Fix broken ET CI on macos builds (#1458)

Summary:
Pull Request resolved: https://github.com/pytorch/executorch/pull/1458

OSS CI for cmake on macos was building when trying to make the xnn_executor_runner

```
  Undefined symbols for architecture arm64:
    "torch::executor::get_out_numel(at::Tensor const&, std::__1::optional<c10::ArrayRef<long long> > const&)", referenced from:
        torch::executor::native::quantize_per_channel_out(at::Tensor const&, at::Tensor const&, at::Tensor const&, long long, long long, long long, c10::ScalarType, at::Tensor&) in op_quantize.cpp.o
    "torch::executor::get_init_index(at::Tensor const&, std::__1::optional<c10::ArrayRef<long long> > const&, unsigned long)", referenced from:
        torch::executor::native::quantize_per_channel_out(at::Tensor const&, at::Tensor const&, at::Tensor const&, long long, long long, long long, c10::ScalarType, at::Tensor&) in op_quantize.cpp.o
    "torch::executor::check_dim_list_is_valid(at::Tensor const&, std::__1::optional<c10::ArrayRef<long long> > const&)", referenced from:
        torch::executor::native::quantize_per_channel_out(at::Tensor const&, at::Tensor const&, at::Tensor const&, long long, long long, long long, c10::ScalarType, at::Tensor&) in op_quantize.cpp.o
    "torch::executor::get_reduced_dim_product(at::Tensor const&, std::__1::optional<c10::ArrayRef<long long> > const&)", referenced from:
        torch::executor::native::quantize_per_channel_out(at::Tensor const&, at::Tensor const&, at::Tensor const&, long long, long long, long long, c10::ScalarType, at::Tensor&) in op_quantize.cpp.o
  ld: symbol(s) not found for architecture arm64
```

I believe the issue is related to the quantized kernels added by larryliu0820 in https://github.com/pytorch/executorch/pull/1323/

Creating this Pull Request to see if the CI passes now

Reviewed By: digantdesai

Differential Revision: D52212686

fbshipit-source-id: 2fa60f91a054d31509cd951b146f6dc238e87495