dedf08166a
Added 863 files from NVIDIA/cccl sparse checkout: - c2h/ (27 files): Catch2 test helpers — generators, validators, runner - nvbench_helper/ (10 files): Benchmark harness utilities - cmake/ (29 files): CMake presets and build helpers - cudax/ (794 files): Experimental CUDA extensions - AGENTS.md: NVIDIA's official AI agent instructions for CCCL - CMakePresets.json: Standardized build configurations - cccl-version.json: Version tracking Also added CCCL_ASSET_MAP.md mapping all 4295 CCCL files to competition value and PRD items. cccl_upstream now covers 100% of competition-critical assets: - 27 tuning headers (SM80/90/100 benchmark data) - 32 dispatch headers (algorithm implementations) - 60 Thrust examples (correctness verification) - 217 CUB Catch2 tests (regression matrix) - 153 CUB benchmarks (parameter space search) - 18 CUB examples (API verification) - 27 test helpers + benchmark harness - 794 cudax experimental extensions
87 lines
2.1 KiB
Plaintext
87 lines
2.1 KiB
Plaintext
//===----------------------------------------------------------------------===//
|
|
//
|
|
// 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 Freeze data in read-only fashion
|
|
*
|
|
*/
|
|
|
|
#include <cuda/experimental/__stf/graph/graph_ctx.cuh>
|
|
#include <cuda/experimental/__stf/stream/stream_ctx.cuh>
|
|
|
|
using namespace cuda::experimental::stf;
|
|
|
|
int X0(int i)
|
|
{
|
|
return 17 * i + 45;
|
|
}
|
|
|
|
__global__ void mult(slice<int> s, int val)
|
|
{
|
|
int tid = blockIdx.x * blockDim.x + threadIdx.x;
|
|
int nthreads = gridDim.x * blockDim.x;
|
|
|
|
for (int i = tid; i < s.size(); i += nthreads)
|
|
{
|
|
s(i) *= val;
|
|
}
|
|
}
|
|
|
|
int main()
|
|
{
|
|
stream_ctx ctx;
|
|
|
|
cudaStream_t stream = ctx.pick_stream();
|
|
|
|
const int N = 16;
|
|
int X[N];
|
|
|
|
for (int i = 0; i < N; i++)
|
|
{
|
|
X[i] = X0(i);
|
|
}
|
|
|
|
auto lX = ctx.logical_data(X).set_symbol("X");
|
|
auto lY = ctx.logical_data(lX.shape()).set_symbol("Y");
|
|
|
|
for (int k = 0; k < 4; k++)
|
|
{
|
|
auto fx = ctx.freeze(lX, access_mode::rw, data_place::current_device());
|
|
|
|
_CCCL_ASSERT(fx.get_access_mode() == access_mode::rw, "invalid access mode");
|
|
|
|
auto dX = fx.get(data_place::current_device(), stream);
|
|
mult<<<8, 4, 0, stream>>>(dX, 4);
|
|
fx.unfreeze(stream);
|
|
|
|
ctx.parallel_for(lX.shape(), lX.read(), lY.write()).set_symbol("Y=X")->*[] __device__(size_t i, auto x, auto y) {
|
|
y(i) = x(i);
|
|
};
|
|
|
|
ctx.parallel_for(lX.shape(), lY.rw()).set_symbol("Y+=1")->*[] __device__(size_t i, auto y) {
|
|
y(i) += 1;
|
|
};
|
|
|
|
// ctx.host_launch(lX.read(), lY.read())->*[](auto x, auto y) {
|
|
// for (int i = 0; i < x.size(); i++) {
|
|
// EXPECT(x(i) == 2*X0(i) + 4);
|
|
// }
|
|
//
|
|
// for (int i = 0; i < y.size(); i++) {
|
|
// EXPECT(y(i) == 2*X0(i) + 4);
|
|
// }
|
|
// };
|
|
}
|
|
|
|
ctx.finalize();
|
|
}
|