message(STATUS  "GGML_SYCL_TARGET=${GGML_SYCL_TARGET}")

if (NOT GGML_SYCL_TARGET MATCHES "^(INTEL|NVIDIA|AMD)$")
    message(FATAL_ERROR "Invalid backend chosen, supported options are INTEL, NVIDIA, or AMD")
endif()

check_cxx_compiler_flag("-fsycl" SUPPORTS_SYCL)

if (DEFINED ENV{ONEAPI_ROOT})
    message(STATUS "Using oneAPI Release SYCL compiler (icpx).")
elseif(SUPPORTS_SYCL)
    message(WARNING "Using open-source SYCL compiler (clang++). Didn't detect ENV {ONEAPI_ROOT}.
        If you expected the oneAPI Release compiler, please install oneAPI & source it, like:
        source /opt/intel/oneapi/setvars.sh")
else()
    message(FATAL_ERROR "C++ compiler lacks SYCL support.")
endif()
message(STATUS "SYCL found")
#todo: AOT

ggml_add_backend_library(ggml-sycl
                         ggml-sycl.cpp
                         ../../include/ggml-sycl.h
                        )

file(GLOB   GGML_HEADERS_SYCL "*.hpp")
file(GLOB   GGML_SOURCES_SYCL "*.cpp")
target_sources(ggml-sycl PRIVATE ${GGML_HEADERS_SYCL} ${GGML_SOURCES_SYCL})

if (WIN32)
    # To generate a Visual Studio solution, using Intel C++ Compiler for ggml-sycl is mandatory
    if( ${CMAKE_GENERATOR} MATCHES "Visual Studio" AND NOT (${CMAKE_GENERATOR_TOOLSET} MATCHES "Intel C"))
        set_target_properties(ggml-sycl PROPERTIES VS_PLATFORM_TOOLSET "Intel C++ Compiler 2025")
        set(CMAKE_CXX_COMPILER "icx")
        set(CMAKE_CXX_COMPILER_ID "IntelLLVM")
    endif()
endif()

find_package(IntelSYCL)
if (IntelSYCL_FOUND)
    # Use oneAPI CMake when possible
    target_link_libraries(ggml-sycl PRIVATE IntelSYCL::SYCL_CXX)
else()
    # Fallback to the simplest way of enabling SYCL when using intel/llvm nightly for instance
    target_compile_options(ggml-sycl PRIVATE "-fsycl")
    target_link_options(ggml-sycl PRIVATE "-fsycl")
endif()

target_compile_options(ggml-sycl PRIVATE "-Wno-narrowing")

# Link against oneDNN
set(GGML_SYCL_DNNL 0)
if(GGML_SYCL_DNN)
    find_package(DNNL)
    if(DNNL_FOUND)
        if (NOT DEFINED DNNL_GPU_VENDOR)
            # default to intel target
            set(DNNL_GPU_VENDOR "INTEL")
            if(NOT "${GGML_SYCL_TARGET}" STREQUAL "INTEL")
                message(WARNING "oneDNN builds bundled with oneapi release only support INTEL target")
            endif()
        endif()

        # Verify oneDNN was compiled for the same target as llama
        if("${GGML_SYCL_TARGET}" STREQUAL "${DNNL_GPU_VENDOR}")
            target_link_libraries(ggml-sycl PRIVATE DNNL::dnnl)
            set(GGML_SYCL_DNNL 1)
            get_target_property(CONFIGS DNNL::dnnl IMPORTED_CONFIGURATIONS)
            foreach(CONFIG ${CONFIGS})
                get_target_property(DNNL_LIB DNNL::dnnl IMPORTED_LOCATION_${CONFIG})
                message(STATUS "Found oneDNN: ${DNNL_LIB}")
            endforeach()
        else()
            message(WARNING
                "oneDNN must be compiled for the same target as llama.cpp.
                 llama.cpp: ${GGML_SYCL_TARGET}, oneDNN: ${DNNL_GPU_VENDOR}.
                 Disabling oneDNN support.")
        endif()
    else()
        message(STATUS "oneDNN not found, disabling oneDNN support")
    endif()
else()
    message(STATUS "oneDNN support disabled by the user")
endif()
target_compile_definitions(ggml-sycl PRIVATE GGML_SYCL_DNNL=${GGML_SYCL_DNNL})

if (GGML_SYCL_F16)
    if (GGML_SYCL_TARGET STREQUAL "AMD")
        message(WARNING "AMD target does not entirely support FP16 in the SYCL backend.")
    endif()
    add_compile_definitions(GGML_SYCL_F16)
endif()

if (GGML_SYCL_TARGET STREQUAL "NVIDIA")
    add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
elseif (GGML_SYCL_TARGET STREQUAL "AMD")
    # INFO: Allowed Sub_group_sizes are not consistent through all
    # hip targets. For example, 64 is used for certain models, but the backend
    # does not support it.
    # Target archs tested working: gfx1030, gfx1031, (Only tested sub_group_size = 32)
    add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
else()
    add_compile_definitions(GGML_SYCL_WARP_SIZE=16)
endif()

if (GGML_SYCL_GRAPH)
    target_compile_definitions(ggml-sycl PRIVATE GGML_SYCL_GRAPH)
endif()

# Link against Intel oneMKL or oneMath
if (GGML_SYCL_TARGET STREQUAL "INTEL")
    # Intel devices use Intel oneMKL directly instead of oneMath to avoid the limitation of linking Intel oneMKL statically
    # See https://github.com/uxlfoundation/oneMath/issues/654
    if (CMAKE_CXX_COMPILER_ID STREQUAL "Clang")
        set(SYCL_COMPILER ON)
    endif()
    find_package(MKL REQUIRED)
    target_link_libraries(ggml-sycl PRIVATE MKL::MKL_SYCL::BLAS)
    target_compile_definitions(ggml-sycl PRIVATE GGML_SYCL_USE_INTEL_ONEMKL)
else()
    find_package(oneMath QUIET)
    if (NOT oneMath_FOUND)
        message(STATUS "oneMath not found: oneMath will be automatically downloaded")
        # Use FetchContent to automatically pull and build oneMath
        include(FetchContent)
        set(BUILD_FUNCTIONAL_TESTS False)
        set(BUILD_EXAMPLES False)
        set(TARGET_DOMAINS blas)
        if (GGML_SYCL_TARGET STREQUAL "NVIDIA")
            set(ENABLE_MKLCPU_BACKEND False)
            set(ENABLE_MKLGPU_BACKEND False)
            set(ENABLE_CUBLAS_BACKEND True)
        elseif (GGML_SYCL_TARGET STREQUAL "AMD")
            set(ENABLE_MKLCPU_BACKEND False)
            set(ENABLE_MKLGPU_BACKEND False)
            set(ENABLE_ROCBLAS_BACKEND True)
            # Ensure setting a string variable here is not overriden by oneMath CACHE variables
            cmake_policy(SET CMP0126 NEW)
            # Setting the device architecture is only needed and useful for AMD devices in oneMath
            set(HIP_TARGETS ${GGML_SYCL_DEVICE_ARCH} CACHE STRING "oneMath HIP target" FORCE)
        endif()
        FetchContent_Declare(
            ONEMATH
            GIT_REPOSITORY https://github.com/uxlfoundation/oneMath.git
            GIT_TAG 8efe85f5aaebb37f1d8c503b7af66315feabf142
        )
        FetchContent_MakeAvailable(ONEMATH)
        # Create alias to match with find_package targets name
        function(onemath_alias target)
            if (TARGET ${target}_obj)
                # Silence verbose warnings from external libraries
                target_compile_options(${target}_obj PRIVATE -w)
            endif()
            if (TARGET ${target})
                add_library(ONEMATH::${target} ALIAS ${target})
            endif()
        endfunction()
        onemath_alias(onemath)
        onemath_alias(onemath_blas_mklcpu)
        onemath_alias(onemath_blas_mklgpu)
        onemath_alias(onemath_blas_cublas)
        onemath_alias(onemath_blas_rocblas)
    endif()

    # Below oneMath compile-time dispatching is used for better performance
    if (GGML_SYCL_TARGET STREQUAL "NVIDIA")
        target_link_libraries(ggml-sycl PRIVATE ONEMATH::onemath_blas_cublas)
        target_compile_options(ggml-sycl PRIVATE "-fsycl-targets=nvptx64-nvidia-cuda")
        target_link_options(ggml-sycl PRIVATE "-fsycl-targets=nvptx64-nvidia-cuda")
        target_compile_definitions(ggml-sycl PRIVATE GGML_SYCL_NVIDIA)
    elseif (GGML_SYCL_TARGET STREQUAL "AMD")
        if (NOT GGML_SYCL_DEVICE_ARCH)
            message(FATAL_ERROR "Can't enable SYCL hip backend, GGML_SYCL_DEVICE_ARCH has not been set.")
        endif()
        target_link_libraries(ggml-sycl PRIVATE ONEMATH::onemath_blas_rocblas)
        target_compile_options(ggml-sycl PRIVATE "-fsycl-targets=amdgcn-amd-amdhsa")
        target_link_options(ggml-sycl PRIVATE "-fsycl-targets=amdgcn-amd-amdhsa")
        target_compile_definitions(ggml-sycl PRIVATE GGML_SYCL_AMD)
    else()
        # Fallback to oneMath runtime dispatcher
        target_link_libraries(ggml-sycl PRIVATE ONEMATH::onemath)
        target_compile_definitions(ggml-sycl PRIVATE GGML_SYCL_GENERIC)
    endif()
endif()

if (GGML_SYCL_DEVICE_ARCH)
    target_compile_options(ggml-sycl PRIVATE -Xsycl-target-backend --offload-arch=${GGML_SYCL_DEVICE_ARCH})
    target_link_options(ggml-sycl PRIVATE -Xsycl-target-backend --offload-arch=${GGML_SYCL_DEVICE_ARCH})
endif()