Merge pull request #15810 from vyasr/branch-24.08-merge-branch-24.06

Branch 24.08 merge branch 24.06
rapidsai · May 22, 2024 · acedfc0 · acedfc0
2 parents 333718a + 6aae787
commit acedfc0
Show file tree

Hide file tree

Showing 15 changed files with 393 additions and 39 deletions.
diff --git a/conda/environments/all_cuda-118_arch-x86_64.yaml b/conda/environments/all_cuda-118_arch-x86_64.yaml
@@ -36,15 +36,15 @@ dependencies:
 - hypothesis
 - identify>=2.5.20
 - ipython
-- libarrow-acero==16.0.0.*
-- libarrow-dataset==16.0.0.*
-- libarrow==16.0.0.*
+- libarrow-acero==16.1.0.*
+- libarrow-dataset==16.1.0.*
+- libarrow==16.1.0.*
 - libcufile-dev=1.4.0.31
 - libcufile=1.4.0.31
 - libcurand-dev=10.3.0.86
 - libcurand=10.3.0.86
 - libkvikio==24.8.*
-- libparquet==16.0.0.*
+- libparquet==16.1.0.*
 - librdkafka>=1.9.0,<1.10.0a0
 - librmm==24.8.*
 - make
@@ -66,7 +66,7 @@ dependencies:
 - pip
 - pre-commit
 - ptxcompiler
-- pyarrow==16.0.0.*
+- pyarrow==16.1.0.*
 - pydata-sphinx-theme!=0.14.2
 - pytest-benchmark
 - pytest-cases>=3.8.2

diff --git a/conda/environments/all_cuda-122_arch-x86_64.yaml b/conda/environments/all_cuda-122_arch-x86_64.yaml
@@ -37,13 +37,13 @@ dependencies:
 - hypothesis
 - identify>=2.5.20
 - ipython
-- libarrow-acero==16.0.0.*
-- libarrow-dataset==16.0.0.*
-- libarrow==16.0.0.*
+- libarrow-acero==16.1.0.*
+- libarrow-dataset==16.1.0.*
+- libarrow==16.1.0.*
 - libcufile-dev
 - libcurand-dev
 - libkvikio==24.8.*
-- libparquet==16.0.0.*
+- libparquet==16.1.0.*
 - librdkafka>=1.9.0,<1.10.0a0
 - librmm==24.8.*
 - make
@@ -63,7 +63,7 @@ dependencies:
 - pandoc
 - pip
 - pre-commit
-- pyarrow==16.0.0.*
+- pyarrow==16.1.0.*
 - pydata-sphinx-theme!=0.14.2
 - pynvjitlink
 - pytest-benchmark

diff --git a/conda/recipes/cudf/meta.yaml b/conda/recipes/cudf/meta.yaml
@@ -64,7 +64,7 @@ requirements:
     - scikit-build-core >=0.7.0
     - dlpack >=0.8,<1.0
     - numpy 1.23
-    - pyarrow ==16.0.0.*
+    - pyarrow ==16.1.0.*
     - libcudf ={{ version }}
     - rmm ={{ minor_version }}
     {% if cuda_major == "11" %}
@@ -82,7 +82,7 @@ requirements:
     - cupy >=12.0.0
     - numba >=0.57
     - {{ pin_compatible('numpy', max_pin='x') }}
-    - {{ pin_compatible('pyarrow', max_pin='x') }}
+    - {{ pin_compatible('pyarrow', max_pin='x.x') }}
     - libcudf ={{ version }}
     - {{ pin_compatible('rmm', max_pin='x.x') }}
     - fsspec >=0.6.0

diff --git a/conda/recipes/libcudf/conda_build_config.yaml b/conda/recipes/libcudf/conda_build_config.yaml
@@ -20,7 +20,7 @@ cmake_version:
   - ">=3.26.4"
 
 libarrow_version:
-  - "==16.0.0"
+  - "==16.1.0"
 
 dlpack_version:
   - ">=0.8,<1.0"

diff --git a/conda/recipes/libcudf/meta.yaml b/conda/recipes/libcudf/meta.yaml
@@ -86,9 +86,6 @@ outputs:
         {% else %}
         - {{ compiler('cuda') }}
         {% endif %}
-        # TODO: start taking libarrow's run exports again wwhen they're correct for 16.0
-        # ref: https://github.com/conda-forge/arrow-cpp-feedstock/issues/1418
-        - libarrow
     requirements:
       build:
         - cmake {{ cmake_version }}
@@ -108,12 +105,6 @@ outputs:
         - librmm ={{ minor_version }}
         - libkvikio ={{ minor_version }}
         - dlpack {{ dlpack_version }}
-        # TODO: start taking libarrow's run exports again wwhen they're correct for 16.0
-        # ref: https://github.com/conda-forge/arrow-cpp-feedstock/issues/1418
-        - libarrow>=16.0.0,<16.1.0a0
-        - libarrow-acero>=16.0.0,<16.1.0a0
-        - libarrow-dataset>=16.0.0,<16.1.0a0
-        - libparquet>=16.0.0,<16.1.0a0
     test:
       commands:
         - test -f $PREFIX/lib/libcudf.so

diff --git a/cpp/benchmarks/CMakeLists.txt b/cpp/benchmarks/CMakeLists.txt
@@ -256,6 +256,11 @@ ConfigureNVBench(
   PARQUET_READER_NVBENCH io/parquet/parquet_reader_input.cpp io/parquet/parquet_reader_options.cpp
 )
 
+# ##################################################################################################
+# * parquet multithread reader benchmark
+# ----------------------------------------------------------------------
+ConfigureNVBench(PARQUET_MULTITHREAD_READER_NVBENCH io/parquet/parquet_reader_multithread.cpp)
+
 # ##################################################################################################
 # * orc reader benchmark --------------------------------------------------------------------------
 ConfigureNVBench(ORC_READER_NVBENCH io/orc/orc_reader_input.cpp io/orc/orc_reader_options.cpp)

diff --git a/cpp/benchmarks/io/cuio_common.hpp b/cpp/benchmarks/io/cuio_common.hpp
@@ -39,6 +39,10 @@ class cuio_source_sink_pair {
     // delete the temporary file
     std::remove(file_name.c_str());
   }
+  // move constructor
+  cuio_source_sink_pair(cuio_source_sink_pair&& ss)            = default;
+  cuio_source_sink_pair& operator=(cuio_source_sink_pair&& ss) = default;
+
   /**
    * @brief Created a source info of the set type
    *