Add Static Dispatch Kernels (#4927)

Summary:
X-link: facebookresearch/FBGEMM#1951


X-link: pytorch/pytorch#163676

Add a few missing static dispatch kernels for remote_ro.

Reviewed By: henryoier

Differential Revision: D83028503

Author: kqfu

fbgemm_gpu/include/fbgemm_gpu/sparse_ops.h

@@ -632,6 +632,12 @@ at::Tensor jagged_to_padded_dense(
     const c10::SymIntArrayRef max_lengths,
     const double padding_value);
 
+at::Tensor jagged_to_padded_dense_forward_cpu(
+    const at::Tensor& values,
+    const std::vector<at::Tensor>& offsets,
+    c10::SymIntArrayRef max_lengths,
+    const double padding_value);
+
 at::Tensor jagged_dense_elementwise_add(
     const at::Tensor& x_values,
     const std::vector<at::Tensor>& x_offsets,

fbgemm_gpu/src/jagged_tensor_ops/jagged_tensor_ops_cpu.cpp

@@ -959,6 +959,15 @@ Tensor jagged_acc_weights_and_counts_cpu(
 
 } // namespace
 
+at::Tensor jagged_to_padded_dense_forward_cpu(
+    const at::Tensor& values,
+    const std::vector<at::Tensor>& offsets,
+    c10::SymIntArrayRef max_lengths,
+    const double padding_value) {
+  return jagged_to_padded_dense_forward(
+      values, offsets, max_lengths, padding_value);
+}
+
 std::tuple<Tensor, Tensor> masked_select_jagged_1d(
     const Tensor& values,
     const Tensor& lengths,