dylanyunlong/project_6
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

[CCCL] 瘦身 + 补全: 移除 cudax/python/libcudacxx-tests 冗余文件, 新增 c2h 测试助手 + cmake 构建系统 + 8 个 CUDA thrust examples

Browse Source 
变更摘要:
- 删除: cudax/ (783 files, 7.2M) — 实验性组件，竞赛不需要
- 删除: python/ (226 files, 2.0M) — Python 绑定，竞赛不需要
- 删除: libcudacxx/{test,benchmarks,codegen,cmake,share} (4432 files, 31M)
  保留: libcudacxx/include/ (1463 headers, cuda::std 编译依赖)
- 新增: c2h/ (27 files) — CUB Catch2 测试辅助头文件，编译 243 个测试必需
- 新增: cmake/ (29 files) — CCCL 原生 CMake 构建系统
- 新增: thrust/examples/cuda/ (7 files) + cpp_integration/ (1 file)
  async_reduce, custom_temporary_allocation, explicit_cuda_stream,
  global_device_vector, range_view, unwrap_pointer, wrap_pointer, device

结果: cccl_upstream 从 74M→35M (瘦身 53%), 核心内容 100% 保留:
  27/27 tuning headers, 78 benchmarks, 243 tests,
  60 thrust examples, 18 CUB examples, 全部编译头文件
This commit is contained in:
muh-bot
2026-08-03 12:39:26 +00:00
parent a2a5dd8f00
commit 24ef6a91b5
5439 changed files with 0 additions and 719516 deletions
Download Patch File Download Diff File Expand all files Collapse all files

115
cccl_upstream/cudax/test/stf/threads/axpy-threads-2.cu
View File

@@ -1,115 +0,0 @@
//===----------------------------------------------------------------------===//
//
// Part of CUDASTF in CUDA C++ Core Libraries,
// under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// SPDX-FileCopyrightText: Copyright (c) 2022-2024 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//
/**
* @file
*
* @brief An AXPY kernel implemented with a task of the CUDA stream backend
*
*/
#include <cuda/experimental/__stf/stream/stream_ctx.cuh>
#include <mutex>
#include <thread>
using namespace cuda::experimental::stf;
static __global__ void cuda_sleep_kernel(long long int clock_cnt)
{
long long int start_clock = clock64();
long long int clock_offset = 0;
while (clock_offset < clock_cnt)
{
clock_offset = clock64() - start_clock;
}
}
void cuda_sleep(double ms, cudaStream_t stream)
{
int device;
cudaGetDevice(&device);
// cudaDevAttrClockRate: Peak clock frequency in kilohertz;
int clock_rate;
cudaDeviceGetAttribute(&clock_rate, cudaDevAttrClockRate, device);
long long int clock_cnt = (long long int) (ms * clock_rate);
cuda_sleep_kernel<<<1, 1, 0, stream>>>(clock_cnt);
}
__global__ void axpy(double a, slice<const double> x, slice<double> y)
{
int tid = blockIdx.x * blockDim.x + threadIdx.x;
int nthreads = gridDim.x * blockDim.x;
for (int i = tid; i < x.size(); i += nthreads)
{
y(i) += a * x(i);
}
}
double X0(int i)
{
return sin((double) i);
}
double Y0(int i)
{
return cos((double) i);
}
void mytask(stream_ctx ctx, int /*id*/, logical_data<slice<double>> lX)
{
// std::cout << "Thread " << id << " is executing.\n";
const size_t N = 16;
double alpha = 3.14;
auto lY = ctx.logical_data<double>(N);
ctx.task(lY.write())->*[](cudaStream_t, auto) {};
/* Compute Y = Y + alpha X */
for (size_t i = 0; i < 10; i++)
{
ctx.task(lX.read(), lY.rw())->*[&](cudaStream_t s, auto dX, auto dY) {
axpy<<<16, 128, 0, s>>>(alpha, dX, dY);
cuda_sleep(100.0, s);
};
}
}
int main()
{
stream_ctx ctx;
const size_t N = 16;
auto lX = ctx.logical_data<double>(N);
ctx.task(lX.write())->*[](cudaStream_t, auto) {};
std::vector<std::thread> threads;
// Launch 8 threads.
for (int i = 0; i < 10; ++i)
{
threads.emplace_back(mytask, ctx, i, lX);
}
// Wait for all threads to complete.
for (auto& th : threads)
{
th.join();
}
ctx.task(lX.rw())->*[&](cudaStream_t s, auto) {
cuda_sleep(100.0, s);
};
ctx.finalize();
}

85
cccl_upstream/cudax/test/stf/threads/axpy-threads-graph-capture.cu
View File

@@ -1,85 +0,0 @@
//===----------------------------------------------------------------------===//
//
// Part of CUDASTF in CUDA C++ Core Libraries,
// under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// SPDX-FileCopyrightText: Copyright (c) 2022-2025 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//
/**
* @file
*
* @brief Ensure a graph_ctx can be used concurrently
*
*/
#include <cuda/experimental/stf.cuh>
#include <mutex>
#include <thread>
using namespace cuda::experimental::stf;
__global__ void axpy(slice<int> y, slice<const int> x, int a)
{
int tid = blockIdx.x * blockDim.x + threadIdx.x;
int nthreads = gridDim.x * blockDim.x;
for (int i = tid; i < x.size(); i += nthreads)
{
y(i) += a * x(i);
}
}
void mytask(graph_ctx ctx, int /*id*/)
{
const size_t N = 16;
int alpha = 3;
auto lX = ctx.logical_data<int>(N);
auto lY = ctx.logical_data<int>(N);
ctx.parallel_for(lX.shape(), lX.write(), lY.write())->*[] __device__(size_t i, auto x, auto y) {
x(i) = (1 + i);
y(i) = (2 + i * i);
};
/* Compute Y = Y + alpha X */
for (size_t i = 0; i < 200; i++)
{
ctx.task(lY.rw(), lX.read())->*[alpha](cudaStream_t stream, auto dY, auto dX) {
axpy<<<128, 64, 0, stream>>>(dY, dX, alpha);
};
}
ctx.host_launch(lX.read(), lY.read())->*[alpha](auto x, auto y) {
for (size_t i = 0; i < N; i++)
{
EXPECT(x(i) == 1 + i);
EXPECT(y(i) == 2 + i * i + 200 * alpha * x(i));
}
};
}
int main()
{
graph_ctx ctx;
::std::vector<::std::thread> threads;
// Launch threads
for (int i = 0; i < 10; ++i)
{
threads.emplace_back(mytask, ctx, i);
}
// Wait for all threads to complete.
for (auto& th : threads)
{
th.join();
}
ctx.finalize();
}

74
cccl_upstream/cudax/test/stf/threads/axpy-threads-graph.cu
View File

@@ -1,74 +0,0 @@
//===----------------------------------------------------------------------===//
//
// Part of CUDASTF in CUDA C++ Core Libraries,
// under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// SPDX-FileCopyrightText: Copyright (c) 2022-2025 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//
/**
* @file
*
* @brief Ensure a graph_ctx can be used concurrently
*
*/
#include <cuda/experimental/stf.cuh>
#include <mutex>
#include <thread>
using namespace cuda::experimental::stf;
void mytask(graph_ctx ctx, int /*id*/)
{
const size_t N = 16;
int alpha = 3;
auto lX = ctx.logical_data<int>(N);
auto lY = ctx.logical_data<int>(N);
ctx.parallel_for(lX.shape(), lX.write(), lY.write())->*[] __device__(size_t i, auto x, auto y) {
x(i) = (1 + i);
y(i) = (2 + i * i);
};
/* Compute Y = Y + alpha X */
for (size_t i = 0; i < 200; i++)
{
ctx.parallel_for(lY.shape(), lY.rw(), lX.read())->*[alpha] __device__(size_t i, auto dY, auto dX) {
dY(i) += alpha * dX(i);
};
}
ctx.host_launch(lX.read(), lY.read())->*[alpha](auto x, auto y) {
for (size_t i = 0; i < N; i++)
{
EXPECT(x(i) == 1 + i);
EXPECT(y(i) == 2 + i * i + 200 * alpha * x(i));
}
};
}
int main()
{
graph_ctx ctx;
::std::vector<::std::thread> threads;
// Launch threads
for (int i = 0; i < 10; ++i)
{
threads.emplace_back(mytask, ctx, i);
}
// Wait for all threads to complete.
for (auto& th : threads)
{
th.join();
}
ctx.finalize();
}

114
cccl_upstream/cudax/test/stf/threads/axpy-threads-pfor.cu
View File

@@ -1,114 +0,0 @@
//===----------------------------------------------------------------------===//
//
// Part of CUDASTF in CUDA C++ Core Libraries,
// under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// SPDX-FileCopyrightText: Copyright (c) 2022-2024 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//
/**
* @file
*
* @brief An AXPY kernel implemented with a task of the CUDA stream backend
*
*/
#include <cuda/experimental/__stf/stream/stream_ctx.cuh>
#include <mutex>
#include <thread>
using namespace cuda::experimental::stf;
static __global__ void cuda_sleep_kernel(long long int clock_cnt)
{
long long int start_clock = clock64();
long long int clock_offset = 0;
while (clock_offset < clock_cnt)
{
clock_offset = clock64() - start_clock;
}
}
void cuda_sleep(double ms, cudaStream_t stream)
{
int device;
cudaGetDevice(&device);
// cudaDevAttrClockRate: Peak clock frequency in kilohertz;
int clock_rate;
cudaDeviceGetAttribute(&clock_rate, cudaDevAttrClockRate, device);
long long int clock_cnt = (long long int) (ms * clock_rate);
cuda_sleep_kernel<<<1, 1, 0, stream>>>(clock_cnt);
}
__global__ void axpy(double a, slice<const double> x, slice<double> y)
{
int tid = blockIdx.x * blockDim.x + threadIdx.x;
int nthreads = gridDim.x * blockDim.x;
for (int i = tid; i < x.size(); i += nthreads)
{
y(i) += a * x(i);
}
}
double X0(int i)
{
return sin((double) i);
}
double Y0(int i)
{
return cos((double) i);
}
void mytask(stream_ctx ctx, int /*id*/)
{
// std::cout << "Thread " << id << " is executing.\n";
const size_t N = 16;
double alpha = 3.14;
auto lX = ctx.logical_data<double>(N);
auto lY = ctx.logical_data<double>(N);
ctx.parallel_for(lX.shape(), lX.write())->*[] __device__(size_t i, auto x) {
x(i) = 1.0;
};
ctx.task(lY.write())->*[](cudaStream_t, auto) {};
/* Compute Y = Y + alpha X */
for (size_t i = 0; i < 10; i++)
{
ctx.task(lX.read(), lY.rw())->*[&](cudaStream_t s, auto dX, auto dY) {
axpy<<<16, 128, 0, s>>>(alpha, dX, dY);
cuda_sleep(100.0, s);
};
}
}
int main()
{
stream_ctx ctx;
std::vector<std::thread> threads;
// Launch 8 threads.
for (int i = 0; i < 10; ++i)
{
threads.emplace_back(mytask, ctx, i);
}
// Wait for all threads to complete.
for (auto& th : threads)
{
th.join();
}
ctx.finalize();
}

112
cccl_upstream/cudax/test/stf/threads/axpy-threads.cu
View File

@@ -1,112 +0,0 @@
//===----------------------------------------------------------------------===//
//
// Part of CUDASTF in CUDA C++ Core Libraries,
// under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
// SPDX-FileCopyrightText: Copyright (c) 2022-2024 NVIDIA CORPORATION & AFFILIATES.
//
//===----------------------------------------------------------------------===//
/**
* @file
*
* @brief An AXPY kernel implemented with a task of the CUDA stream backend
*
*/
#include <cuda/experimental/__stf/stream/stream_ctx.cuh>
#include <mutex>
#include <thread>
using namespace cuda::experimental::stf;
static __global__ void cuda_sleep_kernel(long long int clock_cnt)
{
long long int start_clock = clock64();
long long int clock_offset = 0;
while (clock_offset < clock_cnt)
{
clock_offset = clock64() - start_clock;
}
}
void cuda_sleep(double ms, cudaStream_t stream)
{
int device;
cudaGetDevice(&device);
// cudaDevAttrClockRate: Peak clock frequency in kilohertz;
int clock_rate;
cudaDeviceGetAttribute(&clock_rate, cudaDevAttrClockRate, device);
long long int clock_cnt = (long long int) (ms * clock_rate);
cuda_sleep_kernel<<<1, 1, 0, stream>>>(clock_cnt);
}
__global__ void axpy(double a, slice<const double> x, slice<double> y)
{
int tid = blockIdx.x * blockDim.x + threadIdx.x;
int nthreads = gridDim.x * blockDim.x;
for (int i = tid; i < x.size(); i += nthreads)
{
y(i) += a * x(i);
}
}
double X0(int i)
{
return sin((double) i);
}
double Y0(int i)
{
return cos((double) i);
}
void mytask(stream_ctx ctx, int /*id*/)
{
// std::cout << "Thread " << id << " is executing.\n";
const size_t N = 16;
double alpha = 3.14;
auto lX = ctx.logical_data<double>(N);
auto lY = ctx.logical_data<double>(N);
ctx.task(lX.write())->*[](cudaStream_t, auto) {};
ctx.task(lY.write())->*[](cudaStream_t, auto) {};
/* Compute Y = Y + alpha X */
for (size_t i = 0; i < 10; i++)
{
ctx.task(lX.read(), lY.rw())->*[&](cudaStream_t s, auto dX, auto dY) {
axpy<<<16, 128, 0, s>>>(alpha, dX, dY);
cuda_sleep(100.0, s);
};
}
}
int main()
{
stream_ctx ctx;
std::vector<std::thread> threads;
// Launch 8 threads.
for (int i = 0; i < 10; ++i)
{
threads.emplace_back(mytask, ctx, i);
}
// Wait for all threads to complete.
for (auto& th : threads)
{
th.join();
}
ctx.finalize();
}