• State TI Thinks Resolved
  • Locked Locked
  • Replies 14 replies
  • Answers 2 answers
  • Subscribers 100 subscribers
  • Views 628 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA4VM: Linux SDK for edge AI applications environment

Ethan Tian
Intellectual 1301 points
Part Number: TDA4VM
Other Parts Discussed in Thread: J721EXSOMXEVM,

Hi experts

I have download the 

ti-processor-sdk-linux-sk-tda4vm-etcher-image.zip

and because I work behind a VPN, so I run a docker container

But when I run ./setup_script.sh

There's a error appear.

How can I solve this problem

  • 0
    TI__Intellectual 1301 points 

    -- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- CMAKE_BUILD_TYPE = Release PROJECT_NAME = ti_dl_inferer
-- CMAKE_BUILD_TYPE = Release PROJECT_NAME = edgeai_utils
-- CMAKE_BUILD_TYPE = Release PROJECT_NAME = edgeai_common
-- CMAKE_BUILD_TYPE = Release PROJECT_NAME = app_edgeai
-- CMAKE_BUILD_TYPE = Release PROJECT_NAME = app_dump_model_info
-- CMAKE_BUILD_TYPE = Release PROJECT_NAME = app_config_checker
-- Configuring done
-- Generating done
-- Build files have been written to: /opt/edge_ai_apps/apps_cpp/build
Scanning dependencies of target ti_dl_inferer
Scanning dependencies of target edgeai_utils
[  7%] Building CXX object dl_inferer/CMakeFiles/ti_dl_inferer.dir/src/ti_dl_inferer.cpp.o
[  7%] Building CXX object utils/CMakeFiles/edgeai_utils.dir/src/edgeai_perfstats.cpp.o
In file included from /opt/edge_ai_apps/apps_cpp/dl_inferer/src/ti_dl_inferer.cpp:37:
/opt/edge_ai_apps/apps_cpp/dl_inferer/../dl_inferer/include/ti_tflite_inferer.h:36:10: fatal error: tensorflow/lite/model.h: No such file or directory
   36 | #include <tensorflow/lite/model.h>
      |          ^~~~~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
make[2]: *** [dl_inferer/CMakeFiles/ti_dl_inferer.dir/build.make:63: dl_inferer/CMakeFiles/ti_dl_inferer.dir/src/ti_dl_inferer.cpp.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:199: dl_inferer/CMakeFiles/ti_dl_inferer.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
[ 10%] Building CXX object utils/CMakeFiles/edgeai_utils.dir/src/ti_logger.cpp.o
[ 14%] Linking CXX static library ../../lib/Release/libedgeai_utils.a
[ 14%] Built target edgeai_utils
make: *** [Makefile:130: all] Error 2
Setup Done!
[docker] root@j7-evm:/opt/edge_ai_apps#

    This is the log of error

  • 0 in reply to Ethan Tian
    TI__Intellectual 1301 points

    And I need to restart a number of times to load, the following error is reported every time 

    U-Boot SPL 2021.01-g53e79d0e89 (Sep 02 2021 - 18:20:00 +0000)
Model: Texas Instruments K3 J721E SoC
Board: J721EX-PM2-SOM rev E8
SYSFW ABI: 3.1 (firmware rev 0x0015 '21.5.0--v2021.05 (Terrific Llam')
Trying to boot from MMC2
spl_load_image_fat: error reading image tispl.bin, err - -6
spl_load_image_ext: ext4fs mount err - 0
SPL: failed to boot from all boot devices
### ERROR ### Please RESET the board ###

  • 0 in reply to Ethan Tian
    TI__Mastermind 25850 points

    Hi Ethan,

    For the error that occurs when running ./setup_script.sh, it looks like the required dependencies are not being downloaded properly.

    Could you check three things:

    1. Is the /opt/tensorflow directory empty?
    2. Are you able to run git clone https://github.com/TexasInstruments/tensorflow.git inside docker?
    3. Are you behind TI company proxy? If so, could you do a ping test to ti.com: ping ti.com ?

    A workaround would be to run the ./setup_script.sh while the board is not behind a VPN. 


    As for the problem with having to restart the board a number of time before loading, I have not encountered this problem before. I will see if I can reproduce the issue on my side and get back to you next week.

    Regards,

    Takuma

  • 0 in reply to Takuma Fujiwara
    TI__Mastermind 25850 points

    Additional to my questions above, is this on the SK-TDA4VM or the J721EXSOMXEVM?

    Regards, Takuma

  • 0 in reply to Takuma Fujiwara
    TI__Intellectual 1301 points

    Hi Takuma 

    This is on the J721EXSOMXEVM

    For the question 1,  the /opt/tensorflow directory is empty

    And for question2 and 3,I will get back to you next week.

  • 0 in reply to Takuma Fujiwara
    TI__Intellectual 1301 points

    Hi Takuma

    I try to run git clone https://github.com/TexasInstruments/tensorflow.git 

    This did solve my problem,but a new one came up

    Scanning dependencies of target ti_dl_inferer
Scanning dependencies of target edgeai_utils
[  3%] Building CXX object utils/CMakeFiles/edgeai_utils.dir/src/edgeai_perfstats.cpp.o
[  7%] Building CXX object dl_inferer/CMakeFiles/ti_dl_inferer.dir/src/ti_dl_inferer.cpp.o
In file included from /opt/tensorflow/tensorflow/lite/core/api/op_resolver.h:20,
                 from /opt/tensorflow/tensorflow/lite/model_builder.h:27,
                 from /opt/tensorflow/tensorflow/lite/model.h:22,
                 from /opt/edge_ai_apps/apps_cpp/dl_inferer/../dl_inferer/include/ti_tflite_inferer.h:36,
                 from /opt/edge_ai_apps/apps_cpp/dl_inferer/src/ti_dl_inferer.cpp:37:
/opt/tensorflow/tensorflow/lite/schema/schema_generated.h:20:10: fatal error: flatbuffers/flatbuffers.h: No such file or directory
   20 | #include "flatbuffers/flatbuffers.h"
      |          ^~~~~~~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
make[2]: *** [dl_inferer/CMakeFiles/ti_dl_inferer.dir/build.make:63: dl_inferer/CMakeFiles/ti_dl_inferer.dir/src/ti_dl_inferer.cpp.o1
make[1]: *** [CMakeFiles/Makefile2:199: dl_inferer/CMakeFiles/ti_dl_inferer.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
[ 10%] Building CXX object utils/CMakeFiles/edgeai_utils.dir/src/ti_logger.cpp.o
[ 14%] Linking CXX static library ../../lib/Release/libedgeai_utils.a
[ 14%] Built target edgeai_utils
make: *** [Makefile:130: all] Error 2
Setup Done!

  • 0 in reply to Ethan Tian
    TI__Intellectual 1301 points

    Hi Takuma

    I have saw the setup_script.sh and install_tensorflow.sh,I found that I can't archive this page,and I go to search flatbuffers in git, https://github.com/google/flatbuffers

    I found that in version2.0,there's no folder name archive

    wget https://github.com/google/flatbuffers/archive/${FLATBUFFERS_VERSION}.tar.gz

    And then I do git clone https://github.com/google/flatbuffers, and put flatbuffers folder into /opt

    This error resolved.

    But still a new one came up,

    Scanning dependencies of target ti_dl_inferer
Scanning dependencies of target edgeai_utils
[  3%] Building CXX object utils/CMakeFiles/edgeai_utils.dir/src/edgeai_perfstats.cpp.o
[  7%] Building CXX object dl_inferer/CMakeFiles/ti_dl_inferer.dir/src/ti_dl_inferer.cpp.o
In file included from /opt/tensorflow/tensorflow/lite/core/subgraph.h:27,
                 from /opt/tensorflow/tensorflow/lite/interpreter.h:30,
                 from /opt/tensorflow/tensorflow/lite/interpreter_builder.h:27,
                 from /opt/tensorflow/tensorflow/lite/model.h:27,
                 from /opt/edge_ai_apps/apps_cpp/dl_inferer/../dl_inferer/include/ti_tflite_inferer.h:36,
                 from /opt/edge_ai_apps/apps_cpp/dl_inferer/src/ti_dl_inferer.cpp:37:
/opt/tensorflow/tensorflow/lite/delegates/nnapi/nnapi_delegate.h:23:10: fatal error: absl/types/optional.h: No such file or directory
   23 | #include "absl/types/optional.h"
      |          ^~~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
make[2]: *** [dl_inferer/CMakeFiles/ti_dl_inferer.dir/build.make:63: dl_inferer/CMakeFiles/ti_dl_inferer.dir/src/ti_dl_inferer.cpp.o1
make[1]: *** [CMakeFiles/Makefile2:199: dl_inferer/CMakeFiles/ti_dl_inferer.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
[ 10%] Building CXX object utils/CMakeFiles/edgeai_utils.dir/src/ti_logger.cpp.o
[ 14%] Linking CXX static library ../../lib/Release/libedgeai_utils.a
[ 14%] Built target edgeai_utils
make: *** [Makefile:130: all] Error 2
Setup Done!

    I think this is because the version of tools, could you pls reinstall this image and do the ./setup_script.sh, I found that in install_tensorflow.sh, the version of flatbuffers is 1.12, but now the source in git have changed into 2.0

  • 0 in reply to Ethan Tian
    TI__Mastermind 25850 points

    Hi Ethan,

    Could you try the following:

    1. Start docker with ./docker_run.sh
    2. Remove the tensorflow directory with the command rm -r /opt/tensorflow/
    3. Run ./setup_script.sh while inside docker

    I have tried a fresh install of Linux SDK for edge AI development, and was able to reproduce the initial error of "fatal error: tensorflow/lite/model.h: No such file or directory".

    The issue seems to be that only for the tensorflow dependency, a directory is created regardless of whether the installation was successful. This causes the setup_script to not install the tensorflow dependency because the script only installs missing dependencies, and it checks if the dependency is missing by the existence of the dependency's installation directory.

    Regards, 

    Takuma

  • 0 in reply to Ethan Tian
    TI__Mastermind 25850 points

    Hi Ethan,

    Ethan Tian said:
    And I need to restart a number of times to load, the following error is reported every time

    If you are still facing this issue, could you provide the following information:

    • Power supply info (voltage/amperage)
    • micro SD card info (speed class and application perfomance class)

    Regards,

    Takuma

  • 0 in reply to Takuma Fujiwara
    TI__Intellectual 1301 points

    Sorry Takuma, I've done some other EVM experiments recently, I will respond to you later when the test is complete

  • 0 in reply to Takuma Fujiwara
    TI__Intellectual 1301 points

    Hi Takuma

    Takuma Fujiwara said:
    • Start docker with ./docker_run.sh
    • Remove the tensorflow directory with the command rm -r /opt/tensorflow/
    • Run ./setup_script.sh while inside docker

    This still not solve my issue, I tried to run ./setup_script.sh twice, it still reporting this error

  • 0 in reply to Ethan Tian
    TI__Mastermind 25850 points

    Hi Ethan,

    Could you try the following:

    1. Install to micro SD card a fresh installation of the SK SDK
    2. Create and run docker container
    3. Run setup_script
    4. Post any error messages that print out from the script

    I would like to make sure the things we tried in the past is not the cause of the error, and we both have the same software setup.

    Regards,

    Takuma

  • 0 in reply to Takuma Fujiwara
    TI__Intellectual 1301 points

    Hi Takuma

    It looks like another problem has arisen

    Message: Hotdoc not found, not building the documentation
Build targets in project: 17

GstTIOVX 0.4.0

  User defined options
    pkg_config_path: pkgconfig
    prefix         : /usr

Found ninja-1.10.0 at /usr/bin/ninja
ninja: Entering directory `build'
[1/59] Compiling C object gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxallocator.c.o
FAILED: gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxallocator.c.o 
cc -Igst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p -Igst-libs/gst/tiovx -I../gst-libs/gst/tiovx -I. -I.. -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/aarch64-linux-gnu/glib-2.0/include -c
cc1: error: /usr/include/processor_sdk/imaging/algos/ae/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/algos/awb/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/ti_2a_wrapper/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
[2/59] Compiling C object gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpoolutils.c.o
FAILED: gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpoolutils.c.o 
cc -Igst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p -Igst-libs/gst/tiovx -I../gst-libs/gst/tiovx -I. -I.. -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/aarch64-linux-gnu/glib-2.0/include -c
cc1: error: /usr/include/processor_sdk/imaging/algos/ae/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/algos/awb/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/ti_2a_wrapper/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
[3/59] Compiling C object gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpool.c.o
FAILED: gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpool.c.o 
cc -Igst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p -Igst-libs/gst/tiovx -I../gst-libs/gst/tiovx -I. -I.. -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/aarch64-linux-gnu/glib-2.0/include -c
cc1: error: /usr/include/processor_sdk/imaging/algos/ae/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/algos/awb/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/ti_2a_wrapper/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
ninja: build stopped: subcommand failed.
ninja: Entering directory `build'
[1/60] Compiling C object gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxallocator.c.o
FAILED: gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxallocator.c.o 
cc -Igst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p -Igst-libs/gst/tiovx -I../gst-libs/gst/tiovx -I. -I.. -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/aarch64-linux-gnu/glib-2.0/include -c
cc1: error: /usr/include/processor_sdk/imaging/algos/ae/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/algos/awb/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/ti_2a_wrapper/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
[2/60] Compiling C object gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpoolutils.c.o
FAILED: gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpoolutils.c.o 
cc -Igst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p -Igst-libs/gst/tiovx -I../gst-libs/gst/tiovx -I. -I.. -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/aarch64-linux-gnu/glib-2.0/include -c
cc1: error: /usr/include/processor_sdk/imaging/algos/ae/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/algos/awb/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/ti_2a_wrapper/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
[3/60] Compiling C object gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpool.c.o
FAILED: gst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p/gsttiovxbufferpool.c.o 
cc -Igst-libs/gst/tiovx/libgsttiovx-1.0.so.1.0.p -Igst-libs/gst/tiovx -I../gst-libs/gst/tiovx -I. -I.. -I/usr/include/gstreamer-1.0 -I/usr/include/glib-2.0 -I/usr/lib/aarch64-linux-gnu/glib-2.0/include -c
cc1: error: /usr/include/processor_sdk/imaging/algos/ae/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/algos/awb/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: error: /usr/include/processor_sdk/imaging/ti_2a_wrapper/include/: No such file or directory [-Werror=missing-include-dirs]
cc1: all warnings being treated as errors
ninja: build stopped: subcommand failed.

    The entire error_log I have placed in the attachmenterror_log

  • 0 in reply to Ethan Tian
    TI__Mastermind 25850 points

    Hi Ethan,

    This is a new error introduced in the git repository installed through the setup_script.

    Could you try the workaround in this E2E: https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1050063/sk-tda4vm-problem-with-first-install-of-sk-tda4vm

    Regards,

    Takuma