Skip to content

Support tensorrt compiled in cuda13 #75695

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
swgu98 merged 1 commit into from
Oct 13, 2025
Merged

Support tensorrt compiled in cuda13 #75695

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 7 additions & 7 deletions paddle/fluid/inference/tensorrt/plugin/anchor_generator_op_plugin.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -163,7 +163,7 @@ size_t AnchorGeneratorPlugin::getWorkspaceSize(int max_batch_size) const
return 0;
}

#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
template <typename T>
__global__ void GenAnchors(T* out,
const T* aspect_ratios,
Expand Down Expand Up @@ -233,7 +233,7 @@ int AnchorGeneratorPlugin::enqueue_impl(int batch_size,
const T* aspect_ratios_device = static_cast<const T*>(aspect_ratios_device_);
const T* stride_device = static_cast<const T*>(stride_device_);
const T* variances_device = static_cast<const T*>(variances_device_);
#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
GenAnchors<T><<<gen_anchor_grid, block, 0, stream>>>(anchors,
aspect_ratios_device,
aspect_ratios_.size(),
Expand All @@ -258,7 +258,7 @@ int AnchorGeneratorPlugin::enqueue_impl(int batch_size,
offset_);
#endif
const int var_grid = (box_num_ * 4 + block - 1) / block;
#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
SetVariance<T><<<var_grid, block, 0, stream>>>(
vars, variances_device, variances_.size(), box_num_ * 4);
#else
Expand Down Expand Up @@ -592,7 +592,7 @@ int AnchorGeneratorPluginDynamic::enqueue_impl(
const T* aspect_ratios_device = static_cast<const T*>(aspect_ratios_device_);
const T* stride_device = static_cast<const T*>(stride_device_);
const T* variances_device = static_cast<const T*>(variances_device_);
#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
GenAnchors<T><<<gen_anchor_grid, block, 0, stream>>>(anchors,
aspect_ratios_device,
aspect_ratios_.size(),
Expand All @@ -617,7 +617,7 @@ int AnchorGeneratorPluginDynamic::enqueue_impl(
offset_);
#endif
const int var_grid = (box_num * 4 + block - 1) / block;
#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
SetVariance<T><<<var_grid, block, 0, stream>>>(
vars, variances_device, variances_.size(), box_num * 4);
#else
Expand Down Expand Up @@ -894,7 +894,7 @@ int PIRAnchorGeneratorPluginDynamic::enqueue_impl(
const T* aspect_ratios_device = static_cast<const T*>(aspect_ratios_device_);
const T* stride_device = static_cast<const T*>(stride_device_);
const T* variances_device = static_cast<const T*>(variances_device_);
#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
GenAnchors<T><<<gen_anchor_grid, block, 0, stream>>>(anchors,
aspect_ratios_device,
aspect_ratios_.size(),
Expand All @@ -919,7 +919,7 @@ int PIRAnchorGeneratorPluginDynamic::enqueue_impl(
offset_);
#endif
const int var_grid = (box_num * 4 + block - 1) / block;
#ifdef _WIN32
#if defined(CUDA_VERSION) && CUDA_VERSION >= 13000 || defined(_WIN32)
SetVariance<T><<<var_grid, block, 0, stream>>>(
vars, variances_device, variances_.size(), box_num * 4);
#else
Expand Down
Loading