Skip to content

review feedback, fix race conditions #1110

review feedback, fix race conditions

review feedback, fix race conditions #1110

Workflow file for this run

name: CI (GPU)
on:
push:
branches:
- main
- "member-function-customization"
- "pull-request/[0-9]+"
concurrency:
group: ${{ github.workflow }}-on-${{ github.event_name }}-from-${{ github.ref_name }}
cancel-in-progress: true
jobs:
build-gpu:
if: github.repository == 'NVIDIA/stdexec'
name: GPU (${{ matrix.name }}, CUDA ${{ matrix.cuda }}, ${{ matrix.build }})
strategy:
fail-fast: false
matrix:
include:
- { name: "clang 18", cuda: "12.0", cxx: "clang++", build: "Release", tag: "llvm18-cuda12.0", gpu: "v100", sm: "70", driver: "latest", arch: "amd64" }
- { name: "clang 18", cuda: "12.4", cxx: "clang++", build: "Release", tag: "llvm18-cuda12.4", gpu: "v100", sm: "70", driver: "latest", arch: "amd64" }
- { name: "nvc++ 24.3", cuda: "12.4", cxx: "mpic++", build: "Release", tag: "nvhpc24.3", gpu: "v100", sm: "70", driver: "latest", arch: "amd64" }
- { name: "nvc++ 24.3", cuda: "12.4", cxx: "mpic++", build: "Debug", tag: "nvhpc24.3", gpu: "v100", sm: "70", driver: "latest", arch: "amd64" }
runs-on: linux-${{ matrix.arch }}-gpu-${{ matrix.gpu }}-${{ matrix.driver }}-1
container:
options: -u root
image: rapidsai/devcontainers:24.04-cpp-${{ matrix.tag }}-ubuntu22.04
env:
SCCACHE_REGION: "us-east-2"
SCCACHE_BUCKET: "rapids-sccache-devs"
SCCACHE_S3_KEY_PREFIX: "nvidia-stdexec-dev"
NVIDIA_VISIBLE_DEVICES: "${{ env.NVIDIA_VISIBLE_DEVICES }}"
permissions:
id-token: write # This is required for configure-aws-credentials
contents: read # This is required for actions/checkout
defaults:
run:
shell: su coder {0}
working-directory: /home/coder
steps:
- name: Run nvidia-smi to make sure GPU is working
run: nvidia-smi
- name: Checkout stdexec
uses: actions/checkout@v4
with:
path: stdexec
persist-credentials: false
- if: github.repository_owner == 'NVIDIA'
name: Get AWS credentials for sccache bucket
uses: aws-actions/configure-aws-credentials@v4
with:
aws-region: us-east-2
role-duration-seconds: 28800 # 8 hours
role-to-assume: arn:aws:iam::279114543810:role/gha-oidc-NVIDIA
- name: Build and test GPU schedulers
run: |
set -ex;
# Copy source folder into ~/stdexec
cp -r "${GITHUB_WORKSPACE}"/stdexec ~/;
chown -R coder:coder ~/stdexec;
cd ~/stdexec;
# Configure
cmake -S . -B build -GNinja \
-DSTDEXEC_ENABLE_CUDA=ON \
-DSTDEXEC_ENABLE_IO_URING_TESTS=OFF \
-DCMAKE_BUILD_TYPE=${{ matrix.build }} \
-DCMAKE_CXX_COMPILER=${{ matrix.cxx }} \
-DCMAKE_CUDA_COMPILER=${{ matrix.cxx }} \
-DCMAKE_CUDA_ARCHITECTURES=${{ matrix.sm }} \
;
# Compile
cmake --build build -v;
# Print sccache stats
sccache -s;
# Tests
ctest --test-dir build --verbose --output-on-failure --timeout 60;
# Examples
./build/examples/nvexec/maxwell_cpu_st --iterations=1000 --N=512 --run-cpp --run-inline-scheduler;
./build/examples/nvexec/maxwell_cpu_mt --iterations=1000 --N=512 --run-std --run-stdpar --run-thread-pool-scheduler;
./build/examples/nvexec/maxwell_gpu_s --iterations=1000 --N=512 --run-cuda --run-stdpar --run-stream-scheduler;
ci-gpu:
runs-on: ubuntu-latest
name: CI (GPU)
needs:
- build-gpu
steps:
- run: echo "CI (GPU) success"