[GLUTEN-9577][VL] Enable VCPKG for MacOS build

.github/workflows/velox_backend_arm.yml

@@ -195,3 +195,26 @@ jobs:
         with:
           name: ${{ github.job }}-test-log
           path: "**/target/*.log"
+
+  build-on-macos:
+    strategy:
+      fail-fast: false
+    if: ${{ startsWith(github.repository, 'apache/') }}
+    runs-on: macos-15
+    env:
+      INSTALL_PREFIX: /tmp/deps-install
+      NUM_THREADS: 4
+    steps:
+      - uses: actions/checkout@v5
+      - name: Set up Java 17
+        uses: actions/setup-java@v4
+        with:
+          distribution: temurin
+          java-version: 17
+      - name: Install Dependencies
+        run: brew install wget pkgconfig ccache cmake autoconf automake libtool autoconf-archive bison flex
+      - name: Build
+        run: ./dev/buildbundle-veloxbe.sh --enable_vcpkg=ON --run_setup_script=OFF --build_arrow=ON --build_tests=ON --build_benchmarks=ON --spark_version=3.5
+      - name: Run CPP unit test
+        run: |
+          cd ./cpp/build && ctest -V

cpp/core/CMakeLists.txt

@@ -230,6 +230,7 @@ endif()
 
 target_link_libraries(gluten PUBLIC Arrow::arrow
                                     Arrow::arrow_bundled_dependencies)
+add_compile_definitions(GLOG_CUSTOM_PREFIX=gluten)
 target_link_libraries(gluten PRIVATE google::glog)
 
 install(TARGETS gluten DESTINATION ${CMAKE_INSTALL_LIBDIR})

cpp/core/symbols.map

@@ -11,6 +11,5 @@
       *google::*;
       fL*::FLAGS_*;
       gflags_mutex_namespace::*;
-      glog_internal_namespace_::*;
     };
 };

dev/build-arrow.sh

@@ -56,10 +56,11 @@ function build_arrow_cpp() {
        -DARROW_DEPENDENCY_SOURCE=BUNDLED \
        -DARROW_WITH_THRIFT=ON \
        -DARROW_WITH_LZ4=ON \
-       -DARROW_WITH_SNAPPY=ON \
+       -DARROW_WITH_SNAPPY=OFF \
        -DARROW_WITH_ZLIB=${ARROW_WITH_ZLIB} \
        -DARROW_WITH_ZSTD=ON \
        -DARROW_JEMALLOC=OFF \
+       -DARROW_USE_JEMALLOC=OFF \
        -DARROW_SIMD_LEVEL=NONE \
        -DARROW_RUNTIME_SIMD_LEVEL=NONE \
        -DARROW_WITH_UTF8PROC=OFF \

dev/vcpkg/env.sh

@@ -27,6 +27,7 @@ SCRIPT_ROOT="$(realpath "$(dirname "${BASH_SOURCE[0]}")")"
 export VCPKG_ROOT="$SCRIPT_ROOT/.vcpkg"
 export VCPKG="$SCRIPT_ROOT/.vcpkg/vcpkg"
 export VCPKG_TRIPLET=$([ "${CPU_TARGET:-}" = "aarch64" ] && echo "arm64-linux-neon" || echo "x64-linux-avx")
+export VCPKG_TRIPLET=$([ "$(uname)" == "Darwin" ] && echo "arm64-osx-release" || echo ${VCPKG_TRIPLET})
 export VCPKG_TRIPLET_INSTALL_DIR=${SCRIPT_ROOT}/vcpkg_installed/${VCPKG_TRIPLET}
 
 ${SCRIPT_ROOT}/init.sh "$@"

dev/vcpkg/ports/grpc/00001-fix-uwp.patch

@@ -0,0 +1,73 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 25990a5d8a..4bec4e1e72 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -219,6 +219,9 @@ if(UNIX)
+ endif()
+ if(WIN32)
+   set(_gRPC_PLATFORM_WINDOWS ON)
++  if(CMAKE_SYSTEM_NAME STREQUAL "WindowsStore")
++    set(_gRPC_PLATFORM_UWP ON)
++  endif()
+ endif()
+
+  # Use C11 standard
+@@ -263,6 +266,9 @@ if(MSVC)
+   set(_gRPC_C_CXX_FLAGS "${_gRPC_C_CXX_FLAGS} /wd4267")
+   # TODO(jtattermusch): needed to build boringssl with VS2017, revisit later
+   set(_gRPC_C_CXX_FLAGS "${_gRPC_C_CXX_FLAGS} /wd4987 /wd4774 /wd4819 /wd4996 /wd4619")
++  if(_gRPC_PLATFORM_UWP)
++    add_definitions(-DGRPC_ARES=0)
++  endif()
+   # Silences thousands of trucation warnings
+   set(_gRPC_C_CXX_FLAGS "${_gRPC_C_CXX_FLAGS} /wd4503")
+   # Tell MSVC to build grpc using utf-8
+@@ -430,6 +436,10 @@ file(MAKE_DIRECTORY ${_gRPC_PROTO_GENS_DIR})
+ #     ``.proto`` files
+ #
+ function(protobuf_generate_grpc_cpp)
++  if(_gRPC_PLATFORM_UWP)
++    return()
++  endif()
++
+   if(NOT ARGN)
+     message(SEND_ERROR "Error: PROTOBUF_GENERATE_GRPC_CPP() called without any proto files")
+     return()
+@@ -552,6 +562,7 @@ if (gRPC_BUILD_GRPC_RUBY_PLUGIN)
+   list(APPEND _gRPC_PLUGIN_LIST grpc_ruby_plugin)
+ endif ()
+
++if(NOT _gRPC_PLATFORM_UWP)
+ add_custom_target(plugins
+   DEPENDS ${_gRPC_PLUGIN_LIST}
+ )
+@@ -567,6 +578,7 @@ add_custom_target(tools_cxx
+
+ add_custom_target(tools
+   DEPENDS tools_c tools_cxx)
++endif()
+
+ protobuf_generate_grpc_cpp_with_import_path_correction(
+   src/proto/grpc/channelz/channelz.proto src/proto/grpc/channelz/channelz.proto
+diff --git a/src/core/lib/security/credentials/alts/check_gcp_environment_windows.cc b/src/core/lib/security/credentials/alts/check_gcp_environment_windows.cc
+index 5d2bdc14de..e9870c2656 100644
+--- a/src/core/lib/security/credentials/alts/check_gcp_environment_windows.cc
++++ b/src/core/lib/security/credentials/alts/check_gcp_environment_windows.cc
+@@ -39,6 +39,7 @@ bool check_bios_data(const char*) { return false; }
+ bool check_windows_registry_product_name(HKEY root_key,
+                                          const char* reg_key_path,
+                                          const char* reg_key_name) {
++#if !defined(WINAPI_FAMILY) || WINAPI_FAMILY == WINAPI_FAMILY_DESKTOP_APP
+   const size_t kProductNameBufferSize = 256;
+   char const expected_substr[] = "Google";
+
+@@ -71,6 +72,9 @@ bool check_windows_registry_product_name(HKEY root_key,
+   }
+
+   return strstr(buffer, expected_substr) != nullptr;
++#else
++  return false;
++#endif
+ }
+
+ }  // namespace internal

dev/vcpkg/ports/grpc/00002-static-linking-in-linux.patch

@@ -0,0 +1,16 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 8f78306f77..e09f8fcc1e 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -211,6 +211,11 @@ if (NOT DEFINED CMAKE_POSITION_INDEPENDENT_CODE)
+ endif()
+ list(APPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake/modules")
+
++if (gRPC_STATIC_LINKING AND NOT _gRPC_PLATFORM_WINDOWS)
++  # Force to static link
++  set(CMAKE_EXE_LINKER_FLAGS "-Bstatic")
++endif()
++
+ if(MSVC)
+   include(cmake/msvc_static_runtime.cmake)
+   add_definitions(-D_WIN32_WINNT=0x600 -D_SCL_SECURE_NO_WARNINGS -D_CRT_SECURE_NO_WARNINGS -D_WINSOCK_DEPRECATED_NO_WARNINGS)

dev/vcpkg/ports/grpc/00003-undef-base64-macro.patch

@@ -0,0 +1,13 @@
+diff --git a/src/core/lib/transport/transport.cc b/src/core/lib/transport/transport.cc
+index 5d74ec2..89494b5 100644
+--- a/src/core/lib/transport/transport.cc
++++ b/src/core/lib/transport/transport.cc
+@@ -77,6 +77,8 @@ void grpc_stream_ref_init(grpc_stream_refcount* refcount, int /*initial_refs*/,
+                                                              : nullptr);
+ }
+
++#undef move64
++
+ static void move64bits(uint64_t* from, uint64_t* to) {
+   *to += *from;
+   *from = 0;

dev/vcpkg/ports/grpc/00004-link-gdi32-on-windows.patch

@@ -0,0 +1,13 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 25990a5d8a..8a632d2289 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -404,7 +404,7 @@ if (NOT EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/third_party/xds)
+ endif()
+
+ if(WIN32)
+-  set(_gRPC_BASELIB_LIBRARIES ws2_32 crypt32)
++  set(_gRPC_BASELIB_LIBRARIES wsock32 ws2_32 crypt32 gdi32)
+ endif()
+
+ # Create directory for proto source files

dev/vcpkg/ports/grpc/00005-fix-uwp-error.patch

@@ -0,0 +1,47 @@
+diff --git a/src/core/ext/transport/chttp2/transport/hpack_parser.cc b/src/core/ext/transport/chttp2/transport/hpack_parser.cc
+index b0d3a6465b..5c48f1aa30 100644
+--- a/src/core/ext/transport/chttp2/transport/hpack_parser.cc
++++ b/src/core/ext/transport/chttp2/transport/hpack_parser.cc
+@@ -1037,7 +1037,7 @@ class HPackParser::Parser {
+
+  private:
+   void GPR_ATTRIBUTE_NOINLINE LogHeader(const HPackTable::Memento& memento) {
+-    const char* type;
++    const char* type = nullptr;
+     switch (log_info_.type) {
+       case LogInfo::kHeaders:
+         type = "HDR";
+diff --git a/src/core/lib/slice/slice.cc b/src/core/lib/slice/slice.cc
+index 898a62823c..6b31cdc082 100644
+--- a/src/core/lib/slice/slice.cc
++++ b/src/core/lib/slice/slice.cc
+@@ -188,6 +188,7 @@ grpc_slice grpc_slice_from_moved_buffer(grpc_core::UniquePtr<char> p,
+                                         size_t len) {
+   uint8_t* ptr = reinterpret_cast<uint8_t*>(p.get());
+   grpc_slice slice;
++  memset(&slice, 0, sizeof(grpc_slice));
+   if (len <= sizeof(slice.data.inlined.bytes)) {
+     slice.refcount = nullptr;
+     slice.data.inlined.length = len;
+@@ -206,7 +207,7 @@ grpc_slice grpc_slice_from_moved_string(grpc_core::UniquePtr<char> p) {
+ }
+
+ grpc_slice grpc_slice_from_cpp_string(std::string str) {
+-  grpc_slice slice;
++  grpc_slice slice = { 0 };
+   if (str.size() <= sizeof(slice.data.inlined.bytes)) {
+     slice.refcount = nullptr;
+     slice.data.inlined.length = str.size();
+diff --git a/src/core/lib/surface/server.cc b/src/core/lib/surface/server.cc
+index 141b16e345..89d9d6dafd 100644
+--- a/src/core/lib/surface/server.cc
++++ b/src/core/lib/surface/server.cc
+@@ -902,7 +902,7 @@ grpc_call_error Server::QueueRequestedCall(size_t cq_idx, RequestedCall* rc) {
+     FailCall(cq_idx, rc, GRPC_ERROR_CREATE("Server Shutdown"));
+     return GRPC_CALL_OK;
+   }
+-  RequestMatcherInterface* rm;
++  RequestMatcherInterface* rm = nullptr;
+   switch (rc->type) {
+     case RequestedCall::Type::BATCH_CALL:
+       rm = unregistered_request_matcher_.get();

dev/vcpkg/ports/grpc/00009-use-system-upb.patch

@@ -0,0 +1,91 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 23098aa578..a8e8bc274b 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -85,6 +85,9 @@ set_property(CACHE gRPC_SSL_PROVIDER PROPERTY STRINGS "module" "package")
+ set(gRPC_PROTOBUF_PROVIDER "module" CACHE STRING "Provider of protobuf library")
+ set_property(CACHE gRPC_PROTOBUF_PROVIDER PROPERTY STRINGS "module" "package")
+
++set(gRPC_UPB_PROVIDER "module" CACHE STRING "Provider of upb library")
++set_property(CACHE gRPC_UPB_PROVIDER PROPERTY STRINGS "module" "package")
++
+ set(gRPC_PROTOBUF_PACKAGE_TYPE "" CACHE STRING "Algorithm for searching protobuf package")
+ set_property(CACHE gRPC_PROTOBUF_PACKAGE_TYPE PROPERTY STRINGS "CONFIG" "MODULE")
+
+@@ -1631,6 +1634,7 @@ target_link_libraries(gpr
+   absl::time
+   absl::optional
+   absl::variant
++  ${_gRPC_UPB_LIBRARIES}
+ )
+ if(_gRPC_PLATFORM_ANDROID)
+   target_link_libraries(gpr
+@@ -2435,7 +2439,6 @@ target_link_libraries(grpc
+   gpr
+   ${_gRPC_SSL_LIBRARIES}
+   address_sorting
+-  upb
+ )
+ if(_gRPC_PLATFORM_IOS OR _gRPC_PLATFORM_MAC)
+   target_link_libraries(grpc "-framework CoreFoundation")
+@@ -2979,7 +2982,6 @@ target_link_libraries(grpc_unsecure
+   absl::utility
+   gpr
+   address_sorting
+-  upb
+ )
+ if(_gRPC_PLATFORM_IOS OR _gRPC_PLATFORM_MAC)
+   target_link_libraries(grpc_unsecure "-framework CoreFoundation")
+@@ -4251,6 +4253,7 @@ endif()
+
+ endif()
+
++if (gRPC_UPB_PROVIDER STREQUAL "module")
+ add_library(upb
+   third_party/upb/third_party/utf8_range/naive.c
+   third_party/upb/third_party/utf8_range/range2-neon.c
+@@ -4319,7 +4322,7 @@ if(gRPC_INSTALL)
+     ARCHIVE DESTINATION ${gRPC_INSTALL_LIBDIR}
+   )
+ endif()
+-
++endif()
+
+ if(gRPC_BUILD_TESTS)
+
+diff --git a/cmake/gRPCConfig.cmake.in b/cmake/gRPCConfig.cmake.in
+index 3623f4aa5e..df6ced560e 100644
+--- a/cmake/gRPCConfig.cmake.in
++++ b/cmake/gRPCConfig.cmake.in
+@@ -8,6 +8,7 @@ list(APPEND CMAKE_MODULE_PATH ${CMAKE_CURRENT_LIST_DIR}/modules)
+ @_gRPC_FIND_CARES@
+ @_gRPC_FIND_ABSL@
+ @_gRPC_FIND_RE2@
++@_gRPC_FIND_UPB@
+
+ # Targets
+ include(${CMAKE_CURRENT_LIST_DIR}/gRPCTargets.cmake)
+diff --git a/cmake/upb.cmake b/cmake/upb.cmake
+index f2a0e508c3..09751f5ef0 100644
+--- a/cmake/upb.cmake
++++ b/cmake/upb.cmake
+@@ -12,9 +12,19 @@
+ # See the License for the specific language governing permissions and
+ # limitations under the License.
+
++set(_gRPC_UPB_GRPC_GENERATED_DIR "${CMAKE_CURRENT_SOURCE_DIR}/src/core/ext/upb-generated" "${CMAKE_CURRENT_SOURCE_DIR}/src/core/ext/upbdefs-generated")
++if (gRPC_UPB_PROVIDER STREQUAL "module")
++
+ set(UPB_ROOT_DIR ${CMAKE_CURRENT_SOURCE_DIR}/third_party/upb)
+
+ set(_gRPC_UPB_INCLUDE_DIR "${UPB_ROOT_DIR}")
+ set(_gRPC_UPB_GRPC_GENERATED_DIR "${CMAKE_CURRENT_SOURCE_DIR}/src/core/ext/upb-generated" "${CMAKE_CURRENT_SOURCE_DIR}/src/core/ext/upbdefs-generated")
+
+ set(_gRPC_UPB_LIBRARIES upb)
++
++elseif(gRPC_UPB_PROVIDER STREQUAL "package")
++    find_package(upb CONFIG REQUIRED)
++    set(_gRPC_UPB_LIBRARIES upb::fastdecode upb::json upb::upb upb::utf8_range upb::textformat upb::reflection upb::descriptor_upb_proto)
++    set(_gRPC_UPB_INCLUDE_DIR)
++    set(_gRPC_FIND_UPB "if(NOT upb_FOUND)\n  find_package(upb CONFIG REQUIRED)\nendif()")
++endif()

dev/vcpkg/ports/grpc/00012-fix-use-cxx17.patch

@@ -0,0 +1,16 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 25990a5d8a..ba8df92858 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -326,6 +326,11 @@ include(cmake/xxhash.cmake)
+ include(cmake/zlib.cmake)
+ include(cmake/download_archive.cmake)
+
++if (ABSL_USE_CXX17)
++    message(STATUS "Found absl uses CXX17, enable CXX17 feature.")
++    set(CMAKE_CXX_STANDARD 17)
++endif()
++
+ # Setup external proto library at third_party/envoy-api with 2 download URLs
+ if (NOT EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/third_party/envoy-api)
+   # Download the archive via HTTP, validate the checksum, and extract to third_party/envoy-api.

dev/vcpkg/ports/grpc/00014-pkgconfig-upbdefs.patch

@@ -0,0 +1,13 @@
+diff --git a/CMakeLists.txt b/CMakeLists.txt
+index 48019cce95..1eda700ae8 100644
+--- a/CMakeLists.txt
++++ b/CMakeLists.txt
+@@ -23186,7 +23186,7 @@ generate_pkgconfig(
+   "high performance general RPC framework"
+   "${gRPC_CORE_VERSION}"
+   "gpr openssl absl_any_invocable absl_base absl_bind_front absl_cleanup absl_cord absl_core_headers absl_flat_hash_map absl_flat_hash_set absl_function_ref absl_hash absl_inlined_vector absl_memory absl_optional absl_random_random absl_span absl_status absl_statusor absl_str_format absl_strings absl_synchronization absl_time absl_type_traits absl_utility absl_variant"
+-  "-lgrpc -laddress_sorting -lre2 -lupb -lcares -lz"
++  "-lgrpc -laddress_sorting -lre2 -lupb_textformat -lupb_mini_table -lupb -lupb_collections -lupb_reflection -lupb_extension_registry -lupb_json -lupb_fastdecode -lupb_utf8_range -ldescriptor_upb_proto -lcares -lz"
+   ""
+   "grpc.pc")
+

dev/vcpkg/ports/grpc/00015-disable-download-archive.patch

@@ -0,0 +1,12 @@
+diff --git a/cmake/download_archive.cmake b/cmake/download_archive.cmake
+index 820aafafb7..a59b785c7e 100644
+--- a/cmake/download_archive.cmake
++++ b/cmake/download_archive.cmake
+@@ -19,6 +19,7 @@ file(MAKE_DIRECTORY ${_download_archive_TEMPORARY_DIR})
+ # Note that strip_prefix strips the directory path prefix of the extracted
+ # archive content, and it may strip multiple directories.
+ function(download_archive destination url hash strip_prefix)
++  return()
+   # Fetch and validate
+   set(_TEMPORARY_FILE ${_download_archive_TEMPORARY_DIR}/${strip_prefix}.tar.gz)
+   message(STATUS "Downloading from ${url}, if failed, please try configuring again")

dev/vcpkg/ports/grpc/00016_add_header.patch

@@ -0,0 +1,12 @@
+diff --git a/src/core/lib/iomgr/tcp_posix.cc b/src/core/lib/iomgr/tcp_posix.cc
+index 72e1b66..8dc1fd1 100644
+--- a/src/core/lib/iomgr/tcp_posix.cc
++++ b/src/core/lib/iomgr/tcp_posix.cc
+@@ -47,6 +47,7 @@
+ #include <grpc/support/string_util.h>
+ #include <grpc/support/sync.h>
+ #include <grpc/support/time.h>
++#include <absl/strings/str_cat.h>
+
+ #include "src/core/lib/address_utils/sockaddr_utils.h"
+ #include "src/core/lib/debug/event_log.h"

dev/vcpkg/ports/grpc/gRPCTargets-vcpkg-tools.cmake

@@ -0,0 +1,10 @@
+file(GLOB GRPC_PLUGINS "${_IMPORT_PREFIX}/../@HOST_TRIPLET@/tools/grpc/grpc_*_plugin*")
+
+foreach(PLUGIN ${GRPC_PLUGINS})
+  get_filename_component(PLUGIN_NAME "${PLUGIN}" NAME_WE)
+  add_executable(gRPC::${PLUGIN_NAME} IMPORTED)
+  set_property(TARGET gRPC::${PLUGIN_NAME} APPEND PROPERTY IMPORTED_CONFIGURATIONS RELEASE)
+  set_target_properties(gRPC::${PLUGIN_NAME} PROPERTIES
+    IMPORTED_LOCATION_RELEASE "${PLUGIN}"
+  )
+endforeach()