ARROW-17767: [Java][ORC] Move JNI build configuration from cpp/ to ja…

…va/ (#14163) Authored-by: Sutou Kouhei <[email protected]> Signed-off-by: Sutou Kouhei <[email protected]>
apache · Sep 19, 2022 · 59b5728 · 59b5728
1 parent a866f2f
commit 59b5728
Show file tree

Hide file tree

Showing 18 changed files with 98 additions and 233 deletions.
diff --git a/ci/docker/java-jni-manylinux-201x.dockerfile b/ci/docker/java-jni-manylinux-201x.dockerfile
@@ -40,6 +40,6 @@ ENV JAVA_HOME=/usr/lib/jvm/java-$java-openjdk/
 # For ci/scripts/{cpp,java}_*.sh
 ENV ARROW_HOME=/tmp/local \
     ARROW_JAVA_CDATA=ON \
-    ARROW_JNI=ON \
+    ARROW_JAVA_JNI=ON \
     ARROW_PLASMA=ON \
     ARROW_USE_CCACHE=ON
diff --git a/ci/docker/linux-apt-jni.dockerfile b/ci/docker/linux-apt-jni.dockerfile
@@ -76,7 +76,7 @@ ENV ARROW_BUILD_TESTS=ON \
     ARROW_GANDIVA=ON \
     ARROW_HOME=/usr/local \
     ARROW_JAVA_CDATA=ON \
-    ARROW_JNI=ON \
+    ARROW_JAVA_JNI=ON \
     ARROW_ORC=ON \
     ARROW_PARQUET=ON \
     ARROW_PLASMA_JAVA_CLIENT=ON \

diff --git a/ci/scripts/cpp_build.sh b/ci/scripts/cpp_build.sh
@@ -97,14 +97,12 @@ cmake \
   -DARROW_HDFS=${ARROW_HDFS:-ON} \
   -DARROW_INSTALL_NAME_RPATH=${ARROW_INSTALL_NAME_RPATH:-ON} \
   -DARROW_JEMALLOC=${ARROW_JEMALLOC:-ON} \
-  -DARROW_JNI=${ARROW_JNI:-OFF} \
   -DARROW_JSON=${ARROW_JSON:-ON} \
   -DARROW_LARGE_MEMORY_TESTS=${ARROW_LARGE_MEMORY_TESTS:-OFF} \
   -DARROW_MIMALLOC=${ARROW_MIMALLOC:-OFF} \
   -DARROW_NO_DEPRECATED_API=${ARROW_NO_DEPRECATED_API:-OFF} \
   -DARROW_ORC=${ARROW_ORC:-OFF} \
   -DARROW_PARQUET=${ARROW_PARQUET:-OFF} \
-  -DARROW_PLASMA_JAVA_CLIENT=${ARROW_PLASMA_JAVA_CLIENT:-OFF} \
   -DARROW_PLASMA=${ARROW_PLASMA:-OFF} \
   -DARROW_PYTHON=${ARROW_PYTHON:-OFF} \
   -DARROW_RUNTIME_SIMD_LEVEL=${ARROW_RUNTIME_SIMD_LEVEL:-MAX} \

diff --git a/ci/scripts/java_jni_build.sh b/ci/scripts/java_jni_build.sh
@@ -53,11 +53,12 @@ cmake \
   -DARROW_JAVA_JNI_ENABLE_DATASET=${ARROW_DATASET:-ON} \
   -DARROW_JAVA_JNI_ENABLE_GANDIVA=${ARROW_GANDIVA:-ON} \
   -DARROW_JAVA_JNI_ENABLE_ORC=${ARROW_ORC:-ON} \
+  -DARROW_JAVA_JNI_ENABLE_PLASMA=${ARROW_PLASMA:-ON} \
   -DBUILD_TESTING=${ARROW_JAVA_BUILD_TESTS} \
   -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE} \
   -DCMAKE_PREFIX_PATH=${arrow_install_dir} \
-  -DCMAKE_INSTALL_PREFIX=${prefix_dir} \
   -DCMAKE_INSTALL_LIBDIR=lib \
+  -DCMAKE_INSTALL_PREFIX=${prefix_dir} \
   -DCMAKE_UNITY_BUILD=${CMAKE_UNITY_BUILD:-OFF} \
   -GNinja \
   ${JAVA_JNI_CMAKE_ARGS:-} \

diff --git a/ci/scripts/java_jni_macos_build.sh b/ci/scripts/java_jni_macos_build.sh
@@ -58,18 +58,15 @@ pushd "${build_dir}/cpp"
 cmake \
   -DARROW_BUILD_SHARED=OFF \
   -DARROW_BUILD_TESTS=${ARROW_BUILD_TESTS} \
-  -DARROW_BUILD_UTILITIES=OFF \
   -DARROW_CSV=${ARROW_DATASET} \
   -DARROW_DATASET=${ARROW_DATASET} \
   -DARROW_DEPENDENCY_USE_SHARED=OFF \
   -DARROW_FILESYSTEM=${ARROW_FILESYSTEM} \
   -DARROW_GANDIVA=${ARROW_GANDIVA} \
   -DARROW_GANDIVA_STATIC_LIBSTDCPP=ON \
-  -DARROW_JNI=ON \
   -DARROW_ORC=${ARROW_ORC} \
   -DARROW_PARQUET=${ARROW_PARQUET} \
   -DARROW_PLASMA=${ARROW_PLASMA} \
-  -DARROW_PLASMA_JAVA_CLIENT=${ARROW_PLASMA_JAVA_CLIENT} \
   -DARROW_S3=${ARROW_S3} \
   -DARROW_USE_CCACHE=${ARROW_USE_CCACHE} \
   -DAWSSDK_SOURCE=BUNDLED \
@@ -112,12 +109,8 @@ if [ "${ARROW_USE_CCACHE}" == "ON" ]; then
   ccache -s
 fi
 
-echo "=== Copying libraries to the distribution folder ==="
-mkdir -p "${dist_dir}"
-cp -L ${build_dir}/cpp/*/libplasma_java.dylib ${dist_dir}
 
 echo "=== Checking shared dependencies for libraries ==="
-
 pushd ${dist_dir}
 archery linking check-dependencies \
   --allow CoreFoundation \

diff --git a/ci/scripts/java_jni_manylinux_build.sh b/ci/scripts/java_jni_manylinux_build.sh
@@ -66,7 +66,6 @@ pushd "${build_dir}/cpp"
 cmake \
   -DARROW_BUILD_SHARED=OFF \
   -DARROW_BUILD_TESTS=ON \
-  -DARROW_BUILD_UTILITIES=OFF \
   -DARROW_CSV=${ARROW_DATASET} \
   -DARROW_DATASET=${ARROW_DATASET} \
   -DARROW_DEPENDENCY_SOURCE="VCPKG" \
@@ -75,10 +74,8 @@ cmake \
   -DARROW_GANDIVA_PC_CXX_FLAGS=${GANDIVA_CXX_FLAGS} \
   -DARROW_GANDIVA=${ARROW_GANDIVA} \
   -DARROW_JEMALLOC=${ARROW_JEMALLOC} \
-  -DARROW_JNI=ON \
   -DARROW_ORC=${ARROW_ORC} \
   -DARROW_PARQUET=${ARROW_PARQUET} \
-  -DARROW_PLASMA_JAVA_CLIENT=${ARROW_PLASMA_JAVA_CLIENT} \
   -DARROW_PLASMA=${ARROW_PLASMA} \
   -DARROW_RPATH_ORIGIN=${ARROW_RPATH_ORIGIN} \
   -DARROW_S3=${ARROW_S3} \
@@ -133,11 +130,7 @@ if [ "${ARROW_USE_CCACHE}" == "ON" ]; then
 fi
 
 
-echo "=== Copying libraries to the distribution folder ==="
-cp -L ${build_dir}/cpp/*/libplasma_java.so ${dist_dir}
-
 echo "=== Checking shared dependencies for libraries ==="
-
 pushd ${dist_dir}
 archery linking check-dependencies \
   --allow ld-linux-x86-64 \

diff --git a/ci/scripts/java_test.sh b/ci/scripts/java_test.sh
@@ -38,20 +38,36 @@ pushd ${source_dir}
 
 ${mvn} test
 
-if [ "${ARROW_JNI}" = "ON" ]; then
-  ${mvn} test -Parrow-jni -pl adapter/orc,gandiva,dataset -Darrow.cpp.build.dir=${java_jni_dist_dir}
+projects=()
+if [ "${ARROW_DATASET}" = "ON" ]; then
+  projects+=(gandiva)
 fi
+if [ "${ARROW_GANDIVA}" = "ON" ]; then
+  projects+=(gandiva)
+fi
+if [ "${ARROW_ORC}" = "ON" ]; then
+  projects+=(adapter/orc)
+fi
+if [ "${ARROW_PLASMA}" = "ON" ]; then
+  projects+=(plasma)
+fi
+if [ "${#projects[@]}" -gt 0 ]; then
+  ${mvn} test \
+         -Parrow-jni \
+         -pl $(IFS=,; echo "${projects[*]}") \
+         -Darrow.cpp.build.dir=${java_jni_dist_dir}
 
-if [ "${ARROW_JAVA_CDATA}" = "ON" ]; then
-  ${mvn} test -Parrow-c-data -pl c -Darrow.c.jni.dist.dir=${java_jni_dist_dir}
+  if [ "${ARROW_PLASMA}" = "ON" ]; then
+    pushd ${source_dir}/plasma
+    java -cp target/test-classes:target/classes \
+         -Djava.library.path=${java_jni_dist_dir} \
+         org.apache.arrow.plasma.PlasmaClientTest
+    popd
+  fi
 fi
 
-if [ "${ARROW_PLASMA}" = "ON" ]; then
-  pushd ${source_dir}/plasma
-  java -cp target/test-classes:target/classes \
-       -Djava.library.path=${java_jni_dist_dir} \
-       org.apache.arrow.plasma.PlasmaClientTest
-  popd
+if [ "${ARROW_JAVA_CDATA}" = "ON" ]; then
+  ${mvn} test -Parrow-c-data -pl c -Darrow.c.jni.dist.dir=${java_jni_dist_dir}
 fi
 
 popd
diff --git a/cpp/CMakeLists.txt b/cpp/CMakeLists.txt
@@ -425,10 +425,6 @@ if(MSVC_TOOLCHAIN)
   set(ARROW_USE_GLOG OFF)
 endif()
 
-if(ARROW_JNI)
-  set(ARROW_BUILD_STATIC ON)
-endif()
-
 if(ARROW_ORC)
   set(ARROW_WITH_LZ4 ON)
   set(ARROW_WITH_SNAPPY ON)

diff --git a/cpp/cmake_modules/DefineOptions.cmake b/cpp/cmake_modules/DefineOptions.cmake
@@ -269,8 +269,6 @@ takes precedence over ccache if a storage backend is configured" ON)
     define_option(ARROW_JEMALLOC ${ARROW_JEMALLOC_DESCRIPTION} ON)
   endif()
 
-  define_option(ARROW_JNI "Build the Arrow JNI lib" OFF)
-
   define_option(ARROW_JSON "Build Arrow with JSON support (requires RapidJSON)" OFF)
 
   define_option(ARROW_MIMALLOC "Build the Arrow mimalloc-based allocator" OFF)
@@ -281,8 +279,6 @@ takes precedence over ccache if a storage backend is configured" ON)
 
   define_option(ARROW_PLASMA "Build the plasma object store along with Arrow" OFF)
 
-  define_option(ARROW_PLASMA_JAVA_CLIENT "Build the plasma object store java client" OFF)
-
   define_option(ARROW_PYTHON "Build the Arrow CPython extensions" OFF)
 
   define_option(ARROW_S3 "Build Arrow with S3 support (requires the AWS SDK for C++)" OFF)

diff --git a/cpp/cmake_modules/ThirdpartyToolchain.cmake b/cpp/cmake_modules/ThirdpartyToolchain.cmake
@@ -1626,10 +1626,6 @@ if(ARROW_WITH_PROTOBUF)
   if(ARROW_WITH_GRPC)
     # FlightSQL uses proto3 optionals, which require 3.15 or later.
     set(ARROW_PROTOBUF_REQUIRED_VERSION "3.15.0")
-  elseif(ARROW_GANDIVA_JAVA)
-    # google::protobuf::MessageLite::ByteSize() is deprecated since
-    # Protobuf 3.4.0.
-    set(ARROW_PROTOBUF_REQUIRED_VERSION "3.4.0")
   elseif(ARROW_SUBSTRAIT)
     # Substrait protobuf files use proto3 syntax
     set(ARROW_PROTOBUF_REQUIRED_VERSION "3.0.0")

diff --git a/cpp/src/arrow/gpu/ArrowCUDAConfig.cmake.in b/cpp/src/arrow/gpu/ArrowCUDAConfig.cmake.in
@@ -28,6 +28,11 @@
 
 include(CMakeFindDependencyMacro)
 find_dependency(Arrow)
+if(CMAKE_VERSION VERSION_LESS 3.17)
+  find_package(CUDA REQUIRED)
+else()
+  find_package(CUDAToolkit REQUIRED)
+endif()
 
 include("${CMAKE_CURRENT_LIST_DIR}/ArrowCUDATargets.cmake")
 

diff --git a/cpp/src/plasma/CMakeLists.txt b/cpp/src/plasma/CMakeLists.txt
@@ -171,44 +171,6 @@ install(TARGETS plasma-store-server ${INSTALL_IS_OPTIONAL}
         EXPORT plasma_targets
         DESTINATION ${CMAKE_INSTALL_BINDIR})
 
-if(ARROW_PLASMA_JAVA_CLIENT)
-  # Plasma java client support
-  find_package(JNI REQUIRED)
-  # add jni support
-  include_directories(${JAVA_INCLUDE_PATH})
-  include_directories(${JAVA_INCLUDE_PATH2})
-  if(JNI_FOUND)
-    message(STATUS "JNI_INCLUDE_DIRS = ${JNI_INCLUDE_DIRS}")
-    message(STATUS "JNI_LIBRARIES = ${JNI_LIBRARIES}")
-  else()
-    message(WARNING "Could not find JNI")
-  endif()
-
-  add_compile_options("-I$ENV{JAVA_HOME}/include/")
-  if(WIN32)
-    add_compile_options("-I$ENV{JAVA_HOME}/include/win32")
-  elseif(APPLE)
-    add_compile_options("-I$ENV{JAVA_HOME}/include/darwin")
-  else() # linux
-    add_compile_options("-I$ENV{JAVA_HOME}/include/linux")
-  endif()
-
-  include_directories("${CMAKE_CURRENT_LIST_DIR}/lib/java")
-
-  file(GLOB PLASMA_LIBRARY_EXT_java_SRC lib/java/*.cc lib/*.cc)
-  add_library(plasma_java SHARED ${PLASMA_LIBRARY_EXT_java_SRC})
-
-  if(APPLE)
-    target_link_libraries(plasma_java
-                          plasma_static
-                          ${PLASMA_STATIC_LINK_LIBS}
-                          "-undefined dynamic_lookup"
-                          ${PTHREAD_LIBRARY})
-  else()
-    target_link_libraries(plasma_java plasma_static ${PLASMA_STATIC_LINK_LIBS}
-                          ${PTHREAD_LIBRARY})
-  endif()
-endif()
 #
 # Unit tests
 #

diff --git a/cpp/src/plasma/lib/java/org_apache_arrow_plasma_PlasmaClientJNI.h b/cpp/src/plasma/lib/java/org_apache_arrow_plasma_PlasmaClientJNI.h