resync arrow-cpp

dev/tasks/conda-recipes/arrow-cpp/activate.sh

@@ -7,33 +7,70 @@
 # doesn't come with a deactivate script, because the symlink
 # is benign and doesn't need to be deleted.
 
+_la_log() {
+    if [ "${CF_LIBARROW_ACTIVATE_LOGGING:-}" = "1" ]; then
+        # The following loop is necessary to handle multi-line strings
+        # like for the output of `ls -al`.
+        printf '%s\n' "$*" | while IFS= read -r line
+        do
+            echo "$CONDA_PREFIX/etc/conda/activate.d/libarrow_activate.sh DEBUG: $line"
+        done
+    fi
+}
+
+_la_log "Beginning libarrow activation."
+
 # where the GDB wrappers get installed
-GDB_PREFIX="$CONDA_PREFIX/share/gdb/auto-load"
+_la_gdb_prefix="$CONDA_PREFIX/share/gdb/auto-load"
 
 # this needs to be in sync with ARROW_GDB_INSTALL_DIR in build.sh
-PLACEHOLDER="replace_this_section_with_absolute_slashed_path_to_CONDA_PREFIX"
+_la_placeholder="replace_this_section_with_absolute_slashed_path_to_CONDA_PREFIX"
 # the paths here are intentionally stacked, see #935, resp.
 # https://github.com/apache/arrow/blob/master/docs/source/cpp/gdb.rst#manual-loading
-WRAPPER_DIR="$GDB_PREFIX/$CONDA_PREFIX/lib"
+_la_symlink_dir="$_la_gdb_prefix/$CONDA_PREFIX/lib"
+_la_orig_install_dir="$_la_gdb_prefix/$_la_placeholder/lib"
 
-mkdir -p "$WRAPPER_DIR" || true
-# If the directory is not writable, nothing can be done
-if [ ! -w "$WRAPPER_DIR" ]; then
-    return
-fi
+_la_log "          _la_gdb_prefix: $_la_gdb_prefix"
+_la_log "         _la_placeholder: $_la_placeholder"
+_la_log "         _la_symlink_dir: $_la_symlink_dir"
+_la_log "    _la_orig_install_dir: $_la_orig_install_dir"
+_la_log "  content of that folder:"
+_la_log "$(ls -al "$_la_orig_install_dir" | sed 's/^/      /')"
 
-# there's only one lib in that folder, but the libname changes
+# there's only one lib in the _la_orig_install_dir folder, but the libname changes
 # based on the version so use a loop instead of hardcoding it.
-for f in "$GDB_PREFIX/$PLACEHOLDER/lib/"*.py; do
-    if [ ! -e "$f" ]; then
+for _la_target in "$_la_orig_install_dir/"*.py; do
+    if [ ! -e "$_la_target" ]; then
         # If the file doesn't exist, skip this iteration of the loop.
         # (This happens when no files are found, in which case the
-        # loop runs with f equal to the pattern itself.)
+        # loop runs with target equal to the pattern itself.)
+        _la_log 'Folder $_la_orig_install_dir seems to not contain .py files, skipping'
+        continue
+    fi
+    _la_symlink="$_la_symlink_dir/$(basename "$_la_target")"
+    _la_log "   _la_target: $_la_target"
+    _la_log "  _la_symlink: $_la_symlink"
+    if [ -L "$_la_symlink" ] && [ "$(readlink "$_la_symlink")" = "$_la_target" ]; then
+        _la_log 'symlink $_la_symlink already exists and points to $_la_target, skipping.'
+        continue
+    fi
+    _la_log 'Creating symlink $_la_symlink pointing to $_la_target'
+    mkdir -p "$_la_symlink_dir" || true
+    # this check also creates the symlink; if it fails, we enter the if-branch.
+    if ! ln -sf "$_la_target" "$_la_symlink"; then
+        echo -n "${BASH_SOURCE[0]} ERROR: Failed to create symlink from "
+        echo -n "'$_la_target' to '$_la_symlink'"
+        echo
         continue
     fi
-    target="$WRAPPER_DIR/$(basename "$f")"
-    # We have write permissions to WRAPPER_DIR but not necessarily target.
-    # Thus it's safest to delete the target in case it already exists.
-    rm -f "$target"
-    ln -s "$f" "$WRAPPER_DIR/$(basename "$f")"
 done
+
+_la_log "Libarrow activation complete."
+
+unset _la_gdb_prefix
+unset _la_log
+unset _la_orig_install_dir
+unset _la_placeholder
+unset _la_symlink
+unset _la_symlink_dir
+unset _la_target

dev/tasks/conda-recipes/arrow-cpp/build-arrow.bat

@@ -1,16 +1,12 @@
 @echo on
 
-mkdir "%SRC_DIR%"\cpp\build
-pushd "%SRC_DIR%"\cpp\build
+mkdir cpp\build
+pushd cpp\build
 
 :: Enable CUDA support
 if "%cuda_compiler_version%"=="None" (
     set "EXTRA_CMAKE_ARGS=-DARROW_CUDA=OFF"
 ) else (
-    REM this should move to nvcc-feedstock
-    set "CUDA_PATH=%CUDA_PATH:\=/%"
-    set "CUDA_HOME=%CUDA_HOME:\=/%"
-
     set "EXTRA_CMAKE_ARGS=-DARROW_CUDA=ON"
 )
 
@@ -70,3 +66,6 @@ cmake --build . --target install --config Release
 if %ERRORLEVEL% neq 0 exit 1
 
 popd
+
+:: clean up between builds (and to save space)
+rmdir /s /q cpp\build

dev/tasks/conda-recipes/arrow-cpp/build-arrow.sh

@@ -30,18 +30,6 @@ fi
 # Enable CUDA support
 if [[ ! -z "${cuda_compiler_version+x}" && "${cuda_compiler_version}" != "None" ]]
 then
-    if [[ -z "${CUDA_HOME+x}" ]]
-    then
-        echo "cuda_compiler_version=${cuda_compiler_version} CUDA_HOME=$CUDA_HOME"
-        CUDA_GDB_EXECUTABLE=$(which cuda-gdb || exit 0)
-        if [[ -n "$CUDA_GDB_EXECUTABLE" ]]
-        then
-            CUDA_HOME=$(dirname $(dirname $CUDA_GDB_EXECUTABLE))
-        else
-            echo "Cannot determine CUDA_HOME: cuda-gdb not in PATH"
-            return 1
-        fi
-    fi
     EXTRA_CMAKE_ARGS=" ${EXTRA_CMAKE_ARGS} -DARROW_CUDA=ON -DCUDA_TOOLKIT_ROOT_DIR=${CUDA_HOME} -DCMAKE_LIBRARY_PATH=${CONDA_BUILD_SYSROOT}/lib"
 else
     EXTRA_CMAKE_ARGS=" ${EXTRA_CMAKE_ARGS} -DARROW_CUDA=OFF"
@@ -56,7 +44,7 @@ if [[ "${build_platform}" != "${target_platform}" ]]; then
     EXTRA_CMAKE_ARGS="${EXTRA_CMAKE_ARGS} -DCLANG_EXECUTABLE=${BUILD_PREFIX}/bin/${CONDA_TOOLCHAIN_HOST}-clang"
     EXTRA_CMAKE_ARGS="${EXTRA_CMAKE_ARGS} -DLLVM_LINK_EXECUTABLE=${BUILD_PREFIX}/bin/llvm-link"
     sed -ie "s;protoc-gen-grpc.*$;protoc-gen-grpc=${BUILD_PREFIX}/bin/grpc_cpp_plugin\";g" ../src/arrow/flight/CMakeLists.txt
-    sed -ie 's;"--with-jemalloc-prefix\=je_arrow_";"--with-jemalloc-prefix\=je_arrow_" "--with-lg-page\=14";g' ../cmake_modules/ThirdpartyToolchain.cmake
+    sed -ie 's;"--with-jemalloc-prefix\=je_arrow_";"--with-jemalloc-prefix\=je_arrow_" "--with-lg-page\=16";g' ../cmake_modules/ThirdpartyToolchain.cmake
 fi
 
 # disable -fno-plt, which causes problems with GCC on PPC
@@ -65,17 +53,17 @@ if [[ "$target_platform" == "linux-ppc64le" ]]; then
   CXXFLAGS="$(echo $CXXFLAGS | sed 's/-fno-plt //g')"
 fi
 
-# Limit number of threads used to avoid hardware oversubscription
 if [[ "${target_platform}" == "linux-aarch64" ]] || [[ "${target_platform}" == "linux-ppc64le" ]]; then
-     export CMAKE_BUILD_PARALLEL_LEVEL=3
+    # Limit number of threads used to avoid hardware oversubscription
+    export CMAKE_BUILD_PARALLEL_LEVEL=3
 fi
 
 # reusable variable for dependencies we cannot yet unvendor
 export READ_RECIPE_META_YAML_WHY_NOT=OFF
 
 # for available switches see
 # https://github.com/apache/arrow/blame/apache-arrow-12.0.0/cpp/cmake_modules/DefineOptions.cmake
-# placeholder in ARROW_GDB_INSTALL_DIR must match what's used for replacement in activate.sh
+# placeholder in ARROW_GDB_INSTALL_DIR must match _la_placeholder in activate.sh
 cmake -GNinja \
     -DARROW_ACERO=ON \
     -DARROW_BOOST_USE_SHARED=ON \
@@ -131,3 +119,6 @@ cmake -GNinja \
 cmake --build . --target install --config Release
 
 popd
+
+# clean up between builds (and to save space)
+rm -rf cpp/build

dev/tasks/conda-recipes/arrow-cpp/build-pyarrow.sh

@@ -38,9 +38,9 @@ if [[ "${target_platform}" == osx-* ]]; then
     CXXFLAGS="${CXXFLAGS} -D_LIBCPP_DISABLE_AVAILABILITY"
 fi
 
-# Limit number of threads used to avoid hardware oversubscription
 if [[ "${target_platform}" == "linux-aarch64" ]] || [[ "${target_platform}" == "linux-ppc64le" ]]; then
-     export CMAKE_BUILD_PARALLEL_LEVEL=4
+    # Limit number of threads used to avoid hardware oversubscription
+    export CMAKE_BUILD_PARALLEL_LEVEL=4
 fi
 
 cd python