From bc8686d3659566f3fe2942c7afc66e654383292e Mon Sep 17 00:00:00 2001
From: Ye Wang <52801275+wangyems@users.noreply.github.com>
Date: Wed, 1 Feb 2023 10:45:48 -0800
Subject: [PATCH] Fix prefast warning 11263 (#14508)

### Description
<!-- Describe your changes. -->
https://dev.azure.com/aiinfra/ONNX%20Runtime/_workitems/edit/11263/


### Motivation and Context
<!-- - Why is this change required? What problem does it solve?
- If it fixes an open issue, please link to the issue here. -->
---
 .../contrib_ops/cuda/transformers/sampling_cuda_helper.h       | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/onnxruntime/contrib_ops/cuda/transformers/sampling_cuda_helper.h b/onnxruntime/contrib_ops/cuda/transformers/sampling_cuda_helper.h
index f9046ec049..d82648890f 100644
--- a/onnxruntime/contrib_ops/cuda/transformers/sampling_cuda_helper.h
+++ b/onnxruntime/contrib_ops/cuda/transformers/sampling_cuda_helper.h
@@ -140,8 +140,9 @@ Status Sample(AllocatorPtr& allocator,
   // Multinomial sampling
   gsl::span<float>& d_sampled = sampling_state->d_sampled;
   gsl::span<float>& h_sampled_all = sampling_state->h_sampled_all;
+  size_t sample_offset = (static_cast<size_t>(step) - 1) * static_cast<size_t>(parameters->batch_size);
   CUDA_RETURN_IF_ERROR(cudaMemcpyAsync(d_sampled.data(),
-                                       h_sampled_all.data() + (step - 1) * parameters->batch_size,
+                                       h_sampled_all.data() + sample_offset,
                                        sizeof(float) * parameters->batch_size,
                                        cudaMemcpyHostToDevice,
                                        cuda_stream));