train.py configs/kie/sdmgr/sdmgr_unet16_60e_wildreceipt.py

Reproduces the problem - command or script

cd mmocr-1.0.1
python tools/train.py configs/kie/sdmgr/sdmgr_unet16_60e_wildreceipt.py

Reproduces the problem - error message

mmocr内部模型sdmgr:
Traceback (most recent call last):
File "tools/train-Copy1.py", line 118, in
main()
File "tools/train-Copy1.py", line 114, in main
runner.train()
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmengine/runner/runner.py", line 1777, in train
model = self.train_loop.run() # type: ignore
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmengine/runner/loops.py", line 96, in run
self.run_epoch()
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmengine/runner/loops.py", line 112, in run_epoch
self.run_iter(idx, data_batch)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmengine/runner/loops.py", line 129, in run_iter
data_batch, optim_wrapper=self.runner.optim_wrapper)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmengine/model/base_model/base_model.py", line 114, in train_step
losses = self._run_forward(data, mode='loss') # type: ignore
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmengine/model/base_model/base_model.py", line 361, in _run_forward
results = self(**data, mode=mode)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmocr/models/kie/extractors/sdmgr.py", line 120, in forward
return self.loss(inputs, data_samples, **kwargs)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmocr/models/kie/extractors/sdmgr.py", line 144, in loss
[data_sample.gt_instances.bboxes for data_sample in data_samples])
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmocr/models/kie/extractors/sdmgr.py", line 83, in extract_feat
feats = self.maxpool(self.extractor([x], bbox2roi(gt_bboxes)))
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmdet/models/roi_heads/roi_extractors/single_level_roi_extractor.py", line 95, in forward
return self.roi_layers[0](feats[0], rois)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
return forward_call(*input, **kwargs)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmcv/ops/roi_align.py", line 211, in forward
self.sampling_ratio, self.pool_mode, self.aligned)
File "/home/ma-user/anaconda3/envs/PyTorch-1.11.0/lib/python3.7/site-packages/mmcv/ops/roi_align.py", line 101, in forward
aligned=ctx.aligned)
RuntimeError: roi_align_forward_impl: implementation for device xla:1 not found.
类似还有drrg模型:
RuntimeError: roi_align_rotated_forward_impl: implementation for device xla:1 not found.
mask-rcnn模型:
RuntimeError: nms_impl: implementation for device xla:1 not found.

Additional information

dpnet、dpnetpp、master等模型可以正常进行训练，import mmcv; import mmcv.ops也没有报错，但是在训练上述报错模型的时候就不行了

[GenerallyCovetous]

Pytorch version: 1.11.0 Torch_npu version: 1.11.0.post1.dev20230719 CANN version: 6.3.RC2

[zhouzaida]

Some operators have not yet provided an implementation for the NPU.