Support int4 weight in quantized matmul/linear

[lsy323]

int4 weight can be enabled by torch.ops.xla.quantized_matmul(x, weight, weight_scaler, int4_weight=True), XlaQuantizedLinear(...,int4_weight=True)

The matmul w/ int4 workflow is:

1. The int4 weight is stored in int8 container (unpacked)
2. During HLO lowering, xla::Literal will be created for the int4 weights
3. F.linear on the activation and int4 weight

Original plan was to pack int4 values in int8 container, and do reinterpret cast, but reinterpret cast does't work on TPU now.

Test:
Added tests for quantized op and linear module.

[lsy323]

Removed pack/unpack logic and test since not used now.