Add the support of handle Bias being nullptr for torch.ops.quantized.fbgemm_linear (#22403)
Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/22403
- C10 Operator Registration (https://github.com/pytorch/pytorch/blob/master/aten/src/ATen/core/op_registration/op_registration.cpp) supports None type.
- ATen has None Tensor support, e.g., https://github.com/pytorch/pytorch/blob/master/aten/src/ATen/native/native_functions.yaml#L1078
Reviewed By: zafartahirov
Differential Revision: D16069522
fbshipit-source-id: 3acaec783fc138ff36b14ffc0582d0764be4ad34