CMakeLists.txt

# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.
# Copyright 2024-2026 Arm Limited and/or its affiliates.
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

#
# Simple CMake build system for runtime components.
#
# ### One-time setup ###
#
# Configure the CMake build system. It's good practice to do this whenever
# cloning or pulling the upstream repo. Once this is done, you don't need to do
# it again until you pull from the upstream repo again.
#
# NOTE: Build options can be configured by passing arguments to cmake. For
# example, to enable the EXECUTORCH_BUILD_XNNPACK option, change the cmake
# command to 'cmake -DEXECUTORCH_BUILD_XNNPACK=ON ..'.
#[[
  (rm -rf cmake-out \
    && mkdir cmake-out \
    && cd cmake-out \
    && cmake ..)
]]
#
# ### Build ###
#
# NOTE: The `-j` argument specifies how many jobs/processes to use when
# building, and tends to speed up the build significantly. It's typical to use
# "core count + 1" as the `-j` value.
# ~~~
# cmake --build cmake-out -j9
# ~~~
#
# ### Editing this file ###
#
# This file should be formatted with
# ~~~
# cmake-format -i CMakeLists.txt
# ~~~
# It should also be checked with a linter via
# ~~~
# cmake-lint CMakeLists.txt
# ~~~
#

cmake_minimum_required(VERSION 3.24)

set(EXECUTORCH_ROOT ${CMAKE_CURRENT_SOURCE_DIR})

# --- ExecuTorch Version ---
# Parse version from version.txt (single source of truth)
file(READ "${EXECUTORCH_ROOT}/version.txt" ET_VERSION_STRING)
string(STRIP "${ET_VERSION_STRING}" ET_VERSION_STRING)

# Extract major.minor.patch (handles formats like "1.2.0a0")
string(REGEX MATCH "^([0-9]+)\\.([0-9]+)\\.([0-9]+)" ET_VERSION_MATCH
             "${ET_VERSION_STRING}"
)
set(ET_VERSION_MAJOR "${CMAKE_MATCH_1}")
set(ET_VERSION_MINOR "${CMAKE_MATCH_2}")
set(ET_VERSION_PATCH "${CMAKE_MATCH_3}")

# Validate that version components were successfully extracted
if("${ET_VERSION_MAJOR}" STREQUAL ""
   OR "${ET_VERSION_MINOR}" STREQUAL ""
   OR "${ET_VERSION_PATCH}" STREQUAL ""
)
  message(
    FATAL_ERROR
      "Failed to parse version from ${EXECUTORCH_ROOT}/version.txt. "
      "Expected format: MAJOR.MINOR.PATCH (e.g., '1.2.0' or '1.2.0a0'), "
      "but got: '${ET_VERSION_STRING}'"
  )
endif()

project(executorch
        VERSION "${ET_VERSION_MAJOR}.${ET_VERSION_MINOR}.${ET_VERSION_PATCH}"
)

if(WIN32)
  add_compile_definitions(NOMINMAX)
endif()

# Hexagon toolchain with release build complains about code in third party
# libraries. Must come after project(), which runs the toolchain file that sets
# CMAKE_SYSTEM_PROCESSOR, and before add_subdirectory(third-party).
if("${CMAKE_SYSTEM_PROCESSOR}" MATCHES "Hexagon" AND "${CMAKE_BUILD_TYPE}"
                                                     STREQUAL "Release"
)
  add_compile_options(
    -Wno-error=format -Wno-error=implicit-int-conversion
    -Wno-error=unused-variable -Wno-error=unused-function
  )
endif()

message(
  STATUS
    "ExecuTorch version: ${ET_VERSION_MAJOR}.${ET_VERSION_MINOR}.${ET_VERSION_PATCH}"
)

# Generate version.h from template
configure_file(
  "${EXECUTORCH_ROOT}/runtime/core/version.h.in"
  "${CMAKE_CURRENT_BINARY_DIR}/include/executorch/runtime/core/version.h" @ONLY
)

include(${PROJECT_SOURCE_DIR}/tools/cmake/common/preset.cmake)
include(${PROJECT_SOURCE_DIR}/tools/cmake/Codegen.cmake)
include(${PROJECT_SOURCE_DIR}/tools/cmake/Utils.cmake)
include(CMakeDependentOption)
include(ExternalProject)
include(GNUInstallDirs)

if(NOT CMAKE_CXX_STANDARD)
  # The vendored c10 is mirrored byte for byte from PyTorch, which builds at
  # C++20, so it is free to use C++20 in headers we compile. safe_numerics.h
  # reaches std::countl_zero and std::bit_width on the branch taken when
  # __builtin_*_overflow is unavailable, which is MSVC only -- every other
  # toolchain takes the builtin path and stays within C++17. Raising the
  # standard on Windows alone keeps the embedded targets, where C++20 support is
  # not a given, on C++17.
  if(WIN32)
    set(CMAKE_CXX_STANDARD 20)
  else()
    set(CMAKE_CXX_STANDARD 17)
  endif()
endif()
announce_configured_options(CMAKE_CXX_STANDARD)

if(NOT CMAKE_SYSTEM_PROCESSOR)
  set(CMAKE_SYSTEM_PROCESSOR ${CMAKE_HOST_SYSTEM_PROCESSOR})
endif()
announce_configured_options(CMAKE_SYSTEM_PROCESSOR)

if(NOT CMAKE_BUILD_TYPE)
  set(CMAKE_BUILD_TYPE Debug)
endif()
announce_configured_options(CMAKE_BUILD_TYPE)

# Sanitizer support (ASAN + UBSAN)
if(EXECUTORCH_USE_SANITIZER)
  if(MSVC)
    message(WARNING "Sanitizers are not fully supported on MSVC, skipping")
  else()
    set(EXECUTORCH_SANITIZER_FLAGS
        "-fsanitize=address,undefined -fno-omit-frame-pointer"
    )
    # Suppress deprecation warnings on macOS (third-party code like flatcc uses
    # deprecated sprintf)
    if(APPLE)
      set(EXECUTORCH_SANITIZER_FLAGS
          "${EXECUTORCH_SANITIZER_FLAGS} -Wno-deprecated-declarations"
      )
    endif()
    set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} ${EXECUTORCH_SANITIZER_FLAGS}")
    set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} ${EXECUTORCH_SANITIZER_FLAGS}")
    set(CMAKE_EXE_LINKER_FLAGS
        "${CMAKE_EXE_LINKER_FLAGS} ${EXECUTORCH_SANITIZER_FLAGS}"
    )
    set(CMAKE_SHARED_LINKER_FLAGS
        "${CMAKE_SHARED_LINKER_FLAGS} ${EXECUTORCH_SANITIZER_FLAGS}"
    )
    message(STATUS "Sanitizers enabled: address, undefined")
  endif()
endif()
announce_configured_options(EXECUTORCH_USE_SANITIZER)

if(NOT PYTHON_EXECUTABLE)
  resolve_python_executable()
endif()
announce_configured_options(PYTHON_EXECUTABLE)

# pybind11 >=3.0 uses find_package(Python) which looks for Python_EXECUTABLE,
# not the legacy PYTHON_EXECUTABLE variable. Bridge the two so pybind11 picks up
# the interpreter that setup.py (or the user) specified.
if(PYTHON_EXECUTABLE AND NOT Python_EXECUTABLE)
  set(Python_EXECUTABLE "${PYTHON_EXECUTABLE}")
endif()

announce_configured_options(CMAKE_CXX_COMPILER_ID)
announce_configured_options(CMAKE_TOOLCHAIN_FILE)
announce_configured_options(BUILD_TESTING)

load_build_preset()
include(${PROJECT_SOURCE_DIR}/tools/cmake/preset/default.cmake)

# Keep bare-metal installs enabled only when ExecuTorch owns the top-level
# build. Standalone consumers (e.g., the runner) set
# EXECUTORCH_BAREMETAL_SKIP_INSTALL=ON but still add ExecuTorch as a subproject,
# which cannot satisfy our install() export dependencies until their own targets
# are configured.
if(DEFINED EXECUTORCH_BAREMETAL_SKIP_INSTALL
   AND EXECUTORCH_BAREMETAL_SKIP_INSTALL
   AND NOT (CMAKE_SOURCE_DIR STREQUAL PROJECT_SOURCE_DIR)
)
  set(CMAKE_SKIP_INSTALL_RULES
      ON
      CACHE BOOL
            "Skip install() rules when ExecuTorch is consumed as a subproject"
            FORCE
  )
endif()

if(EXECUTORCH_BUILD_SHARED)
  # Refused rather than silently reconciled, for the same reason the platform
  # check below exists: a configuration this mode cannot honour should stop
  # here, not produce something that looks built.
  #
  # This mode consolidates the component static archives into one shared
  # runtime, extracting each archive's contents so a single library owns the
  # operator and backend registries. BUILD_SHARED_LIBS makes those same
  # components shared objects instead, and --whole-archive against a shared
  # object absorbs nothing and reports nothing: measured, the consolidated
  # library then links successfully while containing none of the core, which is
  # the split-registry failure this library exists to prevent. The untyped core
  # target also claims the same libexecutorch file name under that option, so
  # Ninja stops with "multiple rules generate libexecutorch.so" while Unix
  # Makefiles does not notice and writes one of the two at random.
  if(BUILD_SHARED_LIBS)
    message(
      FATAL_ERROR
        "EXECUTORCH_BUILD_SHARED requires BUILD_SHARED_LIBS=OFF. It builds one consolidated "
        "libexecutorch shared runtime out of the component static archives, so those components "
        "have to be archives. Choose one: EXECUTORCH_BUILD_SHARED=ON for the consolidated runtime "
        "the wheel ships, or BUILD_SHARED_LIBS=ON on its own for CMake's per-target shared layout."
    )
  endif()
  # Said here rather than left to fail somewhere downstream, where packaging
  # looked for a library the build never emitted. The runtime carries no export
  # annotations, so on Windows each shipped component DLL exports all of its
  # symbols instead; see executorch_target_shipped_runtime_path.
  if(NOT CMAKE_SYSTEM_NAME STREQUAL "Linux"
     AND NOT APPLE
     AND NOT (WIN32 AND MSVC)
  )
    message(
      FATAL_ERROR
        "EXECUTORCH_BUILD_SHARED is supported on Linux, macOS and Windows (MSVC) only, not "
        "${CMAKE_SYSTEM_NAME}."
    )
  endif()
  # The Windows DLLs take their components as objects through $<COMPILE_ONLY>,
  # which CMake added in 3.27. Older versions fail at generate time with an
  # expression error that does not name the cause.
  if(MSVC AND CMAKE_VERSION VERSION_LESS 3.27)
    message(
      FATAL_ERROR
        "EXECUTORCH_BUILD_SHARED on Windows needs CMake 3.27 or newer, found "
        "${CMAKE_VERSION}. Upgrade CMake (pip install 'cmake>=3.27,<4') or "
        "build with -DEXECUTORCH_BUILD_SHARED=OFF."
    )
  endif()
  set(CMAKE_POSITION_INDEPENDENT_CODE ON)
endif()

executorch_try_enabling_ccache()
announce_configured_options(CCACHE_PROGRAM)

set(CMAKE_EXPORT_COMPILE_COMMANDS ON)

if(NOT EXECUTORCH_ENABLE_LOGGING)
  # Avoid pulling in the logging strings, which can be large. Note that this
  # will set the compiler flag for all targets in this directory, and for all
  # subdirectories included after this point.
  add_definitions(-DET_LOG_ENABLED=0)
endif()

add_definitions(-DET_MIN_LOG_LEVEL=${ET_MIN_LOG_LEVEL})

if(NOT EXECUTORCH_ENABLE_PROGRAM_VERIFICATION)
  # Avoid pulling in the flatbuffer data verification logic, which can add about
  # 20kB. Note that this will set the compiler flag for all targets in this
  # directory, and for all subdirectories included after this point.
  add_definitions(-DET_ENABLE_PROGRAM_VERIFICATION=0)
endif()

# Disable the deprecated constant_buffer path.
add_definitions(-DET_ENABLE_DEPRECATED_CONSTANT_BUFFER=0)

if(EXECUTORCH_ENABLE_EVENT_TRACER)
  add_definitions(-DET_EVENT_TRACER_ENABLED)
endif()

if(EXECUTORCH_ENABLE_BUNDLE_IO)
  add_definitions(-DET_BUNDLE_IO_ENABLED)
endif()

if(EXECUTORCH_BUILD_ROCM)
  add_definitions(-DCUDA_AVAILABLE=1 -DEXECUTORCH_USE_HIP=1)
elseif(EXECUTORCH_BUILD_CUDA)
  add_definitions(-DCUDA_AVAILABLE=1)
endif()

option(ET_ENABLE_VGF_INPUT_DUMP
       "Enable Arm/VGF delegate-boundary input dump flags in executor_runner"
       OFF
)
announce_configured_options(ET_ENABLE_VGF_INPUT_DUMP)

# -ffunction-sections -fdata-sections: breaks function and data into sections so
# they can be properly gc'd. -s: strip symbol.
if(WIN32)
  set(CMAKE_CXX_FLAGS_RELEASE "/Gy /Gw ${CMAKE_CXX_FLAGS_RELEASE}")
else()
  set(CMAKE_CXX_FLAGS_RELEASE
      "-ffunction-sections -fdata-sections ${CMAKE_CXX_FLAGS_RELEASE}"
  )
endif()
if(CMAKE_CXX_COMPILER_ID STREQUAL "GNU")
  set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -s")
endif()

if(EXECUTORCH_OPTIMIZE_SIZE)
  # -Os: Optimize for size.
  set(CMAKE_CXX_FLAGS_RELEASE "${CMAKE_CXX_FLAGS_RELEASE} -Os")
else()
  # -O2: Moderate opt.
  set(CMAKE_CXX_FLAGS_RELEASE "-O2 ${CMAKE_CXX_FLAGS_RELEASE}")
endif()

if(EXECUTORCH_BUILD_TESTS)
  include(CTest)
else()
  # It looks like some of our third-party deps will try to turn this on if it's
  # not explicitly set, leading to confusing behavior.
  set(BUILD_TESTING OFF)
endif()

if(EXECUTORCH_BUILD_NATIVE)
  set(JSON_Install ON)
endif()
add_subdirectory(third-party)

# Size-optimized builds disable exceptions, RTTI, and unwind tables.
# ExecuTorch's runtime uses Result<T>/Error instead of exceptions, so
# -fno-exceptions is safe. Suppressing .eh_frame via -fno-unwind-tables is safe
# because exceptions are disabled. Placed after add_subdirectory(third-party) to
# avoid leaking into third-party configure checks (e.g. gflags).
if(EXECUTORCH_OPTIMIZE_SIZE
   AND NOT WIN32
   AND NOT MSVC
)
  set(CMAKE_CXX_FLAGS_RELEASE
      "${CMAKE_CXX_FLAGS_RELEASE} -fno-exceptions -fno-rtti -fno-asynchronous-unwind-tables -fno-unwind-tables"
  )
  set(CMAKE_C_FLAGS_RELEASE
      "${CMAKE_C_FLAGS_RELEASE} -fno-asynchronous-unwind-tables -fno-unwind-tables"
  )
endif()

if(NOT DEFINED FXDIV_SOURCE_DIR)
  set(ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG
      ${CMAKE_POSITION_INDEPENDENT_CODE}
  )
  set(FXDIV_SOURCE_DIR "backends/xnnpack/third-party/FXdiv")
  add_subdirectory("${FXDIV_SOURCE_DIR}")
  executorch_move_interface_include_directories_to_build_time_only(fxdiv)
  set(CMAKE_POSITION_INDEPENDENT_CODE
      ${ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG}
  )
endif()

if(EXECUTORCH_BUILD_CPUINFO)
  # --- cpuinfo
  set(ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG
      ${CMAKE_POSITION_INDEPENDENT_CODE}
  )
  set(CMAKE_POSITION_INDEPENDENT_CODE ON)
  set(CPUINFO_SOURCE_DIR
      "${CMAKE_CURRENT_LIST_DIR}/backends/xnnpack/third-party/cpuinfo"
  )
  set(CPUINFO_BUILD_TOOLS
      OFF
      CACHE BOOL ""
  )
  set(CPUINFO_BUILD_UNIT_TESTS
      OFF
      CACHE BOOL ""
  )
  set(CPUINFO_BUILD_MOCK_TESTS
      OFF
      CACHE BOOL ""
  )
  set(CPUINFO_BUILD_BENCHMARKS
      OFF
      CACHE BOOL ""
  )
  set(CPUINFO_LIBRARY_TYPE
      "static"
      CACHE STRING ""
  )
  set(CPUINFO_LOG_LEVEL
      "error"
      CACHE STRING ""
  )
  set(CLOG_SOURCE_DIR "${CPUINFO_SOURCE_DIR}/deps/clog")
  add_subdirectory("${CPUINFO_SOURCE_DIR}")
  set(CMAKE_POSITION_INDEPENDENT_CODE
      ${ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG}
  )
  executorch_add_prefix_to_public_headers(cpuinfo "${CPUINFO_SOURCE_DIR}/")
  install(
    TARGETS cpuinfo
    EXPORT ExecuTorchTargets
    DESTINATION ${CMAKE_INSTALL_LIBDIR}
    INCLUDES
    DESTINATION ${_common_include_directories}
  )
endif()

if(EXECUTORCH_BUILD_PTHREADPOOL)
  # --- pthreadpool
  set(ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG
      ${CMAKE_POSITION_INDEPENDENT_CODE}
  )
  set(CMAKE_POSITION_INDEPENDENT_CODE ON)
  set(PTHREADPOOL_SOURCE_DIR "backends/xnnpack/third-party/pthreadpool")
  set(PTHREADPOOL_BUILD_TESTS
      OFF
      CACHE BOOL ""
  )
  set(PTHREADPOOL_BUILD_BENCHMARKS
      OFF
      CACHE BOOL ""
  )
  set(PTHREADPOOL_LIBRARY_TYPE
      "static"
      CACHE STRING ""
  )
  set(PTHREADPOOL_ALLOW_DEPRECATED_API
      ON
      CACHE BOOL ""
  )
  if(APPLE)
    set(PTHREADPOOL_SYNC_PRIMITIVE
        "condvar"
        CACHE STRING ""
    )
  endif()
  add_subdirectory("${PTHREADPOOL_SOURCE_DIR}")
  executorch_move_interface_include_directories_to_build_time_only(pthreadpool)
  executorch_move_interface_include_directories_to_build_time_only(
    pthreadpool_interface
  )

  if(APPLE)
    # Use hidden visibility for pthreadpool on Apple platforms to avoid issues
    # with pthreadpool symbols from libtorch_cpu taking precedence over the ones
    # from the pthreadpool library statically linked in the Python extension.
    # The pthreadpool public APIs are marked as weak by default on some Apple
    # platforms, so setting to hidden visibility works around this by not
    # putting the symbol in the indirection table. See
    # https://github.com/pytorch/executorch/issues/14321 for more details.
    target_compile_options(pthreadpool PRIVATE -fvisibility=hidden)
  endif()

  install(
    TARGETS pthreadpool pthreadpool_interface fxdiv
    EXPORT ExecuTorchTargets
    DESTINATION ${CMAKE_INSTALL_LIBDIR}
    INCLUDES
    DESTINATION ${_common_include_directories}
  )
  set(CMAKE_POSITION_INDEPENDENT_CODE
      ${ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG}
  )
endif()

if(EXECUTORCH_BUILD_TESTS)
  set(EXECUTORCH_BUILD_EXTENSION_FLAT_TENSOR ON)
  include(CTest)
endif()

# TODO(dbort): Fix these warnings and remove this flag.
list(APPEND _common_compile_options $<$<CXX_COMPILER_ID:MSVC>:/wd4996>
     $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-Wno-deprecated-declarations>
)
# Set default CMAKE_POSITION_INDEPENDENT_CODE behavior if ON or not set
# (default) (and not for MSVC compiler)
if(NOT DEFINED CMAKE_POSITION_INDEPENDENT_CODE
   OR CMAKE_POSITION_INDEPENDENT_CODE
)
  list(APPEND _common_compile_options $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-fPIC>)
endif()

# Let files say "include <executorch/path/to/header.h>".
# TODO(#6475): This requires/assumes that the repo lives in a directory named
# exactly `executorch`. Check the assumption first. Remove this check once we
# stop relying on the assumption.
cmake_path(GET CMAKE_CURRENT_SOURCE_DIR FILENAME _repo_dir_name)
if(NOT "${_repo_dir_name}" STREQUAL "executorch")
  message(
    FATAL_ERROR
      "The ExecuTorch repo must be cloned into a directory named exactly "
      "`executorch`; found `${_repo_dir_name}`. See "
      "https://github.com/pytorch/executorch/issues/6475 for progress on a "
      "fix for this restriction."
  )
endif()
set(_common_include_directories
    $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/..>
    $<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/runtime/core/portable_type/c10>
    $<BUILD_INTERFACE:${CMAKE_CURRENT_BINARY_DIR}/include>
    $<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>
    $<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/core/portable_type/c10>
)

#
# The `_<target>_srcs` lists are defined by executorch_load_build_variables.
#
if(EXECUTORCH_SRCS_FILE)
  message(
    WARNING
      "EXECUTORCH_SRCS_FILE is no longer necessary and will not affect the build."
  )
endif()
executorch_load_build_variables()

# Detect if an iOS toolchain is set.
if(CMAKE_TOOLCHAIN_FILE MATCHES ".*(iOS|ios\.toolchain)\.cmake$")
  set(CMAKE_TOOLCHAIN_IOS ON)
else()
  set(CMAKE_TOOLCHAIN_IOS OFF)
endif()

# Detect if an Android toolchain is set.
if(CMAKE_TOOLCHAIN_FILE MATCHES ".*android\.toolchain\.cmake$")
  set(CMAKE_TOOLCHAIN_ANDROID ON)
  if(NOT ANDROID_PLATFORM)
    set(ANDROID_PLATFORM android-30)
  endif()
else()
  set(CMAKE_TOOLCHAIN_ANDROID OFF)
endif()

# Add code coverage flags to supported compilers
if(EXECUTORCH_USE_CPP_CODE_COVERAGE)
  if("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
    string(APPEND CMAKE_C_FLAGS " --coverage -fprofile-abs-path")
    string(APPEND CMAKE_CXX_FLAGS " --coverage -fprofile-abs-path")
  elseif("${CMAKE_CXX_COMPILER_ID}" MATCHES "Clang")
    string(APPEND CMAKE_C_FLAGS " -fprofile-instr-generate -fcoverage-mapping")
    string(APPEND CMAKE_CXX_FLAGS
           " -fprofile-instr-generate -fcoverage-mapping"
    )
  else()
    message(
      FATAL_ERROR
        "Code coverage for compiler ${CMAKE_CXX_COMPILER_ID} is unsupported"
    )
  endif()
endif()

#
# program_schema: Generated .h files from schema/*.fbs inputs
#
add_subdirectory(schema)

#
# executorch_core: Minimal runtime library
#
# The bare-minimum runtime library, supporting the Program and Method
# interfaces. Does not contain any operators, including primitive ops. Does not
# contain any backends.
#

# Remove any PAL-definition files from the sources.
list(FILTER _executorch_core__srcs EXCLUDE REGEX
     "runtime/platform/default/[^/]*.cpp$"
)

# Add the source file that maps to the requested default PAL implementation.
list(APPEND _executorch_core__srcs ${EXECUTORCH_PAL_DEFAULT_FILE_PATH})

add_library(executorch_core ${_executorch_core__srcs})

# Legacy name alias.
add_library(executorch_no_prim_ops ALIAS executorch_core)

# A list of all configured backends.
set(_executorch_backends "")

# A list of all configured extensions.
set(_executorch_extensions "")

# A list of all configured kernel libraries.
set(_executorch_kernels "")

target_link_libraries(executorch_core PRIVATE program_schema)
if(ANDROID)
  target_link_libraries(executorch_core PUBLIC log)
endif()
# Skip when cross-compiling: find_library() resolves the host libdl even for a
# bare-metal target (e.g. arm-none-eabi), which then links -ldl into the
# baremetal runner and fails ("cannot find -ldl"). dladdr() isn't used on those
# targets anyway. Per-target presets also set EXECUTORCH_USE_DL OFF explicitly.
if(EXECUTORCH_USE_DL AND NOT CMAKE_CROSSCOMPILING)
  # Check if dl exists for this toolchain and only then link it.
  find_library(DL_LIBRARY_EXISTS NAMES dl)
  # Check if the library was found
  if(DL_LIBRARY_EXISTS)
    target_link_libraries(executorch_core PRIVATE dl) # For dladdr()
  endif()
endif()
target_include_directories(
  executorch_core PUBLIC ${_common_include_directories}
)
target_compile_definitions(
  executorch_core PUBLIC C10_USING_CUSTOM_GENERATED_MACROS
)
if(NOT EXECUTORCH_ENABLE_LOGGING)
  target_compile_definitions(executorch_core PUBLIC ET_LOG_ENABLED=0)
endif()
target_compile_options(executorch_core PUBLIC ${_common_compile_options})
if(WIN32)
  # This target carries the vendored c10 headers, and safe_numerics.h reaches
  # std::countl_zero on the branch MSVC takes. Stating it as a compile feature
  # rather than relying on CMAKE_CXX_STANDARD is what makes it reach every
  # consumer: features raise a target's standard even where CMAKE_CXX_STANDARD
  # says 17, and install(EXPORT) carries them into ExecuTorchTargets.cmake, so
  # projects configured separately from this one pick it up too.
  target_compile_features(executorch_core PUBLIC cxx_std_20)
endif()
if(MAX_KERNEL_NUM)
  target_compile_definitions(
    executorch_core PRIVATE MAX_KERNEL_NUM=${MAX_KERNEL_NUM}
  )
endif()

# Build devtools first if needed - some backends depend on protobuf from
# devtools
if(EXECUTORCH_BUILD_DEVTOOLS)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/devtools)
endif()

if(EXECUTORCH_BUILD_PYBIND AND APPLE)
  # shared version
  add_library(executorch_core_shared SHARED ${_executorch_core__srcs})
  target_link_libraries(executorch_core_shared PRIVATE program_schema)
  if(DL_LIBRARY_EXISTS)
    # For dladdr()
    target_link_libraries(executorch_core_shared PRIVATE dl)
  endif()
  target_include_directories(
    executorch_core_shared PUBLIC ${_common_include_directories}
  )
  target_compile_definitions(
    executorch_core_shared PUBLIC C10_USING_CUSTOM_GENERATED_MACROS
  )
  target_compile_options(
    executorch_core_shared PUBLIC ${_common_compile_options}
  )
  if(MAX_KERNEL_NUM)
    target_compile_definitions(
      executorch_core_shared PRIVATE MAX_KERNEL_NUM=${MAX_KERNEL_NUM}
    )
  endif()
endif()

#
# executorch: Primary runtime library with primitive operators.
#
# Provides the Program and Method interfaces, along with primitive operators.
# Does not contain portable kernels or other full operators. Does not contain
# any backends.
#
add_library(executorch ${_executorch__srcs})
target_link_libraries(executorch PRIVATE executorch_core)
target_include_directories(executorch PUBLIC ${_common_include_directories})
target_compile_definitions(executorch PUBLIC C10_USING_CUSTOM_GENERATED_MACROS)
if(NOT EXECUTORCH_ENABLE_LOGGING)
  target_compile_definitions(executorch PUBLIC ET_LOG_ENABLED=0)
endif()
target_compile_options(executorch PUBLIC ${_common_compile_options})
executorch_target_link_options_shared_lib(executorch)

#
# portable_ops_lib: A library to register core ATen ops using portable kernels,
# see kernels/portable/CMakeLists.txt.
#
# Real integrations should supply their own YAML file that only lists the
# operators necessary for the models that will run.
#
if(EXECUTORCH_BUILD_KERNELS_OPTIMIZED)
  # find pytorch lib here to make it available to all sub-directories. Find it
  # before including portable so that optimized_portable_kernels can use it.
  find_package_torch_headers()
endif()

add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/kernels/portable/cpu/util)

# Custom ops AOT needs Torch. Find it at root scope before processing the
# portable subdirectory so that imported targets (e.g. MKL::MKL) are created at
# root scope rather than in a child directory scope.
if(EXECUTORCH_BUILD_KERNELS_CUSTOM_AOT AND EXECUTORCH_BUILD_PORTABLE_OPS)
  find_package_torch()
endif()

if(EXECUTORCH_BUILD_PORTABLE_OPS)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/kernels/portable)
endif()

if(EXECUTORCH_BUILD_KERNELS_OPTIMIZED)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/kernels/optimized)
endif()

add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/configurations)

# Install `executorch` library as well as `executorch-config.cmake` under
# ${CMAKE_INSTALL_PREFIX}/
install(
  DIRECTORY runtime/core/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/core
  FILES_MATCHING
  PATTERN "*.h"
)
# Install generated version.h
install(
  FILES "${CMAKE_CURRENT_BINARY_DIR}/include/executorch/runtime/core/version.h"
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/core
)
install(
  DIRECTORY runtime/executor/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/executor
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  DIRECTORY runtime/kernel/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/kernel
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  DIRECTORY runtime/backend/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/backend
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  DIRECTORY runtime/platform/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/runtime/platform
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  DIRECTORY extension/kernel_util/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/kernel_util
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  DIRECTORY extension/tensor/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/tensor
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  DIRECTORY extension/threadpool/
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/threadpool
  FILES_MATCHING
  PATTERN "*.h"
)
install(
  TARGETS executorch executorch_core
  EXPORT ExecuTorchTargets
  INCLUDES
  DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
)
install(FILES tools/cmake/executorch-config.cmake
        DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/ExecuTorch
)

# Always overwrite backend dependency metadata to avoid stale requirements.
set(_executorch_backend_dependencies "")
if(EXECUTORCH_BUILD_ROCM)
  string(APPEND _executorch_backend_dependencies
         "find_dependency(hip CONFIG)\n"
  )
elseif(EXECUTORCH_BUILD_CUDA)
  string(APPEND _executorch_backend_dependencies
         "find_dependency(CUDAToolkit)\n"
  )
endif()
if(EXECUTORCH_BUILD_NATIVE)
  string(APPEND _executorch_backend_dependencies
         "find_dependency(libzip CONFIG)\n"
         "find_dependency(nlohmann_json CONFIG)\n" "find_dependency(Threads)\n"
  )
endif()
if(EXECUTORCH_BUILD_NATIVE_GGML)
  string(
    APPEND _executorch_backend_dependencies
    "find_dependency(ggml 0.27 CONFIG)\n" "if(NOT TARGET executorch_ggml)\n"
    "  add_library(executorch_ggml ALIAS ggml::ggml)\n" "endif()\n"
  )
endif()
if(EXECUTORCH_BUILD_VULKAN)
  string(APPEND _executorch_backend_dependencies "find_dependency(Threads)\n")
endif()
configure_file(
  tools/cmake/executorch-backend-dependencies.cmake.in
  ${CMAKE_CURRENT_BINARY_DIR}/executorch-backend-dependencies.cmake @ONLY
)
install(FILES ${CMAKE_CURRENT_BINARY_DIR}/executorch-backend-dependencies.cmake
        DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/ExecuTorch
)
if(EXECUTORCH_BUILD_VULKAN
   OR EXECUTORCH_BUILD_VGF
   OR EXECUTORCH_BUILD_NATIVE_VULKAN
)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/vulkan_shared)
endif()

if(EXECUTORCH_BUILD_ARM_BAREMETAL
   OR EXECUTORCH_BUILD_ARM_ETHOSU_LINUX
   OR EXECUTORCH_BUILD_VGF
)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/arm)
endif()

if(EXECUTORCH_BUILD_ARM_BAREMETAL OR EXECUTORCH_BUILD_ARM_ETHOSU_LINUX)
  list(APPEND _executorch_backends executorch_delegate_ethos_u)
endif()

if(EXECUTORCH_BUILD_CADENCE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/cadence)
endif()

if(EXECUTORCH_BUILD_NXP_NEUTRON)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/nxp)
  list(APPEND _executorch_backends executorch_delegate_neutron)
endif()

if(EXECUTORCH_BUILD_NXP_NEUTRON_RUNNER)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/examples/nxp/executor_runner)
endif()

if(EXECUTORCH_BUILD_COREML)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/apple/coreml)
  list(APPEND _executorch_backends coremldelegate)
endif()

if(EXECUTORCH_BUILD_MLX)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/mlx)
  list(APPEND _executorch_backends mlxdelegate)
endif()

if(EXECUTORCH_BUILD_NEURON)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/mediatek)
  list(APPEND _executorch_backends neuron_backend)
endif()

if(EXECUTORCH_BUILD_OPENVINO)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/openvino)
  list(APPEND _executorch_backends openvino_backend)
endif()

if(EXECUTORCH_BUILD_QNN)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/qualcomm)
  list(APPEND _executorch_backends qnn_executorch_backend)
endif()

if(EXECUTORCH_BUILD_ENN)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/samsung)
  list(APPEND _executorch_backends enn_backend)
endif()

if(EXECUTORCH_BUILD_XNNPACK)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/xnnpack)
  list(APPEND _executorch_backends xnnpack_backend)
endif()

if(EXECUTORCH_BUILD_CORTEX_M OR EXECUTORCH_BUILD_CMSIS_NN_PYBINDS)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/cortex_m)
endif()

# Find Torch at root scope when any enabled component needs it so imported
# targets (e.g. MKL::MKL) aren't created independently in child scopes.
if(EXECUTORCH_BUILD_CUDA
   OR EXECUTORCH_BUILD_ROCM
   OR EXECUTORCH_BUILD_METAL
   OR EXECUTORCH_BUILD_KERNELS_LLM_AOT
   OR EXECUTORCH_BUILD_KERNELS_QUANTIZED_AOT
)
  find_package_torch()
elseif(EXECUTORCH_BUILD_EXTENSION_LLM_RUNNER AND EXECUTORCH_BUILD_PYBIND)
  find_package_torch()
elseif(EXECUTORCH_BUILD_EXTENSION_TRAINING AND EXECUTORCH_BUILD_PYBIND)
  find_package_torch()
endif()

# Backend-neutral caller-stream guard consumed by the CUDA AOTI backend (and the
# vendored torch-tensorrt delegate). Built before backends/aoti and
# backends/cuda, which link it.
if(EXECUTORCH_BUILD_CUDA OR EXECUTORCH_BUILD_ROCM)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/cuda)
  install(
    DIRECTORY extension/cuda/
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/cuda
    FILES_MATCHING
    PATTERN "*.h"
  )
  list(APPEND _executorch_extensions extension_cuda)
endif()

# Build common AOTI functionality if needed by CUDA, ROCm, or Metal backends.
if(EXECUTORCH_BUILD_CUDA
   OR EXECUTORCH_BUILD_ROCM
   OR EXECUTORCH_BUILD_METAL
)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/aoti)
endif()

if(EXECUTORCH_BUILD_CUDA OR EXECUTORCH_BUILD_ROCM)
  # Build CUDA-specific AOTI functionality
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/cuda)
  # Add aoti_cuda_backend to backends - it transitively includes aoti_cuda_shims
  # and cuda_platform
  list(APPEND _executorch_backends aoti_cuda_backend)
endif()

if(EXECUTORCH_BUILD_METAL)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/apple/metal)
  list(APPEND _executorch_backends metal_backend)
endif()

if(EXECUTORCH_BUILD_EXTENSION_APPLE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/apple)
endif()

if(EXECUTORCH_BUILD_EXTENSION_DATA_LOADER)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/data_loader)
  install(
    DIRECTORY extension/data_loader/
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/data_loader
    FILES_MATCHING
    PATTERN "*.h"
  )
  list(APPEND _executorch_extensions extension_data_loader)
endif()

if(EXECUTORCH_BUILD_EXTENSION_EVALUE_UTIL)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/evalue_util)
  install(
    DIRECTORY extension/evalue_util/
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/evalue_util
    FILES_MATCHING
    PATTERN "*.h"
  )
endif()

if(EXECUTORCH_BUILD_EXTENSION_FLAT_TENSOR)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/flat_tensor)
  install(
    DIRECTORY extension/flat_tensor/
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/flat_tensor
    FILES_MATCHING
    PATTERN "*.h"
  )
  list(APPEND _executorch_extensions extension_flat_tensor)
endif()

if(EXECUTORCH_BUILD_EXTENSION_MODULE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/module)
  install(
    DIRECTORY extension/module/
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/module
    FILES_MATCHING
    PATTERN "*.h"
  )
  list(APPEND _executorch_extensions extension_module_static)
endif()

if(EXECUTORCH_BUILD_EXTENSION_NAMED_DATA_MAP)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/named_data_map)
  list(APPEND _executorch_extensions extension_named_data_map)
endif()

if(EXECUTORCH_BUILD_EXTENSION_LLM)
  if(EXECUTORCH_BUILD_EXTENSION_LLM_RUNNER
     OR EXECUTORCH_BUILD_NATIVE_LLM_RUNNER
  )
    # Both LLM runners need tokenizers with regex lookahead support.
    set(SUPPORT_REGEX_LOOKAHEAD ON)
    # llama/runner/CMakeLists.txt builds a shared library libllama_runner.so
    # that transitively depends on tokenizers. Need to build tokenizers with
    # -fPIC.
    set(ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG
        ${CMAKE_POSITION_INDEPENDENT_CODE}
    )
    set(CMAKE_POSITION_INDEPENDENT_CODE ON)
  endif()
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/tokenizers)
  if(EXECUTORCH_BUILD_EXTENSION_LLM_RUNNER
     OR EXECUTORCH_BUILD_NATIVE_LLM_RUNNER
  )
    set(CMAKE_POSITION_INDEPENDENT_CODE
        ${ORIGINAL_CMAKE_POSITION_INDEPENDENT_CODE_FLAG}
    )
  endif()
  list(APPEND _executorch_extensions tokenizers)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/cache)
  list(APPEND _executorch_extensions extension_llm_cache)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/batching)
  list(APPEND _executorch_extensions extension_llm_batching
       extension_llm_batching_util extension_llm_batching_module
  )
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/serving)
  list(APPEND _executorch_extensions extension_llm_serving)
  if(EXECUTORCH_BUILD_CUDA)
    # The CUDA batching executor implements extension/llm/batching's seam, so it
    # builds here rather than with backends/cuda.
    add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/cuda/batching)
  endif()
endif()

if(EXECUTORCH_BUILD_EXTENSION_RUNNER_UTIL)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/runner_util)
  install(
    DIRECTORY extension/runner_util/
    DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/executorch/extension/runner_util
    FILES_MATCHING
    PATTERN "*.h"
  )
  list(APPEND _executorch_extensions extension_runner_util)
endif()

if(EXECUTORCH_BUILD_EXTENSION_TENSOR)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/tensor)
  list(APPEND _executorch_extensions extension_tensor)
endif()

if(EXECUTORCH_BUILD_EXTENSION_IMAGE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/image)
  list(APPEND _executorch_extensions extension_image)
endif()

if(EXECUTORCH_BUILD_PTHREADPOOL AND EXECUTORCH_BUILD_CPUINFO)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/threadpool)
endif()

# Consolidated shared library: bundles executorch_core plus commonly used
# extensions into a single libexecutorch.so. Defined before the pybind and
# kernel targets below so they can link this one runtime instead of embedding a
# private copy of the core, which would give the process a second backend
# registry.
if(EXECUTORCH_BUILD_SHARED)
  executorch_add_shared_library(executorch_shared)
  set_target_properties(
    executorch_shared
    PROPERTIES OUTPUT_NAME executorch
               ARCHIVE_OUTPUT_NAME executorch_shared
               EXPORT_NAME executorch-shared
  )
  # Ships in the wheel's lib/ directory beside the libraries that link it.
  executorch_target_shipped_runtime_path(executorch_shared)
  target_include_directories(
    executorch_shared PUBLIC ${_common_include_directories}
  )
  target_compile_definitions(
    executorch_shared PUBLIC C10_USING_CUSTOM_GENERATED_MACROS
  )

  # Outside MSVC, link executorch without WHOLE_ARCHIVE because its INTERFACE
  # link options (from executorch_target_link_options_shared_lib) already force
  # whole-archive. Everything else is pulled in through link options rather than
  # the WHOLE_ARCHIVE link feature, because these archives also reference each
  # other plainly and CMake before 3.29 refuses to mix a feature with a plain
  # reference to the same item.
  if(MSVC)
    # Bundled as objects like the rest; linking it would bring its interface
    # /WHOLEARCHIVE, which lld-link processes ahead of the bundled objects.
    set(_executorch_shared_whole_archive executorch executorch_core)
    # The objects come through $<COMPILE_ONLY>, which does not pass on
    # executorch_core's cxx_std_20. The public headers need it on Windows (the
    # vendored c10 reaches std::countl_zero on the MSVC branch), so a consumer
    # linking this target directly has to be told.
    target_compile_features(executorch_shared PUBLIC cxx_std_20)
    # A DLL resolves its own references at link time, so a weak PAL cannot be
    # overridden through it anyway, and the export list
    # WINDOWS_EXPORT_ALL_SYMBOLS generates skips weak symbols: without this the
    # runtime DLL exports none of the et_pal_* functions, and the public
    # clock.h, whose inline ticks_to_ns() calls et_pal_ticks_to_ns_multiplier(),
    # fails to link against it.
    set_source_files_properties(
      ${EXECUTORCH_PAL_DEFAULT_FILE_PATH} PROPERTIES COMPILE_DEFINITIONS
                                                     ET_PAL_STRONG_SYMBOLS
    )
  else()
    target_link_libraries(executorch_shared PRIVATE executorch)
    set(_executorch_shared_whole_archive executorch_core)
  endif()
  foreach(_ext_target
          extension_data_loader extension_flat_tensor extension_named_data_map
          extension_module_static extension_tensor
  )
    if(TARGET ${_ext_target})
      list(APPEND _executorch_shared_whole_archive ${_ext_target})
    endif()
  endforeach()
  foreach(_whole_target ${_executorch_shared_whole_archive})
    # EXPORTED: this is the runtime API every other component links.
    executorch_target_whole_archive(executorch_shared ${_whole_target} EXPORTED)
  endforeach()
  configure_file(
    tools/cmake/executorch.pc.in ${CMAKE_CURRENT_BINARY_DIR}/executorch.pc
    @ONLY
  )
  install(FILES ${CMAKE_CURRENT_BINARY_DIR}/executorch.pc
          DESTINATION ${CMAKE_INSTALL_LIBDIR}/pkgconfig
  )
  # pip decides where the wheel lands, so the wheel's copy of this file locates
  # everything relative to its own location. Not on Windows: its link flags are
  # GNU-style (-Wl,-rpath) and a DLL records no search path to point them at.
  if(EXECUTORCH_BUILD_WHEEL_DO_NOT_USE AND NOT WIN32)
    set(_executorch_pc_definitions "-DC10_USING_CUSTOM_GENERATED_MACROS")
    if(EXECUTORCH_ENABLE_EVENT_TRACER)
      string(APPEND _executorch_pc_definitions " -DET_EVENT_TRACER_ENABLED")
    endif()
    # Matches the CMake package, which adds both to the runtime whenever the
    # thread pool ships. Without the definition, parallel_for runs serially.
    set(_executorch_pc_libs "-lexecutorch")
    if(TARGET extension_threadpool)
      string(APPEND _executorch_pc_definitions " -DET_USE_THREADPOOL")
      string(APPEND _executorch_pc_libs " -lexecutorch_threadpool")
    endif()
    # Separate arguments, because Meson only keeps a dependency's rpath at
    # install when the argument starts with -Wl,-rpath.
    if(APPLE)
      set(_executorch_pc_rpath "-Wl,-rpath,\${libdir}")
    else()
      set(_executorch_pc_rpath "-Wl,--enable-new-dtags -Wl,-rpath,\${libdir}")
    endif()
    configure_file(
      tools/cmake/executorch-wheel.pc.in
      ${CMAKE_CURRENT_BINARY_DIR}/executorch-wheel.pc @ONLY
    )
  endif()
endif()

# The macro changes the layout of the tracer scope objects, so a consumer
# compiling without it while linking libraries built with it has a different
# view of those structs, with no diagnostic. add_definitions is directory scoped
# and does not survive export, and attaching it to the core alone does not reach
# a consumer either: executorch links the core PRIVATE and portable_kernels
# exports it under LINK_ONLY, so usage requirements do not travel. It has to sit
# on every target a consumer is handed, at the outer scope so it applies to a
# static install too, and after the shared block so the shared target exists
# when it is named.
if(EXECUTORCH_ENABLE_EVENT_TRACER)
  target_compile_definitions(executorch_core INTERFACE ET_EVENT_TRACER_ENABLED)
  target_compile_definitions(executorch INTERFACE ET_EVENT_TRACER_ENABLED)
  if(TARGET executorch_shared)
    target_compile_definitions(
      executorch_shared INTERFACE ET_EVENT_TRACER_ENABLED
    )
  endif()
endif()

if(EXECUTORCH_BUILD_KERNELS_TORCHAO)
  if(NOT TARGET cpuinfo)
    message(
      FATAL_ERROR
        "EXECUTORCH_BUILD_KERNELS_TORCHAO requires EXECUTORCH_BUILD_CPUINFO be set ON"
    )
  endif()
  if(NOT TARGET pthreadpool)
    message(
      FATAL_ERROR
        "EXECUTORCH_BUILD_KERNELS_TORCHAO requires EXECUTORCH_BUILD_PTHREADPOOL be set ON"
    )
  endif()

  # Configure TorchAO kernels
  set(TORCHAO_BUILD_ATEN_OPS OFF)
  set(TORCHAO_BUILD_EXECUTORCH_OPS ON)
  set(TORCHAO_BUILD_CPU_AARCH64 ON)
  set(TORCHAO_ENABLE_ARM_NEON_DOT ON)
  set(TORCHAO_BUILD_KLEIDIAI ON)

  # TorchAO requires find_package(Torch) for include dirs; add its cmake path.
  add_torch_to_cmake_prefix_path()

  # TorchAO kernels look for EXECUTORCH_INCLUDE_DIRS
  if(DEFINED EXECUTORCH_INCLUDE_DIRS)
    message(FATAL_ERROR "EXECUTORCH_INCLUDE_DIRS is already defined")
  endif()
  set(EXECUTORCH_INCLUDE_DIRS
      ${EXECUTORCH_ROOT}/backends/xnnpack/third-party/pthreadpool/include
      ${EXECUTORCH_ROOT}/backends/xnnpack/third-party/cpuinfo/include
  )
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/third-party/ao/torchao/csrc/cpu)
  unset(EXECUTORCH_INCLUDE_DIRS)

  executorch_target_link_options_shared_lib(torchao_ops_executorch)
  list(APPEND _executorch_kernels torchao_ops_executorch)

  # The wheel ships one shared library per component so a process holds a single
  # copy of each. These kernels are static archives declared in the TorchAO
  # submodule, so wrap them the way the quantized kernels are wrapped rather
  # than changing a third-party target. extension_threadpool is named explicitly
  # because the kernels call into it and the wheel ships it as its own library,
  # so this resolves against that one copy instead of bundling a second.
  if(EXECUTORCH_BUILD_SHARED)
    executorch_add_shared_library(
      executorch_kernels_torchao torchao_ops_executorch extension_threadpool
      executorch_shared
    )
    # Holds registration constructors and no symbol a consumer names, so a link
    # with --as-needed would drop it and register nothing.
    executorch_target_link_options_shared_lib(executorch_kernels_torchao)
    # Ships beside the runtime in the wheel's lib/ directory, and links it, so
    # it has to be able to find it from wherever the package is installed.
    executorch_target_shipped_runtime_path(executorch_kernels_torchao)
  endif()

  install(
    TARGETS torchao_ops_executorch torchao_kernels_aarch64
    EXPORT ExecuTorchTargets
    DESTINATION ${CMAKE_INSTALL_LIBDIR}
    INCLUDES
    DESTINATION ${_common_include_directories}
  )
  # If using KleidiAI and XNNPACK has not installed it already, install it
  if(TORCHAO_BUILD_KLEIDIAI AND NOT (EXECUTORCH_BUILD_XNNPACK
                                     AND EXECUTORCH_XNNPACK_ENABLE_KLEIDI)
  )
    install(
      TARGETS kleidiai
      EXPORT ExecuTorchTargets
      DESTINATION ${CMAKE_INSTALL_LIBDIR}
      INCLUDES
      DESTINATION ${_common_include_directories}
    )
  endif()

endif()

if(TARGET cpublas
   AND TARGET kleidiai
   AND TARGET cpuinfo
)
  if(CMAKE_SYSTEM_PROCESSOR MATCHES "aarch64|arm64"
     OR ANDROID_ABI STREQUAL "arm64-v8a"
     OR CMAKE_OSX_ARCHITECTURES MATCHES "arm64"
  )
    target_compile_definitions(cpublas PRIVATE ET_BUILD_WITH_KLEIDIAI)
    if(CMAKE_SYSTEM_NAME STREQUAL "Darwin")
      target_compile_definitions(cpublas PRIVATE ET_KLEIDIAI_DISABLE_NEON_BF16)
    endif()
    target_link_libraries(cpublas PRIVATE kleidiai)
  endif()
endif()

# The shared build ships the profiler as one of its libraries, and the Python
# extension records a hard dependency on it, and the Apple ETDump extension
# links it, so the target has to exist whenever any of those is being built
# rather than only when devtools is asked for.
if((EXECUTORCH_BUILD_PYBIND
    OR EXECUTORCH_BUILD_SHARED
    OR EXECUTORCH_BUILD_EXTENSION_ETDUMP_APPLE)
   AND NOT EXECUTORCH_BUILD_DEVTOOLS
)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/devtools)
endif()

if(EXECUTORCH_BUILD_PYBIND)

  if(NOT EXECUTORCH_BUILD_EXTENSION_DATA_LOADER)
    add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/data_loader)
  endif()

  # Add codegen tools subdirectory for selective_build pybind module
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/codegen/tools)

  # Create bundled_module target only for pybindings when bundled_program exists
  # This target has hard dependencies on devtools generated headers
  if(TARGET bundled_program)
    add_library(
      bundled_module STATIC
      ${CMAKE_CURRENT_SOURCE_DIR}/extension/module/bundled_module.cpp
    )

    # Ensure bundled_module waits for bundled_program's generated headers
    add_dependencies(bundled_module bundled_program)

    # extension_module_static and the data loader are bundled into
    # libexecutorch.so, so link that instead of pulling private static copies in
    # through this target's PUBLIC interface.
    if(EXECUTORCH_BUILD_SHARED)
      target_link_libraries(
        bundled_module PUBLIC executorch_shared bundled_program
      )
    else()
      target_link_libraries(bundled_module PRIVATE extension_data_loader)
      target_link_libraries(
        bundled_module PUBLIC extension_module_static bundled_program
      )
    endif()

    target_include_directories(
      bundled_module PUBLIC ${_common_include_directories}
    )
    target_compile_options(
      bundled_module
      PUBLIC $<$<CXX_COMPILER_ID:MSVC>:/wd4996>
             $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-Wno-deprecated-declarations
             -fPIC>
    )
  endif()

  # When the consolidated shared runtime is built, the pybind extension links it
  # instead of whole-archiving the static core, so Python and C++ consumers
  # share one backend registry. `executorch` and the extensions bundled into
  # libexecutorch.so must stay off this list: their INTERFACE link options force
  # whole-archive, which would give this module a private second registry.
  # executorch_shared is named here for its include directories and compile
  # definitions; executorch_target_link_shared_runtime below is what fixes its
  # position on the link line.
  if(EXECUTORCH_BUILD_SHARED)
    set(_dep_libs bundled_program etdump flatccrt executorch_shared)
  else()
    set(_dep_libs bundled_program etdump flatccrt executorch
                  extension_data_loader extension_tensor
    )
  endif()
  # The bindings call the threadpool control APIs directly, so static builds
  # must retain this library rather than relying on transitive linkage.
  if(TARGET extension_threadpool)
    list(APPEND _dep_libs extension_threadpool)
  endif()

  # Build common AOTI functionality if needed by CUDA, ROCm, or Metal backends.
  if(EXECUTORCH_BUILD_CUDA OR EXECUTORCH_BUILD_ROCM)
    # CUDA uses SlimTensor-based shims. Not on Windows: a CUDA program cannot be
    # lowered there, only run, so the delegate ships for C++ applications and
    # the Python extension stays free of any CUDA dependency.
    if(NOT WIN32)
      list(APPEND _dep_libs aoti_cuda_backend)
    endif()
  elseif(EXECUTORCH_BUILD_METAL)
    # Metal still uses ETensor-based shims (for now)
    list(APPEND _dep_libs aoti_common)
  endif()

  # RPATH for the wheel's own lib/ directory. Backend paths are appended below.
  set(_python_extension_rpath "$ORIGIN/../../src/executorch/lib")

  if(EXECUTORCH_BUILD_EXTENSION_MODULE)
    # extension_module_static is already bundled into libexecutorch.so; linking
    # it again here would whole-archive a second copy.
    if(NOT EXECUTORCH_BUILD_SHARED)
      # Always use static linking for pybindings to avoid runtime symbol
      # resolution issues
      list(APPEND _dep_libs extension_module_static)
    endif()
    # Add bundled_module if available
    if(TARGET bundled_module)
      list(APPEND _dep_libs bundled_module)
    endif()
  endif()

  if(EXECUTORCH_BUILD_TESTS)
    list(APPEND _dep_libs test_backend_compiler_lib)
  endif()

  if(EXECUTORCH_BUILD_KERNELS_OPTIMIZED)
    list(APPEND _dep_libs optimized_native_cpu_ops_lib)
  else()
    list(APPEND _dep_libs portable_ops_lib)
  endif()

  if(EXECUTORCH_BUILD_COREML AND APPLE)
    list(APPEND _dep_libs coremldelegate)
  endif()

  if(EXECUTORCH_BUILD_VGF)
    list(APPEND _dep_libs vgf_backend)
  endif()

  if(EXECUTORCH_BUILD_MLX)
    list(APPEND _dep_libs mlxdelegate)
  endif()

  if(EXECUTORCH_BUILD_OPENVINO)
    # Under a shared build this resolves to the shipped library rather than a
    # static archive, so the delegate is not copied into the extension and the
    # process registers it once.
    list(APPEND _dep_libs openvino_backend)
  endif()

  if(EXECUTORCH_BUILD_CUDA OR EXECUTORCH_BUILD_ROCM)
    string(APPEND _python_extension_rpath ":$ORIGIN/../../backends/cuda")
  endif()

  if(EXECUTORCH_BUILD_QNN)
    list(APPEND _dep_libs qnn_executorch_backend)
    string(APPEND _python_extension_rpath ":$ORIGIN/../../backends/qualcomm")
  endif()

  if(EXECUTORCH_BUILD_ENN)
    list(APPEND _dep_libs enn_backend)
  endif()

  if(EXECUTORCH_BUILD_XNNPACK)
    if(EXECUTORCH_BUILD_SHARED)
      # The delegate bundles XNNPACK and its microkernels, so naming them again
      # here would ship a second copy.
      list(APPEND _dep_libs xnnpack_backend)
    else()
      # need to explicitly specify XNNPACK and xnnpack-microkernels-prod here
      # otherwise uses XNNPACK and microkernel-prod symbols from libtorch_cpu
      list(APPEND _dep_libs xnnpack_backend XNNPACK xnnpack-microkernels-prod)
    endif()
  endif()

  if(EXECUTORCH_BUILD_VULKAN)
    list(APPEND _dep_libs vulkan_backend)
  endif()

  if(EXECUTORCH_BUILD_WEBGPU)
    list(APPEND _dep_libs webgpu_backend)
  endif()

  # compile options for pybind
  set(_pybind_compile_options
      $<$<CXX_COMPILER_ID:MSVC>:/EHsc
      /GR
      /wd4996>
      $<$<NOT:$<CXX_COMPILER_ID:MSVC>>:-Wno-deprecated-declarations
      -fPIC
      -frtti
      -fexceptions>
  )

  # The training pybinding accepts native PyTorch tensors and still needs the
  # ATen bridge. Keep that bridge on the ATen-consuming target rather than on
  # the portable pybinding.
  if(EXECUTORCH_BUILD_EXTENSION_TRAINING)
    find_library(
      TORCH_PYTHON_LIBRARY torch_python PATHS "${TORCH_INSTALL_PREFIX}/lib"
    )
    add_library(
      aten_bridge STATIC
      ${CMAKE_CURRENT_SOURCE_DIR}/extension/aten_util/aten_bridge.cpp
    )
    set_target_properties(aten_bridge PROPERTIES CXX_STANDARD 20)
    target_include_directories(
      aten_bridge PUBLIC ${_common_include_directories} ${TORCH_INCLUDE_DIRS}
    )
    target_compile_definitions(
      aten_bridge PUBLIC C10_USING_CUSTOM_GENERATED_MACROS
    )
    target_compile_options(aten_bridge PUBLIC ${_pybind_compile_options})
    # aten_bridge is static, so _training_lib gets this link. torch would record
    # libtorch_cuda with a CUDA build of PyTorch; torch_cpu_library does not.
    if(EXECUTORCH_BUILD_SHARED)
      target_link_libraries(
        aten_bridge PRIVATE torch_cpu_library c10 executorch_shared
      )
    else()
      target_link_libraries(
        aten_bridge PRIVATE torch_cpu_library c10 executorch extension_tensor
      )
    endif()
  endif()

  # pybind portable_lib
  executorch_add_pybind_module(portable_lib extension/pybindings/pybindings.cpp)
  # The actual output file needs a leading underscore so it can coexist with
  # portable_lib.py in the same python package.
  set_target_properties(
    portable_lib PROPERTIES OUTPUT_NAME "_C" CXX_STANDARD 17
  )
  target_compile_definitions(
    portable_lib PUBLIC EXECUTORCH_PYTHON_MODULE_NAME=_C
  )
  target_compile_options(portable_lib PUBLIC ${_pybind_compile_options})
  if(CMAKE_SYSTEM_NAME STREQUAL "Linux" AND CMAKE_CXX_COMPILER_ID STREQUAL
                                            "GNU"
  )
    # Keep inline helpers local so GCC LTO retains definitions needed by
    # manylinux's libstdc++_nonshared.a instead of resolving them through the
    # shared runtime.
    target_compile_options(portable_lib PRIVATE -fvisibility-inlines-hidden)
  endif()
  target_link_libraries(portable_lib PRIVATE ${_dep_libs})
  if(TARGET extension_cuda)
    target_link_libraries(portable_lib PRIVATE extension_cuda)
    target_compile_definitions(portable_lib PRIVATE EXECUTORCH_PYBINDINGS_CUDA)
  endif()
  executorch_target_link_shared_runtime(portable_lib)
  # These libraries register their operators or their backend from a static
  # initializer, so nothing here references a symbol from them and some linkers
  # drop them from DT_NEEDED. That surfaces at runtime as a missing kernel or an
  # unregistered backend rather than as a link error. A library reached
  # transitively is still subject to --as-needed and gets dropped, so the thread
  # pool and the profiler need naming here too even though the extension does
  # reference symbols from them. A distro toolchain that defaults to --as-needed
  # would otherwise leave them out of DT_NEEDED entirely.
  #
  # Only the libraries that do not already retain themselves are named. A
  # delegate that calls executorch_target_link_options_shared_lib exports
  # --no-as-needed around its own file to every consumer, so the CUDA and
  # OpenVINO backends are covered by that and are deliberately absent here. That
  # export is emitted only for NOT APPLE, so it does nothing for an Apple-only
  # delegate like Core ML. In the wheel's shared build that dylib is instead
  # kept because it is named directly on the link line, which retains its
  # namespace-scope initializers without an --as-needed equivalent; only
  # -dead_strip_dylibs would drop it, and nothing here passes that. In a static
  # build the delegate is an archive kept by force_load instead.
  foreach(_retained_component optimized_native_cpu_ops_lib xnnpack_backend
                              extension_threadpool etdump
  )
    if(TARGET ${_retained_component})
      executorch_target_retain_shared_library(
        portable_lib ${_retained_component}
      )
    endif()
  endforeach()

  # Mach-O receives its runtime paths from the helpers below. The explicit
  # backend paths above use ELF's $ORIGIN spelling and only apply elsewhere.
  if(NOT APPLE)
    set_target_properties(
      portable_lib PROPERTIES BUILD_RPATH "${_python_extension_rpath}"
                              INSTALL_RPATH "${_python_extension_rpath}"
    )
  endif()
  executorch_target_shared_runtime_path(
    portable_lib "extension/pybindings" "executorch/extension/pybindings"
  )

  install(
    TARGETS portable_lib
    EXPORT ExecuTorchTargets
    LIBRARY DESTINATION executorch/extension/pybindings
  )

  # pybind data_loader module - provides PyDataLoader type for external
  # pybinding extensions to create custom data loaders
  executorch_add_pybind_module(
    data_loader extension/pybindings/pybindings_data_loader.cpp
  )
  target_include_directories(data_loader PRIVATE ${_common_include_directories})
  target_compile_options(data_loader PUBLIC ${_pybind_compile_options})
  # This module only exposes a pybind type and calls into no runtime symbols.
  # The static target force-links every registration object, which would give
  # this module its own operator registry alongside the one in the shared
  # runtime, so resolve against the shared runtime instead when there is one.
  if(TARGET executorch_shared AND MSVC)
    # Compile settings only. Linking would add the runtime's /INCLUDE anchor,
    # the module's one import from executorch.dll, and nothing on this module's
    # import path registers lib/ for the Windows loader, so importing it first
    # failed to load. A runtime call added later fails to link instead.
    target_link_libraries(data_loader PRIVATE $<COMPILE_ONLY:executorch_shared>)
  elseif(TARGET executorch_shared)
    target_link_libraries(data_loader PRIVATE executorch_shared)
    executorch_target_shared_runtime_path(
      data_loader "extension/pybindings" "executorch/extension/pybindings"
    )
  else()
    target_link_libraries(data_loader PRIVATE executorch)
  endif()
  install(TARGETS data_loader
          LIBRARY DESTINATION executorch/extension/pybindings
  )

  # Copy the MLX metallib next to the Python extension for editable installs.
  # MLX uses dladdr() to find the directory containing the library with MLX
  # code, then looks for mlx.metallib in that directory. This covers the case
  # where the delegate is static and therefore absorbed into the extension; a
  # shared delegate carries MLX itself and is served by the copy installed
  # beside it instead.
  executorch_target_copy_mlx_metallib(portable_lib)
endif()

if(EXECUTORCH_BUILD_TOKENIZERS_WASM)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/wasm/tokenizers)
endif()

if(EXECUTORCH_BUILD_EXTENSION_TRAINING)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/training)
  list(APPEND _executorch_extensions extension_training)
endif()

if(EXECUTORCH_BUILD_EXTENSION_LLM_RUNNER)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/memory_allocator)
  list(APPEND _executorch_extensions extension_memory_allocator)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/runner)
  list(APPEND _executorch_extensions extension_llm_runner)
endif()

if(EXECUTORCH_BUILD_EXTENSION_ASR_RUNNER)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/asr/runner)
  list(APPEND _executorch_extensions extension_asr_runner)
endif()

if(EXECUTORCH_BUILD_EXTENSION_ETDUMP_APPLE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/apple/dump)
endif()

if(EXECUTORCH_BUILD_EXTENSION_LLM_APPLE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/apple)
endif()

if(EXECUTORCH_BUILD_KERNELS_LLM)
  # TODO: move all custom kernels to ${CMAKE_CURRENT_SOURCE_DIR}/kernels/custom
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/llm/custom_ops)
  list(APPEND _executorch_kernels custom_ops_aot_lib)
endif()

if(EXECUTORCH_BUILD_KERNELS_QUANTIZED)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/kernels/quantized)
  executorch_target_link_options_shared_lib(quantized_ops_lib)
  list(APPEND _executorch_kernels quantized_ops_lib)
endif()

if(EXECUTORCH_BUILD_VULKAN)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/vulkan)
  list(APPEND _executorch_backends vulkan_backend vulkan_schema)
endif()

if(EXECUTORCH_BUILD_NATIVE)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/native)
endif()

if(EXECUTORCH_BUILD_WEBGPU)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/backends/webgpu)
  list(APPEND _executorch_backends webgpu_backend)
endif()

if(EXECUTORCH_BUILD_VGF)
  list(APPEND _executorch_backends vgf_backend)
endif()

# Added after the backends so EXECUTORCH_BUILD_WASM_WEBGPU can link
# webgpu_backend, which has to already be defined at that point.
if(EXECUTORCH_BUILD_WASM)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/wasm)
endif()

# Top-level interface targets.

# A target containing all configured backends.
add_library(executorch_backends INTERFACE)
add_library(executorch::backends ALIAS executorch_backends)
target_link_libraries(executorch_backends INTERFACE ${_executorch_backends})

# A target containing all configured extensions.
add_library(executorch_extensions INTERFACE)
add_library(executorch::extensions ALIAS executorch_extensions)
target_link_libraries(executorch_extensions INTERFACE ${_executorch_extensions})

# A target containing all configured kernels, with selective build, if enabled.
add_library(executorch_kernels INTERFACE)
add_library(executorch::kernels ALIAS executorch_kernels)
if(NOT EXECUTORCH_SELECT_OPS_YAML STREQUAL ""
   OR NOT EXECUTORCH_SELECT_OPS_LIST STREQUAL ""
   OR NOT EXECUTORCH_SELECT_OPS_MODEL STREQUAL ""
)
  # Use optimized kernels when available.
  if(EXECUTORCH_BUILD_KERNELS_OPTIMIZED)
    set(_selected_kernel_functions_yaml
        ${CMAKE_CURRENT_BINARY_DIR}/configurations/merged.yaml
    )
    set(_selected_kernel_libs "optimized_kernels" "optimized_portable_kernels")
  else()
    set(_selected_kernel_functions_yaml
        ${EXECUTORCH_ROOT}/kernels/portable/functions.yaml
    )
    set(_selected_kernel_libs "portable_kernels")
  endif()

  gen_selected_ops(
    LIB_NAME
    "executorch_selected_kernels"
    OPS_SCHEMA_YAML
    "${EXECUTORCH_SELECT_OPS_YAML}"
    ROOT_OPS
    "${EXECUTORCH_SELECT_OPS_LIST}"
    INCLUDE_ALL_OPS
    FALSE
    OPS_FROM_MODEL
    "${EXECUTORCH_SELECT_OPS_MODEL}"
    DTYPE_SELECTIVE_BUILD
    "${EXECUTORCH_ENABLE_DTYPE_SELECTIVE_BUILD}"
  )

  generate_bindings_for_kernels(
    LIB_NAME
    "executorch_selected_kernels"
    FUNCTIONS_YAML
    ${_selected_kernel_functions_yaml}
    CUSTOM_OPS_YAML
    ""
    DTYPE_SELECTIVE_BUILD
    "${EXECUTORCH_ENABLE_DTYPE_SELECTIVE_BUILD}"
  )

  gen_operators_lib(
    LIB_NAME
    "executorch_selected_kernels"
    KERNEL_LIBS
    ${_selected_kernel_libs}
    DEPS
    executorch_core
    DTYPE_SELECTIVE_BUILD
    "${EXECUTORCH_ENABLE_DTYPE_SELECTIVE_BUILD}"
  )
  list(APPEND _executorch_kernels executorch_selected_kernels)

  # Auto-right-size the kernel registry unless the user has pinned
  # MAX_KERNEL_NUM.
  if(NOT DEFINED CACHE{MAX_KERNEL_NUM} AND NOT DEFINED MAX_KERNEL_NUM)
    gen_selected_max_kernel_num(
      LIB_NAME "executorch_selected_kernels" OPLIST_YAMLS
      ${gen_selected_ops_output_yaml}
    )
    target_include_directories(
      executorch_core
      PRIVATE ${executorch_selected_kernels_max_kernel_num_include_dir}
    )
    add_dependencies(
      executorch_core executorch_selected_kernels_max_kernel_num_header
    )
    if(TARGET executorch_core_shared)
      target_include_directories(
        executorch_core_shared
        PRIVATE ${executorch_selected_kernels_max_kernel_num_include_dir}
      )
      add_dependencies(
        executorch_core_shared
        executorch_selected_kernels_max_kernel_num_header
      )
    endif()
  endif()

  install(
    TARGETS executorch_selected_kernels
    EXPORT ExecuTorchTargets
    DESTINATION ${CMAKE_INSTALL_LIBDIR}
  )
else()
  # No selective build - link the full library.
  if(EXECUTORCH_BUILD_KERNELS_OPTIMIZED)
    list(APPEND _executorch_kernels optimized_native_cpu_ops_lib)
  else()
    list(APPEND _executorch_kernels portable_ops_lib)
  endif()
endif()
target_link_libraries(executorch_kernels INTERFACE ${_executorch_kernels})

install(TARGETS executorch_backends executorch_extensions executorch_kernels
        EXPORT ExecuTorchTargets
)

if(EXECUTORCH_BUILD_EXECUTOR_RUNNER)
  # Baseline libraries that executor_runner will link against.
  set(_executor_runner_libs executorch extension_evalue_util
                            extension_runner_util gflags executorch_backends
  )

  if(EXECUTORCH_BUILD_EXTENSION_FLAT_TENSOR)
    list(APPEND _executor_runner_libs extension_flat_tensor)
  endif()

  if(EXECUTORCH_BUILD_KERNELS_OPTIMIZED)
    list(APPEND _executor_runner_libs optimized_native_cpu_ops_lib)
  elseif(EXECUTORCH_BUILD_CADENCE)
    list(APPEND _executor_runner_libs cadence_ops_lib)
  else()
    list(APPEND _executor_runner_libs portable_ops_lib)
  endif()

  # Generate lib to register quantized ops
  if(EXECUTORCH_BUILD_KERNELS_QUANTIZED)
    list(APPEND _executor_runner_libs quantized_ops_lib)
  endif()

  if(EXECUTORCH_BUILD_KERNELS_LLM)
    list(APPEND _executor_runner_libs $<LINK_LIBRARY:WHOLE_ARCHIVE,custom_ops>)
  endif()

  if(EXECUTORCH_ENABLE_EVENT_TRACER)
    list(APPEND _executor_runner_libs etdump flatccrt)
  endif()

  if(EXECUTORCH_ENABLE_BUNDLE_IO)
    list(APPEND _executor_runner_libs bundled_program)
  endif()

  add_executable(executor_runner ${_executor_runner__srcs})
  if(NOT CMAKE_BUILD_TYPE STREQUAL "Debug")
    target_link_options_gc_sections(executor_runner)
  endif()
  target_link_libraries(executor_runner PRIVATE ${_executor_runner_libs})
  target_compile_options(executor_runner PUBLIC ${_common_compile_options})
  if(ET_ENABLE_VGF_INPUT_DUMP)
    target_compile_definitions(executor_runner PRIVATE ET_ENABLE_VGF_INPUT_DUMP)
  endif()
  if(EXECUTORCH_BUILD_ARM_ETHOSU_LINUX)
    target_compile_definitions(
      executor_runner PRIVATE EXECUTORCH_BUILD_ARM_ETHOSU_LINUX=1
    )
    target_link_libraries(
      executor_runner
      PRIVATE $<LINK_LIBRARY:WHOLE_ARCHIVE,executorch_delegate_ethos_u>
    )
    # Wrap static linking like the delegate_runner to keep images
    # self-contained.
    target_link_options(
      executor_runner PRIVATE -static-libstdc++ -static-libgcc
    )
  endif()

  # Automatically set when using `emcmake cmake` for Wasm build.
  if(EMSCRIPTEN)
    # Directory of model pte files to embed in the wasm binary.
    if(NOT DEFINED WASM_MODEL_DIR)
      set(WASM_MODEL_DIR "${CMAKE_SOURCE_DIR}/models/")
    endif()

    set(CMAKE_EXECUTABLE_SUFFIX ".html")
    target_link_options(
      executor_runner PUBLIC -sALLOW_MEMORY_GROWTH --embed-file
      "${WASM_MODEL_DIR}@/"
    )
  endif()
endif()

if(EXECUTORCH_BUILD_ANDROID_JNI)
  add_subdirectory(${CMAKE_CURRENT_SOURCE_DIR}/extension/android)
endif()

include(Test.cmake)

install(
  EXPORT ExecuTorchTargets
  FILE ExecuTorchTargets.cmake
  DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/ExecuTorch
)

# Print all the configs that were called with announce_configured_options.
print_configured_options()