CI: Measure complete PR workflow timing

.github/workflows/pr.yaml

@@ -1,4 +1,5 @@
 name: pr
+# Temporary no-op marker for a complete PR-workflow timing run.
 on:
   push:
     branches:
@@ -98,19 +99,28 @@ jobs:
           - 'cpp/libcudf_kafka/include/**'
           - 'ci/build_docs.sh'
           - 'python/**'
-        test_cpp:
+        test_cpp_common:
           - 'conda/recipes/libcudf/**'
           - 'RAPIDS_BRANCH'
           - 'build.sh'
           - 'dependencies.yaml'
           - 'ci/build_cpp.sh'
           - 'ci/test_cpp.sh'
           - 'ci/test_cpp_common.sh'
+          - '.github/workflows/pr.yaml'
+        test_cpp_libcudf:
           - 'ci/run_cudf_ctests.sh'
+          - 'cpp/tests/**'
+        test_cpp_examples:
           - 'ci/run_cudf_examples.sh'
+          - 'cpp/examples/**'
+        test_cpp_kafka:
           - 'ci/run_cudf_kafka_ctests.sh'
+          - 'cpp/libcudf_kafka/tests/**'
+        test_cpp_streaming:
           - 'ci/run_cudf_streaming_ctests.sh'
-          - '.github/workflows/pr.yaml'
+          - 'cpp/libcudf_streaming/examples/**'
+          - 'cpp/libcudf_streaming/tests/**'
         cpp_test_files:
           - 'cpp/examples/**'
           - 'cpp/libcudf_kafka/tests/**'
@@ -169,20 +179,23 @@ jobs:
           - '.github/workflows/pr.yaml'
         test_python_cudf:
           - 'python/cudf/**'
-        test_python_other:
+        test_python_dask_cudf:
           - 'python/dask_cudf/**'
+          - 'ci/run_dask_cudf_pytests.sh'
+        test_python_cudf_kafka:
           - 'python/cudf_kafka/**'
-          - 'python/custreamz/**'
-          - 'ci/test_python_other.sh'
           - 'ci/run_cudf_kafka_pytests.sh'
-          - 'ci/run_cudf_streaming_pytests.sh'
+        test_python_custreamz:
+          - 'python/custreamz/**'
           - 'ci/run_custreamz_pytests.sh'
-          - 'ci/run_dask_cudf_pytests.sh'
+        test_python_other_common:
+          - 'ci/test_python_other.sh'
         test_python_cudf_polars:
           - 'python/cudf_polars/**'
         test_python_cudf_streaming:
           - 'python/cudf_streaming/**'
           - 'python/libcudf_streaming/**'
+          - 'ci/run_cudf_streaming_pytests.sh'
         test_python_conda_builds:
           - 'conda/**'
           - 'build.sh'
@@ -306,10 +319,16 @@ jobs:
       pull-requests: read
     secrets: inherit # zizmor: ignore[secrets-inherit]
     uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@main
-    if: fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_test_files || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp
+    if: fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_libcudf || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_examples || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_kafka || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_streaming
     with:
       build_type: pull-request
-      script: ci/test_cpp.sh
+      script: >-
+        env
+        RUN_LIBCUDF_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_libcudf }}
+        RUN_EXAMPLES_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_examples }}
+        RUN_LIBCUDF_KAFKA_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_kafka }}
+        RUN_LIBCUDF_STREAMING_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp_streaming }}
+        ci/test_cpp.sh
       # https://github.com/NVIDIA/cudf/issues/23498
       matrix_filter: map(select(.GPU != "gb300" and .GPU != "gh200"))
   conda-python-build:
@@ -370,12 +389,19 @@ jobs:
       pull-requests: read
     secrets: inherit # zizmor: ignore[secrets-inherit]
     uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@main
-    if: fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_polars || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_streaming || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_polars_runner
+    if: fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_dask_cudf || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_kafka || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_custreamz || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_polars || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_streaming || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_polars_runner
     with:
       build_type: pull-request
       # https://github.com/NVIDIA/cudf/pull/22381/changes#r3196736965
       container-options: "--cap-add CAP_SYS_PTRACE --shm-size=8g --ulimit=nofile=1000000:1000000"
-      script: "ci/test_python_other.sh"
+      script: >-
+        env
+        RUN_DASK_CUDF_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_dask_cudf || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other_common }}
+        RUN_CUDF_KAFKA_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_kafka || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other_common }}
+        RUN_CUSTREAMZ_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_custreamz || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other_common }}
+        RUN_CUDF_POLARS_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_polars || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_polars_runner || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other_common }}
+        RUN_CUDF_STREAMING_TESTS=${{ fromJSON(needs.changed-files.outputs.changed_file_groups).cpp_build_inputs || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_all || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_cudf_streaming || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_builds || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda_common || fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_other_common }}
+        ci/test_python_other.sh
       # https://github.com/NVIDIA/cudf/issues/23498
       matrix_filter: map(select(.GPU != "gb300" and .GPU != "gh200"))
   # Build and test the cuDF Java JAR for every Maven classifier.
@@ -723,7 +749,7 @@ jobs:
     steps:
       - id: shards
         env:
-          NUM_SHARDS: 2
+          NUM_SHARDS: 3
         run: |
           {
             echo "num_shards=${NUM_SHARDS}"

ci/test_cpp.sh

@@ -9,30 +9,39 @@ cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../
 
 source ./ci/test_cpp_common.sh
 
+RUN_LIBCUDF_TESTS="${RUN_LIBCUDF_TESTS:-true}"
+RUN_EXAMPLES_TESTS="${RUN_EXAMPLES_TESTS:-true}"
+RUN_LIBCUDF_KAFKA_TESTS="${RUN_LIBCUDF_KAFKA_TESTS:-true}"
+RUN_LIBCUDF_STREAMING_TESTS="${RUN_LIBCUDF_STREAMING_TESTS:-true}"
+
 EXITCODE=0
 trap "EXITCODE=1" ERR
 set +e
 
 # Run gtests from the libcudf-tests and libcudf-streaming-tests packages
 export GTEST_OUTPUT=xml:${RAPIDS_TESTS_DIR}/
 
-rapids-logger "Run libcudf gtests"
-timeout 30m ./ci/run_cudf_ctests.sh -j20
-SUITEERROR=$?
+SUITEERROR=0
+
+if [[ "${RUN_LIBCUDF_TESTS}" == "true" ]]; then
+    rapids-logger "Run libcudf gtests"
+    timeout 30m ./ci/run_cudf_ctests.sh -j20
+    SUITEERROR=$?
+fi
 
-if (( SUITEERROR == 0 )); then
+if (( SUITEERROR == 0 )) && [[ "${RUN_EXAMPLES_TESTS}" == "true" ]]; then
     rapids-logger "Run libcudf examples"
     timeout 30m ./ci/run_cudf_examples.sh
     SUITEERROR=$?
 fi
 
-if (( SUITEERROR == 0 )); then
+if (( SUITEERROR == 0 )) && [[ "${RUN_LIBCUDF_KAFKA_TESTS}" == "true" ]]; then
     rapids-logger "Run libcudf_kafka gtests"
     timeout 30m ./ci/run_cudf_kafka_ctests.sh -j20
     SUITEERROR=$?
 fi
 
-if (( SUITEERROR == 0 )); then
+if (( SUITEERROR == 0 )) && [[ "${RUN_LIBCUDF_STREAMING_TESTS}" == "true" ]]; then
     rapids-logger "Run libcudf_streaming gtests"
     # cudf_streaming contains distributed tests, and running tests in
     # parallel results in resource starvation CI env.

ci/test_python_other.sh

@@ -10,53 +10,69 @@ cd "$(dirname "$(realpath "${BASH_SOURCE[0]}")")"/../
 # Common setup steps shared by Python test jobs
 source ./ci/test_python_common.sh test_python_other
 
+RUN_DASK_CUDF_TESTS="${RUN_DASK_CUDF_TESTS:-true}"
+RUN_CUDF_KAFKA_TESTS="${RUN_CUDF_KAFKA_TESTS:-true}"
+RUN_CUSTREAMZ_TESTS="${RUN_CUSTREAMZ_TESTS:-true}"
+RUN_CUDF_POLARS_TESTS="${RUN_CUDF_POLARS_TESTS:-true}"
+RUN_CUDF_STREAMING_TESTS="${RUN_CUDF_STREAMING_TESTS:-true}"
+
 rapids-logger "Check GPU usage"
 nvidia-smi
 rapids-print-env
 EXITCODE=0
 trap "EXITCODE=1" ERR
 set +e
 
-rapids-logger "pytest dask_cudf"
-timeout 30m ./ci/run_dask_cudf_pytests.sh \
-  --junitxml="${RAPIDS_TESTS_DIR}/junit-dask-cudf.xml" \
-  --numprocesses=8 \
-  --dist=worksteal \
-  --cov-config=../.coveragerc \
-  --cov=dask_cudf \
-  --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/dask-cudf-coverage.xml" \
-  --cov-report=term
-
-rapids-logger "pytest cudf_kafka"
-timeout 30m ./ci/run_cudf_kafka_pytests.sh \
-  --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf-kafka.xml"
-
-rapids-logger "pytest custreamz"
-timeout 30m ./ci/run_custreamz_pytests.sh \
-  --junitxml="${RAPIDS_TESTS_DIR}/junit-custreamz.xml" \
-  --numprocesses=8 \
-  --dist=worksteal \
-  --cov-config=../.coveragerc \
-  --cov=custreamz \
-  --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/custreamz-coverage.xml" \
-  --cov-report=term
-
-rapids-logger "pytest cudf-polars"
-# Fail fast (-x) rather than trying to continue because failed tests pollute the state
-./ci/run_cudf_polars_pytests.sh \
-  -x \
-  --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf-polars.xml" \
-  --numprocesses=4 \
-  --dist=worksteal \
-  --cov-config=./pyproject.toml \
-  --cov=cudf_polars \
-  --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/cudf-polars-coverage.xml" \
-  --cov-report=term \
-  --durations=10 --durations-min=10
-
-rapids-logger "pytest cudf_streaming"
-timeout 30m ./ci/run_cudf_streaming_pytests.sh \
-  --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf-streaming.xml"
+if [[ "${RUN_DASK_CUDF_TESTS}" == "true" ]]; then
+  rapids-logger "pytest dask_cudf"
+  timeout 30m ./ci/run_dask_cudf_pytests.sh \
+    --junitxml="${RAPIDS_TESTS_DIR}/junit-dask-cudf.xml" \
+    --numprocesses=8 \
+    --dist=worksteal \
+    --cov-config=../.coveragerc \
+    --cov=dask_cudf \
+    --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/dask-cudf-coverage.xml" \
+    --cov-report=term
+fi
+
+if [[ "${RUN_CUDF_KAFKA_TESTS}" == "true" ]]; then
+  rapids-logger "pytest cudf_kafka"
+  timeout 30m ./ci/run_cudf_kafka_pytests.sh \
+    --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf-kafka.xml"
+fi
+
+if [[ "${RUN_CUSTREAMZ_TESTS}" == "true" ]]; then
+  rapids-logger "pytest custreamz"
+  timeout 30m ./ci/run_custreamz_pytests.sh \
+    --junitxml="${RAPIDS_TESTS_DIR}/junit-custreamz.xml" \
+    --numprocesses=8 \
+    --dist=worksteal \
+    --cov-config=../.coveragerc \
+    --cov=custreamz \
+    --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/custreamz-coverage.xml" \
+    --cov-report=term
+fi
+
+if [[ "${RUN_CUDF_POLARS_TESTS}" == "true" ]]; then
+  rapids-logger "pytest cudf-polars"
+  # Fail fast (-x) rather than trying to continue because failed tests pollute the state
+  ./ci/run_cudf_polars_pytests.sh \
+    -x \
+    --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf-polars.xml" \
+    --numprocesses=4 \
+    --dist=worksteal \
+    --cov-config=./pyproject.toml \
+    --cov=cudf_polars \
+    --cov-report=xml:"${RAPIDS_COVERAGE_DIR}/cudf-polars-coverage.xml" \
+    --cov-report=term \
+    --durations=10 --durations-min=10
+fi
+
+if [[ "${RUN_CUDF_STREAMING_TESTS}" == "true" ]]; then
+  rapids-logger "pytest cudf_streaming"
+  timeout 30m ./ci/run_cudf_streaming_pytests.sh \
+    --junitxml="${RAPIDS_TESTS_DIR}/junit-cudf-streaming.xml"
+fi
 
 rapids-logger "Test script exiting with value: $EXITCODE"
 exit ${EXITCODE}

cpp/CMakeLists.txt

@@ -7,6 +7,8 @@
 
 cmake_minimum_required(VERSION 4.0 FATAL_ERROR)
 
+# Temporary no-op marker for a complete PR-workflow timing run.
+
 include(../cmake/rapids_config.cmake)
 include(rapids-cmake)
 include(rapids-cpm)

python/cudf/cudf/pandas/__init__.py

@@ -1,6 +1,8 @@
-# SPDX-FileCopyrightText: Copyright (c) 2023-2026, NVIDIA CORPORATION & AFFILIATES.  All rights reserved.
+# SPDX-FileCopyrightText: Copyright (c) 2023-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 # SPDX-License-Identifier: Apache-2.0
 
+# Temporary no-op marker for a complete PR-workflow timing run.
+
 import os
 import warnings