From 12d91173c4478e0975771c06fd9d062a33c46339 Mon Sep 17 00:00:00 2001
From: Misha Chornyi <99709299+mc-nv@users.noreply.github.com>
Date: Sat, 11 Feb 2023 16:14:40 -0800
Subject: [PATCH] Add extra include to fix build w/ CUDA 12 (#14659)

Signed-off-by: Cliff Woolley <jwoolley@nvidia.com>

### Description
Including file to fix build w/CUDA 12



### Motivation and Context
It should allow users to compile against CUDA 12

Signed-off-by: Cliff Woolley <jwoolley@nvidia.com>
Co-authored-by: Cliff Woolley <jwoolley@nvidia.com>
---
 onnxruntime/contrib_ops/cuda/bert/attention_impl.cu | 1 +
 1 file changed, 1 insertion(+)

diff --git a/onnxruntime/contrib_ops/cuda/bert/attention_impl.cu b/onnxruntime/contrib_ops/cuda/bert/attention_impl.cu
index fcf8663735..972b2d4ade 100644
--- a/onnxruntime/contrib_ops/cuda/bert/attention_impl.cu
+++ b/onnxruntime/contrib_ops/cuda/bert/attention_impl.cu
@@ -28,6 +28,7 @@ limitations under the License.
 
 #include <cassert>
 #include <cuda_fp16.h>
+#include <cub/cub.cuh>
 #include "core/providers/cuda/cu_inc/common.cuh"
 #include "core/providers/cuda/cuda_common.h"
 #include "core/providers/cuda/shared_inc/fpgeneric.h"