Support non-standard bools in CUDA mode (#79393)
Closes #54789
For the `fused_mode` kernel, this just uses `c10::load` but the
`apply_mode` function is a bit harder because it uses `thrust`.
Instead, I've added a second dedicated path for bool which also
only uses 2 thrust calls instead of the normal 6, by exploiting
the fact that bools only have two possible values.
In the following `timeit` benchmark which calls the `apply_mode`
version, I see execution time drop from 16.9 ms to 2.2 ms (which is
still terrible, but my main goal is fixing the bool handling).
```python
import torch a = torch.randint(
0, 2, size=(100, 4096), device='cuda', dtype=torch.bool)
%timeit a.mode(1)
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/79393
Approved by: https://github.com/ngimel
