Add stream parameter to Set Operations (Public List APIs)

[SurajAralihalli]

This PR marks the conclusion of the List API series. The PR introduces the stream parameter to the Set operations (Public List Comparison and Intersection APIs).

Comparison and Intersection (set_operations.hpp)

have_overlap
intersect_distinct
union_distinct
difference_distinct

Reference 13744

Checklist

• I am familiar with the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[ttnghia]

Those are not quite "list" operations. Indeed, they are set operations (set is part of list, but saying "list" means something else).

[SurajAralihalli]

Thanks @ttnghia. I have updated the PR title and description.

[ttnghia]

/ok to test

[shrshi]

Looks good to me, thanks!

[ttnghia]

/ok to test

[SurajAralihalli]

The test for pr / conda-cpp-build / build (11.8.0, ubuntu22.04, amd64, 3.10) failed with the following error. This error didn't show up during local build and testing because it originates from the benchmarks suite. I'm moving this PR to draft and will investigate the issue further.

$SRC_DIR/cpp/benchmarks/lists/set_operations.cpp:53:10: error: use of deleted function 'constexpr rmm::cuda_stream_view::cuda_stream_view(int)'
   53 |     bfunc(cudf::lists_column_view{*lhs},
      |     ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   54 |           cudf::lists_column_view{*rhs},
      |           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   55 |           cudf::null_equality::EQUAL,
      |           ~~~~~~~~~~~~~~~~~~~~~~~~~~~
   56 |           cudf::nan_equality::ALL_EQUAL,
      |           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   57 |           rmm::mr::get_current_device_resource());
      |           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from $PREFIX/include/rmm/device_buffer.hpp:18,
                 from $SRC_DIR/cpp/include/cudf/utilities/span.hpp:19,
                 from $SRC_DIR/cpp/include/cudf/column/column_view.hpp:20,
                 from $SRC_DIR/cpp/include/cudf/column/column.hpp:18,
                 from $SRC_DIR/cpp/include/cudf/table/table.hpp:18,
                 from $SRC_DIR/cpp/benchmarks/common/generate_input.hpp:19,
                 from $SRC_DIR/cpp/benchmarks/lists/set_operations.cpp:17:
$PREFIX/include/rmm/cuda_stream_view.hpp:51:13: note: declared here
   51 |   constexpr cuda_stream_view(int)            = delete;  //< Prevent cast from 0

[davidwendt]

To reproduce locally, add the -DBUILD_BENCHMARKS=ON option to your local cmake build.

[ttnghia]

Just to update the benchmark code to add stream param.
This is when CI/CD shows its advantages.

[SurajAralihalli]

Thanks @davidwendt @ttnghia, I'll do that!

[davidwendt]

/ok to test

[ttnghia]

/merge