Test WSL in CI

.github/workflows/build.yaml

@@ -52,7 +52,7 @@ jobs:
   cpp-build:
     needs: [telemetry-setup]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@wsl
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -63,7 +63,7 @@ jobs:
   python-build:
     needs: [telemetry-setup, cpp-build]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@wsl
     with:
       # Build a conda package for each CUDA x ARCH x minimum supported Python version
       matrix_filter: group_by({CUDA_VER, ARCH}) | map(min_by(.PY_VER | split(".") | map(tonumber)))
@@ -75,7 +75,7 @@ jobs:
   upload-conda:
     needs: [cpp-build, python-build]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-upload-packages.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-upload-packages.yaml@wsl
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -85,7 +85,7 @@ jobs:
     if: github.ref_type == 'branch'
     needs: [python-build]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@wsl
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -98,7 +98,7 @@ jobs:
   wheel-build-cpp:
     needs: [telemetry-setup]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@wsl
     with:
       matrix_filter: group_by([.ARCH, (.CUDA_VER|split(".")|map(tonumber)|.[0])]) | map(max_by(.PY_VER|split(".")|map(tonumber)))
       build_type: ${{ inputs.build_type || 'branch' }}
@@ -112,7 +112,7 @@ jobs:
   wheel-build-python:
     needs: [telemetry-setup, wheel-build-cpp]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@wsl
     with:
       # Build a wheel for each CUDA x ARCH x minimum supported Python version
       matrix_filter: group_by({CUDA_VER, ARCH}) | map(min_by(.PY_VER | split(".") | map(tonumber)))
@@ -127,7 +127,7 @@ jobs:
   wheel-publish-cpp:
     needs: wheel-build-cpp
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-publish.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-publish.yaml@wsl
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -138,7 +138,7 @@ jobs:
   wheel-publish-python:
     needs: wheel-build-python
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-publish.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-publish.yaml@wsl
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}

.github/workflows/pr.yaml

@@ -26,7 +26,7 @@ jobs:
       - devcontainer
       - telemetry-setup
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/pr-builder.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/pr-builder.yaml@wsl
     if: always()
     with:
       needs: ${{ toJSON(needs) }}
@@ -70,7 +70,7 @@ jobs:
     needs:
       - telemetry-setup
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/changed-files.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/changed-files.yaml@wsl
     with:
       files_yaml: |
         build_docs:
@@ -188,30 +188,30 @@ jobs:
     secrets: inherit
     needs:
       - telemetry-setup
-    uses: rapidsai/shared-workflows/.github/workflows/checks.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/checks.yaml@wsl
     with:
       enable_check_generated_files: false
       ignored_pr_jobs: "telemetry-summarize conda-python-tests-integration-optional wheel-tests-integration-optional"
   conda-cpp-build:
     needs: checks
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@wsl
     with:
       build_type: pull-request
       node_type: cpu8
       script: ci/build_cpp.sh
   conda-cpp-tests:
     needs: [conda-cpp-build, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp
     with:
       build_type: pull-request
       script: ci/test_cpp.sh
   conda-cpp-debug-tests:
     needs: [checks, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_cpp
     with:
       build_type: pull-request
@@ -222,7 +222,7 @@ jobs:
   conda-python-build:
     needs: conda-cpp-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@wsl
     with:
       build_type: pull-request
       script: ci/build_python.sh
@@ -231,23 +231,23 @@ jobs:
   conda-python-tests:
     needs: [conda-python-build, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda
     with:
       build_type: pull-request
       script: ci/test_python.sh
   conda-python-tests-integration-optional:
     needs: [conda-python-build, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_conda
     with:
       build_type: pull-request
       script: ci/test_python_integrations.sh
   docs-build:
     needs: [conda-python-build, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).build_docs
     with:
       build_type: pull-request
@@ -258,7 +258,7 @@ jobs:
   wheel-build-cpp:
     needs: checks
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@wsl
     with:
       matrix_filter: group_by([.ARCH, (.CUDA_VER|split(".")|map(tonumber)|.[0])]) | map(max_by(.PY_VER|split(".")|map(tonumber)))
       build_type: pull-request
@@ -269,7 +269,7 @@ jobs:
   wheel-build-python:
     needs: wheel-build-cpp
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@wsl
     with:
       # Build a wheel for each CUDA x ARCH x minimum supported Python version
       matrix_filter: group_by({CUDA_VER, ARCH}) | map(min_by(.PY_VER | split(".") | map(tonumber)))
@@ -281,15 +281,15 @@ jobs:
   wheel-tests:
     needs: [wheel-build-python, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_wheels
     with:
       build_type: pull-request
       script: ci/test_wheel.sh
   wheel-tests-integration-optional:
     needs: [wheel-build-python, changed-files]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@wsl
     if: fromJSON(needs.changed-files.outputs.changed_file_groups).test_python_wheels
     with:
       build_type: pull-request
@@ -298,7 +298,7 @@ jobs:
     secrets: inherit
     needs:
       - telemetry-setup
-    uses: rapidsai/shared-workflows/.github/workflows/build-in-devcontainer.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/build-in-devcontainer.yaml@wsl
     with:
       arch: '["amd64", "arm64"]'
       cuda: '["13.1"]'

.github/workflows/test.yaml

@@ -25,7 +25,7 @@ on:
 jobs:
   cpp-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@wsl
     with:
       build_type: ${{ inputs.build_type }}
       branch: ${{ inputs.branch }}
@@ -34,7 +34,7 @@ jobs:
       sha: ${{ inputs.sha }}
   cpp-debug-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@wsl
     with:
       build_type: ${{ inputs.build_type }}
       branch: ${{ inputs.branch }}
@@ -46,7 +46,7 @@ jobs:
       script: "ci/test_cpp_debug.sh"
   python-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@wsl
     with:
       build_type: ${{ inputs.build_type }}
       branch: ${{ inputs.branch }}
@@ -55,7 +55,7 @@ jobs:
       sha: ${{ inputs.sha }}
   wheel-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@main
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@wsl
     with:
       build_type: ${{ inputs.build_type }}
       branch: ${{ inputs.branch }}

ci/test_cpp.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+# SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
 # SPDX-License-Identifier: Apache-2.0
 
 set -euo pipefail
@@ -33,14 +33,19 @@ set -u
 
 rapids-print-env
 
-rapids-logger "Check GPU usage"
+rapids-logger "Check system info"
+uname -a
 nvidia-smi
 
 # Run librmm gtests from librmm-tests package
 rapids-logger "Run gtests"
 
 export GTEST_OUTPUT=xml:${RAPIDS_TESTS_DIR}/
-timeout 15m ./ci/run_ctests.sh -j20 && EXITCODE=$? || EXITCODE=$?;
+CTEST_TIMEOUT=15m
+if grep -qi microsoft /proc/version 2>/dev/null; then
+    CTEST_TIMEOUT=30m  # WSL is slow for large device memory allocations
+fi
+timeout "${CTEST_TIMEOUT}" ./ci/run_ctests.sh -j20 && EXITCODE=$? || EXITCODE=$?;
 
 # Run all examples from librmm-example package
 for example in "${CONDA_PREFIX}"/bin/examples/librmm/*; do

ci/test_cpp_debug.sh

@@ -33,7 +33,8 @@ set -u
 
 rapids-print-env
 
-rapids-logger "Check GPU usage"
+rapids-logger "Check system info"
+uname -a
 nvidia-smi
 
 rapids-logger "Building librmm in Debug mode"

ci/test_python.sh

@@ -39,7 +39,8 @@ RAPIDS_TESTS_DIR=${RAPIDS_TESTS_DIR:-"${PWD}/test-results"}
 RAPIDS_COVERAGE_DIR=${RAPIDS_COVERAGE_DIR:-"${PWD}/coverage-results"}
 mkdir -p "${RAPIDS_TESTS_DIR}" "${RAPIDS_COVERAGE_DIR}"
 
-rapids-logger "Check GPU usage"
+rapids-logger "Check system info"
+uname -a
 nvidia-smi
 
 rapids-logger "pytest rmm"

ci/test_python_integrations.sh

@@ -22,7 +22,8 @@ RAPIDS_TESTS_DIR=${RAPIDS_TESTS_DIR:-"${PWD}/test-results"}
 RAPIDS_COVERAGE_DIR=${RAPIDS_COVERAGE_DIR:-"${PWD}/coverage-results"}
 mkdir -p "${RAPIDS_TESTS_DIR}" "${RAPIDS_COVERAGE_DIR}"
 
-rapids-logger "Check GPU usage"
+rapids-logger "Check system info"
+uname -a
 nvidia-smi
 
 EXITCODE=0

ci/test_wheel_integrations.sh

@@ -28,7 +28,8 @@ PIP_INSTALL_SHARED_ARGS=(
 
 EXITCODE=0
 
-rapids-logger "Check GPU usage"
+rapids-logger "Check system info"
+uname -a
 nvidia-smi
 
 echo "::group::PyTorch Tests"

cpp/include/rmm/mr/arena_memory_resource.hpp

@@ -1,5 +1,5 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2020-2025, NVIDIA CORPORATION.
+ * SPDX-FileCopyrightText: Copyright (c) 2020-2026, NVIDIA CORPORATION.
  * SPDX-License-Identifier: Apache-2.0
  */
 #pragma once
@@ -155,8 +155,13 @@ class arena_memory_resource final : public device_memory_resource {
       void* pointer = arena.allocate_sync(bytes);
       if (pointer == nullptr) {
         if (dump_log_on_failure_) { dump_memory_log(bytes); }
+        auto const [free_memory, total_memory] = rmm::available_device_memory();
         auto const msg = std::string("Maximum pool size exceeded (failed to allocate ") +
-                         rmm::detail::format_bytes(bytes) + "): No room in arena.";
+                         rmm::detail::format_bytes(bytes) +
+                         ", arena size=" + rmm::detail::format_bytes(global_arena_.size()) +
+                         ", device free=" + rmm::detail::format_bytes(free_memory) +
+                         " total=" + rmm::detail::format_bytes(total_memory) +
+                         "): No room in arena.";
         RMM_FAIL(msg.c_str(), rmm::out_of_memory);
       }
       return pointer;

cpp/include/rmm/mr/detail/arena.hpp

@@ -1,5 +1,5 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2019-2025, NVIDIA CORPORATION.
+ * SPDX-FileCopyrightText: Copyright (c) 2019-2026, NVIDIA CORPORATION.
  * SPDX-License-Identifier: Apache-2.0
  */
 
@@ -517,6 +517,13 @@ class global_arena final {
     upstream_mr_.deallocate_sync(upstream_block_.pointer(), upstream_block_.size());
   }
 
+  /**
+   * @brief Returns the total size of the global arena in bytes.
+   *
+   * @return std::size_t The size of the upstream block allocated for the arena.
+   */
+  [[nodiscard]] std::size_t size() const noexcept { return upstream_block_.size(); }
+
   /**
    * @brief Should allocation of `size` bytes be handled by the global arena directly?
    *

cpp/include/rmm/mr/pool_memory_resource.hpp

@@ -239,12 +239,25 @@ class pool_memory_resource final
   block_type try_to_expand(std::size_t try_size, std::size_t min_size, cuda_stream_view stream)
   {
     auto report_error = [&](const char* reason) {
-      RMM_LOG_ERROR("[A][Stream %s][Upstream %zuB][FAILURE maximum pool size exceeded: %s]",
-                    rmm::detail::format_stream(stream),
-                    min_size,
-                    reason);
+      auto const max_size = maximum_pool_size_.value_or(std::numeric_limits<std::size_t>::max());
+      auto const [free_memory, total_memory] = rmm::available_device_memory();
+      RMM_LOG_ERROR(
+        "[A][Stream %s][Upstream %zuB][FAILURE maximum pool size exceeded: %s]"
+        "[Pool current=%s max=%s][Device free=%s total=%s]",
+        rmm::detail::format_stream(stream),
+        min_size,
+        reason,
+        rmm::detail::format_bytes(pool_size()),
+        rmm::detail::format_bytes(max_size),
+        rmm::detail::format_bytes(free_memory),
+        rmm::detail::format_bytes(total_memory));
       auto const msg = std::string("Maximum pool size exceeded (failed to allocate ") +
-                       rmm::detail::format_bytes(min_size) + std::string("): ") + reason;
+                       rmm::detail::format_bytes(min_size) +
+                       ", pool current=" + rmm::detail::format_bytes(pool_size()) +
+                       " max=" + rmm::detail::format_bytes(max_size) +
+                       ", device free=" + rmm::detail::format_bytes(free_memory) +
+                       " total=" + rmm::detail::format_bytes(total_memory) + std::string("): ") +
+                       reason;
       RMM_FAIL(msg.c_str(), rmm::out_of_memory);
     };