Update on "[ET-VK][10/n] copy node, aten.repeat"

1. Introduce a `CopyNode` for generic copy-with-offset operations.
2. `aten.repeat` on all dimensions.
2.1 Use `CopyNode` where possible.
2.2. Specialized `repeat_channel` shader to handle packings
3. Update codegen to support `Methods` variant only operations. Need a new route to trigger the dispatch.

Differential Revision: [D56499329](https://our.internmc.facebook.com/intern/diff/D56499329/)

[ghstack-poisoned]

Author: yipjustin

backends/vulkan/runtime/graph/ops/impl/Repeat.cpp

@@ -28,7 +28,8 @@ void check_args(
 
   int64_t in_dim = in.dim();
   VK_CHECK_COND(
-      in_dim == repeats.size(), "Input tensor dim size must match argument");
+      in_dim <= repeats.size(),
+      "Input tensor dim size must be not greater than the repeat argument's size");
 
   VK_CHECK_COND(
       dim_at<Dim4D::Width>(in.sizes()) * dim_at<Dim4D::Width>(repeats) ==
@@ -98,7 +99,6 @@ void add_repeat_channel_node(
   };
 
   auto shader = VK_KERNEL_FROM_STR(kernel_name);
-  // std::cout << "out tile size: " << shader.out_tile_size << std::endl;
 
   graph.execute_nodes().emplace_back(new ExecuteNode(
       graph,
@@ -110,7 +110,7 @@ void add_repeat_channel_node(
       // Parameter buffers
       {graph.create_params_buffer(repeat_channel_args)},
       // Specialization Constants
-      {}));
+      {SV(t_out->gpu_memory_layout_int())}));
 }
 
 void add_repeat_node(

backends/vulkan/test/op_tests/cases.py

@@ -415,6 +415,9 @@ def get_repeat_inputs():
             ((S1, S2, S2, S2), [1, 3, 1, 3]),
             ((S1, S2, S2, S2), [3, 3, 3, 3]),
             ((S1, S2, S2, S2), [3, 3, 1, 1]),
+            # Expanding cases
+            ((2, 3), [3, 1, 4]),
+            ((2, 3), [3, 3, 2, 4]),
         ]
     )
     test_suite.layouts = [