test passing a const uninitialized_buffer to launch

cudax/test/containers/uninitialized_buffer.cu

@@ -168,13 +168,34 @@ __global__ void kernel(_CUDA_VSTD::span<int> data)
   data[0] = 42;
 }
 
+__global__ void const_kernel(_CUDA_VSTD::span<const int> data)
+{
+  // Touch the memory to be sure it's accessible
+  CUDAX_CHECK(data.size() == 1024);
+}
+
 TEST_CASE("uninitialized_buffer is usable with cudax::launch", "[container]")
 {
-  const int grid_size = 4;
-  cudax::uninitialized_buffer<int, ::cuda::mr::device_accessible> buffer{cuda::mr::device_memory_resource{}, 1024};
-  auto dimensions = cudax::make_hierarchy(cudax::grid_dims(grid_size), cudax::block_dims<256>());
+  SECTION("non-const")
+  {
+    const int grid_size = 4;
+    cudax::uninitialized_buffer<int, ::cuda::mr::device_accessible> buffer{cuda::mr::device_memory_resource{}, 1024};
+    auto dimensions = cudax::make_hierarchy(cudax::grid_dims(grid_size), cudax::block_dims<256>());
 
-  cudax::stream stream;
+    cudax::stream stream;
 
-  cudax::launch(stream, dimensions, kernel, buffer);
+    cudax::launch(stream, dimensions, kernel, buffer);
+  }
+
+  SECTION("const")
+  {
+    const int grid_size = 4;
+    const cudax::uninitialized_buffer<int, ::cuda::mr::device_accessible> buffer{
+      cuda::mr::device_memory_resource{}, 1024};
+    auto dimensions = cudax::make_hierarchy(cudax::grid_dims(grid_size), cudax::block_dims<256>());
+
+    cudax::stream stream;
+
+    cudax::launch(stream, dimensions, const_kernel, buffer);
+  }
 }