Merge branch 'main' into update-readme

.github/actions/compute-matrix/action.yaml

@@ -25,9 +25,9 @@ runs:
         "
 
         export TEST_MATRIX="
-          - { CUDA_VER: '12.0.1', ARCH: 'amd64', PY_VER: '3.9', LINUX_VER: 'ubuntu18.04', gpu: 'v100', driver: 'latest' }
-          - { CUDA_VER: '12.0.1', ARCH: 'amd64', PY_VER: '3.10', LINUX_VER: 'ubuntu18.04', gpu: 'v100', driver: 'latest' }
-          - { CUDA_VER: '12.0.1', ARCH: 'amd64', PY_VER: '3.11', LINUX_VER: 'ubuntu18.04', gpu: 'v100', driver: 'latest' }
+          - { CUDA_VER: '12.0.1', ARCH: 'amd64', PY_VER: '3.9', LINUX_VER: 'ubuntu20.04', gpu: 'v100', driver: 'latest' }
+          - { CUDA_VER: '12.0.1', ARCH: 'amd64', PY_VER: '3.10', LINUX_VER: 'ubuntu20.04', gpu: 'v100', driver: 'latest' }
+          - { CUDA_VER: '12.0.1', ARCH: 'amd64', PY_VER: '3.11', LINUX_VER: 'ubuntu20.04', gpu: 'v100', driver: 'latest' }
           - { CUDA_VER: '12.0.1', ARCH: 'arm64', PY_VER: '3.9', LINUX_VER: 'ubuntu20.04', gpu: 'a100', driver: 'latest' }
           - { CUDA_VER: '12.0.1', ARCH: 'arm64', PY_VER: '3.10', LINUX_VER: 'ubuntu20.04', gpu: 'a100', driver: 'latest' }
           - { CUDA_VER: '12.0.1', ARCH: 'arm64', PY_VER: '3.11', LINUX_VER: 'ubuntu20.04', gpu: 'a100', driver: 'latest' }

.github/workflows/build.yaml

@@ -28,6 +28,14 @@ jobs:
       build_type: branch
       script: "ci/build_wheel.sh"
       matrix_filter: ${{ needs.compute-matrix.outputs.BUILD_MATRIX }}
+  build-conda:
+    needs:
+      - compute-matrix
+    uses: rapidsai/shared-workflows/.github/workflows/[email protected]
+    with:
+      build_type: pull-request
+      build_script: "ci/build_conda.sh"
+      matrix_filter: ${{ needs.compute-matrix.outputs.BUILD_MATRIX }}
   publish-wheels:
     needs:
       - build-wheels
@@ -39,3 +47,13 @@ jobs:
       sha: ${{ inputs.sha }}
       date: ${{ inputs.date }}
       package-name: pynvjitlink
+  publish-conda:
+    needs:
+     - build-conda
+    secrets: inherit
+    uses: rapidsai/shared-workflows/.github/workflows/[email protected]
+    with:
+      build_type: ${{ inputs.build_type || 'branch' }}
+      branch: ${{ inputs.branch }}
+      date: ${{ inputs.date }}
+      sha: ${{ inputs.sha }}

.github/workflows/pr.yaml

@@ -15,6 +15,8 @@ jobs:
     needs:
       - checks
       - compute-matrix
+      - build-conda
+      - test-conda
       - build-wheels
       - test-wheels
     secrets: inherit
@@ -34,6 +36,23 @@ jobs:
       - name: Compute Build Matrix
         id: compute-matrix
         uses: ./.github/actions/compute-matrix
+  build-conda:
+    needs:
+      - compute-matrix
+    uses: rapidsai/shared-workflows/.github/workflows/[email protected]
+    with:
+      build_type: pull-request
+      build_script: "ci/build_conda.sh"
+      matrix_filter: ${{ needs.compute-matrix.outputs.BUILD_MATRIX }}
+  test-conda:
+    needs:
+      - build-conda
+      - compute-matrix
+    uses: rapidsai/shared-workflows/.github/workflows/[email protected]
+    with:
+      build_type: pull-request
+      test_script: "ci/test_conda.sh"
+      matrix_filter: ${{ needs.compute-matrix.outputs.TEST_MATRIX }}
   build-wheels:
     needs:
       - compute-matrix

CMakeLists.txt

@@ -11,7 +11,8 @@ find_package(Python COMPONENTS Interpreter Development REQUIRED)
 Python_add_library(_nvjitlinklib MODULE pynvjitlink/_nvjitlinklib.cpp WITH_SOABI)
 
 find_package(
-  CUDAToolkit 12.3 REQUIRED
+  # Require CUDA 12.2 Update 2 to avoid nvjitlink bugs
+  CUDAToolkit 12.2.140 REQUIRED
 )
 target_link_libraries(_nvjitlinklib PRIVATE CUDA::nvJitLink_static CUDA::nvptxcompiler_static)
 

ci/build_conda.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION
+
+set -euo pipefail
+
+source rapids-env-update
+
+export CMAKE_GENERATOR=Ninja
+
+rapids-print-env
+
+rapids-logger "Begin py build"
+
+rapids-conda-retry mambabuild conda/recipes/pynvjitlink
+
+rapids-upload-conda-to-s3 python

ci/build_wheel.sh

@@ -6,9 +6,18 @@ set -euo pipefail
 rapids-logger "Install CUDA Toolkit"
 source "$(dirname "$0")/install_latest_cuda_toolkit.sh"
 
+RAPIDS_PY_CUDA_SUFFIX="$(rapids-wheel-ctk-name-gen ${RAPIDS_CUDA_VERSION})"
+
+# This is the version of the suffix with a preceding hyphen. It's used
+# everywhere except in the final wheel name.
+PACKAGE_CUDA_SUFFIX="-${RAPIDS_PY_CUDA_SUFFIX}"
+
+# Patch project metadata files to include the CUDA version suffix.
+sed -i "s/^name = \"pynvjitlink\"/name = \"pynvjitlink${PACKAGE_CUDA_SUFFIX}\"/g" pyproject.toml
+
 rapids-logger "Build wheel"
-mkdir -p ./wheel-build
-pip wheel . --wheel-dir=./wheel-build -vvv
+mkdir -p ./dist
+python -m pip wheel . --wheel-dir=./dist -vvv --disable-pip-version-check --no-deps
 
 rapids-logger "Upload Wheel"
-RAPIDS_PY_WHEEL_NAME="pynvjitlink-cu12" rapids-upload-wheels-to-s3 ./wheel-build
+RAPIDS_PY_WHEEL_NAME="pynvjitlink_${RAPIDS_PY_CUDA_SUFFIX}" rapids-upload-wheels-to-s3 ./dist

ci/test_conda.sh

@@ -0,0 +1,54 @@
+#!/bin/bash
+# Copyright (c) 2024, NVIDIA CORPORATION
+
+set -euo pipefail
+
+. /opt/conda/etc/profile.d/conda.sh
+
+rapids-logger "Generate testing dependencies"
+# TODO: Replace with rapids-dependency-file-generator
+rapids-mamba-retry create -n test \
+    c-compiler \
+    cxx-compiler \
+    cuda-nvcc \
+    cuda-version=${RAPIDS_CUDA_VERSION%.*} \
+    "numba>=0.57" \
+    make \
+    pytest \
+    python=${RAPIDS_PY_VERSION}
+
+# Temporarily allow unbound variables for conda activation.
+set +u
+conda activate test
+set -u
+
+PYTHON_CHANNEL=$(rapids-download-conda-from-s3 python)
+RAPIDS_TESTS_DIR=${RAPIDS_TESTS_DIR:-"${PWD}/test-results"}/
+mkdir -p "${RAPIDS_TESTS_DIR}"
+
+rapids-print-env
+
+rapids-mamba-retry install \
+  --channel "${PYTHON_CHANNEL}" \
+  pynvjitlink
+
+rapids-logger "Build Tests"
+pushd test_binary_generation
+make
+popd
+
+rapids-logger "Check GPU usage"
+nvidia-smi
+
+EXITCODE=0
+trap "EXITCODE=1" ERR
+set +e
+
+rapids-logger "pytest pynvjitlink"
+pytest \
+  --cache-clear \
+  --junitxml="${RAPIDS_TESTS_DIR}/junit-pynvjitlink.xml" \
+  -v
+
+rapids-logger "Test script exiting with value: $EXITCODE"
+exit ${EXITCODE}

ci/test_wheel.sh

@@ -4,15 +4,21 @@
 set -euo pipefail
 
 rapids-logger "Download Wheel"
-RAPIDS_PY_WHEEL_NAME="pynvjitlink-cu12" rapids-download-wheels-from-s3 ./wheel-build/
+RAPIDS_PY_CUDA_SUFFIX="$(rapids-wheel-ctk-name-gen ${RAPIDS_CUDA_VERSION})"
+RAPIDS_PY_WHEEL_NAME="pynvjitlink_${RAPIDS_PY_CUDA_SUFFIX}" rapids-download-wheels-from-s3 ./dist/
+
+# This is the version of the suffix with a preceding hyphen. It's used
+# everywhere except in the final wheel name.
+PACKAGE_CUDA_SUFFIX="-${RAPIDS_PY_CUDA_SUFFIX}"
 
 rapids-logger "Install wheel"
-pip install --find-links ./wheel-build pynvjitlink-cu12
+pip install --find-links ./dist pynvjitlink${PACKAGE_CUDA_SUFFIX}
 
 rapids-logger "Build Tests"
-cd test_binary_generation && make
+pushd test_binary_generation
+make
+popd
 
 rapids-logger "Run Tests"
-cd ..
 pip install pytest
 pytest pynvjitlink/tests

conda/recipes/pynvjitlink/conda_build_config.yaml

@@ -0,0 +1,11 @@
+c_compiler_version:
+  - 11
+
+cxx_compiler_version:
+  - 11
+
+cuda_compiler:
+  - cuda-nvcc
+
+cuda_compiler_version:
+  - 12.2

conda/recipes/pynvjitlink/meta.yaml

@@ -0,0 +1,56 @@
+# Copyright (c) 2018-2024, NVIDIA CORPORATION.
+
+{% set data = load_file_data("pyproject.toml") %}
+{% set project_data = data.get("project") %}
+{% set project_urls = project_data["urls"] %}
+
+{% if cuda_compiler_version %}
+{% set cuda_major = cuda_compiler_version.split(".")[0]|int %}
+{% set cuda_minor = cuda_compiler_version.split(".")[1]|int %}
+{% else %}
+{% set cuda_major = 0 %}
+{% set cuda_minor = 0 %}
+{% endif %}
+
+package:
+  name: pynvjitlink
+  version: {{ project_data["version"] }}
+
+source:
+  path: ../../..
+
+build:
+  script:
+    - {{ PYTHON }} -m pip install . -vv
+  ignore_run_exports_from:
+    - libnvjitlink-dev
+
+requirements:
+  build:
+    - {{ compiler('c') }}
+    - {{ compiler('cxx') }}
+    - {{ compiler('cuda') }}
+    - cmake >=3.26.4
+    - ninja
+    - sysroot_{{ target_platform }} 2.17
+  host:
+    - libnvjitlink-dev
+    - libnvjitlink-static
+    - cuda-version {{ cuda_compiler_version }}
+    - python
+    - pip
+    - scikit-build-core
+  run:
+    - python
+    - numba >=0.57
+    - {{ pin_compatible('cuda-version', min_pin='x', upper_bound=(cuda_major ~ "." ~ (cuda_minor + 1))) }}
+
+about:
+  home: {{ project_urls["Homepage"] }}
+  dev_url: {{ project_urls["Repository"] }}
+  doc_url: {{ project_urls["Documentation"] }}
+  license: {{ project_data["license"]["text"] }}
+  license_family: Apache
+  license_file: LICENSE
+  license_url: {{ project_urls["License"] }}
+  summary: nvJitLink Python binding

pynvjitlink/tests/test_pynvjitlink_api.py

@@ -94,6 +94,11 @@ def test_add_cubin_with_fatbin_error(device_functions_fatbin):
         nvjitlinker.add_cubin(device_functions_fatbin, name)
 
 
+@pytest.mark.xfail(
+    reason="Fails in conda due to using CUDA 12.2u2. This should be fixed "
+    "when requiring CUDA 12.3 nvjitlink library. See: "
+    "https://github.com/rapidsai/pynvjitlink/pull/33#issuecomment-1885182905"
+)
 def test_add_fatbin_with_cubin_error(device_functions_cubin):
     nvjitlinker = NvJitLinker("-arch=sm_75")
     name = "test_device_functions.cubin"

pyproject.toml

@@ -10,7 +10,7 @@ requires = ["scikit-build-core"]
 build-backend = "scikit_build_core.build"
 
 [project]
-name = "pynvjitlink-cu12"
+name = "pynvjitlink"
 version = "0.1.6"
 description = "nvJitLink Python binding"
 readme = { file = "README.md", content-type = "text/markdown" }
@@ -20,6 +20,12 @@ authors = [
 license = { text = "Apache 2.0" }
 requires-python = ">=3.9"
 
+[project.urls]
+Homepage = "https://rapids.ai/"
+Documentation = "https://github.com/rapidsai/pynvjitlink/blob/main/README.md"
+Repository = "https://github.com/rapidsai/pynvjitlink"
+License = "https://github.com/rapidsai/pynvjitlink/blob/main/LICENSE"
+
 [tool.setuptools]
 license-files = ["LICENSE"]