[RELEASE] rmm v24.02

.clang-tidy

@@ -62,8 +62,8 @@ CheckOptions:
     value:           'alignment'
   - key:             cppcoreguidelines-avoid-magic-numbers.IgnorePowersOf2IntegerValues
     value:           '1'
-  - key:             readability-magic-numbers.IgnorePowersOf2IntegerValues
-    value:           '1'
+  - key:             cppcoreguidelines-avoid-magic-numbers.IgnoredIntegerValues
+    value:           "0;1;2;3;4;50;100"
   - key:             cppcoreguidelines-avoid-do-while.IgnoreMacros
     value:           'true'
 ...

.devcontainer/cuda11.8-conda/devcontainer.json

@@ -5,12 +5,12 @@
     "args": {
       "CUDA": "11.8",
       "PYTHON_PACKAGE_MANAGER": "conda",
-      "BASE": "rapidsai/devcontainers:23.12-cpp-cuda11.8-mambaforge-ubuntu22.04"
+      "BASE": "rapidsai/devcontainers:24.02-cpp-cuda11.8-mambaforge-ubuntu22.04"
     }
   },
   "hostRequirements": {"gpu": "optional"},
   "features": {
-    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:23.12": {}
+    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:24.2": {}
   },
   "overrideFeatureInstallOrder": [
     "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils"

.devcontainer/cuda11.8-pip/devcontainer.json

@@ -5,12 +5,12 @@
     "args": {
       "CUDA": "11.8",
       "PYTHON_PACKAGE_MANAGER": "pip",
-      "BASE": "rapidsai/devcontainers:23.12-cpp-llvm16-cuda11.8-ubuntu22.04"
+      "BASE": "rapidsai/devcontainers:24.02-cpp-llvm16-cuda11.8-ubuntu22.04"
     }
   },
   "hostRequirements": {"gpu": "optional"},
   "features": {
-    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:23.12": {}
+    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:24.2": {}
   },
   "overrideFeatureInstallOrder": [
     "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils"

.devcontainer/cuda12.0-conda/devcontainer.json

@@ -5,12 +5,12 @@
     "args": {
       "CUDA": "12.0",
       "PYTHON_PACKAGE_MANAGER": "conda",
-      "BASE": "rapidsai/devcontainers:23.12-cpp-mambaforge-ubuntu22.04"
+      "BASE": "rapidsai/devcontainers:24.02-cpp-mambaforge-ubuntu22.04"
     }
   },
   "hostRequirements": {"gpu": "optional"},
   "features": {
-    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:23.12": {}
+    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:24.2": {}
   },
   "overrideFeatureInstallOrder": [
     "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils"

.devcontainer/cuda12.0-pip/devcontainer.json

@@ -5,12 +5,12 @@
     "args": {
       "CUDA": "12.0",
       "PYTHON_PACKAGE_MANAGER": "pip",
-      "BASE": "rapidsai/devcontainers:23.12-cpp-llvm16-cuda12.0-ubuntu22.04"
+      "BASE": "rapidsai/devcontainers:24.02-cpp-llvm16-cuda12.0-ubuntu22.04"
     }
   },
   "hostRequirements": {"gpu": "optional"},
   "features": {
-    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:23.12": {}
+    "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils:24.2": {}
   },
   "overrideFeatureInstallOrder": [
     "ghcr.io/rapidsai/devcontainers/features/rapids-build-utils"

.github/CODEOWNERS

@@ -14,5 +14,4 @@ python/            @rapidsai/rmm-python-codeowners
 .github/           @rapidsai/ops-codeowners
 ci/                @rapidsai/ops-codeowners
 conda/             @rapidsai/ops-codeowners
-**/Dockerfile      @rapidsai/ops-codeowners
-**/.dockerignore   @rapidsai/ops-codeowners
+dependencies.yaml  @rapidsai/ops-codeowners

.github/workflows/build.yaml

@@ -28,7 +28,7 @@ concurrency:
 jobs:
   cpp-build:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@branch-24.02
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -37,7 +37,7 @@ jobs:
   python-build:
     needs: [cpp-build]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@branch-24.02
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -46,7 +46,7 @@ jobs:
   upload-conda:
     needs: [cpp-build, python-build]
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-upload-packages.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-upload-packages.yaml@branch-24.02
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -56,7 +56,7 @@ jobs:
     if: github.ref_type == 'branch'
     needs: python-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@branch-24.02
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -68,7 +68,7 @@ jobs:
       run_script: "ci/build_docs.sh"
   wheel-build:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@branch-24.02
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}
@@ -78,7 +78,7 @@ jobs:
   wheel-publish:
     needs: wheel-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-publish.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-publish.yaml@branch-24.02
     with:
       build_type: ${{ inputs.build_type || 'branch' }}
       branch: ${{ inputs.branch }}

.github/workflows/pr.yaml

@@ -22,40 +22,40 @@ jobs:
       - wheel-tests
       - devcontainer
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/pr-builder.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/pr-builder.yaml@branch-24.02
   checks:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/checks.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/checks.yaml@branch-24.02
     with:
       enable_check_generated_files: false
   conda-cpp-build:
     needs: checks
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-build.yaml@branch-24.02
     with:
       build_type: pull-request
   conda-cpp-tests:
     needs: conda-cpp-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@branch-24.02
     with:
       build_type: pull-request
   conda-python-build:
     needs: conda-cpp-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-build.yaml@branch-24.02
     with:
       build_type: pull-request
   conda-python-tests:
     needs: conda-python-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@branch-24.02
     with:
       build_type: pull-request
   docs-build:
     needs: conda-python-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/custom-job.yaml@branch-24.02
     with:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
@@ -65,20 +65,20 @@ jobs:
   wheel-build:
     needs: checks
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-build.yaml@branch-24.02
     with:
       build_type: pull-request
       script: ci/build_wheel.sh
   wheel-tests:
     needs: wheel-build
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@branch-24.02
     with:
       build_type: pull-request
       script: ci/test_wheel.sh
   devcontainer:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/build-in-devcontainer.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/build-in-devcontainer.yaml@branch-24.02
     with:
       build_command: |
         sccache -z;

.github/workflows/test.yaml

@@ -16,23 +16,23 @@ on:
 jobs:
   cpp-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-cpp-tests.yaml@branch-24.02
     with:
       build_type: nightly
       branch: ${{ inputs.branch }}
       date: ${{ inputs.date }}
       sha: ${{ inputs.sha }}
   python-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/conda-python-tests.yaml@branch-24.02
     with:
       build_type: nightly
       branch: ${{ inputs.branch }}
       date: ${{ inputs.date }}
       sha: ${{ inputs.sha }}
   wheel-tests:
     secrets: inherit
-    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@branch-23.12
+    uses: rapidsai/shared-workflows/.github/workflows/wheels-test.yaml@branch-24.02
     with:
       build_type: nightly
       branch: ${{ inputs.branch }}

.gitignore

@@ -22,6 +22,7 @@ rmm.egg-info/
 python/build
 python/*/build
 python/rmm/**/_lib/**/*.cpp
+!python/rmm/_lib/_torch_allocator.cpp
 python/rmm/**/_lib/**/*.h
 python/rmm/**/_lib/.nfs*
 python/rmm/_cuda/*.cpp

.pre-commit-config.yaml

@@ -6,7 +6,7 @@ repos:
           - id: trailing-whitespace
           - id: end-of-file-fixer
     - repo: https://github.com/rapidsai/dependency-file-generator
-      rev: v1.5.1
+      rev: v1.8.0
       hooks:
           - id: rapids-dependency-file-generator
             args: ["--clean"]

CMakeLists.txt

@@ -23,7 +23,7 @@ include(rapids-find)
 
 project(
   RMM
-  VERSION 23.12.00
+  VERSION 24.02.00
   LANGUAGES CXX)
 
 # Write the version header
@@ -70,8 +70,7 @@ rapids_cpm_init()
 
 include(cmake/thirdparty/get_fmt.cmake)
 include(cmake/thirdparty/get_spdlog.cmake)
-include(cmake/thirdparty/get_libcudacxx.cmake)
-include(cmake/thirdparty/get_thrust.cmake)
+include(cmake/thirdparty/get_cccl.cmake)
 
 # ##################################################################################################
 # * library targets --------------------------------------------------------------------------------
@@ -90,8 +89,7 @@ else()
   target_link_libraries(rmm INTERFACE CUDA::cudart)
 endif()
 
-target_link_libraries(rmm INTERFACE libcudacxx::libcudacxx)
-target_link_libraries(rmm INTERFACE rmm::Thrust)
+target_link_libraries(rmm INTERFACE CCCL::CCCL)
 target_link_libraries(rmm INTERFACE fmt::fmt-header-only)
 target_link_libraries(rmm INTERFACE spdlog::spdlog_header_only)
 target_link_libraries(rmm INTERFACE dl)
@@ -152,20 +150,12 @@ The goal of the [RMM](https://github.com/rapidsai/rmm) is to provide:
   A collection of data structures that use the interface for memory allocation
 ]=])
 
-set(code_string
-    [=[
-if(NOT TARGET rmm::Thrust)
-  thrust_create_target(rmm::Thrust FROM_OPTIONS)
-endif()
-]=])
-
 rapids_export(
   INSTALL rmm
   EXPORT_SET rmm-exports
   GLOBAL_TARGETS rmm
   NAMESPACE rmm::
-  DOCUMENTATION doc_string
-  FINAL_CODE_BLOCK code_string)
+  DOCUMENTATION doc_string)
 
 # ##################################################################################################
 # * build export -----------------------------------------------------------------------------------
@@ -175,8 +165,7 @@ rapids_export(
   EXPORT_SET rmm-exports
   GLOBAL_TARGETS rmm
   NAMESPACE rmm::
-  DOCUMENTATION doc_string
-  FINAL_CODE_BLOCK code_string)
+  DOCUMENTATION doc_string)
 
 # ##################################################################################################
 # * make documentation -----------------------------------------------------------------------------
@@ -197,6 +186,6 @@ add_custom_target(
 # * make gdb helper scripts ------------------------------------------------------------------------
 
 # optionally assemble Thrust pretty-printers
-if(Thrust_SOURCE_DIR)
+if(CCCL_SOURCE_DIR)
   configure_file(scripts/load-pretty-printers.in load-pretty-printers @ONLY)
 endif()

CONTRIBUTING.md

@@ -37,7 +37,7 @@ are required.
 To set up a development environment, follow the steps in the [README](https://github.com/rapidsai/rmm/blob/main/README.md) for cloning the repository and creating the conda environment.
 Once the environment is created, you can build and install RMM using
 ```bash
-$ python setup.py develop
+$ python -m pip install ./python
 ```
 This command will build the RMM Python library inside the clone and automatically make it importable when running Python anywhere on your machine.
 Remember, if you are unsure about anything, don't hesitate to comment on issues

README.md

@@ -61,13 +61,15 @@ Compiler requirements:
 
 CUDA/GPU requirements:
 
-* CUDA 11.4+
-* Pascal architecture or better
+* CUDA 11.4+. You can obtain CUDA from
+  [https://developer.nvidia.com/cuda-downloads](https://developer.nvidia.com/cuda-downloads)
 
-You can obtain CUDA from [https://developer.nvidia.com/cuda-downloads](https://developer.nvidia.com/cuda-downloads)
+GPU Support:
+* RMM is tested and supported only on Volta architecture and newer (Compute Capability 7.0+). It
+  may work on earlier architectures.
 
 Python requirements:
-* `scikit-build`
+* `scikit-build-core`
 * `cuda-python`
 * `cython`
 
@@ -125,8 +127,7 @@ $ make test
 
 - Build, install, and test the `rmm` python package, in the `python` folder:
 ```bash
-$ python setup.py build_ext --inplace
-$ python setup.py install
+$ python -m pip install -e ./python
 $ pytest -v
 ```
 
@@ -331,7 +332,9 @@ Accessing and modifying the default resource is done through two functions:
 ```c++
 rmm::mr::cuda_memory_resource cuda_mr;
 // Construct a resource that uses a coalescing best-fit pool allocator
-rmm::mr::pool_memory_resource<rmm::mr::cuda_memory_resource> pool_mr{&cuda_mr};
+// With the pool initially half of available device memory
+auto initial_size = rmm::percent_of_free_device_memory(50);
+rmm::mr::pool_memory_resource<rmm::mr::cuda_memory_resource> pool_mr{&cuda_mr, initial_size};
 rmm::mr::set_current_device_resource(&pool_mr); // Updates the current device resource pointer to `pool_mr`
 rmm::mr::device_memory_resource* mr = rmm::mr::get_current_device_resource(); // Points to `pool_mr`
 ```
@@ -350,11 +353,13 @@ per-device resources. Here is an example loop that creates `unique_ptr`s to `poo
 objects for each device and sets them as the per-device resource for that device.
 
 ```c++
-std::vector<unique_ptr<pool_memory_resource>> per_device_pools;
+using pool_mr = rmm::mr::pool_memory_resource<rmm::mr::cuda_memory_resource>;
+std::vector<unique_ptr<pool_mr>> per_device_pools;
 for(int i = 0; i < N; ++i) {
   cudaSetDevice(i); // set device i before creating MR
   // Use a vector of unique_ptr to maintain the lifetime of the MRs
-  per_device_pools.push_back(std::make_unique<pool_memory_resource>());
+  // Note: for brevity, omitting creation of upstream and computing initial_size
+  per_device_pools.push_back(std::make_unique<pool_mr>(upstream, initial_size));
   // Set the per-device resource for device i
   set_per_device_resource(cuda_device_id{i}, &per_device_pools.back());
 }

VERSION

@@ -1 +1 @@
-23.12.00
+24.02.00