remove special casing for sparse CSR shape comparison
Fixes https://github.com/pytorch/pytorch/pull/74264#discussion_r845780445.
The shape check works with or without the extras added in #74264.
```py
>>> a = torch.rand(2, 2).to_sparse_csr()
>>> b = torch.rand(2, 3).to_sparse_csr()
>>> torch.testing.assert_close(a, b)
AssertionError: The values for attribute 'shape' do not match: torch.Size([2, 2]) != torch.Size([2, 3]).
```
Tensor comparison is split into two parts:
1. Attribute comparison.
2. Value comparison.
https://github.com/pytorch/pytorch/blob/bcf6974c207ac0339bfb8bdfdb0b0ec348f7a22f/torch/testing/_comparison.py#L611-L616
The attribute comparison happens in
https://github.com/pytorch/pytorch/blob/bcf6974c207ac0339bfb8bdfdb0b0ec348f7a22f/torch/testing/_comparison.py#L618
The check for the matching shape
https://github.com/pytorch/pytorch/blob/bcf6974c207ac0339bfb8bdfdb0b0ec348f7a22f/torch/testing/_comparison.py#L647-L648
is one of the few checks that cannot be disabled through keyword arguments. Thus, there is no need for this check in `_compare_sparse_csr_values` since the comparison will fail before if the shapes mismatch.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/75593
Approved by: https://github.com/cpuhrsch
