Merge branch 'branch-25.02' into high-lvl-comp-api

.github/workflows/build.yaml

@@ -62,7 +62,7 @@ jobs:
       arch: "amd64"
       branch: ${{ inputs.branch }}
       build_type: ${{ inputs.build_type || 'branch' }}
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       date: ${{ inputs.date }}
       node_type: "gpu-v100-latest-1"
       run_script: "ci/build_docs.sh"

.github/workflows/pr.yaml

@@ -186,7 +186,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       run_script: "ci/test_java.sh"
   static-configure:
     needs: checks
@@ -207,7 +207,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       run_script: "ci/test_notebooks.sh"
   docs-build:
     needs: conda-python-build
@@ -217,7 +217,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       run_script: "ci/build_docs.sh"
   wheel-build-libcudf:
     needs: checks

.github/workflows/test.yaml

@@ -41,7 +41,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       run_script: "ci/test_cpp_memcheck.sh"
   static-configure:
     secrets: inherit
@@ -94,7 +94,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       run_script: "ci/test_java.sh"
   conda-notebook-tests:
     secrets: inherit
@@ -106,7 +106,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
+      container_image: "rapidsai/ci-conda:latest"
       run_script: "ci/test_notebooks.sh"
   wheel-tests-cudf:
     secrets: inherit

conda/environments/all_cuda-118_arch-x86_64.yaml

@@ -55,7 +55,7 @@ dependencies:
 - nbsphinx
 - ninja
 - notebook
-- numba-cuda>=0.0.13,<0.0.18
+- numba-cuda>=0.2.0,<0.3.0
 - numpy>=1.23,<3.0a0
 - numpydoc
 - nvcc_linux-64=11.8
@@ -66,7 +66,7 @@ dependencies:
 - pandas
 - pandas>=2.0,<2.2.4dev0
 - pandoc
-- polars>=1.11,<1.15
+- polars>=1.11,<1.18
 - pre-commit
 - ptxcompiler
 - pyarrow>=14.0.0,<19.0.0a0

conda/environments/all_cuda-125_arch-x86_64.yaml

@@ -54,7 +54,7 @@ dependencies:
 - nbsphinx
 - ninja
 - notebook
-- numba-cuda>=0.0.13,<0.0.18
+- numba-cuda>=0.2.0,<0.3.0
 - numpy>=1.23,<3.0a0
 - numpydoc
 - nvcomp==4.1.0.6
@@ -64,7 +64,7 @@ dependencies:
 - pandas
 - pandas>=2.0,<2.2.4dev0
 - pandoc
-- polars>=1.11,<1.15
+- polars>=1.11,<1.18
 - pre-commit
 - pyarrow>=14.0.0,<19.0.0a0
 - pydata-sphinx-theme!=0.14.2

conda/recipes/cudf-polars/meta.yaml

@@ -1,4 +1,4 @@
-# Copyright (c) 2024, NVIDIA CORPORATION.
+# Copyright (c) 2024-2025, NVIDIA CORPORATION.
 
 {% set version = environ['RAPIDS_PACKAGE_VERSION'].lstrip('v') %}
 {% set minor_version = version.split('.')[0] + '.' + version.split('.')[1] %}
@@ -43,7 +43,7 @@ requirements:
   run:
     - python
     - pylibcudf ={{ version }}
-    - polars >=1.11,<1.15
+    - polars >=1.11,<1.18
     - {{ pin_compatible('cuda-version', max_pin='x', min_pin='x') }}
 
 test:

conda/recipes/cudf/meta.yaml

@@ -1,4 +1,4 @@
-# Copyright (c) 2018-2024, NVIDIA CORPORATION.
+# Copyright (c) 2018-2025, NVIDIA CORPORATION.
 
 {% set version = environ['RAPIDS_PACKAGE_VERSION'].lstrip('v') %}
 {% set minor_version = version.split('.')[0] + '.' + version.split('.')[1] %}
@@ -80,7 +80,7 @@ requirements:
     - typing_extensions >=4.0.0
     - pandas >=2.0,<2.2.4dev0
     - cupy >=12.0.0
-    - numba-cuda >=0.0.13,<0.0.18
+    - numba-cuda >=0.2.0,<0.3.0
     - numpy >=1.23,<3.0a0
     - pyarrow>=14.0.0,<18.0.0a0
     - libcudf ={{ version }}

cpp/include/cudf/detail/utilities/integer_utils.hpp

@@ -1,7 +1,7 @@
 /*
  * Copyright 2019 BlazingDB, Inc.
  *     Copyright 2019 Eyal Rozenberg <[email protected]>
- * Copyright (c) 2020-2024, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -23,6 +23,8 @@
  */
 
 #include <cudf/fixed_point/temporary.hpp>
+#include <cudf/types.hpp>
+#include <cudf/utilities/error.hpp>
 
 #include <cmath>
 #include <cstdlib>
@@ -44,13 +46,17 @@ namespace util {
  * `modulus` is positive. The safety is in regard to rollover.
  */
 template <typename S>
-constexpr S round_up_safe(S number_to_round, S modulus)
+CUDF_HOST_DEVICE constexpr S round_up_safe(S number_to_round, S modulus)
 {
   auto remainder = number_to_round % modulus;
   if (remainder == 0) { return number_to_round; }
   auto rounded_up = number_to_round - remainder + modulus;
   if (rounded_up < number_to_round) {
-    throw std::invalid_argument("Attempt to round up beyond the type's maximum value");
+#ifndef __CUDA_ARCH__
+    CUDF_FAIL("Attempt to round up beyond the type's maximum value", cudf::data_type_error);
+#else
+    CUDF_UNREACHABLE("Attempt to round up beyond the type's maximum value");
+#endif
   }
   return rounded_up;
 }

cpp/include/cudf/utilities/span.hpp

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020-2024, NVIDIA CORPORATION.
+ * Copyright (c) 2020-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -197,11 +197,16 @@ struct host_span : public cudf::detail::span_base<T, Extent, host_span<T, Extent
 
   constexpr host_span() noexcept : base() {}  // required to compile on centos
 
-  /// Constructor from pointer and size
-  /// @param data Pointer to the first element in the span
-  /// @param size The number of elements in the span
-  /// @param is_device_accessible Whether the data is device accessible (e.g. pinned memory)
-  constexpr host_span(T* data, std::size_t size, bool is_device_accessible)
+  /**
+   * @brief Constructor from pointer and size
+   *
+   * @note This needs to be host-device , as it's used by a host-device function in base_2dspan
+   *
+   * @param data Pointer to the first element in the span
+   * @param size The number of elements in the span
+   * @param is_device_accessible Whether the data is device accessible (e.g. pinned memory)
+   */
+  CUDF_HOST_DEVICE constexpr host_span(T* data, std::size_t size, bool is_device_accessible)
     : base(data, size), _is_device_accessible{is_device_accessible}
   {
   }
@@ -311,8 +316,8 @@ struct host_span : public cudf::detail::span_base<T, Extent, host_span<T, Extent
    * @param count The number of elements in the subspan
    * @return A subspan of the sequence, of requested count and offset
    */
-  [[nodiscard]] constexpr host_span subspan(typename base::size_type offset,
-                                            typename base::size_type count) const noexcept
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr host_span subspan(
+    typename base::size_type offset, typename base::size_type count) const noexcept
   {
     return host_span{this->data() + offset, count, _is_device_accessible};
   }
@@ -434,8 +439,8 @@ struct device_span : public cudf::detail::span_base<T, Extent, device_span<T, Ex
    * @param count The number of elements in the subspan
    * @return A subspan of the sequence, of requested count and offset
    */
-  [[nodiscard]] constexpr device_span subspan(typename base::size_type offset,
-                                              typename base::size_type count) const noexcept
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr device_span subspan(
+    typename base::size_type offset, typename base::size_type count) const noexcept
   {
     return device_span{this->data() + offset, count};
   }
@@ -475,28 +480,28 @@ class base_2dspan {
    *
    * @return A pointer to the first element of the span
    */
-  [[nodiscard]] constexpr auto data() const noexcept { return _flat.data(); }
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr auto data() const noexcept { return _flat.data(); }
 
   /**
    * @brief Returns the size in the span as pair.
    *
    * @return pair representing rows and columns size of the span
    */
-  [[nodiscard]] constexpr auto size() const noexcept { return _size; }
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr auto size() const noexcept { return _size; }
 
   /**
    * @brief Returns the number of elements in the span.
    *
    * @return Number of elements in the span
    */
-  [[nodiscard]] constexpr auto count() const noexcept { return _flat.size(); }
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr auto count() const noexcept { return _flat.size(); }
 
   /**
    * @brief Checks if the span is empty.
    *
    * @return True if the span is empty, false otherwise
    */
-  [[nodiscard]] constexpr bool is_empty() const noexcept { return count() == 0; }
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr bool is_empty() const noexcept { return count() == 0; }
 
   /**
    * @brief Returns a reference to the row-th element of the sequence.
@@ -507,7 +512,7 @@ class base_2dspan {
    * @param row the index of the element to access
    * @return A reference to the row-th element of the sequence, i.e., `data()[row]`
    */
-  constexpr RowType<T, dynamic_extent> operator[](size_t row) const
+  CUDF_HOST_DEVICE constexpr RowType<T, dynamic_extent> operator[](size_t row) const
   {
     return _flat.subspan(row * _size.second, _size.second);
   }
@@ -517,7 +522,10 @@ class base_2dspan {
    *
    * @return A flattened span of the 2D span
    */
-  [[nodiscard]] constexpr RowType<T, dynamic_extent> flat_view() const { return _flat; }
+  [[nodiscard]] CUDF_HOST_DEVICE constexpr RowType<T, dynamic_extent> flat_view() const
+  {
+    return _flat;
+  }
 
   /**
    * @brief Construct a 2D span from another 2D span of convertible type