fix paddle.linalg.matrix_rank

[DongBaiYue]

API：paddle.linalg.matrix_rank
torch执行报错，添加到tester/api_config/torch_error_skip.txt

[paddle-bot]

Thanks for your contribution!

[DongBaiYue]

Torch报错

[torch error] paddle.linalg.matrix_rank(x=Tensor([4, 1073741824],"float32"), tol=None, hermitian=False, ) 
 cusolver error: CUSOLVER_STATUS_INVALID_VALUE, when calling `cusolverDnSgesvdj_bufferSize(handle, jobz, econ, m, n, A, lda, S, U, ldu, V, ldv, lwork, params)`. This error may appear if the input matrix contains NaN. If you keep seeing this error, you may use `torch.backends.cuda.preferred_linalg_library()` to try linear algebra operators with other supported backends. See https://pytorch.org/docs/stable/backends.html#torch.backends.cuda.preferred_linalg_library
Traceback (most recent call last):
  File "/data/PaddleAPITest/tester/accuracy.py", line 93, in test
    exec(code.core_compiled, exec_globals, exec_locals)
  File "<string>", line 8, in <module>
  File "/usr/local/lib/python3.10/dist-packages/torch/utils/_device.py", line 104, in __torch_function__
    return func(*args, **kwargs)
torch._C._LinAlgError: cusolver error: CUSOLVER_STATUS_INVALID_VALUE, when calling `cusolverDnSgesvdj_bufferSize(handle, jobz, econ, m, n, A, lda, S, U, ldu, V, ldv, lwork, params)`. This error may appear if the input matrix contains NaN. If you keep seeing this error, you may use `torch.backends.cuda.preferred_linalg_library()` to try linear algebra operators with other supported backends. See **https://pytorch.org/docs/stable/backends.html#torch.backends.cuda.preferred_linalg_library**

[cangtianhuang]

LGTM