Merge branch 'branch-23.04' into fea/parallel_testing

rapidsai · Mar 22, 2023 · 97ff54b · 97ff54b
2 parents 23ac812 + bf18cea
commit 97ff54b
Show file tree

Hide file tree

Showing 146 changed files with 2,779 additions and 1,876 deletions.
diff --git a/build.sh b/build.sh
@@ -300,8 +300,7 @@ if buildAll || hasArg libcudf; then
     # Record build times
     if [[ "$BUILD_REPORT_METRICS" == "ON" && -f "${LIB_BUILD_DIR}/.ninja_log" ]]; then
         echo "Formatting build metrics"
-        python ${REPODIR}/cpp/scripts/sort_ninja_log.py ${LIB_BUILD_DIR}/.ninja_log --fmt xml > ${LIB_BUILD_DIR}/ninja_log.xml
-        MSG="<p>"
+        MSG=""
         # get some sccache stats after the compile
         if [[ "$BUILD_REPORT_INCL_CACHE_STATS" == "ON" && -x "$(command -v sccache)" ]]; then
            COMPILE_REQUESTS=$(sccache -s | grep "Compile requests \+ [0-9]\+$" | awk '{ print $NF }')
@@ -318,7 +317,9 @@ if buildAll || hasArg libcudf; then
         BMR_DIR=${RAPIDS_ARTIFACTS_DIR:-"${LIB_BUILD_DIR}"}
         echo "Metrics output dir: [$BMR_DIR]"
         mkdir -p ${BMR_DIR}
-        python ${REPODIR}/cpp/scripts/sort_ninja_log.py ${LIB_BUILD_DIR}/.ninja_log --fmt html --msg "$MSG" > ${BMR_DIR}/ninja_log.html
+        MSG_OUTFILE="$(mktemp)"
+        echo "$MSG" > "${MSG_OUTFILE}"
+        python ${REPODIR}/cpp/scripts/sort_ninja_log.py ${LIB_BUILD_DIR}/.ninja_log --fmt html --msg "${MSG_OUTFILE}" > ${BMR_DIR}/ninja_log.html
         cp ${LIB_BUILD_DIR}/.ninja_log ${BMR_DIR}/ninja.log
     fi
 

diff --git a/ci/build_cpp.sh b/ci/build_cpp.sh
@@ -14,29 +14,3 @@ rapids-logger "Begin cpp build"
 rapids-mamba-retry mambabuild conda/recipes/libcudf
 
 rapids-upload-conda-to-s3 cpp
-
-echo "++++++++++++++++++++++++++++++++++++++++++++"
-
-if [[ -d $RAPIDS_ARTIFACTS_DIR ]]; then
-  ls -l ${RAPIDS_ARTIFACTS_DIR}
-fi
-
-echo "++++++++++++++++++++++++++++++++++++++++++++"
-
-FILE=${RAPIDS_ARTIFACTS_DIR}/ninja.log
-if [[ -f $FILE ]]; then
-  echo -e "\x1B[33;1m\x1B[48;5;240m Ninja log for this build available at the following link \x1B[0m"
-  UPLOAD_NAME=cpp_cuda${RAPIDS_CUDA_VERSION%%.*}_$(arch).ninja.log
-  rapids-upload-to-s3 "${UPLOAD_NAME}" "${FILE}"
-fi
-
-echo "++++++++++++++++++++++++++++++++++++++++++++"
-
-FILE=${RAPIDS_ARTIFACTS_DIR}/ninja_log.html
-if [[ -f $FILE ]]; then
-  echo -e "\x1B[33;1m\x1B[48;5;240m Build Metrics Report for this build available at the following link \x1B[0m"
-  UPLOAD_NAME=cpp_cuda${RAPIDS_CUDA_VERSION%%.*}_$(arch).BuildMetricsReport.html
-  rapids-upload-to-s3 "${UPLOAD_NAME}" "${FILE}"
-fi
-
-echo "++++++++++++++++++++++++++++++++++++++++++++"
diff --git a/ci/release/apply_wheel_modifications.sh b/ci/release/apply_wheel_modifications.sh
@@ -6,12 +6,6 @@
 VERSION=${1}
 CUDA_SUFFIX=${2}
 
-# __init__.py versions
-sed -i "s/__version__ = .*/__version__ = \"${VERSION}\"/g" python/cudf/cudf/__init__.py
-sed -i "s/__version__ = .*/__version__ = \"${VERSION}\"/g" python/dask_cudf/dask_cudf/__init__.py
-sed -i "s/__version__ = .*/__version__ = \"${VERSION}\"/g" python/cudf_kafka/cudf_kafka/__init__.py
-sed -i "s/__version__ = .*/__version__ = \"${VERSION}\"/g" python/custreamz/custreamz/__init__.py
-
 # pyproject.toml versions
 sed -i "s/^version = .*/version = \"${VERSION}\"/g" python/cudf/pyproject.toml
 sed -i "s/^version = .*/version = \"${VERSION}\"/g" python/dask_cudf/pyproject.toml

diff --git a/conda/recipes/libcudf/meta.yaml b/conda/recipes/libcudf/meta.yaml
@@ -92,6 +92,7 @@ outputs:
         - test -f $PREFIX/include/cudf/concatenate.hpp
         - test -f $PREFIX/include/cudf/copying.hpp
         - test -f $PREFIX/include/cudf/datetime.hpp
+        - test -f $PREFIX/include/cudf/timezone.hpp
         - test -f $PREFIX/include/cudf/detail/aggregation/aggregation.hpp
         - test -f $PREFIX/include/cudf/detail/aggregation/result_cache.hpp
         - test -f $PREFIX/include/cudf/detail/binaryop.hpp
@@ -128,6 +129,8 @@ outputs:
         - test -f $PREFIX/include/cudf/detail/stream_compaction.hpp
         - test -f $PREFIX/include/cudf/detail/structs/utilities.hpp
         - test -f $PREFIX/include/cudf/detail/tdigest/tdigest.hpp
+        - test -f $PREFIX/include/cudf/detail/timezone.cuh
+        - test -f $PREFIX/include/cudf/detail/timezone.hpp
         - test -f $PREFIX/include/cudf/detail/transform.hpp
         - test -f $PREFIX/include/cudf/detail/transpose.hpp
         - test -f $PREFIX/include/cudf/detail/unary.hpp

diff --git a/conda/recipes/libcudf/post-link.sh b/conda/recipes/libcudf/post-link.sh
@@ -1,6 +1,6 @@
 #!/bin/bash
-# Copyright (c) 2022, NVIDIA CORPORATION.
+# Copyright (c) 2022-2023, NVIDIA CORPORATION.
 # Only add the license notice to libcudf and not our examples / tests
 if [[ "$PKG_NAME" == "libcudf" ]]; then
-  cat ./nvlink.txt >> $PREFIX/.messages.txt
+  cat ./nvcomp.txt >> $PREFIX/.messages.txt
 fi
diff --git a/cpp/CMakeLists.txt b/cpp/CMakeLists.txt
@@ -369,7 +369,7 @@ add_library(
   src/io/orc/stripe_data.cu
   src/io/orc/stripe_enc.cu
   src/io/orc/stripe_init.cu
-  src/io/orc/timezone.cpp
+  src/datetime/timezone.cpp
   src/io/orc/writer_impl.cu
   src/io/parquet/compact_protocol_reader.cpp
   src/io/parquet/compact_protocol_writer.cpp
@@ -890,31 +890,20 @@ install(
   EXPORT cudf-exports
 )
 
-install(DIRECTORY ${CUDF_SOURCE_DIR}/include/cudf ${CUDF_SOURCE_DIR}/include/cudf_test
-                  ${CUDF_SOURCE_DIR}/include/nvtext DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
-)
-
-if(CUDF_BUILD_TESTUTIL)
+set(_components_export_string)
+if(TARGET cudftestutil)
   install(
     TARGETS cudftest_default_stream cudftestutil
     DESTINATION ${lib_dir}
     EXPORT cudf-testing-exports
   )
-
-  install(
-    EXPORT cudf-testing-exports
-    FILE cudf-testing-targets.cmake
-    NAMESPACE cudf::
-    DESTINATION "${lib_dir}/cmake/cudf"
-  )
-
-  include("${rapids-cmake-dir}/export/write_dependencies.cmake")
-  rapids_export_write_dependencies(
-    INSTALL cudf-testing-exports
-    "${PROJECT_BINARY_DIR}/rapids-cmake/cudf/export/cudf-testing-dependencies.cmake"
-  )
+  set(_components_export_string COMPONENTS testing COMPONENTS_EXPORT_SET cudf-testing-exports)
 endif()
 
+install(DIRECTORY ${CUDF_SOURCE_DIR}/include/cudf ${CUDF_SOURCE_DIR}/include/cudf_test
+                  ${CUDF_SOURCE_DIR}/include/nvtext DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
+)
+
 if(CUDF_BUILD_STREAMS_TEST_UTIL)
   install(TARGETS cudf_identify_stream_usage_mode_cudf DESTINATION ${lib_dir})
   install(TARGETS cudf_identify_stream_usage_mode_testing DESTINATION ${lib_dir})
@@ -976,21 +965,15 @@ string(
   [=[
 if(testing IN_LIST cudf_FIND_COMPONENTS)
   enable_language(CUDA)
-  if(EXISTS "${CMAKE_CURRENT_LIST_DIR}/cudf-testing-dependencies.cmake")
-    include("${CMAKE_CURRENT_LIST_DIR}/cudf-testing-dependencies.cmake")
-  endif()
-  if(EXISTS "${CMAKE_CURRENT_LIST_DIR}/cudf-testing-targets.cmake")
-    include("${CMAKE_CURRENT_LIST_DIR}/cudf-testing-targets.cmake")
-  endif()
 endif()
 ]=]
 )
 string(APPEND install_code_string "${common_code_string}")
 
 rapids_export(
   INSTALL cudf
-  EXPORT_SET cudf-exports
-  GLOBAL_TARGETS cudf
+  EXPORT_SET cudf-exports ${_components_export_string}
+  GLOBAL_TARGETS cudf cudftestutil
   NAMESPACE cudf::
   DOCUMENTATION doc_string
   FINAL_CODE_BLOCK install_code_string
@@ -1013,23 +996,13 @@ string(APPEND build_code_string "${common_code_string}")
 
 rapids_export(
   BUILD cudf
-  EXPORT_SET cudf-exports
-  GLOBAL_TARGETS cudf
+  EXPORT_SET cudf-exports ${_components_export_string}
+  GLOBAL_TARGETS cudf cudftestutil
   NAMESPACE cudf::
   DOCUMENTATION doc_string
   FINAL_CODE_BLOCK build_code_string
 )
 
-if(CUDF_BUILD_TESTUTIL)
-  export(
-    EXPORT cudf-testing-exports
-    FILE ${CUDF_BINARY_DIR}/cudf-testing-targets.cmake
-    NAMESPACE cudf::
-  )
-  rapids_export_write_dependencies(
-    BUILD cudf-testing-exports "${CUDF_BINARY_DIR}/cudf-testing-dependencies.cmake"
-  )
-endif()
 # ##################################################################################################
 # * make documentation ----------------------------------------------------------------------------
 

diff --git a/cpp/benchmarks/common/generate_input.cu b/cpp/benchmarks/common/generate_input.cu
@@ -430,8 +430,12 @@ std::unique_ptr<cudf::column> create_random_column(data_profile const& profile,
                    null_mask.begin());
   }
 
-  auto [result_bitmask, null_count] = cudf::detail::valid_if(
-    null_mask.begin(), null_mask.end(), thrust::identity<bool>{}, cudf::get_default_stream());
+  auto [result_bitmask, null_count] =
+    cudf::detail::valid_if(null_mask.begin(),
+                           null_mask.end(),
+                           thrust::identity<bool>{},
+                           cudf::get_default_stream(),
+                           rmm::mr::get_current_device_resource());
 
   return std::make_unique<cudf::column>(
     dtype,
@@ -509,8 +513,12 @@ std::unique_ptr<cudf::column> create_random_utf8_string_column(data_profile cons
                      thrust::make_zip_iterator(offsets.begin(), offsets.begin() + 1),
                      num_rows,
                      string_generator{chars.data(), engine});
-  auto [result_bitmask, null_count] = cudf::detail::valid_if(
-    null_mask.begin(), null_mask.end() - 1, thrust::identity<bool>{}, cudf::get_default_stream());
+  auto [result_bitmask, null_count] =
+    cudf::detail::valid_if(null_mask.begin(),
+                           null_mask.end() - 1,
+                           thrust::identity<bool>{},
+                           cudf::get_default_stream(),
+                           rmm::mr::get_current_device_resource());
   return cudf::make_strings_column(
     num_rows,
     std::move(offsets),
@@ -628,8 +636,11 @@ std::unique_ptr<cudf::column> create_random_column<cudf::struct_view>(data_profi
       auto [null_mask, null_count] = [&]() {
         if (profile.get_null_probability().has_value()) {
           auto valids = valid_dist(engine, num_rows);
-          return cudf::detail::valid_if(
-            valids.begin(), valids.end(), thrust::identity<bool>{}, cudf::get_default_stream());
+          return cudf::detail::valid_if(valids.begin(),
+                                        valids.end(),
+                                        thrust::identity<bool>{},
+                                        cudf::get_default_stream(),
+                                        rmm::mr::get_current_device_resource());
         }
         return std::pair<rmm::device_buffer, cudf::size_type>{};
       }();
@@ -712,9 +723,12 @@ std::unique_ptr<cudf::column> create_random_column<cudf::list_view>(data_profile
     auto offsets_column = std::make_unique<cudf::column>(
       cudf::data_type{cudf::type_id::INT32}, num_rows + 1, offsets.release());
 
-    auto [null_mask, null_count] = cudf::detail::valid_if(
-      valids.begin(), valids.end(), thrust::identity<bool>{}, cudf::get_default_stream());
-    list_column = cudf::make_lists_column(
+    auto [null_mask, null_count] = cudf::detail::valid_if(valids.begin(),
+                                                          valids.end(),
+                                                          thrust::identity<bool>{},
+                                                          cudf::get_default_stream(),
+                                                          rmm::mr::get_current_device_resource());
+    list_column                  = cudf::make_lists_column(
       num_rows,
       std::move(offsets_column),
       std::move(current_child_column),
@@ -840,7 +854,8 @@ std::pair<rmm::device_buffer, cudf::size_type> create_random_null_mask(
     return cudf::detail::valid_if(thrust::make_counting_iterator<cudf::size_type>(0),
                                   thrust::make_counting_iterator<cudf::size_type>(size),
                                   bool_generator{seed, 1.0 - *null_probability},
-                                  cudf::get_default_stream());
+                                  cudf::get_default_stream(),
+                                  rmm::mr::get_current_device_resource());
   }
 }
 

diff --git a/cpp/benchmarks/iterator/iterator.cu b/cpp/benchmarks/iterator/iterator.cu
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2019-2022, NVIDIA CORPORATION.
+ * Copyright (c) 2019-2023, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -140,8 +140,8 @@ void BM_iterator(benchmark::State& state)
   cudf::column_view hasnull_F = wrap_hasnull_F;
 
   // Initialize dev_result to false
-  auto dev_result =
-    cudf::detail::make_zeroed_device_uvector_sync<TypeParam>(1, cudf::get_default_stream());
+  auto dev_result = cudf::detail::make_zeroed_device_uvector_sync<TypeParam>(
+    1, cudf::get_default_stream(), rmm::mr::get_current_device_resource());
   for (auto _ : state) {
     cuda_event_timer raii(state, true);  // flush_l2_cache = true, stream = 0
     if (cub_or_thrust) {
@@ -210,7 +210,7 @@ void BM_pair_iterator(benchmark::State& state)
 
   // Initialize dev_result to false
   auto dev_result = cudf::detail::make_zeroed_device_uvector_sync<thrust::pair<T, bool>>(
-    1, cudf::get_default_stream());
+    1, cudf::get_default_stream(), rmm::mr::get_current_device_resource());
   for (auto _ : state) {
     cuda_event_timer raii(state, true);  // flush_l2_cache = true, stream = 0
     if (cub_or_thrust) {

diff --git a/cpp/benchmarks/join/join_common.hpp b/cpp/benchmarks/join/join_common.hpp
@@ -104,8 +104,11 @@ void BM_join(state_type& state, Join JoinFunc)
     // roughly 75% nulls
     auto validity =
       thrust::make_transform_iterator(thrust::make_counting_iterator(0), null75_generator{});
-    return cudf::detail::valid_if(
-             validity, validity + size, thrust::identity<bool>{}, cudf::get_default_stream())
+    return cudf::detail::valid_if(validity,
+                                  validity + size,
+                                  thrust::identity<bool>{},
+                                  cudf::get_default_stream(),
+                                  rmm::mr::get_current_device_resource())
       .first;
   };