Merge branch 'branch-24.10' into doc/pylibcudf-doc

.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:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       date: ${{ inputs.date }}
       node_type: "gpu-v100-latest-1"
       run_script: "ci/build_docs.sh"

.github/workflows/pr.yaml

@@ -160,7 +160,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_java.sh"
   static-configure:
     needs: checks
@@ -181,7 +181,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_notebooks.sh"
   docs-build:
     needs: conda-python-build
@@ -191,7 +191,7 @@ jobs:
       build_type: pull-request
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       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:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_cpp_memcheck.sh"
   static-configure:
     secrets: inherit
@@ -81,7 +81,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_java.sh"
   conda-notebook-tests:
     secrets: inherit
@@ -93,7 +93,7 @@ jobs:
       sha: ${{ inputs.sha }}
       node_type: "gpu-v100-latest-1"
       arch: "amd64"
-      container_image: "rapidsai/ci-conda:latest"
+      container_image: "rapidsai/ci-conda:cuda12.5.1-ubuntu22.04-py3.11"
       run_script: "ci/test_notebooks.sh"
   wheel-tests-cudf:
     secrets: inherit

cpp/include/cudf/datetime.hpp

@@ -17,9 +17,12 @@
 #pragma once
 
 #include <cudf/types.hpp>
+#include <cudf/utilities/default_stream.hpp>
 #include <cudf/utilities/export.hpp>
 #include <cudf/utilities/memory_resource.hpp>
 
+#include <rmm/cuda_stream_view.hpp>
+
 #include <memory>
 
 /**
@@ -40,97 +43,111 @@ namespace datetime {
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t years
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_year(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Extracts month from any datetime type and returns an int16_t
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t months
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_month(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Extracts day from any datetime type and returns an int16_t
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t days
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_day(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Extracts a weekday from any datetime type and returns an int16_t
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t days
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_weekday(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Extracts hour from any datetime type and returns an int16_t
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t hours
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_hour(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Extracts minute from any datetime type and returns an int16_t
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t minutes
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_minute(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Extracts second from any datetime type and returns an int16_t
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t seconds
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_second(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -141,13 +158,15 @@ std::unique_ptr<cudf::column> extract_second(
  * For example, the millisecond fraction of 1.234567890 seconds is 234.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t milliseconds
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_millisecond_fraction(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -158,13 +177,15 @@ std::unique_ptr<cudf::column> extract_millisecond_fraction(
  * For example, the microsecond fraction of 1.234567890 seconds is 567.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t microseconds
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_microsecond_fraction(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -175,13 +196,15 @@ std::unique_ptr<cudf::column> extract_microsecond_fraction(
  * For example, the nanosecond fraction of 1.234567890 seconds is 890.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of the extracted int16_t nanoseconds
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> extract_nanosecond_fraction(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /** @} */  // end of group
@@ -196,27 +219,31 @@ std::unique_ptr<cudf::column> extract_nanosecond_fraction(
  * cudf::column.
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column containing last day of the month as TIMESTAMP_DAYS
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP
  */
 std::unique_ptr<cudf::column> last_day_of_month(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief  Computes the day number since the start of the year from the datetime and
  * returns an int16_t cudf::column. The value is between [1, {365-366}]
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of datatype INT16 containing the day number since the start of the year
  * @throw cudf::logic_error if input column datatype is not a TIMESTAMP
  */
 std::unique_ptr<cudf::column> day_of_year(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -245,13 +272,15 @@ std::unique_ptr<cudf::column> day_of_year(
  *
  * @param timestamps cudf::column_view of timestamp type
  * @param months cudf::column_view of integer type containing the number of months to add
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of timestamp type containing the computed timestamps
  */
 std::unique_ptr<cudf::column> add_calendrical_months(
   cudf::column_view const& timestamps,
   cudf::column_view const& months,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -280,13 +309,15 @@ std::unique_ptr<cudf::column> add_calendrical_months(
  *
  * @param timestamps cudf::column_view of timestamp type
  * @param months cudf::scalar of integer type containing the number of months to add
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @return cudf::column of timestamp type containing the computed timestamps
  */
 std::unique_ptr<cudf::column> add_calendrical_months(
   cudf::column_view const& timestamps,
   cudf::scalar const& months,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -297,13 +328,15 @@ std::unique_ptr<cudf::column> add_calendrical_months(
  * `output[i] is null` if `column[i]` is null
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @returns cudf::column of datatype BOOL8 truth value of the corresponding date
  * @throw cudf::logic_error if input column datatype is not a TIMESTAMP
  */
 std::unique_ptr<cudf::column> is_leap_year(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -315,11 +348,13 @@ std::unique_ptr<cudf::column> is_leap_year(
  * @throw cudf::logic_error if input column datatype is not a TIMESTAMP
  *
  * @param column cudf::column_view of the input datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  * @return cudf::column of datatype INT16 of days in month of the corresponding date
  */
 std::unique_ptr<cudf::column> days_in_month(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -331,11 +366,13 @@ std::unique_ptr<cudf::column> days_in_month(
  * @throw cudf::logic_error if input column datatype is not a TIMESTAMP
  *
  * @param column The input column containing datetime values
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  * @return A column of INT16 type indicating which quarter the date is in
  */
 std::unique_ptr<cudf::column> extract_quarter(
   cudf::column_view const& column,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
@@ -357,6 +394,7 @@ enum class rounding_frequency : int32_t {
  *
  * @param column cudf::column_view of the input datetime values
  * @param freq rounding_frequency indicating the frequency to round up to
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP.
@@ -365,13 +403,15 @@ enum class rounding_frequency : int32_t {
 std::unique_ptr<cudf::column> ceil_datetimes(
   cudf::column_view const& column,
   rounding_frequency freq,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief Round datetimes down to the nearest multiple of the given frequency.
  *
  * @param column cudf::column_view of the input datetime values
  * @param freq rounding_frequency indicating the frequency to round down to
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP.
@@ -380,13 +420,15 @@ std::unique_ptr<cudf::column> ceil_datetimes(
 std::unique_ptr<cudf::column> floor_datetimes(
   cudf::column_view const& column,
   rounding_frequency freq,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /**
  * @brief Round datetimes to the nearest multiple of the given frequency.
  *
  * @param column cudf::column_view of the input datetime values
  * @param freq rounding_frequency indicating the frequency to round to
+ * @param stream CUDA stream used for device memory operations and kernel launches
  * @param mr Device memory resource used to allocate device memory of the returned column
  *
  * @throw cudf::logic_error if input column datatype is not TIMESTAMP.
@@ -395,6 +437,7 @@ std::unique_ptr<cudf::column> floor_datetimes(
 std::unique_ptr<cudf::column> round_datetimes(
   cudf::column_view const& column,
   rounding_frequency freq,
+  rmm::cuda_stream_view stream      = cudf::get_default_stream(),
   rmm::device_async_resource_ref mr = cudf::get_current_device_resource_ref());
 
 /** @} */  // end of group