Tests

Running the tests

The CI/CD pipelines build and run the tests in Docker containers for a variety of configurations. The most convenient way to build and run the tests without reconstructing the dependencies is to reuse one of our images.

For example, for an x86_64 platform, with a Blackwell120 GPU, targeting a compiler toolchain using nvcc as the CUDA compiler and gcc as the host compiler, the image cuda-gnu-14-nvidia-py3.13:13.1.0-devel-ubuntu24.04 may be used.

Start a container in interactive mode:

docker run --rm -it --gpus=all --cap-add=SYS_ADMIN \
    ghcr.io/uliegecsm/reprospect/cuda-gnu-14-nvidia-py3.13:13.1.0-devel-ubuntu24.04

The GPU architecture enters through CMAKE_CUDA_ARCHITECTURES below. Only the images containing a Kokkos installation used for the ReProspect examples are specific to a GPU architecture.

Then, inside the container:

git clone https://github.com/uliegecsm/reprospect.git /workspace/reprospect
cd /workspace/reprospect

cmake -S . --preset=gnu-nvidia \
    -DCMAKE_CUDA_ARCHITECTURES=120-real \
    -DReProspect_ENABLE_EXAMPLES=OFF \
    -DReProspect_ENABLE_TESTS=ON \
    -DReProspect_ENABLE_DOCS=OFF

cmake --build --preset=gnu-nvidia
ctest --preset=gnu-nvidia

For another configuration, choose the image from the list of images the CI/CD pipelines build, set CMAKE_CUDA_ARCHITECTURES to match the architecture of the GPU (e.g., 70-real for Volta70), and select the preset matching the targeted compiler toolchain, namely, gnu-nvidia for nvcc as the CUDA compiler with gcc as the host compiler, clang-nvidia for nvcc as the CUDA compiler with clang as the host compiler, and clang for clang as the CUDA compiler.

Overview of the tests

The test directory structure mirrors the reprospect package source directory structure.

The tests for the three main subpackages can be found in:

Tests that involve many subpackages are grouped in:

Integration tests

Utilities for testing

ReProspect tests use helpers for diverse tasks such as compiling or extracting random bits from cuBLAS.

Test assets

Test assets are located in:

a subdirectory named assets next to the test file that uses them
the central tests/assets directory if shared across multiple tests

Assets