[MPS] Skip gather/blit calls in case of strided output (#94260)
Skip gather/blit calls in case of strided output - this prevents:
- allocating additional memory for the output
- additional transpose for both the input and output
Fixes:
```
x = torch.rand((256,10), device='mps')
x = x.permute(1,0)
x.exp()
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/94260
Approved by: https://github.com/razarmehr
