Add cuda_parallel to CI. (#2338)

* Add python cuda_parallel to CI. * Fix cuda.parallel installation order * WAR occupancy/V100 failure * Update cuda_cooperative readme --------- Co-authored-by: Georgy Evtushenko <[email protected]>
NVIDIA · Sep 4, 2024 · 1e9125e · 1e9125e
1 parent a154e7b
commit 1e9125e
Show file tree

Hide file tree

Showing 3 changed files with 19 additions and 4 deletions.
diff --git a/c/src/reduce.cu b/c/src/reduce.cu
@@ -301,8 +301,12 @@ cudaError_t InvokePasses(
     // Init regular kernel configuration
     const auto tile_size = policy.block_size * policy.items_per_thread;
 
+    // TODO Should be checking the return code here, but for some reason I'm getting invalid handle error on V100
     int sm_occupancy = 1;
-    check(cuOccupancyMaxActiveBlocksPerMultiprocessor(&sm_occupancy, reduce_kernel, policy.block_size, 0));
+    if (CUDA_SUCCESS != cuOccupancyMaxActiveBlocksPerMultiprocessor(&sm_occupancy, reduce_kernel, policy.block_size, 0))
+    {
+      sm_occupancy = 6;
+    }
 
     int reduce_device_occupancy = sm_occupancy * sm_count;
 

diff --git a/ci/test_pycuda.sh b/ci/test_pycuda.sh
@@ -13,8 +13,19 @@ export PYTHONPATH="${prefix}:${PYTHONPATH:-}"
 
 pushd ../python/cuda_cooperative >/dev/null
 
-run_command "⚙️  Pip install cuda" pip install --force-reinstall --target "${prefix}" .[test]
-run_command "🚀  Pytest cuda" python -m pytest -v ./tests
+run_command "⚙️  Pip install cuda_cooperative" pip install --force-reinstall --upgrade --target "${prefix}" .[test]
+run_command "🚀  Pytest cuda_cooperative" python -m pytest -v ./tests
+
+popd >/dev/null
+
+pushd ../python/cuda_parallel >/dev/null
+
+# Temporarily install the package twice to populate include directory as part of the first installation
+# and to let manifest discover these includes during the second installation. Do not forget to remove the
+# second installation after https://github.com/NVIDIA/cccl/issues/2281 is addressed.
+run_command "⚙️  Pip install cuda_parallel once" pip install --force-reinstall --upgrade --target "${prefix}" .[test]
+run_command "⚙️  Pip install cuda_parallel twice" pip install --force-reinstall --upgrade --target "${prefix}" .[test]
+run_command "🚀  Pytest cuda_parallel" python -m pytest -v ./tests
 
 popd >/dev/null
 

diff --git a/python/cuda_cooperative/README.md b/python/cuda_cooperative/README.md
@@ -8,5 +8,5 @@ Please visit the documentation here: https://nvidia.github.io/cccl/python.html.
 
 ```bash
 pip3 install -e .[test]
-pytest -v ./tests/device/
+pytest -v ./tests/
 ```