Repository navigation

build

Fix out-of-bounds writes from an undersized buffer bound; validate th… #24

	# Build verification across CUDA toolkits. GitHub-hosted runners have no
	# GPU, so this compiles the library, tool, examples, and tests but does not
	# execute them; execution tests (ctest) run on developer machines with a GPU.
	name: build

	on:
	push:
	branches: [main]
	pull_request:
	workflow_dispatch:

	jobs:
	cuda-build:
	runs-on: ubuntu-24.04
	strategy:
	fail-fast: false
	matrix:
	cuda: ["12.4.1", "12.8.1", "13.2.1"]
	container:
	image: nvidia/cuda:${{ matrix.cuda }}-devel-ubuntu22.04
	steps:
	- uses: actions/checkout@v4
	- name: Install build tools
	run: \|
	apt-get update
	apt-get install -y --no-install-recommends cmake make g++ gfortran python3
	- name: Configure
	# No architecture override: CI compiles the library's defaults, which
	# is what a plain clone-and-build gets (sm_80;90, plus sm_100 and PTX
	# when the toolkit knows it).
	run: cmake -S . -B build
	- name: Build
	run: cmake --build build -j
	- name: Python package syntax check
	run: \|
	python3 -m compileall python