rem custom device mlu/npu from build sh (#52373)

7a94e3a4 · Kim Yann · GitHub · 07c7926f · 7a94e3a4
显示空白变更内容
内联并排

Showing with 19 addition and 321 deletion

paddle/scripts/paddle_build.sh paddle/scripts/paddle_build.sh +19 -321

未找到文件。
--- a/paddle/scripts/paddle_build.sh
+++ b/paddle/scripts/paddle_build.sh
@@ -234,15 +234,11 @@ function cmake_base() {
        -DWITH_LITE=${WITH_LITE:-OFF}
        -DWITH_CNCL=${WITH_CNCL:-OFF}
        -DWITH_XPU=${WITH_XPU:-OFF}
-        -DWITH_MLU=${WITH_MLU:-OFF}
        -DWITH_IPU=${WITH_IPU:-OFF}
        -DLITE_GIT_TAG=release/v2.10
        -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF}
        -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF}
        -DWITH_ARM=${WITH_ARM:-OFF}
-        -DWITH_ASCEND=${WITH_ASCEND:-OFF}
-        -DWITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF}
-        -DWITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF}
        -DWITH_STRIP=${WITH_STRIP:-ON}
        -DON_INFER=${ON_INFER:-OFF}
        -DWITH_HETERPS=${WITH_HETERPS:-OFF}
@@ -291,16 +287,12 @@ EOF
        -DWITH_GLOO=${gloo_flag} \
        -DLITE_GIT_TAG=release/v2.10 \
        -DWITH_XPU=${WITH_XPU:-OFF} \
-        -DWITH_MLU=${WITH_MLU:-OFF} \
        -DWITH_IPU=${WITH_IPU:-OFF} \
        -DWITH_CNCL=${WITH_CNCL:-OFF} \
        -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} \
        -DWITH_LITE=${WITH_LITE:-OFF} \
        -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} \
        -DWITH_ARM=${WITH_ARM:-OFF} \
-        -DWITH_ASCEND=${WITH_ASCEND:-OFF} \
-        -DWITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} \
-        -DWITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} \
        -DWITH_STRIP=${WITH_STRIP:-ON} \
        -DON_INFER=${ON_INFER:-OFF} \
        -DWITH_HETERPS=${WITH_HETERPS:-OFF} \
@@ -1328,12 +1320,8 @@ function card_test() {
    # get the CUDA device count, XPU device count is one
    if [ "${WITH_XPU}" == "ON" ];then
        CUDA_DEVICE_COUNT=1
-    elif [ "${WITH_ASCEND_CL}" == "ON" ];then
-        CUDA_DEVICE_COUNT=1
    elif [ "${WITH_ROCM}" == "ON" ];then
        CUDA_DEVICE_COUNT=$(rocm-smi -i | grep GPU | wc -l)
-    elif [ "${WITH_MLU}" == "ON" ];then
-        CUDA_DEVICE_COUNT=1
    elif [ "${WITH_IPU}" == "ON" ];then
        CUDA_DEVICE_COUNT=1
    else
@@ -2280,274 +2268,6 @@ set -x
    fi
 }
-function parallel_test_base_npu() {
-    # skipping if no NPU related files changed
-    if [ ${SKIP_NPU_TEST:-ON} == "ON" ] ; then
-        fetch_upstream_develop_if_not_exist
-        # get npu py or npu cc file changes
-        git diff --name-only remotes/upstream/$BRANCH
-        npu_cc_changes=$(git diff --name-only remotes/upstream/$BRANCH | grep "op_npu.cc" || true)
-        npu_py_changes=$(git diff --name-only remotes/upstream/$BRANCH | grep "op_npu.py" || true)
-        # get PR name
-        npu_pr_tile=$(curl https://github.com/PaddlePaddle/Paddle/pull/${GIT_PR_ID} 2>/dev/null | grep "<title>" | grep "NPU" || true)
-        if [ -z "${npu_cc_changes}" ] && [ -z "${npu_py_changes}" ] && [ -z "${npu_pr_tile}" ]; then
-            echo "NO NPU operators files changed and no '[NPU]' found in PR title, skip NPU unit tests!"
-            exit 0
-        fi
-    fi
-    mkdir -p ${PADDLE_ROOT}/build
-    cd ${PADDLE_ROOT}/build/python/paddle/fluid/tests/unittests/npu
-    if [ ${WITH_TESTING:-ON} == "ON" ] ; then
-    cat <<EOF
-    ========================================
-    Running unit npu tests ...
-    ========================================
-EOF
-set +x
-        test_cases=$(ctest -N -V) # get all test cases
-        get_quickly_disable_ut||disable_ut_quickly='disable_ut'   # indicate whether the case was in quickly disable list
-        while read -r line; do
-            if [[ "$line" == "" ]]; then
-                continue
-            fi
-            read testcase <<< $(echo "$line"|grep -oEi "\w+$")
-            if [[ "$single_card_tests" == "" ]]; then
-                single_card_tests="^$testcase$"
-            else
-                single_card_tests="$single_card_tests|^$testcase$"
-            fi
-        done <<< "$test_cases";
-        ut_actual_total_startTime_s=`date +%s`
-        card_test "$single_card_tests" 1 # run cases 1 job each time with single GPU
-        collect_failed_tests
-        # add unit test retry for NPU
-        rm -f $tmp_dir/*
-        exec_times=0
-        retry_unittests_record=''
-        retry_time=4
-        exec_time_array=('first' 'second' 'third' 'fourth')
-        parallel_failed_tests_exec_retry_threshold=120
-        exec_retry_threshold=30
-        is_retry_execuate=0
-        rerun_ut_startTime_s=`date +%s`
-        if [ -n "$failed_test_lists" ];then
-            if [ ${TIMEOUT_DEBUG_HELP:-OFF} == "ON" ];then
-                bash $PADDLE_ROOT/tools/timeout_debug_help.sh "$failed_test_lists"    # cat logs for tiemout uts which killed by ctest
-            fi
-            need_retry_ut_str=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-            need_retry_ut_arr=(${need_retry_ut_str})
-            need_retry_ut_count=${#need_retry_ut_arr[@]}
-            retry_unittests=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-            while ( [ $exec_times -lt $retry_time ] )
-                do
-                    if [[ "${exec_times}" == "0" ]] ;then
-                        if [ $need_retry_ut_count -lt $parallel_failed_tests_exec_retry_threshold ];then
-                            is_retry_execuate=0
-                        else
-                            is_retry_execuate=1
-                        fi
-                    elif [[ "${exec_times}" == "1" ]] ;then
-                        need_retry_ut_str=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-                        need_retry_ut_arr=(${need_retry_ut_str})
-                        need_retry_ut_count=${#need_retry_ut_arr[@]}
-                        if [ $need_retry_ut_count -lt $exec_retry_threshold ];then
-                            is_retry_execuate=0
-                        else
-                            is_retry_execuate=1
-                        fi
-                    fi
-                    if [[ "$is_retry_execuate" == "0" ]];then
-                        set +e
-                        retry_unittests_record="$retry_unittests_record$failed_test_lists"
-                        failed_test_lists_ult=`echo "${failed_test_lists}" |grep -Po '[^ ].*$'`
-                        set -e
-                        if [[ "${exec_times}" == "1" ]] || [[ "${exec_times}" == "3" ]];then
-                            if [[ "${failed_test_lists}" == "" ]];then
-                                break
-                            else
-                                retry_unittests=$( echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-                            fi
-                        fi
-                        echo "========================================="
-                        echo "This is the ${exec_time_array[$exec_times]} time to re-run"
-                        echo "========================================="
-                        echo "The following unittest will be re-run:"
-                        echo "${retry_unittests}"
-                        for line in ${retry_unittests[@]} ;
-                            do
-                                tmp_one_tmp="$( echo $single_card_tests | grep -oEi $line )"
-                                if [[ "$tmp_one_tmp" != ""  ]]; then
-                                    if [[ "$one_card_retry" == "" ]]; then
-                                        one_card_retry="^$line$"
-                                    else
-                                        one_card_retry="$one_card_retry|^$line$"
-                                    fi
-                                fi
-                            done
-                        if [[ "$one_card_retry" != "" ]]; then
-                            card_test "$one_card_retry" 1 # run cases 1 job each time with single GPU
-                        fi
-                        exec_times=$[$exec_times+1]
-                        failed_test_lists=''
-                        collect_failed_tests
-                        rm -f $tmp_dir/*
-                        one_card_retry=''
-                    else
-                        break
-                    fi
-                done
-        fi
-        rerun_ut_endTime_s=`date +%s`
-        echo "ipipe_log_param_Rerun_TestCases_Total_Time: $[ $rerun_ut_endTime_s - $rerun_ut_startTime_s ]s" >> ${PADDLE_ROOT}/build/build_summary.txt
-        ut_actual_total_endTime_s=`date +%s`
-        echo "ipipe_log_param_actual_TestCases_Total_Time: $[ $ut_actual_total_endTime_s - $ut_actual_total_startTime_s ]s" >> ${PADDLE_ROOT}/build/build_summary.txt
-        if [[ "$EXIT_CODE" != "0" ]]; then
-            show_ut_retry_result
-        fi
-set -ex
-    fi
-}
-function parallel_test_base_mlu() {
-    mkdir -p ${PADDLE_ROOT}/build
-    cd ${PADDLE_ROOT}/build/python/paddle/fluid/tests/unittests/mlu
-    if [ ${WITH_TESTING:-ON} == "ON" ] ; then
-    cat <<EOF
-    ========================================
-    Running unit mlu tests ...
-    ========================================
-EOF
-set +x
-        test_cases=$(ctest -N -V) # get all test cases
-        mlu_card_num=$(cnmon info -t | grep Card | wc -l)
-        if [[ $mlu_card_num == 1 ]]; then
-            get_quickly_disable_ut||disable_ut_quickly='disable_ut'   # indicate whether the case was in quickly disable list
-        else
-            disable_ut_quickly='disable_ut'
-        fi
-        while read -r line; do
-            if [[ "$line" == "" ]]; then
-                continue
-            fi
-            read testcase <<< $(echo "$line"|grep -oEi "\w+$")
-            if [[ "$single_card_tests" == "" ]]; then
-                single_card_tests="^$testcase$"
-            else
-                single_card_tests="$single_card_tests|^$testcase$"
-            fi
-        done <<< "$test_cases";
-        ut_actual_total_startTime_s=`date +%s`
-        card_test "$single_card_tests" 1 # run cases 1 job each time with single MLU
-        collect_failed_tests
-        # add unit test retry for MLU
-        rm -f $tmp_dir/*
-        exec_times=0
-        retry_unittests_record=''
-        retry_time=4
-        exec_time_array=('first' 'second' 'third' 'fourth')
-        parallel_failed_tests_exec_retry_threshold=120
-        exec_retry_threshold=30
-        is_retry_execuate=0
-        rerun_ut_startTime_s=`date +%s`
-        if [ -n "$failed_test_lists" ];then
-            if [ ${TIMEOUT_DEBUG_HELP:-OFF} == "ON" ];then
-                bash $PADDLE_ROOT/tools/timeout_debug_help.sh "$failed_test_lists"    # cat logs for tiemout uts which killed by ctest
-            fi
-            need_retry_ut_str=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-            need_retry_ut_arr=(${need_retry_ut_str})
-            need_retry_ut_count=${#need_retry_ut_arr[@]}
-            retry_unittests=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-            while ( [ $exec_times -lt $retry_time ] )
-                do
-                    if [[ "${exec_times}" == "0" ]] ;then
-                        if [ $need_retry_ut_count -lt $parallel_failed_tests_exec_retry_threshold ];then
-                            is_retry_execuate=0
-                        else
-                            is_retry_execuate=1
-                        fi
-                    elif [[ "${exec_times}" == "1" ]] ;then
-                        need_retry_ut_str=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-                        need_retry_ut_arr=(${need_retry_ut_str})
-                        need_retry_ut_count=${#need_retry_ut_arr[@]}
-                        if [ $need_retry_ut_count -lt $exec_retry_threshold ];then
-                            is_retry_execuate=0
-                        else
-                            is_retry_execuate=1
-                        fi
-                    fi
-                    if [[ "$is_retry_execuate" == "0" ]];then
-                        set +e
-                        retry_unittests_record="$retry_unittests_record$failed_test_lists"
-                        failed_test_lists_ult=`echo "${failed_test_lists}" |grep -Po '[^ ].*$'`
-                        set -e
-                        if [[ "${exec_times}" == "1" ]] || [[ "${exec_times}" == "3" ]];then
-                            if [[ "${failed_test_lists}" == "" ]];then
-                                break
-                            else
-                                retry_unittests=$(echo "$failed_test_lists" | grep -oEi "\-.+\(.+\)" | sed 's/(.\+)//' | sed 's/- //' )
-                            fi
-                        fi
-                        echo "========================================="
-                        echo "This is the ${exec_time_array[$exec_times]} time to re-run"
-                        echo "========================================="
-                        echo "The following unittest will be re-run:"
-                        echo "${retry_unittests}"
-                        for line in ${retry_unittests[@]} ;
-                            do
-                                tmp_one_tmp="$( echo $single_card_tests | grep -oEi $line )"
-                                if [[ "$tmp_one_tmp" != ""  ]]; then
-                                    if [[ "$one_card_retry" == "" ]]; then
-                                        one_card_retry="^$line$"
-                                    else
-                                        one_card_retry="$one_card_retry|^$line$"
-                                    fi
-                                fi
-                            done
-                        if [[ "$one_card_retry" != "" ]]; then
-                            card_test "$one_card_retry" 1 # run cases 1 job each time with single GPU
-                        fi
-                        exec_times=$[$exec_times+1]
-                        failed_test_lists=''
-                        collect_failed_tests
-                        rm -f $tmp_dir/*
-                        one_card_retry=''
-                    else
-                        break
-                    fi
-                done
-        fi
-        rerun_ut_endTime_s=`date +%s`
-        echo "ipipe_log_param_Rerun_TestCases_Total_Time: $[ $rerun_ut_endTime_s - $rerun_ut_startTime_s ]s" >> ${PADDLE_ROOT}/build/build_summary.txt
-        ut_actual_total_endTime_s=`date +%s`
-        echo "ipipe_log_param_actual_TestCases_Total_Time: $[ $ut_actual_total_endTime_s - $ut_actual_total_startTime_s ]s" >> ${PADDLE_ROOT}/build/build_summary.txt
-        if [[ "$EXIT_CODE" != "0" ]]; then
-            show_ut_retry_result
-        fi
-set -ex
-    fi
-}
 function parallel_test_base_gpu_test() {
    if [ ${WITH_TESTING:-ON} == "ON" ] ; then
    cat <<EOF
@@ -2910,10 +2630,6 @@ function parallel_test() {
        parallel_test_base_gpu_test
    elif [ "$WITH_XPU" == "ON" ];then
        parallel_test_base_xpu
-    elif [ "$WITH_ASCEND_CL" == "ON" ];then
-        parallel_test_base_npu
-    elif [ "$WITH_MLU" == "ON" ];then
-        parallel_test_base_mlu
    elif [ "$WITH_IPU" == "ON" ];then
        parallel_test_base_ipu
    else
@@ -3169,13 +2885,13 @@ EOF
    echo "if you use setup.py to compile,please export envs as following in /paddle ..."
    cat << EOF
    ========================================
-    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} WITH_GPU=${WITH_GPU:-OFF} WITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} WITH_TENSORRT=${WITH_TENSORRT:-ON} WITH_ROCM=${WITH_ROCM:-OFF} WITH_CINN=${WITH_CINN:-OFF} WITH_DISTRIBUTE=${distibuted_flag} WITH_MKL=${WITH_MKL:-ON} WITH_AVX=${WITH_AVX:-OFF} CUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} NEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} NEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} NEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} WITH_PYTHON=${WITH_PYTHON:-ON} CUDNN_ROOT=/usr/ WITH_TESTING=${WITH_TESTING:-ON} WITH_COVERAGE=${WITH_COVERAGE:-OFF} WITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} CMAKE_MODULE_PATH=/opt/rocm/hip/cmake CMAKE_EXPORT_COMPILE_COMMANDS=ON WITH_CONTRIB=${WITH_CONTRIB:-ON} WITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} INFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} PY_VERSION=${PY_VERSION:-3.7} CMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} WITH_PSCORE=${pscore_flag} WITH_PSLIB=${pslib_flag} WITH_GLOO=${gloo_flag} LITE_GIT_TAG=release/v2.10 WITH_XPU=${WITH_XPU:-OFF} WITH_MLU=${WITH_MLU:-OFF} WITH_IPU=${WITH_IPU:-OFF} WITH_CNCL=${WITH_CNCL:-OFF} XPU_SDK_ROOT=${XPU_SDK_ROOT:-""} WITH_LITE=${WITH_LITE:-OFF} WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} WITH_ARM=${WITH_ARM:-OFF} WITH_ASCEND=${WITH_ASCEND:-OFF} WITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} WITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} WITH_STRIP=${WITH_STRIP:-ON} ON_INFER=${ON_INFER:-OFF} WITH_HETERPS=${WITH_HETERPS:-OFF} WITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} WITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} CUDA_ARCH_BIN=${CUDA_ARCH_BIN} WITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} WITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} WITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} WITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
+    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} WITH_GPU=${WITH_GPU:-OFF} WITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} WITH_TENSORRT=${WITH_TENSORRT:-ON} WITH_ROCM=${WITH_ROCM:-OFF} WITH_CINN=${WITH_CINN:-OFF} WITH_DISTRIBUTE=${distibuted_flag} WITH_MKL=${WITH_MKL:-ON} WITH_AVX=${WITH_AVX:-OFF} CUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} NEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} NEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} NEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} WITH_PYTHON=${WITH_PYTHON:-ON} CUDNN_ROOT=/usr/ WITH_TESTING=${WITH_TESTING:-ON} WITH_COVERAGE=${WITH_COVERAGE:-OFF} WITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} CMAKE_MODULE_PATH=/opt/rocm/hip/cmake CMAKE_EXPORT_COMPILE_COMMANDS=ON WITH_CONTRIB=${WITH_CONTRIB:-ON} WITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} INFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} PY_VERSION=${PY_VERSION:-3.7} CMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} WITH_PSCORE=${pscore_flag} WITH_PSLIB=${pslib_flag} WITH_GLOO=${gloo_flag} LITE_GIT_TAG=release/v2.10 WITH_XPU=${WITH_XPU:-OFF} WITH_IPU=${WITH_IPU:-OFF} WITH_CNCL=${WITH_CNCL:-OFF} XPU_SDK_ROOT=${XPU_SDK_ROOT:-""} WITH_LITE=${WITH_LITE:-OFF} WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} WITH_ARM=${WITH_ARM:-OFF} WITH_STRIP=${WITH_STRIP:-ON} ON_INFER=${ON_INFER:-OFF} WITH_HETERPS=${WITH_HETERPS:-OFF} WITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} WITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} CUDA_ARCH_BIN=${CUDA_ARCH_BIN} WITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} WITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} WITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} WITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
    ========================================
 EOF
    echo "if you use cmake to compile,please Configuring cmake in /paddle/build ..."
    cat <<EOF
    ========================================
-    cmake .. -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} -DWITH_GPU=${WITH_GPU:-OFF} -DWITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} -DWITH_TENSORRT=${WITH_TENSORRT:-ON} -DWITH_ROCM=${WITH_ROCM:-OFF} -DWITH_CINN=${WITH_CINN:-OFF} -DWITH_DISTRIBUTE=${distibuted_flag} -DWITH_MKL=${WITH_MKL:-ON} -DWITH_AVX=${WITH_AVX:-OFF} -DCUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} -DNEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} -DNEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} -DNEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} -DWITH_PYTHON=${WITH_PYTHON:-ON} -DCUDNN_ROOT=/usr/ -DWITH_TESTING=${WITH_TESTING:-ON} -DWITH_COVERAGE=${WITH_COVERAGE:-OFF} -DWITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} -DCMAKE_MODULE_PATH=/opt/rocm/hip/cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWITH_CONTRIB=${WITH_CONTRIB:-ON} -DWITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} -DINFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} -DPY_VERSION=${PY_VERSION:-3.7} -DCMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} -DWITH_PSCORE=${pscore_flag} -DWITH_PSLIB=${pslib_flag} -DWITH_GLOO=${gloo_flag} -DLITE_GIT_TAG=release/v2.10 -DWITH_XPU=${WITH_XPU:-OFF} -DWITH_MLU=${WITH_MLU:-OFF} -DWITH_IPU=${WITH_IPU:-OFF} -DWITH_CNCL=${WITH_CNCL:-OFF} -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} -DWITH_LITE=${WITH_LITE:-OFF} -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} -DWITH_ARM=${WITH_ARM:-OFF} -DWITH_ASCEND=${WITH_ASCEND:-OFF} -DWITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} -DWITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} -DWITH_STRIP=${WITH_STRIP:-ON} -DON_INFER=${ON_INFER:-OFF} -DWITH_HETERPS=${WITH_HETERPS:-OFF} -DWITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} -DWITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} -DCUDA_ARCH_BIN=${CUDA_ARCH_BIN} -DWITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} -DWITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} -DWITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
+    cmake .. -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} -DWITH_GPU=${WITH_GPU:-OFF} -DWITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} -DWITH_TENSORRT=${WITH_TENSORRT:-ON} -DWITH_ROCM=${WITH_ROCM:-OFF} -DWITH_CINN=${WITH_CINN:-OFF} -DWITH_DISTRIBUTE=${distibuted_flag} -DWITH_MKL=${WITH_MKL:-ON} -DWITH_AVX=${WITH_AVX:-OFF} -DCUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} -DNEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} -DNEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} -DNEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} -DWITH_PYTHON=${WITH_PYTHON:-ON} -DCUDNN_ROOT=/usr/ -DWITH_TESTING=${WITH_TESTING:-ON} -DWITH_COVERAGE=${WITH_COVERAGE:-OFF} -DWITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} -DCMAKE_MODULE_PATH=/opt/rocm/hip/cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWITH_CONTRIB=${WITH_CONTRIB:-ON} -DWITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} -DINFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} -DPY_VERSION=${PY_VERSION:-3.7} -DCMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} -DWITH_PSCORE=${pscore_flag} -DWITH_PSLIB=${pslib_flag} -DWITH_GLOO=${gloo_flag} -DLITE_GIT_TAG=release/v2.10 -DWITH_XPU=${WITH_XPU:-OFF} -DWITH_IPU=${WITH_IPU:-OFF} -DWITH_CNCL=${WITH_CNCL:-OFF} -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} -DWITH_LITE=${WITH_LITE:-OFF} -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} -DWITH_ARM=${WITH_ARM:-OFF} -DWITH_STRIP=${WITH_STRIP:-ON} -DON_INFER=${ON_INFER:-OFF} -DWITH_HETERPS=${WITH_HETERPS:-OFF} -DWITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} -DWITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} -DCUDA_ARCH_BIN=${CUDA_ARCH_BIN} -DWITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} -DWITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} -DWITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
    ========================================
 EOF
    # reset ccache zero stats for collect PR's actual hit rate
@@ -3652,13 +3368,13 @@ function run_setup(){
    echo "if you use setup.py to compile,please export envs as following in /paddle ..."
    cat << EOF
    ========================================
-    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} WITH_GPU=${WITH_GPU:-OFF} WITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} WITH_TENSORRT=${WITH_TENSORRT:-ON} WITH_ROCM=${WITH_ROCM:-OFF} WITH_CINN=${WITH_CINN:-OFF} WITH_DISTRIBUTE=${distibuted_flag} WITH_MKL=${WITH_MKL:-ON} WITH_AVX=${WITH_AVX:-OFF} CUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} NEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} NEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} NEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} WITH_PYTHON=${WITH_PYTHON:-ON} CUDNN_ROOT=/usr/ WITH_TESTING=${WITH_TESTING:-ON} WITH_COVERAGE=${WITH_COVERAGE:-OFF} WITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} CMAKE_MODULE_PATH=/opt/rocm/hip/cmake CMAKE_EXPORT_COMPILE_COMMANDS=ON WITH_CONTRIB=${WITH_CONTRIB:-ON} WITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} INFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} PY_VERSION=${PY_VERSION:-3.7} CMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} WITH_PSCORE=${pscore_flag} WITH_PSLIB=${pslib_flag} WITH_GLOO=${gloo_flag} LITE_GIT_TAG=release/v2.10 WITH_XPU=${WITH_XPU:-OFF} WITH_MLU=${WITH_MLU:-OFF} WITH_IPU=${WITH_IPU:-OFF} WITH_CNCL=${WITH_CNCL:-OFF} XPU_SDK_ROOT=${XPU_SDK_ROOT:-""} WITH_LITE=${WITH_LITE:-OFF} WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} WITH_ARM=${WITH_ARM:-OFF} WITH_ASCEND=${WITH_ASCEND:-OFF} WITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} WITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} WITH_STRIP=${WITH_STRIP:-ON} ON_INFER=${ON_INFER:-OFF} WITH_HETERPS=${WITH_HETERPS:-OFF} WITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} WITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} CUDA_ARCH_BIN=${CUDA_ARCH_BIN} WITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} WITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} WITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} WITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
+    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} WITH_GPU=${WITH_GPU:-OFF} WITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} WITH_TENSORRT=${WITH_TENSORRT:-ON} WITH_ROCM=${WITH_ROCM:-OFF} WITH_CINN=${WITH_CINN:-OFF} WITH_DISTRIBUTE=${distibuted_flag} WITH_MKL=${WITH_MKL:-ON} WITH_AVX=${WITH_AVX:-OFF} CUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} NEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} NEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} NEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} WITH_PYTHON=${WITH_PYTHON:-ON} CUDNN_ROOT=/usr/ WITH_TESTING=${WITH_TESTING:-ON} WITH_COVERAGE=${WITH_COVERAGE:-OFF} WITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} CMAKE_MODULE_PATH=/opt/rocm/hip/cmake CMAKE_EXPORT_COMPILE_COMMANDS=ON WITH_CONTRIB=${WITH_CONTRIB:-ON} WITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} INFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} PY_VERSION=${PY_VERSION:-3.7} CMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} WITH_PSCORE=${pscore_flag} WITH_PSLIB=${pslib_flag} WITH_GLOO=${gloo_flag} LITE_GIT_TAG=release/v2.10 WITH_XPU=${WITH_XPU:-OFF} WITH_IPU=${WITH_IPU:-OFF} WITH_CNCL=${WITH_CNCL:-OFF} XPU_SDK_ROOT=${XPU_SDK_ROOT:-""} WITH_LITE=${WITH_LITE:-OFF} WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} WITH_ARM=${WITH_ARM:-OFF} WITH_STRIP=${WITH_STRIP:-ON} ON_INFER=${ON_INFER:-OFF} WITH_HETERPS=${WITH_HETERPS:-OFF} WITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} WITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} CUDA_ARCH_BIN=${CUDA_ARCH_BIN} WITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} WITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} WITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} WITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
    ========================================
 EOF
    echo "if you use cmake to compile,please Configuring cmake in /paddle/build ..."
    cat <<EOF
    ========================================
-    cmake .. -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} -DWITH_GPU=${WITH_GPU:-OFF} -DWITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} -DWITH_TENSORRT=${WITH_TENSORRT:-ON} -DWITH_ROCM=${WITH_ROCM:-OFF} -DWITH_CINN=${WITH_CINN:-OFF} -DWITH_DISTRIBUTE=${distibuted_flag} -DWITH_MKL=${WITH_MKL:-ON} -DWITH_AVX=${WITH_AVX:-OFF} -DCUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} -DNEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} -DNEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} -DNEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} -DWITH_PYTHON=${WITH_PYTHON:-ON} -DCUDNN_ROOT=/usr/ -DWITH_TESTING=${WITH_TESTING:-ON} -DWITH_COVERAGE=${WITH_COVERAGE:-OFF} -DWITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} -DCMAKE_MODULE_PATH=/opt/rocm/hip/cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWITH_CONTRIB=${WITH_CONTRIB:-ON} -DWITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} -DINFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} -DPY_VERSION=${PY_VERSION:-3.7} -DCMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} -DWITH_PSCORE=${pscore_flag} -DWITH_PSLIB=${pslib_flag} -DWITH_GLOO=${gloo_flag} -DLITE_GIT_TAG=release/v2.10 -DWITH_XPU=${WITH_XPU:-OFF} -DWITH_MLU=${WITH_MLU:-OFF} -DWITH_IPU=${WITH_IPU:-OFF} -DWITH_CNCL=${WITH_CNCL:-OFF} -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} -DWITH_LITE=${WITH_LITE:-OFF} -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} -DWITH_ARM=${WITH_ARM:-OFF} -DWITH_ASCEND=${WITH_ASCEND:-OFF} -DWITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} -DWITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} -DWITH_STRIP=${WITH_STRIP:-ON} -DON_INFER=${ON_INFER:-OFF} -DWITH_HETERPS=${WITH_HETERPS:-OFF} -DWITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} -DWITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} -DCUDA_ARCH_BIN=${CUDA_ARCH_BIN} -DWITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} -DWITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} -DWITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
+    cmake .. -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} -DWITH_GPU=${WITH_GPU:-OFF} -DWITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} -DWITH_TENSORRT=${WITH_TENSORRT:-ON} -DWITH_ROCM=${WITH_ROCM:-OFF} -DWITH_CINN=${WITH_CINN:-OFF} -DWITH_DISTRIBUTE=${distibuted_flag} -DWITH_MKL=${WITH_MKL:-ON} -DWITH_AVX=${WITH_AVX:-OFF} -DCUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} -DNEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} -DNEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} -DNEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} -DWITH_PYTHON=${WITH_PYTHON:-ON} -DCUDNN_ROOT=/usr/ -DWITH_TESTING=${WITH_TESTING:-ON} -DWITH_COVERAGE=${WITH_COVERAGE:-OFF} -DWITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} -DCMAKE_MODULE_PATH=/opt/rocm/hip/cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWITH_CONTRIB=${WITH_CONTRIB:-ON} -DWITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} -DINFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} -DPY_VERSION=${PY_VERSION:-3.7} -DCMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} -DWITH_PSCORE=${pscore_flag} -DWITH_PSLIB=${pslib_flag} -DWITH_GLOO=${gloo_flag} -DLITE_GIT_TAG=release/v2.10 -DWITH_XPU=${WITH_XPU:-OFF} -DWITH_IPU=${WITH_IPU:-OFF} -DWITH_CNCL=${WITH_CNCL:-OFF} -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} -DWITH_LITE=${WITH_LITE:-OFF} -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} -DWITH_ARM=${WITH_ARM:-OFF} -DWITH_STRIP=${WITH_STRIP:-ON} -DON_INFER=${ON_INFER:-OFF} -DWITH_HETERPS=${WITH_HETERPS:-OFF} -DWITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} -DWITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} -DCUDA_ARCH_BIN=${CUDA_ARCH_BIN} -DWITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} -DWITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} -DWITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
    ========================================
 EOF
    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release}
@@ -3692,16 +3408,12 @@ EOF
    export WITH_GLOO=${gloo_flag}
    export LITE_GIT_TAG=release/v2.10
    export WITH_XPU=${WITH_XPU:-OFF}
-    export WITH_MLU=${WITH_MLU:-OFF}
    export WITH_IPU=${WITH_IPU:-OFF}
    export WITH_CNCL=${WITH_CNCL:-OFF}
    export XPU_SDK_ROOT=${XPU_SDK_ROOT:-""}
    export WITH_LITE=${WITH_LITE:-OFF}
    export WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF}
    export WITH_ARM=${WITH_ARM:-OFF}
-    export WITH_ASCEND=${WITH_ASCEND:-OFF}
-    export WITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF}
-    export WITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF}
    export WITH_STRIP=${WITH_STRIP:-ON}
    export ON_INFER=${ON_INFER:-OFF}
    export WITH_HETERPS=${WITH_HETERPS:-OFF}
@@ -3901,13 +3613,13 @@ function run_setup_mac(){
    echo "if you use setup.py to compile,please export envs as following in /paddle ..."
    cat << EOF
    ========================================
-    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} WITH_GPU=${WITH_GPU:-OFF} WITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} WITH_TENSORRT=${WITH_TENSORRT:-ON} WITH_ROCM=${WITH_ROCM:-OFF} WITH_CINN=${WITH_CINN:-OFF} WITH_DISTRIBUTE=${distibuted_flag} WITH_MKL=${WITH_MKL:-ON} WITH_AVX=${WITH_AVX:-OFF} CUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} NEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} NEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} NEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} WITH_PYTHON=${WITH_PYTHON:-ON} CUDNN_ROOT=/usr/ WITH_TESTING=${WITH_TESTING:-ON} WITH_COVERAGE=${WITH_COVERAGE:-OFF} WITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} CMAKE_MODULE_PATH=/opt/rocm/hip/cmake CMAKE_EXPORT_COMPILE_COMMANDS=ON WITH_CONTRIB=${WITH_CONTRIB:-ON} WITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} INFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} PY_VERSION=${PY_VERSION:-3.7} CMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} WITH_PSCORE=${pscore_flag} WITH_PSLIB=${pslib_flag} WITH_GLOO=${gloo_flag} LITE_GIT_TAG=release/v2.10 WITH_XPU=${WITH_XPU:-OFF} WITH_MLU=${WITH_MLU:-OFF} WITH_IPU=${WITH_IPU:-OFF} WITH_CNCL=${WITH_CNCL:-OFF} XPU_SDK_ROOT=${XPU_SDK_ROOT:-""} WITH_LITE=${WITH_LITE:-OFF} WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} WITH_ARM=${WITH_ARM:-OFF} WITH_ASCEND=${WITH_ASCEND:-OFF} WITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} WITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} WITH_STRIP=${WITH_STRIP:-ON} ON_INFER=${ON_INFER:-OFF} WITH_HETERPS=${WITH_HETERPS:-OFF} WITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} WITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} CUDA_ARCH_BIN=${CUDA_ARCH_BIN} WITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} WITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} WITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} WITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
+    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} WITH_GPU=${WITH_GPU:-OFF} WITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} WITH_TENSORRT=${WITH_TENSORRT:-ON} WITH_ROCM=${WITH_ROCM:-OFF} WITH_CINN=${WITH_CINN:-OFF} WITH_DISTRIBUTE=${distibuted_flag} WITH_MKL=${WITH_MKL:-ON} WITH_AVX=${WITH_AVX:-OFF} CUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} NEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} NEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} NEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} WITH_PYTHON=${WITH_PYTHON:-ON} CUDNN_ROOT=/usr/ WITH_TESTING=${WITH_TESTING:-ON} WITH_COVERAGE=${WITH_COVERAGE:-OFF} WITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} CMAKE_MODULE_PATH=/opt/rocm/hip/cmake CMAKE_EXPORT_COMPILE_COMMANDS=ON WITH_CONTRIB=${WITH_CONTRIB:-ON} WITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} INFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} PY_VERSION=${PY_VERSION:-3.7} CMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} WITH_PSCORE=${pscore_flag} WITH_PSLIB=${pslib_flag} WITH_GLOO=${gloo_flag} LITE_GIT_TAG=release/v2.10 WITH_XPU=${WITH_XPU:-OFF} WITH_IPU=${WITH_IPU:-OFF} WITH_CNCL=${WITH_CNCL:-OFF} XPU_SDK_ROOT=${XPU_SDK_ROOT:-""} WITH_LITE=${WITH_LITE:-OFF} WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} WITH_ARM=${WITH_ARM:-OFF} WITH_STRIP=${WITH_STRIP:-ON} ON_INFER=${ON_INFER:-OFF} WITH_HETERPS=${WITH_HETERPS:-OFF} WITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} WITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} CUDA_ARCH_BIN=${CUDA_ARCH_BIN} WITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} WITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} WITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} WITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
    ========================================
 EOF
    echo "if you use cmake to compile,please Configuring cmake in /paddle/build ..."
    cat <<EOF
    ========================================
-    cmake .. -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} -DWITH_GPU=${WITH_GPU:-OFF} -DWITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} -DWITH_TENSORRT=${WITH_TENSORRT:-ON} -DWITH_ROCM=${WITH_ROCM:-OFF} -DWITH_CINN=${WITH_CINN:-OFF} -DWITH_DISTRIBUTE=${distibuted_flag} -DWITH_MKL=${WITH_MKL:-ON} -DWITH_AVX=${WITH_AVX:-OFF} -DCUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} -DNEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} -DNEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} -DNEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} -DWITH_PYTHON=${WITH_PYTHON:-ON} -DCUDNN_ROOT=/usr/ -DWITH_TESTING=${WITH_TESTING:-ON} -DWITH_COVERAGE=${WITH_COVERAGE:-OFF} -DWITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} -DCMAKE_MODULE_PATH=/opt/rocm/hip/cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWITH_CONTRIB=${WITH_CONTRIB:-ON} -DWITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} -DINFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} -DPY_VERSION=${PY_VERSION:-3.7} -DCMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} -DWITH_PSCORE=${pscore_flag} -DWITH_PSLIB=${pslib_flag} -DWITH_GLOO=${gloo_flag} -DLITE_GIT_TAG=release/v2.10 -DWITH_XPU=${WITH_XPU:-OFF} -DWITH_MLU=${WITH_MLU:-OFF} -DWITH_IPU=${WITH_IPU:-OFF} -DWITH_CNCL=${WITH_CNCL:-OFF} -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} -DWITH_LITE=${WITH_LITE:-OFF} -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} -DWITH_ARM=${WITH_ARM:-OFF} -DWITH_ASCEND=${WITH_ASCEND:-OFF} -DWITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF} -DWITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF} -DWITH_STRIP=${WITH_STRIP:-ON} -DON_INFER=${ON_INFER:-OFF} -DWITH_HETERPS=${WITH_HETERPS:-OFF} -DWITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} -DWITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} -DCUDA_ARCH_BIN=${CUDA_ARCH_BIN} -DWITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} -DWITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} -DWITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
+    cmake .. -DCMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release} -DWITH_GPU=${WITH_GPU:-OFF} -DWITH_CUDNN_DSO=${WITH_CUDNN_DSO:-OFF} -DWITH_TENSORRT=${WITH_TENSORRT:-ON} -DWITH_ROCM=${WITH_ROCM:-OFF} -DWITH_CINN=${WITH_CINN:-OFF} -DWITH_DISTRIBUTE=${distibuted_flag} -DWITH_MKL=${WITH_MKL:-ON} -DWITH_AVX=${WITH_AVX:-OFF} -DCUDA_ARCH_NAME=${CUDA_ARCH_NAME:-All} -DNEW_RELEASE_PYPI=${NEW_RELEASE_PYPI:-OFF} -DNEW_RELEASE_ALL=${NEW_RELEASE_ALL:-OFF} -DNEW_RELEASE_JIT=${NEW_RELEASE_JIT:-OFF} -DWITH_PYTHON=${WITH_PYTHON:-ON} -DCUDNN_ROOT=/usr/ -DWITH_TESTING=${WITH_TESTING:-ON} -DWITH_COVERAGE=${WITH_COVERAGE:-OFF} -DWITH_INCREMENTAL_COVERAGE=${WITH_INCREMENTAL_COVERAGE:-OFF} -DCMAKE_MODULE_PATH=/opt/rocm/hip/cmake -DCMAKE_EXPORT_COMPILE_COMMANDS=ON -DWITH_CONTRIB=${WITH_CONTRIB:-ON} -DWITH_INFERENCE_API_TEST=${WITH_INFERENCE_API_TEST:-ON} -DINFERENCE_DEMO_INSTALL_DIR=${INFERENCE_DEMO_INSTALL_DIR} -DPY_VERSION=${PY_VERSION:-3.7} -DCMAKE_INSTALL_PREFIX=${INSTALL_PREFIX:-/paddle/build} -DWITH_PSCORE=${pscore_flag} -DWITH_PSLIB=${pslib_flag} -DWITH_GLOO=${gloo_flag} -DLITE_GIT_TAG=release/v2.10 -DWITH_XPU=${WITH_XPU:-OFF} -DWITH_IPU=${WITH_IPU:-OFF} -DWITH_CNCL=${WITH_CNCL:-OFF} -DXPU_SDK_ROOT=${XPU_SDK_ROOT:-""} -DWITH_LITE=${WITH_LITE:-OFF} -DWITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF} -DWITH_ARM=${WITH_ARM:-OFF} -DWITH_STRIP=${WITH_STRIP:-ON} -DON_INFER=${ON_INFER:-OFF} -DWITH_HETERPS=${WITH_HETERPS:-OFF} -DWITH_GPU_GRAPH=${WITH_GPU_GRAPH:-OFF} -DWITH_FLUID_ONLY=${WITH_FLUID_ONLY:-OFF} -DCUDA_ARCH_BIN=${CUDA_ARCH_BIN} -DWITH_RECORD_BUILDTIME=${WITH_RECORD_BUILDTIME:-OFF} -DWITH_UNITY_BUILD=${WITH_UNITY_BUILD:-OFF} -DWITH_ONNXRUNTIME=${WITH_ONNXRUNTIME:-OFF} -DWITH_CUDNN_FRONTEND=${WITH_CUDNN_FRONTEND:-OFF}
    ========================================
 EOF
    export CMAKE_BUILD_TYPE=${CMAKE_BUILD_TYPE:-Release}
@@ -3941,16 +3653,12 @@ EOF
    export WITH_GLOO=${gloo_flag}
    export LITE_GIT_TAG=release/v2.10
    export WITH_XPU=${WITH_XPU:-OFF}
-    export WITH_MLU=${WITH_MLU:-OFF}
    export WITH_IPU=${WITH_IPU:-OFF}
    export WITH_CNCL=${WITH_CNCL:-OFF}
    export XPU_SDK_ROOT=${XPU_SDK_ROOT:-""}
    export WITH_LITE=${WITH_LITE:-OFF}
    export WITH_XPU_BKCL=${WITH_XPU_BKCL:-OFF}
    export WITH_ARM=${WITH_ARM:-OFF}
-    export WITH_ASCEND=${WITH_ASCEND:-OFF}
-    export WITH_ASCEND_CL=${WITH_ASCEND_CL:-OFF}
-    export WITH_ASCEND_INT64=${WITH_ASCEND_INT64:-OFF}
    export WITH_STRIP=${WITH_STRIP:-ON}
    export ON_INFER=${ON_INFER:-OFF}
    export WITH_HETERPS=${WITH_HETERPS:-OFF}
@@ -4229,16 +3937,6 @@ function main() {
        parallel_test
        check_coverage
        ;;
-      check_npu_coverage)
-        cmake_gen_and_build ${PYTHON_ABI:-""} ${parallel_number}
-        parallel_test
-        check_coverage
-        ;;
-      check_mlu_coverage)
-        cmake_gen_and_build ${PYTHON_ABI:-""} ${parallel_number}
-        parallel_test
-        check_coverage
-        ;;
      check_ipu_coverage)
        cmake_gen_and_build ${PYTHON_ABI:-""} ${parallel_number}
        parallel_test