xnnpack in tensorflow lite in imx8qm

[jimdeway]

HI,

Hardware: IMX8QM-MEK

Vision:LF5.10.72_2.2.0 

I am trying to cmake my program but, I got error.

$. /opt/fsl-imx-xwayland/5.10-hardknott/environment-setup-cortexa72-cortexa53-crypto-poky-linux 

$ cmake ../tensorflow/lite/examples/yolov5/

The terminal displays "Unrecognized XNNPACK_TARGET_PROCESSOR = cortexa72-cortexa53-crypto", but it should be recognizable, right?

terminal：

soc123@jimdeway:~/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build$ cmake ../tensorflow/lite/examples/yolov5/
-- Toolchain file defaulted to '/opt/fsl-imx-xwayland/5.10-hardknott/sysroots/x86_64-pokysdk-linux/usr/share/cmake/OEToolchainConfig.cmake'
-- The C compiler identification is GNU 10.2.0
-- The CXX compiler identification is GNU 10.2.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /opt/fsl-imx-xwayland/5.10-hardknott/sysroots/x86_64-pokysdk-linux/usr/bin/aarch64-poky-linux/aarch64-poky-linux-gcc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /opt/fsl-imx-xwayland/5.10-hardknott/sysroots/x86_64-pokysdk-linux/usr/bin/aarch64-poky-linux/aarch64-poky-linux-g++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Setting build type to Release, for debug builds use'-DCMAKE_BUILD_TYPE=Debug'.
CMake Warning at /home/soc123/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build/abseil-cpp/absl/copts/AbseilConfigureCopts.cmake:30 (message):
  Value of CMAKE_SYSTEM_PROCESSOR (cortexa72-cortexa53-crypto) is unknown and
  cannot be used to set ABSL_RANDOM_RANDEN_COPTS
Call Stack (most recent call first):
  /home/soc123/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build/abseil-cpp/CMake/AbseilHelpers.cmake:18 (include)
  /home/soc123/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build/abseil-cpp/CMakeLists.txt:73 (include)


-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE  
-- Performing Test EIGEN_COMPILER_SUPPORT_CPP11
-- Performing Test EIGEN_COMPILER_SUPPORT_CPP11 - Success
-- Performing Test COMPILER_SUPPORT_std=cpp03
-- Performing Test COMPILER_SUPPORT_std=cpp03 - Success
-- Performing Test standard_math_library_linked_to_automatically
-- Performing Test standard_math_library_linked_to_automatically - Success
-- Standard libraries to link to explicitly: none
-- Performing Test COMPILER_SUPPORT_WERROR
-- Performing Test COMPILER_SUPPORT_WERROR - Success
-- Performing Test COMPILER_SUPPORT_pedantic
-- Performing Test COMPILER_SUPPORT_pedantic - Success
-- Performing Test COMPILER_SUPPORT_Wall
-- Performing Test COMPILER_SUPPORT_Wall - Success
-- Performing Test COMPILER_SUPPORT_Wextra
-- Performing Test COMPILER_SUPPORT_Wextra - Success
-- Performing Test COMPILER_SUPPORT_Wundef
-- Performing Test COMPILER_SUPPORT_Wundef - Success
-- Performing Test COMPILER_SUPPORT_Wcastalign
-- Performing Test COMPILER_SUPPORT_Wcastalign - Success
-- Performing Test COMPILER_SUPPORT_Wcharsubscripts
-- Performing Test COMPILER_SUPPORT_Wcharsubscripts - Success
-- Performing Test COMPILER_SUPPORT_Wnonvirtualdtor
-- Performing Test COMPILER_SUPPORT_Wnonvirtualdtor - Success
-- Performing Test COMPILER_SUPPORT_Wunusedlocaltypedefs
-- Performing Test COMPILER_SUPPORT_Wunusedlocaltypedefs - Success
-- Performing Test COMPILER_SUPPORT_Wpointerarith
-- Performing Test COMPILER_SUPPORT_Wpointerarith - Success
-- Performing Test COMPILER_SUPPORT_Wwritestrings
-- Performing Test COMPILER_SUPPORT_Wwritestrings - Success
-- Performing Test COMPILER_SUPPORT_Wformatsecurity
-- Performing Test COMPILER_SUPPORT_Wformatsecurity - Success
-- Performing Test COMPILER_SUPPORT_Wshorten64to32
-- Performing Test COMPILER_SUPPORT_Wshorten64to32 - Failed
-- Performing Test COMPILER_SUPPORT_Wlogicalop
-- Performing Test COMPILER_SUPPORT_Wlogicalop - Success
-- Performing Test COMPILER_SUPPORT_Wenumconversion
-- Performing Test COMPILER_SUPPORT_Wenumconversion - Failed
-- Performing Test COMPILER_SUPPORT_Wcpp11extensions
-- Performing Test COMPILER_SUPPORT_Wcpp11extensions - Failed
-- Performing Test COMPILER_SUPPORT_Wdoublepromotion
-- Performing Test COMPILER_SUPPORT_Wdoublepromotion - Success
-- Performing Test COMPILER_SUPPORT_Wshadow
-- Performing Test COMPILER_SUPPORT_Wshadow - Success
-- Performing Test COMPILER_SUPPORT_Wnopsabi
-- Performing Test COMPILER_SUPPORT_Wnopsabi - Success
-- Performing Test COMPILER_SUPPORT_Wnovariadicmacros
-- Performing Test COMPILER_SUPPORT_Wnovariadicmacros - Success
-- Performing Test COMPILER_SUPPORT_Wnolonglong
-- Performing Test COMPILER_SUPPORT_Wnolonglong - Success
-- Performing Test COMPILER_SUPPORT_fnochecknew
-- Performing Test COMPILER_SUPPORT_fnochecknew - Success
-- Performing Test COMPILER_SUPPORT_fnocommon
-- Performing Test COMPILER_SUPPORT_fnocommon - Success
-- Performing Test COMPILER_SUPPORT_fstrictaliasing
-- Performing Test COMPILER_SUPPORT_fstrictaliasing - Success
-- Performing Test COMPILER_SUPPORT_wd981
-- Performing Test COMPILER_SUPPORT_wd981 - Failed
-- Performing Test COMPILER_SUPPORT_wd2304
-- Performing Test COMPILER_SUPPORT_wd2304 - Failed
-- Performing Test COMPILER_SUPPORT_STRICTANSI
-- Performing Test COMPILER_SUPPORT_STRICTANSI - Failed
-- Performing Test COMPILER_SUPPORT_Qunusedarguments
-- Performing Test COMPILER_SUPPORT_Qunusedarguments - Failed
-- Performing Test COMPILER_SUPPORT_ansi
-- Performing Test COMPILER_SUPPORT_ansi - Success
-- Performing Test COMPILER_SUPPORT_OPENMP
-- Performing Test COMPILER_SUPPORT_OPENMP - Success
-- Looking for a Fortran compiler
-- Looking for a Fortran compiler - /usr/bin/f95
-- The Fortran compiler identification is GNU 7.5.0
-- Detecting Fortran compiler ABI info
-- Detecting Fortran compiler ABI info - done
-- Check for working Fortran compiler: /usr/bin/f95 - skipped
-- Checking whether /usr/bin/f95 supports Fortran 90
-- Checking whether /usr/bin/f95 supports Fortran 90 - yes
-- 
-- Configured Eigen 3.3.90
-- 
-- Available targets (use: make TARGET):
-- ---------+--------------------------------------------------------------
-- Target   |   Description
-- ---------+--------------------------------------------------------------
-- install  | Install Eigen. Headers will be installed to:
--          |     <CMAKE_INSTALL_PREFIX>/<INCLUDE_INSTALL_DIR>
--          |   Using the following values:
--          |     CMAKE_INSTALL_PREFIX: /usr/local
--          |     INCLUDE_INSTALL_DIR:  include/eigen3
--          |   Change the install location of Eigen headers using:
--          |     cmake . -DCMAKE_INSTALL_PREFIX=yourprefix
--          |   Or:
--          |     cmake . -DINCLUDE_INSTALL_DIR=yourdir
-- doc      | Generate the API documentation, requires Doxygen & LaTeX
-- blas     | Build BLAS library (not the same thing as Eigen)
-- uninstall| Remove files installed by the install target
-- ---------+--------------------------------------------------------------
-- 
-- Performing Test FARMHASH_HAS_BUILTIN_EXPECT
-- Performing Test FARMHASH_HAS_BUILTIN_EXPECT - Success
-- Looking for strtof_l
-- Looking for strtof_l - found
-- Looking for strtoull_l
-- Looking for strtoull_l - found
-- Using toolchain file: /opt/fsl-imx-xwayland/5.10-hardknott/sysroots/x86_64-pokysdk-linux/usr/share/cmake/OEToolchainConfig.cmake.
-- The ASM compiler identification is GNU
-- Found assembler: /opt/fsl-imx-xwayland/5.10-hardknott/sysroots/x86_64-pokysdk-linux/usr/bin/aarch64-poky-linux/aarch64-poky-linux-gcc
CMake Error at /home/soc123/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build/xnnpack/CMakeLists.txt:80 (MESSAGE):
  Unrecognized XNNPACK_TARGET_PROCESSOR = cortexa72-cortexa53-crypto


-- Configuring incomplete, errors occurred!
See also "/home/soc123/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build/CMakeFiles/CMakeOutput.log".
See also "/home/soc123/Desktop/tensorflow-imx-lf-5.10.72_2.2.0/build/CMakeFiles/CMakeError.log".

CMakeLists.txt:

#
# Copyright 2020 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#      https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.


# Builds the minimal Tensorflow Lite example.

cmake_minimum_required(VERSION 3.16)
project(yolov5 C CXX)

set(TENSORFLOW_SOURCE_DIR "" CACHE PATH
  "Directory that contains the TensorFlow project"
)
if(NOT TENSORFLOW_SOURCE_DIR)
  get_filename_component(TENSORFLOW_SOURCE_DIR
    "${CMAKE_CURRENT_LIST_DIR}/../../../../"
    ABSOLUTE
  )
endif()

add_subdirectory(
  "${TENSORFLOW_SOURCE_DIR}/tensorflow/lite"
  "${CMAKE_CURRENT_BINARY_DIR}/tensorflow-lite"
  EXCLUDE_FROM_ALL
)

find_package( OpenCV REQUIRED )
find_package(Threads REQUIRED)
include_directories( ${OpenCV_INCLUDE_DIRS} )
include_directories( include/ncnn2 )



set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++14 -fopenmp")
set(CMAKE_CXX_FLAGS_RELEASE "-O2 -DNDEBUG")
set(CMAKE_CXX_FLAGS_DEBUG "-g")
set(CMAKE_CXX_STANDARD 14)



if (NOT CMAKE_BUILD_TYPE AND NOT CMAKE_CONFIGURATION_TYPES)
    message(STATUS "No build type selected, default to Release")
    set(CMAKE_BUILD_TYPE "Release" CACHE STRING "Build type (default Debug)" FORCE)
endif()





set(THREADS_PREFER_PTHREAD_FLAG ON)

aux_source_directory(./src/ SRC_LIST)

add_executable(${PROJECT_NAME} ${SRC_LIST})
target_link_libraries(yolov5 PUBLIC tensorflow-lite ${OpenCV_LIBS} /home/soc123/try/tensorflow-master/tensorflow/lite/examples/yolov5/lib/libncnn.a Threads::Threads)

log:https://drive.google.com/drive/folders/1Lmi2pgd7w_ZlNZDGCNO4hOKzlejmy-dy?usp=sharing

Regards,

YUN-XIAN HUANG

[Bio_TICFSL · Accepted solution]

Hello,

The MX8QM don't have NPU, That the reason of Tensorflow does not run on it.

Regards

[jimdeway]

Thanks for your reply.

But I still have a small problem here.
How can I stop using NPU?