[quant][core][feature] Implemented masked_fill for QuantizedCPU tensors
Summary:
Previously, there was no suport for masked_fill for quantized tensors.
This PR introduces the feature for quantized CPU tensors. A
corresponding test case was added to test the impl.
Most of the implementation is copied over from the fp kernel and adapted
for quantized tensors.
TODO: implement masked_fill for QuantizedCUDA tensors.
Test Plan:
```
python test/test_quantization.py -k test_qtensor_masked_fill
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/78368
Approved by: https://github.com/vkuzo