Fix CUDA unified memory allocator

- add correct dealloc fence
- add error reporting

core/src/Cuda/Kokkos_CudaSpace.cpp

@@ -195,15 +195,20 @@ void *impl_allocate_common(const int device_id,
   static_assert(CUDART_VERSION >= 12020);
   if (arg_alloc_size) {  // cudaMemAdvise_v2 does not work with nullptr
     ptr = malloc(arg_alloc_size);
-    // TODO: what error do we want to throw here if it fails, i.e. ptr ==
-    // nullptr? One would thing cudaMemLocation{device_id,
+    if(ptr != nullptr) {
+    // One would think cudaMemLocation{device_id,
     // cudaMemLocationTypeDeivce} would work but it doesn't. I.e. the order of
     // members doesn't seem to be defined.
     cudaMemLocation loc;
     loc.id   = device_id;
     loc.type = cudaMemLocationTypeDevice;
     KOKKOS_IMPL_CUDA_SAFE_CALL(cudaMemAdvise_v2(
         ptr, arg_alloc_size, cudaMemAdviseSetPreferredLocation, loc));
+    } else {
+      // I think this is the most logical error to return unless we
+      // want a different mechanism for this code path
+      error_code = cudaErrorMemoryAllocation;
+    }
   }
 #elif (defined(KOKKOS_ENABLE_IMPL_CUDA_MALLOC_ASYNC) && CUDART_VERSION >= 11020)
   if (arg_alloc_size >= memory_threshold_g) {
@@ -373,6 +378,8 @@ void CudaSpace::impl_deallocate(
     KOKKOS_IMPL_CUDA_SAFE_CALL(cudaSetDevice(m_device));
     KOKKOS_IMPL_CUDA_SAFE_CALL(cudaFree(arg_alloc_ptr));
 #elif defined(KOKKOS_ENABLE_IMPL_CUDA_UNIFIED_MEMORY)
+    Impl::cuda_device_synchronize(
+          "Kokkos::Cuda: backend fence before unified memory free");
     free(arg_alloc_ptr);
 #elif (defined(KOKKOS_ENABLE_IMPL_CUDA_MALLOC_ASYNC) && CUDART_VERSION >= 11020)
     if (arg_alloc_size >= memory_threshold_g) {