[ET-VK][ez] Handle zero-element tensors when building Vulkan graph

Pull Request resolved: #11984

## Changes

As title.

## Motivation

Some models may have parameter tensors which are zero-shape (i.e. no elements). In this case, trying to serialize the tensor data will result in a null pointer exception.
ghstack-source-id: 292864904

Differential Revision: [D77281492](https://our.internmc.facebook.com/intern/diff/D77281492/)

Author: SS-JIA

backends/vulkan/serialization/vulkan_graph_serialize.py

@@ -191,19 +191,23 @@ def serialize_constant_tensors(
 
     current_offset = len(raw_bytes)
     for tensor in const_tensors:
-        array_type = ctypes.c_char * tensor.untyped_storage().nbytes()
-        array = ctypes.cast(
-            tensor.untyped_storage().data_ptr(),
-            ctypes.POINTER(array_type),
-        ).contents
-
-        tensor_bytes = bytes(array)
-        # Pad the tensor bytes to the next 16 byte boundary
-        raw_bytes += tensor_bytes
-        raw_bytes += b"\x00" * padding_required(len(tensor_bytes))
-
-        vk_graph.constants.append(VkBytes(current_offset, len(tensor_bytes)))
-        current_offset += aligned_size(len(tensor_bytes))
+        if tensor.numel() == 0:
+            vk_graph.constants.append(VkBytes(current_offset, 0))
+            continue
+        else:
+            array_type = ctypes.c_char * tensor.untyped_storage().nbytes()
+            array = ctypes.cast(
+                tensor.untyped_storage().data_ptr(),
+                ctypes.POINTER(array_type),
+            ).contents
+
+            tensor_bytes = bytes(array)
+            # Pad the tensor bytes to the next 16 byte boundary
+            raw_bytes += tensor_bytes
+            raw_bytes += b"\x00" * padding_required(len(tensor_bytes))
+
+            vk_graph.constants.append(VkBytes(current_offset, len(tensor_bytes)))
+            current_offset += aligned_size(len(tensor_bytes))
 
 
 def serialize_custom_shaders(