Processor SDK Linux Automotive Software Developers Guide
Contents
Processor SDK Linux Automotive Software Developers Guide
Welcome to the DRA7xx Processor SDK Linux Automotive Software Developer’s Guide.
Thank you for choosing the DRA7xx EVM for your application. The purpose of this guide is to get you going with developing software for the DRA7xx on a Linux development host only.
NOTE!
This is alive Wiki page that gets updated for every release
. For the Software Developers Guide that is specific to the release you are using, please refer to the SW Dev Guide link in Processor SDK Linux Automotive Landing Page (http://processors.wiki.ti.com/index.php/Category:Processor_SDK_Linux_Automotive)
Introduction
Some commands are to be executed on the Linux development host, some on the Linux target and some on the u-boot (bootloader) prompt. The following conventions are used todistinguish the commands on a host and on the target:
host $
<this command is to be executed on the host>
target #
<this command is to be executed on the target>
u-boot :>
<this command is to be executed on the u-boot prompt>
Note!
All instructions in this guide are for Ubuntu 14.04 (http://releases.ubuntu.com/14.04/) on 64-bit machine. At this time, this is the only supported Linux host distribution fordevelopment.
Starting your software development
Setup up ARM linux development Environment on the host. Please refer to this link to see how to set one up.
Configuration of ARM Linux development Environment (http://processors.wiki.ti.com/index.php/How_to_Build_a_Ubuntu_Linux_host_under_VirtualBox)
Step 1: Install the Processor SDK Linux Automotive release on the host machine.
- Download the Processor SDK Linux Automotive installer.
-
If necessary make the installer executable manually by executing:
host $ chmod +x ti-processor-sdk-linux-automotive-dra7xx-evm-06.00.00.03-installer.bin - Execute the installer on the host and follow the instructions:
host $ ./ti-processor-sdk-linux-automotive-dra7xx-evm-06.00.00.03-installer.bin
Step 2 :Setup the INSTALL_DIR environment variable to the location where the Processor SDK Linux Automotive is installed (the following assumes that Processor SDK LinuxAutomotive was installed at default location):
host export INSTALL_DIR="{HOME}/ti-processor-sdk-linux-automotive_dra7xx-evm_06_00_00_03"
Step 3: Setup the Processor SDK Linux Automotive on host
Host Setup
- Ubuntu : The recommended Linux distribution is Ubuntu 14.04 or Ubuntu 16.04.
The following build host packages are required for Ubuntu. The following command will install the required tools on the Ubuntu Linux distribution.
For Ubuntu 14.04 and 16.04, please run the following:$ sudo apt-get install git build-essential python diffstat texinfo gawk chrpath dos2unix wget unzip socat doxygen libc6:i386 libncurses5:i386 libstdc++6:i386 libz1:i386The build also requires that bash is configured as the default system shell. The following steps will configure Ubuntu to use bash instead of the default dash.
$ sudo dpkg-reconfigure dash (Select "no" when prompted)NOTE!
This step is required only if you want to use Code Composer Studio IDE. In the latest Processor SDK Linux Automotive releases, M4/DSP firmware image build is handled aspart of SDK Yocto build.Code Composer Studio 6.1.3.00034 must be downloaded manually and placed into either the downloads directory, or placed on a mirror. Please seeCCS v6 Download (http://processors.wiki.ti.com/index.php/Download_CCS#Code_Composer_Studio_Version_6_Downloads) for details and download the Linux off-line installation tarball, e.g.,CCS6.1.3.00034_linux.tar.gz.
The Processor SDK Linux Automotive comes with a script for setting up your Ubuntu 14.04 LTS development host. It is an interactive script, but if you accept the defaults by pressingreturn you will use the recommended settings. This is recommended for first time users. Note that this script requires ethernet access as it will update your Ubuntu Linux developmenthost with the packages required to develop using the Processor SDK Linux Automotive.
Note: Please make sure that the proxy settings are done for http, https, git, ftp and wget before proceeding further.
Execute the script in the Processor SDK Linux Automotive release directory using:host cd{INSTALL_DIR} host $./setup.shThe setup script would perform the following operations:
1.Installs all the necessary package on the host for the SDK.
2.Prepares the UART terminal to communicate with the target over Debug USB on Minicom. If you want to use a windows host for connecting to the target instead, see the #Setting_up_Tera_Term section
3.Setups the linaro cross compiler
4.Installs the dependencies for the repo tool.
5.Initialize the repo by pointing it to Processor SDK Linux Automotive release manifest location.To start minicom on your Linux development host execute
minicom -w(or Tera Term on Windows).
Step 4: Prepare SD card
To install the release image, you need a μSD Card (at least 4GB) with 2 partitions:
- boot (vfat) partition.
- rootfs (ext4 or ext3) partition.
The following procedure prepares the sdcard: (however, the user can choose to do it manually as well)
- Plug an SD card reader to your PC and insert a μSD card. It must be at least 4GB size.
-
Identify which device corresponds to the SD card reader.
sudo fdisk -lcommand can help you find out where the μSD Card is mapped. We will call it/dev/sdYhere.Note : If you are using NFS file system, then edit ${INSTALL_DIR}/board-support/pre-built-images/uenv.txt, add “ip=dhcp” in bootargs.
- Re-format your μSD card with this script mksdboot.sh from the bin directory in the Processor SDK Linux Automotive
sudo{INSTALL_DIR}/bin/mksdboot.sh –device /dev/sdY –sdk ${INSTALL_DIR}
The above script would prepare the SD card with the prebuilt images and yocto filesystem for SD boot.
Step 5: Booting the board
To boot the board with the above created SD card, refer to the Quick Start Guide at the Processor SDK Linux Automotive download location Link (http://software-dl.ti.com/infotainment/esd/jacinto6/processor-sdk-linux-automotive/latest/index_FDS.html) to setup the board.
Then, power cycle the board and login with username as root
.Repo tool Usage
Starting source code development using repo tool
The Processor SDK Linux Automotive release uses the repo tool to effectively manage the different components.
NOTE :
1: The first step to the repo tool is the repo initialization and this is done as part of the $INSTALL_DIR/setup.sh script
2: The repo tool is downloaded into the bin folder in the Processor SDK Linux Automotive directory. Please ensure that this path is updated in the environmentvariable as shown below
host export PATH={INSTALL_DIR}/bin:$PATH
The Processor SDK Linux Automotive release contains a helper script that sets up the development environment. Run the script as shown below:
host cd{INSTALL_DIR}
host $ ./bin/fetch-sources.sh
The script does the following:
- Check for the repo tool.
-
Perform repo sync
-
Create a branch called dev
-
Checkout the branch dev
It is expected that the development is done on the dev branch.
How to get updates
If there are changes in the remote repositories, it could be fetched using the same script. However, please make note of these important points.
1. The script will fetch the latest changes, and switch back to the dev branch.
2. The new updates from the remote, will be available in the master branch.
3. The decision on whether to merge the changes to the local branch or merge the local branch to the master is left to the developer
Building Yocto Filesystem
Please ensure that the setup.sh script is run as described in Starting your software development
section and answer ‘yes’ to fetch-sources prompt. Before building the filesystem, ensure that the svn, http, ftp and git proxies are set correctly. Refer to the following link for these settings
https://wiki.yoctoproject.org/wiki/Working_Behind_a_Network_Proxy
Add the Linaro cross-compile toolchain path in the PATH environment variable.
host export PATH=<Path to Linaro cross-compile toolchain>/bin:PATH
Run this command to as a one-time setup for the yocto build
host cdINSTALL_DIR
NOTE: Prior to running the setup-yocto.sh script, make sure that GCC Linaro cross compiler installation path and the CCS binary is downloaded into a particular directory # Update the ./bin/setup-yocto.sh file to indicate this path, the default path is set to /sdk/tools and this may vary on every machine.
host $ ./bin/setup-yocto.sh
For building core sdk , run the build-core-sdk.sh passing machine name as an argument.
host $ cd yocto-layers
Create a downloads directory (if building using Yocto for the first time), where the Yocto build will place the downloads. Note the path of the directory.
host $ mkdir downloads
NOTE: Please pass the same downloads directory path to the following build-script when prompted
host $ ./build-core-sdk.sh dra7xx-evm
These scripts will build the tisdk-rootfs-image.
After build is complete the generated images can be found in yocto-layers/build/arago-tmp-external-linaro-toolchain/deploy/images/
Note: The build does not generate a uenv.txt. You need to copy it from the prebuilt binaries in the release.
Modifying source code and rebuilding a component
Once the Yocto setup is complete developers would like to modify a certain component source code and rebuild it. The source code for the generic components like omapdrmtest can befound in ${INSTALL_DIR}/yocto-layers/build/arago-tmp-external-linaro-toolchain/work/cortexa15hf-neon-linux-gnueabi/
Steps to rebuild:
- Modify the source of the component in its work area
-
cd to yocto-layers
host cd{INSTALL_DIR}/yocto-layers - export PATH if not done previously:
host export PATH=<Path to Linaro cross-compile toolchain>/bin:PATH - Run one/combination of the folowing tasks:
a. To configure and compile with the latest changes in the work area:
host ./build-specific-recipe.sh-f -c configure ./build-specific-recipe.sh < MACHINE-NAME > -f -c compile < RECIPE-NAME > For example, if the omapdrmtest source is modified and you want to generate only the binaries for dra7xx-evm, the compile task should suffice:
host $ ./build-specific-recipe.sh < MACHINE-NAME > -f -c compile omapdrmtestThe binaries will be present in the same folder as the source files.
b. To generate .ipk package in addition to the binaries, following additional tasks need to be run after above steps:
host ./build-specific-recipe.sh <MACHINE-NAME> -f -c install <RECIPE-NAME> host./build-specific-recipe.sh <MACHINE-NAME> -f -c package <RECIPE-NAME> host $./build-specific-recipe.sh <MACHINE-NAME> -f -c package_write_ipk <RECIPE-NAME>This will generate .ipk packages for the recipe. They can be found in the deploy-ipks folder of the work area of the recipe.
The .ipk packages can be copied and installed on the target by:target $ opkg install <PACKAGE-NAME>This approach is useful in the case of recipes that generate a large number of binaries, that are difficult to copy manually, like gst-plugins-bad.
c. During building and debugging the kernel or kernel modules, the compile_kmodules task needs to be executed that generates kernel modules.
host $ ./build-specific-recipe.sh <MACHINE-NAME> -f -c compile_kmodules <RECIPE-NAME>For a complete set of tasks that a specific recipe executes during its build, please refer to the log.task_order in the temp folder of the component.
Pass the relevant tasks to build-specific-recipe.sh for required outcome. -
For a robust solution , once the change in source area is tested with the above steps, please update the recipe.
a. create a patch of the change
b. copy it in the folder where the recipe is present
c. add the patch name in the SRC_URI variable in the recipeThe build-specific-recipe.sh recognizes the change in the recipe and builds it if required.
To build a specific component after recipe is updated, use the build-specific-recipe.sh
host $./build-specific-recipe.sh <MACHINE-NAME> <RECIPE NAME>For example to build omapdrmtest:
host $./build-specific-recipe.sh <MACHINE-NAME> omapdrmtestMACHINE-NAME can be either omap5-evm or dra7xx-evm
To clean a specific component:
host $ ./clean-specific-recipe.sh <MACHINE-NAME> <RECIPE NAME>Processor SDK Linux Automotive software overview
The Processor SDK Linux Automotive contains many software components. Several components are developed by the open source community (white). A few components that leverageHW acceleration in SoC are developed by Texas Instruments. TI contributes, and sometimes even maintains, some of these open source community projects, but the support model isdifferent from a project developed solely by TI.
Running Examples
Running OMAP DRM DSS Examples
The drmclone, drmextended, and modetest examples demonstrates how to create a CRTC (i.e. FB) and display planes (overlays) on the CRTC. Additionally, drmtest demonstrates similarfunctionality as the previously mentioned demos, along with dynamic plane updates for 2 CRTCs.
Retrieve the omapdrm-tests source
git clone https://github.com/tomba/omapdrm-tests.git
cd omapdrm-tests
Run (or example planescale)
./planescale
Graphics Demos from Command Line
The graphics driver and userspace libraries and binaries are distributed along with the SDK.
Graphic demos can also run from command line. In order to do so, exit Weston by pressing Ctrl-Alt-Backspace from the keyboard which connects to the EVM. Then, if the LCD screenstays in “Please wait…”, press Ctrl-Alt-F1 to go to the command line on LCD console. After that, the command line can be used from serial console, SSH console, or LCD console. |Graphic demos can also run from command line. In order to do so, exit Weston by pressing Ctrl-Alt-Backspace from the keyboard which connects to the EVM. Then, if the LCD screenstays in “Please wait…”, press Ctrl-Alt-F1 to go to the command line on LCD console. After that, the command line can be used from serial console, SSH console, or LCD console.
Please make sure the board is connected to atleast one display before running these demos
Finding Connector ID
Note: Most of the applications used in the Demos would require the user to pass a connector id. A connector id is a number that is assigned to each of the display devices connected to the system. To get the list of the display devices connected and the corresponding connector id one can use the modetest application (shipped with the file system) as mentioned below:
target # modetest
Look for the display device for which the connector ID is required – such as HDMI, LCD etc.
Connectors:
id encoder status type size (mm) modes encoders
4 3 connected unknown 0x0 1 3
modes:
name refresh (Hz) hdisp hss hse htot vdisp vss vse vtot)
1280x800 60 1280 1328 1360 1404 800 804 811 823 flags: nhsync, nvsync; type: preferred, driver
...
16 11 connected HDMI-A 700x390 31 11
modes:
name refresh (Hz) hdisp hss hse htot vdisp vss vse vtot)
1280x720 60 1280 1390 1430 1650 720 725 730 750 flags: phsync, pvsync; type: preferred, driver
Usually LCD is assigned 4 (800×480), HDMI is assigned 16 (multiple resolutions).
Finding Plane ID
To find the Plane ID, run the modetest command:
target # modetest
Look for the section called Planes. (Sample truncated output of the Planes section is given below:
Planes:
id crtc fb CRTC x,y x,y gamma size
19 0 0 0,0 0,0 0
formats: RG16 RX12 XR12 RA12 AR12 XR15 AR15 RG24 RX24 XR24 RA24 AR24 NV12 YUYV UYVY
props:
...
20 0 0 0,0 0,0 0
formats: RG16 RX12 XR12 RA12 AR12 XR15 AR15 RG24 RX24 XR24 RA24 AR24 NV12 YUYV UYVY
props:
...
kmscube
Run kmscube on default display (LCD):
target # kmscube
Run kmscube on secondary display (HDMI):
target # kmscube -c <connector-id>
target # kmscube -c 16 #Usually, the connector id for HDMI is 16.
Run kmscube on all connected displays (LCD & HDMI & FPDLink(optional)):
target # kmscube -a
kmscube with video
This demo allows a video frame to be applied as a texture onto the surface of the kmscube. The user can invoke the
demo by following the syntax below:
target # viddec3test <path_to_the_file> --kmscube --connector <connector_number>
Run kmscube with video on default display (LCD):
target # viddec3test <path_to_the_file> --kmscube
Run kmscube with video on secondary display (HDMI):
target # viddec3test <path_to_the_file> --kmscube --connector 16 #Usually, the connector id for HDMI is 16.
Additionally, to change the field of view of the rotating cube, the user can specify the same on the command line like
below:
target # viddec3test <path_to_the_file> --kmscube --connector <connector_number> --fov <number>
Wayland/Weston
The supported Wayland/Weston version brings in the multiple display support in extended desktop mode and the ability to drag-and-drop windows from one display to the other.
To execute the demos, the graphics driver must be initialized by running start weston, if this has not been done earlier.
target # /etc/init.d/weston start
To launch weston without using systemd init scripts, do the following:
On all connected displays (LCD, HDMI and FPDLink):
target # weston --tty=1 --backend=drm-backend.so
By default, the screensaver timeout is configured to 300 seconds.
The user can change the screensaver timeout using a command line option
--idle-time=<number of seconds>
To disable the screen timeout and to configure weston configured to display on all connectors, use the option with “0” as the input:
--idle-time=0
The filesystem comes with a preconfigured weston.ini file which will be located in /etc/weston.ini
Running weston clients
There is one icon on the top right hand corner of the weston desktop window which has been configured for
– weston-terminal
Clicking this icon should launch the applications on the Weston Desktop.
It is possible to add other icons by editing the weston.ini file.
There are several other applications that are included in the default filesystem. To invoke these applications, the user should launch the weston-terminal (top right hand corner of the desktop) and then invoke the client apps as described below from within the terminal window:
wayland sh # /usr/bin/weston-flower
wayland sh # /usr/bin/weston-clickdot
wayland sh # /usr/bin/weston-cliptest
wayland sh # /usr/bin/weston-dnd
wayland sh # /usr/bin/weston-editor
wayland sh # /usr/bin/weston-eventdemo
wayland sh # /usr/bin/weston-image /usr/share/weston/terminal.png
wayland sh # /usr/bin/weston-resizor
wayland sh # /usr/bin/weston-simple-egl
wayland sh # /usr/bin/weston-simple-shm
wayland sh # /usr/bin/weston-simple-touch
wayland sh # /usr/bin/weston-smoke
wayland sh # /usr/bin/weston-info
wayland sh # /usr/bin/weston-terminal
Running multimedia with Wayland sink
The GStreamer video sink for Wayland is the waylandsink. To use this video-sink for video playback:
target # gst-launch-1.0 playbin uri=file://<path-to-file-name> video-sink=waylandsink
Exiting weston
Terminate all Weston clients before exiting Weston. If you have invoked Weston from the serial console, exit Weston by pressing Ctrl-C.
It is also possible to invoke Weston from the native console, exit Weston by using pressing Ctrl-Alt-Backspace.
IMG PowerVR Demos
The Processor SDK Linux Automotive filesystem comes packaged with example OpenGLES applications. Both DRM and Wayland based applications are packaged as part of the filesystem.
The examples running on Wayland can be invoked using the below commands.
target # /usr/bin/SGX/demos/Wayland/OGLES2ChameleonMan
target # /usr/bin/SGX/demos/Wayland/OGLES2Navigation
The examples running on DRM/KMS can be invoked using the below commands.
target # /usr/bin/SGX/demos/Raw/OGLES2ChameleonMan
target # /usr/bin/SGX/demos/Raw/OGLES2Navigation
After you see the output on the display interface, hit q to terminate the application.
Using the PowerVR Tools
Please refer to http://community.imgtec.com/developers/powervr/graphics-sdk/ for additional details on the tools and detailed documentation.
The target file system includes tools such as PVRScope and PVRTrace recorder libraries from Imagination PowerVR SDK to profile and trace SGX activities. In addition, it also includes PVRPerfServerDeveloper toolfor Jacinto6 platform.
PVRTune
PVRPerfServerDeveloper tool can be used along with the PVRTune running on the PC to gather data on the SGX loading and activity threads. You can invoke the tool with the below command:
target # /opt/img-powervr-sdk/PVRHub/PVRPerfServer/PVRPerfServerDeveloper
PVRTrace
The default filesystem contains helper scripts to obtain the PVRTrace of the graphics application. This trace can then be played back on the PC using the PVRTrace Utility.
To start tracing, use the below commands as reference:
target # cp /opt/img-powervr-sdk/PVRHub/Scripts/start_tracing.sh ~/.
target # ./start_tracing.sh <log-filename> <application-to-be-traced>
Example:
target # ./start_tracing.sh westonapp weston-simple-egl
The above command will do the following:
1. Setup the required environment for the tracing
2. Create a directory under the current working directory called pvrtrace
3. Launch the application specified by the user
4. Start tracing the PVR Interactions and record the same to the log-filename
To end the tracing, user can invoke the Ctrl-C and the trace file path will be displayed.
The trace file can then be transferred to a PC and we can visualize the application using the host side PVRTrace utility. Please refer to the link at the beginning of this section for more details.
Testing DSS WB pipeline
Memory to Memory (M2M)
- Identify the WB pipeline M2M device.
# ls /sys/class/video4linux/
Video0 video10 video11
# cat /sys/class/video4linux/video10/name
omapwb-m2m
- Look at list of formats supported
# v4l2-ctl -d /dev/video10 --list-formats
ioctl: VIDIOC_ENUM_FMT
Index : 0
Type : Video Capture Multiplanar
Pixel Format: 'NV12'
Name : Y/CbCr 4:2:0
Index : 1
Type : Video Capture Multiplanar
Pixel Format: 'YUYV'
Name : YUYV 4:2:2
Index : 2
Type : Video Capture Multiplanar
Pixel Format: 'UYVY'
Name : UYVY 4:2:2
Index : 3
Type : Video Capture Multiplanar
Pixel Format: 'XR24'
Name : 32-bit BGRX 8-8-8-8
- Use v4l2-ctl command to test the input output. Below command converts from NV12 to YUYV using WB pipeline in M2M mode.
# v4l2-ctl -d /dev/video10 --set-fmt-video-out=width=1920,height=1080,pixelformat=NV12 \
--stream-from=test/BigBuckBunny_1920_1080_24fps_100frames.nv12 \
--set-fmt-video=width=1920,height=1080,pixelformat=YUYV \
--stream-to=out/video_test_file.yuyv --stream-mmap=3 --stream-out-mmap=3 --stream-count=70 --stream-poll
Capture
- Identify the WB pipeline capture device.
# ls /sys/class/video4linux/
Video0 video10 video11
# cat /sys/class/video4linux/video11/name
omapwb-cap
- Look at list of formats supported.
# v4l2-ctl -d /dev/video11 --list-formats
ioctl: VIDIOC_ENUM_FMT
Index : 0
Type : Video Capture Multiplanar
Pixel Format: 'NV12'
Name : Y/CbCr 4:2:0
Index : 1
Type : Video Capture Multiplanar
Pixel Format: 'YUYV'
Processor SDK Linux Automotive Software Developers Guide 13
Name : YUYV 4:2:2
Index : 2
Type : Video Capture Multiplanar
Pixel Format: 'UYVY'
Name : UYVY 4:2:2
Index : 3
Type : Video Capture Multiplanar
Pixel Format: 'XR24'
Name : 32-bit BGRX 8-8-8-8
- Use v4l2-ctl command to test the input output. Below command converts from NV12 to YUYV using WB pipeline in M2M mode.
# v4l2-ctl -d /dev/video11 -i 0 --set-fmt-video=pixelformat=NV12 \
--stream-to=/test/video_test_file.yuv --stream-mmap=6 --stream-count=10 --stream-poll
Video input set to 0 (CRTC#0 - LCD1: ok)
<<<<<<<<< 7.84 fps
<
# v4l2-ctl -d /dev/video11 -i 1 --set-fmt-video=pixelformat=NV12 --stream-to=/test/video
_test_file.yuv --stream-mmap=6 --stream-count=10 --stream-poll
Video input set to 1 (CRTC#1 - DIGIT/TV: ok)
<<<<<<<<<< 8.65 fps
Running aplay and arecord application
To playback/record on the evm via headset/mic, please make sure the following amixer settings are done:
– For playback via headset, enter the following at prompt target#
amixer sset 'Left DAC Mux',0 'DAC_L2'
amixer sset 'Right DAC Mux',0 'DAC_R2'
amixer cset name='HP Playback Switch' On
amixer cset name='Line Playback Switch' Off
amixer cset name='PCM Playback Volume' 127
Once these settings are successful, use aplay application for playback:
target # aplay <path_to_example_audio>.wav
- For recording via Mic In
amixer cset name='Left PGA Mixer Mic3L Switch' On
amixer cset name='Right PGA Mixer Mic3L Switch' On
amixer cset name='Left PGA Mixer Line1L Switch' off
amixer cset name='Right PGA Mixer Line1R Switch' off
amixer cset name='PGA Capture Switch' on
amixer cset name='PGA Capture Volume' 6
Once these settings are successful, use arecord to record
target # arecord -r 44.1 > <path_to_example_audio>.wav
Running GC320 application
GC320 is a 2D Graphic accelerator in DRA7xx. This IP can be utilized for the usecases like alpha blending, overlaying, bitBlit, color conversion, scaling, rotation etc.
SDK provides two sample GC320 testcases in the root filesystem. Before running the test, the gc320 kernel module needs to be inserted into system.
On 1.5GB RAM configuration
target# insmod /lib/modules/4.4.xx-gyyyyyyyy/extra/galcore.ko baseAddress=0x80000000 physSize=0x60000000
On 2GB RAM configuration
target# insmod /lib/modules/4.4.xx-gyyyyyyyy/extra/galcore.ko baseAddress=0x80000000 physSize=0x80000000
Now follow these instrctions to execute the applications
target# cd /usr/bin/GC320/tests/unit_test
target# export LD_LIBRARY_PATH=$PWD
target# ./runtest.sh
This script executes two sample unit test cases of filling rectangles and GC320 rendered results will be stored in .bmp file in a directory “result” under /usr/bin/GC320/tests/unit_test.
Note: To run all GC320 unit testcases, clone ti-gc320-test package from git://git.ti.com/graphics/ti-gc320-test.git : branch:ti-5.0.11.p7, rebuild test application, libraries etc and install the package on target.
Running viddec3test application
viddec3test is a demo application for decoder/video playback using hardware accelerators. The application currently runs on the kms display. The application requires the connector information for display. One can get the information of the display connected to the board by running the modetest application in the filesystem, as described above. To execute the application “modetest” make sure the display is connected to the board.
Running a decode on a display
To run a hardware decode on a display connected to the board, execute the following command:
target # viddec3test -s <connector_id>:<display resolution> filename --fps 30
e.g.: target # viddec3test -s 4:1920x1080 file.h264 --fps 30
Running single decode on dual displays
To run the output of a single decode on the dual displays. Please make sure both the displays are connected and get the information about the connectors and the resolution associated with it for both the displays from the modetest application.
target # viddec3test -s <connector_id_1>:<display resolution> -s <connector_id_2>:<display resolution> filename --fps 30
e.g.: target # viddec3test -s 4:1920×1080 -s 12:1024×768 file.h264 –fps 30
Running dual decode on dual displays
One can also run a dual decode and display their output on two different displays. Please make sure both the displays are connected and get the information about the connectors and the resolution associated with it for both the displays from the modetest application.
target # viddec3test -s <connector_id_1>:<display resolution> filename1 -s <connector_id_2>:<display resolution> filename2
e.g.: target # viddec3test -s 4:1920×1080 file1.h264 — -s 12:1024×768 file2.h264
Running DSP sample applications
DCE framework running on DSP1 supports C66x codecs with VIDDEC2 interface. In the current version of dspdce package, Universal_copy algorithm (with IUNIVERSAL interface) is invoked from dce VIDDEC2 interfaces. Universal_copy algorithm does simple input buffer to outbuffer memcopy. To integrate C66x codecs with VIDDEC2 interface, iuniversal interfaces needs to be replaced with ividdec2 interfaces in DCE framework.
The following are the applications that one could run from A15 to exercise the above on DSP:
copycodectest: This application is to test Universal_copy with VIDDEC2 interface. This application fills the input buffer with a number entered as argument and after process output buffer is tested for same pattern.
usage: copycodectest pattern.
Example:
target # copycodectest 123
yuvcopytest: This applciation reads one frame of yuv of known width and height into input buffer. Luma output buffer after process call is sent to display to check if universal_copy running on DSP is properly copied to output buffer.
usage: yuvcopytest -s
Example:
target # yuvcopytest -s 4:800x480 input.yuv -w 1920 -h 1080
Running a gstreamer pipeline
GStreamer v1.6 is supported in Processor SDK Linux Automotive 3.01.
One can run an audio video file using the gstreamer playbin from the console. Currently, the supported Audio/video sink is kmssink, waylandsink and alsassink.
kmssink:
target # gst-launch-1.0 playbin uri=file:///<path_to_file> video-sink=kmssink audio-sink=alsasink
waylandsink:
1. refer #Wayland/Weston to start the weston
2. target # gst-launch-1.0 playbin uri=file:///<path_to_file> video-sink=waylandsink audio-sink=alsasink
The following pipelines show how to use vpe for scaling and color space conversion.
1. Decode-> Scale->Display
target # gst-launch-1.0 -v filesrc location=example_h264.mp4 ! qtdemux ! h264parse ! \
ducatih264dec ! vpe ! 'video/x-raw, format=(string)NV12, width=(int)720, height=(int)480' ! kmssink
2. Color space conversion:
target # gst-launch-1.0 -v videotestsrc ! 'video/x-raw, format=(string)YUY2, width= \
(int)1280, height=(int)720' ! vpe ! 'video/x-raw, format=(string)NV12, width=(int)720, height=(int)480' \
! kmssink
Note:
1. While using playbin for playing the stream, vpe plugin is automatically picked up. However vpe cannot be used with playbin for scaling.
For utilizing scaling capabilities of vpe, using manual pipeline given above is recommended.
2. Waylandsink and Kmssink uses the cropping metadata set on buffers and does not require vpe plugin for cropping
The following pipelines show how to use v4l2src and ducatimpeg4enc elements to capture video from VIP and encode captured video respectively.
Capture and Display Fullscreen
target # gst-launch-1.0 v4l2src device=/dev/video1 num-buffers=1000 io-mode=4 ! 'video/x-raw, \
format=(string)YUY2, width=(int)1280, height=(int)720' ! vpe num-input-buffers=8 ! queue ! kmssink
Note:
The following pipelines can also be used for NV12 capture-display usecase.
Dmabuf is allocated by v4l2src if io-mode=4 and by kmssink and imported by v4l2src if io-mode=5
target # gst-launch-1.0 v4l2src device=/dev/video1 num-buffers=1000 io-mode=4 ! 'video/x-raw, \
format=(string)NV12, width=(int)1280, height=(int)720' ! kmssink
target # gst-launch-1.0 v4l2src device=/dev/video1 num-buffers=1000 io-mode=5 ! 'video/x-raw, \
format=(string)NV12, width=(int)1280, height=(int)720' ! kmssink
Capture and Display to a window in wayland
1. refer #Wayland/Weston to start the weston
2. target # gst-launch-1.0 v4l2src device=/dev/video1 num-buffers=1000 io-mode=4 ! 'video/x-raw, \
format=(string)YUY2, width=(int)1280, height=(int)720' ! vpe num-input-buffers=8 ! queue ! waylandsink
Note:
The following pipelines can also be used for NV12 capture-display usecase. Dmabuf is allocated by v4l2src
if io-mode=4 and by waylandsink and imported by v4l2src if io-mode=5.
Waylandsink supports both shm and drm. A new property use-drm is added to specify drm allocator based bufferpool to be used.
When using ducati or vpe plugins, use-drm is set in caps as true.
target # gst-launch-1.0 v4l2src device=/dev/video1 num-buffers=1000 io-mode=4 ! 'video/x-raw, \
format=(string)NV12, width=(int)1280, height=(int)720' ! waylandsink use-drm=true
target # gst-launch-1.0 v4l2src device=/dev/video1 num-buffers=1000 io-mode=5 ! 'video/x-raw, \
format=(string)NV12, width=(int)1280, height=(int)720' ! waylandsink use-drm=true
Capture and Encode into a MP4 file.
target # gst-launch-1.0 -e v4l2src device=/dev/video1 num-buffers=1000 io-mode=4 ! 'video/x-raw, \
format=(string)YUY2, width=(int)1280, height=(int)720, framerate=(fraction)30/1' ! vpe num-input-buffers=8 ! \
queue ! ducatimpeg4enc bitrate=4000 ! queue ! mpeg4videoparse ! qtmux ! filesink location=x.mp4
Note:
The following pipeline can be used in usecases where vpe processing is not required.
target # gst-launch-1.0 -e v4l2src device=/dev/video1 num-buffers=1000 io-mode=5 ! 'video/x-raw, \
format=(string)NV12, width=(int)1280, height=(int)720, framerate=(fraction)30/1' ! ducatimpeg4enc bitrate=4000 ! \
queue ! mpeg4videoparse ! qtmux ! filesink location=x.mp4
Capture and Encode and Display in parallel.
target # gst-launch-1.0 -e v4l2src device=/dev/video1 num-buffers=1000 io-mode=4 ! 'video/x-raw, \
format=(string)YUY2, width=(int)1280, height=(int)720, framerate=(fraction)30/1' ! vpe num-input-buffers=8 ! tee name=t ! \
queue ! ducatimpeg4enc bitrate=4000 ! queue ! mpeg4videoparse ! qtmux ! filesink location=x.mp4 t. ! queue ! kmssink
Running VIP/VPE/CAL application
Video Input Port
Video Input Port is used to capture video frames from BT56/ BT601 Camera. Currently the VIP driver supports following features.
For more information on VIP driver and other features, please refer to http://processors.wiki.ti.com/index.php/
Processor_SDK_VIP_Driver
• Standard V4L2 capture driver
• Supports single planar buffers
• Supports MMAP buffering method
• Supports DMABUF based buffering method
• Supports V4L2 endpoint standard way of specifying camera nodes
• Supports captures upto 60FPS
• Multi instance capture – All slices, ports supported
• Capture from a YUYV camera(8bit)
• NV12 capture format
Camera Adapter Layer
Camera Adapter Layer is used to capture video from CSI Camera. Currently the CAL driver supports following features.
• Standard V4L2 capture driver
• Supports single planar buffers
• Supports MMAP buffering method
• Supports DMABUF based buffering method
• Supports V4L2 endpoint standard way of specifying camera nodes (CSI bindings)
• Multi instance capture – 4data lane phy0 + 2data lane phy1
Supported cameras
Camera Adapter Layer is used to capture video from CSI Camera. Currently the CAL driver supports following features. Processor SDK Linux Automotive release supports following sensors/cameras/video inputs:
• OV10635 sensor – YUYV sensor on Vision board
• OV10635 sensor – YUYV sensor connected through LVDS
• OV10633 sensor – YUYV sensor connected on J6 EVM
• TVP5158 decoder – Support for decoding single channel analog video
• OV10640/OV490 – 720p CSI2 raw camera connected to OV490 ISP in YUYV format
Processor SDK supports following sensors/cameras
• mt9t111 camera sensor
Running dmabuftest
dmabuftest is a user space application which demonstrates capture display loopback. It can support multiple captures at the same time.
Video buffers are allocated by libdrm and they are shared to VIP through dmabuf.
It interfaces with the VIP through standard v4l2 ioctls.
Filesystem from release has dmabuftest app preinstalled.
To capture and display on the LCD screen, run following command target# dmabuftest -s 4:800×480 -d /dev/video1 -c 1280×720@YUYV
To capture and display on the HDMI display, run following command target# dmabuftest -s 32:1920×1200 -d /dev/video1 -c 1280×720@YUYV
To capture video in NV12 format, run following command target# dmabuftest -s 32:1920×1200 -d /dev/video1 -c
1280×720@NV12
To capture and display on KMScube backend (Video on a rotating cube), run following command
target# dmabuftest --kmscube --fov 20 -d /dev/video1 -c 1280x720@YUYV
This feature is currently not supported
To capture and display on wayland backend (Video in a wayland client window), run following command
target# dmabuftest -w 640x480 --pos 100x400 /dev/video1 -c 1280x720@YUYV
Capturing from OV10633 onboard camera
Linux kernel driver for OV1063x cameras support OV10633 sensor.
Video capture can be verified from the OV10633 sensor as follows
– Connect OV10633 sensor to the Leopard Imaging port on the J6 EVM
– Reboot the board and enable i2c2 as given above
– I2C device on Bus 2 slave address 0x37 should be probed successfully
– VIP should register a V4L2 video device (e.g. /dev/video1) using this i2c device
– Run dmabuftest with ‘1280×720@YUYV’ as capture format
Capturing from OV10635 Vision board camera
Linux kernel driver for OV1063x cameras support OV10635 sensor.
Video capture can be verified from the OV10635 sensor as follows
• Connect OV10635 sensor to the OVcam port on the Vision board
• Change the SW3 switch setting on Vision board as SW3[1-8] = 01010101
• Reboot the board and enable i2c2 as given above
• I2C device on Bus 2 slave address 0x30 should be probed successfully
• VIP should register a V4L2 video device (e.g. /dev/video1) using this i2c device
• Run dmabuftest with ‘1280×720@YUYV’ as capture format
Capturing through TVP decoder
Linux kernel supports TVP5158 NTSC/PAL decoder.
TVP5158 decoder is a TI chip which can decode upto 4 channels of NTSC/PAL analog video and multiplex it.
Video capture from 1 channel TVP5158 can be verified as follows.
• Connect analog camera to the Vin1 port of the JAMR3 board
• Change the SW2 switch setting on JAMR board as SW2[1-2] = [OFF, ON] – This is to select i2c4 for the IO
expander
• Reboot the board and enable i2c2 as given above
• I2C device on Bus 2 slave address 0x58 should be probed successfully
• VIP should register a V4L2 video device (e.g. /dev/video1) using this i2c device
• Run dmabuftest with capture format of the analog camera (e.g. ‘720×240@YUYV’)
Capturing through LVDS camera
LVDS camera is also a camera connected through a serializer and deserializer Linux kernel has driver for FPDlink serializers and deserializers
For interfacing every LVDS camera with J6, an I2C slave for ser, deser and camera is needed. By default, all of the device tree nodes are disabled.
Following table shows mapping between all LVDS cameras on multi deserializer duaghter card for Vision Board.
| LVDS camera | Camera address alias | Serializer address alias | Derializer address | VIP port |
|---|---|---|---|---|
| cam1 | 0x38 | 0x74 | 0x60 | Vin1a(VIP1 slice0 port A) |
| cam2 | 0x39 | 0x75 | 0x64 | Vin2a(VIP1 slice1 port A) |
| cam3 | 0x3A | 0x76 | 0x68 | Vin3a(VIP2 slice0 port A) |
| cam4 | 0x3B | 0x77 | 0x6C | Vin5a(VIP3 slice0 port A) |
| cam5 | 0x3C | 0x78 | 0x61 | Vin4b(VIP2 slice1 port B) |
| cam6 | 0x3D | 0x79 | 0x69 | Vin6a(VIP3 slice1 port A) |
Video capture from LVDS camera can be verified as follows.
• Connect a LVDS camera to cam1/2/3/4 port of Multides board.
• Change the SW3 switch setting on Vision board as SW3[1-8] = 00100101
• Apply attached patches on linux kernel and rebuild
• I2C device on Bus 2 slave address (e.g. 0x38 for cam1) should be probed successfully
• VIP should register a V4L2 video device (e.g. /dev/video1) using this i2c device
• Run dmabuftest with ‘1280×720@YUYV’ as capture format
Capturing through OV10640/OV490 CSI camera/ISP
Linux kernel supports CSI capture from OV10640 RAW camera and OV490 ISP.
CAL works on the CSI2 protocol and supports both raw and YUYV capture. It is verified with the OV10640 raw camera and OV490 ISP. TI-EVM has support for capturing via two CSI phys. – phy0 (4data lanes) and phy1 (2data lanes)
Video capture from OV490 can be verified as follows.
• Connect OV10640 camera to the OV490 board
• Connect the OV490 board to the TI-EVM via the CSI2 dual 490 adaptor board
• I2C device on Bus 4 slave address 0x24 should be probed successfully
• VIP should register a V4L2 video device (e.g. /dev/video1) using this i2c device
Processor SDK Linux Automotive Software Developers Guide 20
• Run dmabuftest with capture format of the analog camera (e.g. ‘1280×720@YUYV’)
Video Processing Engine(VPE)
VPE supports Scalar, Colour Space Conversion and Deinterlace.It uses V4L2 mem2mem API.
- Supported Input formats: nv12, yuyv, uyvy
- Supported Output formats: nv12, yuyv,uyvy, rgb24, bgr24, argb24, abgr24
- Not Supported formats: yuv444, yvyu, vyuy, nv16, nv61, nv21
File to File
testvpe
Usage:
<SRCfilename> <SRCWidth> <SRCHeight> <SRCFormat> <DSTfilename> <DSTWidth> <DSTHeight> <DSTformat> <interlace> <translen>
Note:
input height.
Deinterlace(DI):-
target# testvpe frame-176-144-nv12-inp.yuv 176 144 nv12 progressive_output.nv12 176 288 nv12 1 1
Scalar(SC):-
target# testvpe frame-176-144-nv12-inp.yuv 176 144 nv12 frame-1920-1080-nv12-out.nv12 1920 1080 nv12 0 1
Colour Space Conversion(CSC):-
target# testvpe frame-720-240-yuyv-inp.yuv 720 240 yuyv frame-720-240-argb32-out.argb32 720 240 argb32 0 1
SC+CSC+DI:-
target# testvpe frame-720-240-yuyv-inp.yuv 720 240 yuyv frame-1920-1080-rgb24-dei-out.rgb24 1920 1080 rgb24 1 1
File to Display
filevpedisplay
Usage:
<src_filename> <src_w> <src_h> <src_format> <dst_w> <dst_h> <dst_format> <top> <left> <w> <h> <inter> <trans> -s <conn_id>:<mode>
Input without crop:
target# filevpedisplay frame-176-144-nv12-inp.yuv 176 144 nv12 800 480 yuyv 0 0 176 144 0 1 -s 4:800x480
Input with crop:
target# filevpedisplay frame-176-144-nv12-inp.yuv 176 144 nv12 800 480 yuyv 16 32 128 128 0 1 -s 4:800x480
VIP-VPE-Display
Camera captures the frames, which are processed by VPE(SC, CSC, Dei) then displays on LCD/HDMI.
capturevpedisplay
Usage:
<src_w> <src_h> <src_format> <dst_w> <dst_h> <dst_format> <inter> <trans> -s <conn_id>:<mode>
target# capturevpedisplay 640 480 yuyv 320 240 uyvy 0 1 -s 4:640×480
Running DSS application
DSS applications are omapdrm based. These will demonstrate the clone mode, extended mode, overlay window,z-order and alpha blending features. To demonstrate clone and extended mode, HDMI display must be connected to board. Application requires the supported mode information of connected displays and plane ids. One can get these information by running the modetest application in the filesystem.
target # modetest
Running drmclone application
This displays same test pattern on both LCD and HDMI (clone). Overlay window also displayed on LCD. To test
clone mode, execute the following command:
target # drmclone -l <lcd_w>x<lcd_h> -p <plane_w>x<plane_h>:<x>+<y> -h <hdmi_w>x<hdmi_h>
e.g.: target # drmclone -l 1280x800 -p 320x240:0+0 -h 640x480
We can change position of overlay window by changing x+y values. eg. 240+120 will show @ center
Running drmextended application
This displays different test pattern on LCD and HDMI. Overlay window also displayed on LCD. To test extended mode, execute the following command:
target # drmextended -l <lcd_w>x<lcd_h> -p <plane_w>x<plane_h>:<x>+<y> -h <hdmi_w>x<hdmi_h>
e.g.: target # drmextended -l 1280x800 -p 320x240:0+0 -h 640x480
Running drmzalpha application
Z-order:
It determines, which overlay window appears on top of the other.
Range: 0 to 3
lowest value for bottom
highest value for top
Alpha Blend:
It determines transparency level of image as a result of both global alpha & pre multiplied alpha value.
Global alpha range: 0 to 255
0 – fully transparent
127 – semi transparent
255 – fully opaque
Pre multipled alpha value: 0 or 1
Processor SDK Linux Automotive Software Developers Guide 22
0 – source is not premultiply with alpha
1 – source is premultiply with alpha
To test drmzalpha, execute the following command:
target # drmzalpha -s <crtc_w>x<crtc_h> -w <plane1_id>:<z_val>:<glo_alpha>:<pre_mul_alpha> -w <plane2_id>:<z_val>:<glo_alpha>:<pre_mul_alpha>
e.g.: target # drmzalpha -s 1280x800 -w 19:1:255:1 -w 20:2:255:1
Testing with FPDLink Display setup
H/W setup
• DRA7xx EVM + 12V supply for the EVM.
• FPDLink Cable between DRA7xx and De-serilzer board (DS90UB928Q).
• 5V power supply for De-serializer board.
• LCD Adapter board (DS90UH928Q) that sits on De-serializer board.
• LCD Adapter cable which is between LCD panel and the Adapter board.
• 12V power supply for LCD Adapter board.
• The actual LCD panel (LG101(10.1in) or AUO (7.1 in))
• Please refer the picture which shows the overall setup attached here
Kernel Config shouldn’t be necessary since the supported panels and fpdlink are built into the kernel
Kernel Configuration
-> Device Drivers
-> Graphics support
-> OMAP2+ Display Subsystem support (OMAP2_DSS [=y])
-> OMAP2/3 Display Device Drivers
<*> LG101 LCD Panel
-> Device Drivers
-> Graphics support
-> OMAP2+ Display Subsystem support (OMAP2_DSS [=y])
-> OMAP2/3 Display Device Drivers
<*> SERLINK
-> Device Drivers
-> Graphics support
-> OMAP2+ Display Subsystem support (OMAP2_DSS [=y])
-> OMAP2/3 Display Device Drivers
<*> DSERLINK
Make CONFIG_NUM_CRTCS=3 and add the following line to dra7-evm.dts and dra7xx-evm-lcd10.dtsi (or
dra7x-evm-lcd-osd.dtsi)
aliases {
display0 = &hdmi;
display1 = &fpd_disp;
...
Change status of disp_ser in dra7-evm.dts from “disabled” to “ok”
disp_ser: serializer@1b {
status = "ok";
Processor SDK Linux Automotive Software Developers Guide 23
...
Testing
On power-cycle, the EVM should boot and show the Tux Logo.
Given below is a snippet from the modetest – it should show the FPDLink Display connected via the following path.
The connector-id, crtc-id or plane-ids may change based on the release or kernel version.
Encoders:
id crtc type possible crtcs possible clones
3 10 TMDS 0x00000003 0x00000000
11 14 TMDS 0x00000002 0x00000000 ---------------------------> FPDLink
Connectors:
id encoder status type size (mm) modes encoders
4 3 connected unknown 0x0 1 3
modes:
name refresh (Hz) hdisp hss hse htot vdisp vss vse vtot)
800x480 60 800 841 890 931 480 493 497 526 flags: nhsync, nvsync; type: preferred, driver
props:
1 EDID:
flags: immutable blob
blobs:
value: 2 DPMS:
flags: enum
enums: On=0 Standby=1 Suspend=2 Off=3
value: 0
12 11 connected unknown 0x0 1 11 -----------> FPDLink
modes:
name refresh (Hz) hdisp hss hse htot vdisp vss vse vtot)
1280x800 57 1280 1312 1360 1440 800 804 807 814 flags: nhsync, nvsync; type: preferred, driver
props:
1 EDID:
flags: immutable blob
blobs:
value: 2 DPMS:
flags: enum
enums: On=0 Standby=1 Suspend=2 Off=3
value: 0
Configuring OSD 10.1 Display (Rev H board)
Set device tree file to “dra7-evm-lcd-osd.dtb”
Enable EDT_FT5x06 Touchscreen Controller
Edit file
board-support/linux/ti_config_fragments/connectivity.cfg
-CONFIG_TOUCHSCREEN_EDT_FT5X06=m + CONFIG_TOUCHSCREEN_EDT_FT5X06=y
Gsttestplayer
gsttestplayer is a gstreamer test application useful for testing some features not testable with gst-launch-1.0 such as:
1. Seek – Seeking to random points in a stream
2. Trick play – Playback at different speeds (fast forward, rewind)
3. Pause, Resume
4. Playing multiple streams simultaneously in the same process, in a loop or one after another.
Running gsttestplayer
Command line options:
target # gsttestplayer -h
Usage: gsttestplayer <options>
-s <sinkname> Specify the video sink name to be used, default: kmssink
-n Do not use VPE, implies no scaling
-r <width>x<height> Resize the output to widthxheight, no scaling if left blank
-a Play with no A/V Sync
-c <cmds file> Non-interactive mode, reading commands from <cmds file>
--help-gst Show GStreamer Options
Example: To use waylandsink and resize the output video to 800×400.
target # gsttestplayer -s waylandsink -r 800x400
In normal mode, when -c option is not used, the application enters an command prompt at which the user enter
various commands. Type “help” to print out the list of possible commands:
target # gsttestplayer -s waylandsink -r 800x400
Scaling output to 800x400
Using videosink=waylandsink
<Enter ip> help
Commands available:
start <instance num> <filename/capture device>
stop <instance num>
pause <instance num>
resume <instance num>
seek <instance num> <seek to time in seconds> <optional: playback speed>
sleep <sleep time in seconds>
msleep <sleep time in milliseconds>
rewind <line number>
exit
<Enter ip>
Example commands:
start 0 KMS_MPEG4_D1.MP4 # Start playing the file "KMS_MPEG4_D1.MP4", using instance 0.
start 1 NTSC_h264.mp4 # Start playing the file "NTSC_h264.mp4" (simultaneously) using instance 1.
stop 0 # Stop playback of instance 0.
seek 0 0 2 # Seek to "0"th second mark of the stream playing in instance 0,
# and start playing back at speed 2x.
seek 0 300 -1 # Seek to "300"th second mark of the stream playing in instance 0,
# and start playing back reserve at speed 1x.
start 2 /dev/video1 # Start capturing from /dev/video1 using the v4l2src plugin
All these commands could be put into a text file and given as input to gsttestplayer with the “-c” option. In this case, gsttestplayer runs non-interactively, reading commands from the text file one line after another. The commands sleep and rewind are useful for this mode, to introduce delays or to create a loop respectively.
Notes:
1. This application plays video only. Audio path is not used.
2. The input filename should have the correct file extension to indicate the type of file. The supported extensions are
“mp4”, “avi”, “ts”, “asf” & “wmv”.
3. The input filename should contain the string “264”, “mpeg2”, “mpeg4” or “vc1″/”wmv” to indicate which video
codec should be used for decoding – H.264, MPEG-2, MPEG-4 or Windows Media Video.
4. If the input filename is a video device which matches /dev/videoX pattern, v4l2src plugin would be used for
video capture instead of playback.
5. Decode and capture can be run in parallel depending on the sink being used.
Running IPC examples
Processor SDK Linux Automotive includes IPC examples are part of the target filesystem.
User space sample application
MessageQ is the user space API provided for IPC. The sample application for MessageQ consists of an application
“MessageQApp” running on the A15 and corresponding binaries running on the remotecore. The below table shows
the paths under which the remotecore binaries can be found on the target filesystem. To ensure that these binaries are
loaded by the kernel, please symbolic link them to the location shown in the below table.
| Core | Binary path on target relative to /lib/firmware | Binary should be symlinked to |
|---|---|---|
| DSP2 | ./ipc/ti_platforms_evmDRA7XX_dsp1/messageq_single.xe66 | /lib/firmware/dra7-dsp2-fw.xe66 |
| IPU2 | ./ipc/ti_platforms_evmDRA7XX_ipu2/messageq_single.xem4 | /lib/firmware/dra7-ipu2-fw.xem4 |
| IPU1 | ./ipc/ti_platforms_evmDRA7XX_ipu1/messageq_single.xem4 | /lib/firmware/dra7-ipu1-fw.xem4 |
| DSP1 | ./ipc/ti_platforms_evmDRA7XX_dsp2/messageq_single.xe66 | /lib/firmware/dra7-dsp1-fw.xe66 |
Boot the target and ensure that the lad daemon is running. Use the ps command to check if the lad daemon is
already running. If not, please start the lad daemon.
target # /usr/bin/lad_dra7xx
Start the MessageQApp
target # /usr/bin/MessageQApp <number of messages> <core id>
The core id to be used is 1,2,3,4 for IPU2,IPU1,DSP2 and DSP1 respectively.
target # MessageQApp 10 3
Using numLoops: 10; procId : 3
Entered MessageQApp_execute
Local MessageQId: 0x80
Remote queueId [0x30080]
Exchanging 10 messages with remote processor DSP2...
MessageQ_get #1 Msg = 0xb6400468
MessageQ_get #2 Msg = 0xb6400468
MessageQ_get #3 Msg = 0xb6400468
MessageQ_get #4 Msg = 0xb6400468
MessageQ_get #5 Msg = 0xb6400468
MessageQ_get #6 Msg = 0xb6400468
MessageQ_get #7 Msg = 0xb6400468
MessageQ_get #8 Msg = 0xb6400468
MessageQ_get #9 Msg = 0xb6400468
MessageQ_get #10 Msg = 0xb6400468
Exchanged 10 messages with remote processor DSP2
Sample application successfully completed!
Leaving MessageQApp_execute
RPMsg client sample application
RPMsg is the kernel space IPC and the building block for the user space MessageQ IPC API. The below wiki page
illustrates how to build and run an rpmsg Linux kernel space client to communicate with a slave processor (e.g. DSP,
IPU, etc) using IPC’s RPMsg module.
RPMsg_Kernel_Client_ApplicationRPMsg is the kernel space IPC and the building block for the user space MessageQ IPC API. The below wiki page
illustrates how to build and run an rpmsg Linux kernel space client to communicate with a slave processor (e.g. DSP,
IPU, etc) using IPC’s RPMsg module.
RPMsg_Kernel_Client_Application.
Running basic Wifi tests
To run this test, you would need to have the Wilink COM module connected to the EVM.
Check if the wlan0 interface is showing up:
target # ifconfig -a
Bring up the wlan0 inteface:
target # ifconfig wlan0 up
Search for the available Wifi networks:
target # iw wlan0 scan | grep -i ssid
Running basic Bluetooth tests
To run this test, you would need to have the Wilink COM module connected to the EVM. Make sure that this module supports the Bluetooth.
target # hciconfig hci0 up
target # hciconfig -a
Turn on the Bluetooth on the device that you want to pair and make it discoverable, then run the following command:
target # hcitool scan
How to bring up the GNSS driver and sample application
WL8 GNSS driver that is compatible with SDK is now available as part of the click wrap license at the following location. http://www.ti.com/tool/wilink-sw
Users are requested to register and obtain the package.
The package contains the driver source and the required documentation.
The document “Bring up manual for WiLink8 GNSS driver on Linux” is the starting point that contains the instructions for compiling and trying the sample application.
NOTE: These instructions are known to work if the user starts with the Processor SDK Linux Automotive installer and compiles the Linux kernel using the instructions provided in the Software Developers Guide.
Supported Boot modes
The same U-Boot image shipped with the SDK can be used to boot the system in following different modes based on the boot switch settings:
• MMC/SD
• eMMC
• Uart
• QSPI
• NAND
Booting EVM with different modes
Choosing the correct device tree
Booting Linux kernel needs to have a kernel image(zImage) and device tree blob(DTB) file. DTB file describes the hierarchy of the devices, and also describes various parameters about the devices.
Depending on which CPU board and application board you are trying to boot, there are different set of devices. For each permutation, there is a different DTB file describing all devices on baseboard as well as application board.
Release filesystem comes with most commonly used permutations of device trees. Following table shows the name of the device tree file to be used for each of the combination.
Device tree file (DTB) options
| Board | DTB name |
|---|---|
| J6 EVM + 10inch LG LCD | dra7-evm-lcd-lg.dtb (Base board DTB works for JAMR3 as well) |
| J6 EVM + 10inch OSD LCD | dra7-evm-lcd-osd.dtb (Base board DTB works for JAMR3 as well) |
| J6 EVM + Vision app board | dra7-evm-vision.dtb |
| J6eco EVM upto revision C | dra72-evm-revc.dtb (Base board DTB works for JAMR3 as well) |
| J6eco EVM revision C + 10inch standard OSD panel | dra72-evm-revc-lcd-osd101t2045.dtb (Base board DTB works for JAMR3 as well) |
| J6eco EVM revision C + 10inch new OSD panel | dra72-evm-revc-lcd-osd101t2587.dtb (Base board DTB works for JAMR3 as well) |
| J6entry EVM | dra71-evm.dtb (Base board DTB works for JAMR3 as well) |
| J6entry EVM + 10inch AUO LCD panel | dra71-evm-lcd-auo-g101evn01.0.dtb (Base board DTB works for JAMR3 as well) |
U-boot tries to detect the platform (J6/J6eco) and choose the appropriate dtb. For the DRA7xx boards, it only detects the baseboard. If you are using a stackup of baseboard plus LCD, then you must manually specify the proper dtb by defining fdtfile inside uenv.txt. For example, if you were using J6 EVM + 10inch LG LCD then you would add the following line to uenv.txt:
fdtfile=dra7-evm-lcd-lg.dtb
Identifying LCD Panels
| Panel name | Resolution | Images | Remarks |
|---|---|---|---|
| LG LCD | 1280×800 | Media:lcd-lg-front.jpg, Media:lcd-lg-reverse.jpg, Media:lcd-lg-push-button.jpg | Has a circular hole on top for mounting camera. Has push buttons on the side. Device tree files end with lcd-lg.dtb |
| OSD Panel | 1920×1200 | Media:osd_2045_forward.jpg, Media:osd_2045_reverse.jpg | Back has notation “OSD 1080P Touch Display I/F board”. Serial number start with OSD_. Device tree files end with lcd-osd.dtb or lcd-osd101t2045.dtb |
| New OSD Panel | 1920×1200 | TBD | Device tree files end with lcd-osd101t2587.dtb |
| AUO Panel | 1280×800 | Media:Auo-forward.jpg, Media:Auo_notation_reverse.jpg | Serial numbers on the back start with AU. Device tree files have AUO in name. |
Choosing the correct bootloader config
All the padmux is performed from the first stage bootloader (MLO) instead of the kernel. Also, the bootloader takes care of configuring all the required PADs as well as the IOdelay configuration. Depending on the use case, boot loader needs to know the required pads to be configured at the boot time. Kernel would know the device information from DTB file but bootloader does not know which DTB file would be used by the kernel.
For this reason, we need to rebuild the bootloader (MLO) with a different config so that the appropriate PAD and IODELAY configuration is performed as required by the use case. Release prebuilt bootloader works for most of the use cases.
When using vision use cases, rebuild the bootloader with extra Kconfig CONFIG_TARGET_DRA7XX_EVM_VISION enabled.
Using QSPI Boot
QSPI is a serial peripheral interface like SPI the major difference being the support for Quad read, uses 4 data lines for read compared to 2 lines used by the traditional SPI.
Supported boot modes
• QSPI Production Boot mode
• QSPI Development Boot mode
QSPI Production Boot Mode
This boot mode also called as ‘spl_early_boot’ in Processor SDK Linux Automotive.
• In this boot mode SPL(first stage of Uboot) directly boots the Linux kernel.
• The executables – MLO, dtb, uImage & IPU are stored in QSPI flash memory. Refer the “Memory Layout” section for offset details.
• eMMC partition 2 contains the filesystem & is mounted as rootfs.
• Build MLO and u-boot.img.
Note: By continuous press of character ‘c’ key on keyboard and resetting the EVM, SPL will load u-boot and enter into second stage boot loader.
SYS BOOT Switch Settings
change the boot switches to QSPI boot mode as:
SW2[5:0] = 110110 for development (jump to u-boot)
SW2[5:0] = 110111 for production mode (jump to MLO->kernel)
Also
SW5.4 = 0 (OFF)
Memory Layout of QSPI Flash
+----------------+ 0x00000
| MLO |
| |
+----------------+ 0x040000
| u-boot.img |
| |
+----------------+ 0x140000
| DRA7-evm.dtb |
+----------------+ 0x1c0000
| u-boot env |
+----------------+ 0x1d0000
| u-boot env |
| (backup) |
+----------------+ 0x1e0000
| |
| uImage |
| |
| |
+----------------+ 0x9e0000
| |
Processor SDK Linux Automotive Software Developers Guide 30
| IPU exe |
| |
+----------------+
Build kernel uImage for QSPI boot mode
In ‘spl_early_boot’ mode, While building kernel uImage, set CONFIG_CMDLINE bootargs appropriately and set rootfs=/dev/mmcblk0p2 (eMMC device) and set kernel CONFIG_CMDLINE bootargs @menuconfig->Boot options->”Default kernel command string”.
for example:
CONFIG_CMDLINE = "elevator=noop console=ttyO0,115200n8
root=<rootfs> rw rootwait earlyprintk fixrtc omapdrm.num_crtc=2
consoleblank=0 cma=64M rootfstype=ext4";
where <rootfs> can be /dev/mmcblk0p2 for eMMC or /dev/mmcblk1p2
for mmc/sd.
alternate option: Add the bootargs in chosen node in DTB file, using fdtput utility.
#fdtput -v -t s <DTB-FILE-PATH> "/chosen" bootargs
"elevator=noop console=ttyO0,115200n8 root=<rootfs> rw rootwait
earlyprintk fixrtc omapdrm.num_crtc=2 consoleblank=0 cma=64M
rootfstype=ext4"
where <rootfs> can be /dev/mmcblk0p2 for eMMC or /dev/mmcblk1p2
for mmc/sd.
Flashing the Image to QSPI from Linux kernel
The mk-qspi-boot.sh will be available at home direcotry of “Processor SDK Linux Automotive” filesystem. The mk-qspi-boot.sh runs on the target. The scripts reads the binaries (MLO, u-boot.img, kernel/dtb, ipu images) from MMC/SD and flashes into QSPI flash memory at appropriate partition described in above table. Also format &creates rootfs partition in eMMC and copies the filesystem the eMMC rootfs parition.
Usage:
target#mk-qspi-boot.sh --device1 --device2 --bootmode
--device1 - devfs entry for qspi flash as char device node
e.g /dev/mtd
--device2 - devfs entry for eMMC flash as block device node
e.g /dev/mmcblk0
--bootmode - 'spl_early_boot' & 'two_stage_boot'
spl_early_boot - ROM=>SPL=>uImage
two_stage_boot - ROM=>SPL=>u-boot.img=>uImage
e.g
target#: mk-qspi-boot.sh --device1 /dev/mtd --device2 /dev/mmcblk0 --bootmode spl_early_boot
Note: Defualt DTB is dra7-evm-lcd-osd.dtb. Do change DTB in script if you want to boot the board with different device tree.
You can modify the mk-qspi-boot.sh for custom changes (like change of DTB file etc).
QSPI Development Boot Mode
This boot mode also called as ‘two_stage_boot’ in “Processor SDK Linux Automotive”
• In this boot mode SPL(first stage of Uboot) brings-up u-boot.img(second part of u-boot).
• The second stage of u-boot, then, loads & boots the Linux kernel.
• Build MLO and u-boot.img for QSPI boot mode (use “dra7xx_evm_defconfig”).
• Only MLO and u-boot.img are stored in QSPI flash memory.
• DTB & uImage are stored in MMC/SD boot parition.
• The rootfs is mounted from MMC/SD partition 2
Note: By continuous press of character ‘c’ key on keyboard and resetting the EVM, SPL will load u-boot and enter into second stage boot loader.
SYS BOOT Switch Settings
change the boot switches to QSPI boot mode as:
SW2[5:0] = 100110 for development (jump to u-boot)
SW2[5:0] = 100111 for production mode (single stage)
Also,
SW5.4 = 0 (OFF)
Flashing UBoot to QSPI from Linux kernel
The mk-qspi-boot.sh will be available at home direcotry of Processor SDK Linux Automotive filesystem. The mk-qspi-boot.sh runs on the target . It reads the executables from MMC/SD & flashes into QSPI flash memory for ‘two_stage_boot’
target#mk-qspi-boot.sh --device1 --device2 --bootmode
--device1 - devfs entry for qspi flash as char device node
e.g /dev/mtd
--device2 - devfs entry for eMMC flash as block device node
e.g /dev/mmcblk0
--bootmode - 'spl_early_boot' & 'two_stage_boot'
spl_early_boot - ROM=>SPL=>uImage
two_stage_boot - ROM=>SPL=>u-boot.img=>uImage
e.g
target#: mk-qspi-boot.sh --device1 /dev/mtd --device2 /dev/mmcblk0 --bootmode two_stage_boot
Note: Defualt DTB is dra7-evm-lcd-osd.dtb. Do change DTB in script if you want to boot the board with different device tree.
Also QSPI flash can be accessed from kernel through MTD device.
Following commands demonstrate how to flash MLO and u-boot from SD card to QSPI.
target# mount /dev/mmcblk1p1 /mnt
target# cat /proc/mtd
target# flash_erase -N /dev/mtd0 0 4
target# flash_erase -N /dev/mtd1 0 16
target# mtd_debug write /dev/mtd0 0x0 (ls -l /mnt/MLO {{!}} awk '{ print5 }') /mnt/MLO
target# mtd_debug write /dev/mtd1 0x0 (ls -l /mnt/u-boot.img {{!}} awk '{ print5 }') /mnt/u-boot.img
Processor SDK Linux Autom
Flashing UBoot to QSPI from UBoot
UBoot can be flashed to QSPI from any boot mode, here we are using MMC boot mode as an example Boot from SD card and at the U-Boot prompt, Choose the mmc card to read from
uboot# mmc dev 0
uboot# mmc0 is current device
• Probe the flash to see if there is a device connected,
NOTE: Do NOT copy-paste the commands, please type them at the prompt
# sf probe 0
SF: Detected S25FL256S with page size 64 KiB, total 32 MiB, mapped at 5c000000
• Erase flash before writing the bin
NOTE: Do NOT copy-paste the commands, please type them at the prompt
# sf erase <offset> +<size>
where,
offset - qspi flash offset address start from 0.
size - erase region size. This is automatically rounded up to the block size.
# sf write <memory address> <offset> <len>
where,
memory-address - source data to read from
offset - qspi offset location to write
len - length of the data to be written
• Load the MLO (to offset 0x0) and u-boot.img(to offset 0x40000) to QSPI flash
Flashing MLO to QSPI
uboot# fatload mmc 0 0x82000000 MLO.qspi
uboot# sf erase 0 +{filesize}
uboot# sf write 0x82000000 0x0{filesize}
Flashing u-boot.img to QSPI
uboot# fatload mmc 0 0x83000000 u-boot.img
uboot# sf erase 40000 60000;
uboot# sf write 0x83000000 0x40000 0x60000
• Build kernel to create uImage and dtb files. Load the uImage(to offset 0x1e0000) and DTB files (to offset 0x140000) to QSPI flash
Flashing kernel(uImage) to QSPI
uboot# fatload mmc 0 0x82000000 uImage
uboot# sf erase 1e0000 500000;
uboot# sf write 82000000 1e0000 500000;
Flashing dtb file to QSPI
uboot# fatload mmc 0 0x83000000 <dtb file>
uboot# sf erase 140000 20000;
uboot# sf write 0x83000000 0x140000 0x20000
After reset, you should be able to boot from QSPI flash.
Using eMMC Boot
eMMC boot method is the same as MMC/SD flashing and booting procedure. Format MMC/SD card and create boot/rootfs partition (refer to Software development setup section).
$ sudo ${INSTALL_DIR}/bin/mksdboot.sh --device /dev/sdY --sdk ${INSTALL_DIR}
The MMC/SD boot partition contains MLO,u-boot.img and uenv.txt. The rootfs partition contains ext4 file system with kernel & DTBs in rootfs/boot directory. Formatting of eMMC device can be done by using mk-eMMC-boot.sh script. Boot the EVM with MMC/SD boot mode first and then run the mk-eMMC-boot.sh script to create the eMMC partitions.
Partitioning and formatting eMMC
• Boot kernel using MMC/SD boot mode (refer to Software development setup section).
• mount the MMC/SD boot partition, create uenv-emmc.txt.
# mount /dev/mmcblk1p1 /mnt
# cp /mnt/uenv.txt /mnt/uenv-emmc.txt
Note: edit /mnt/uenv-emmc.txt and set rootfs=/dev/mmcblk0p2 in bootargs.
• consider /dev/mmcblk0 is eMMC device, then use the following script to partition the eMMC device.The mk-eMMC-boot.sh will create boot, rootfs partition on eMMC device and copy the all files from MMC/SD card to eMMC.
#./mk-eMMC-boot.sh --device /dev/mmcblk0
• reboot EVM to u-boot prompt, and set environment variables “mmcdev” and the “bootpart” point to eMMC device.
#env default -a
#setenv bootpart 1:2
#setenv mmcdev 1
#saveenv
• Set boot switches to EMMC boot mode.
Set SW2[7..0] = 00111000
• Remove MMC/SD card and reboot the EVM, now should boot from eMMC device.
Partitioning and formatting eMMC from Host PC through USB
The ums (USB Mass Storage feature) command in U-Boot can be used to expose the MMC as USB storage drive (/dev/sdX) to Host PC.
Syntax: ums
• Setup: Connect USB0 port of the EVM to Ubuntu host PC.
From U-boot prompt
• Exposing MMC/SD as storage media
=> ums 0 mmc 0
• Exposing eMMC as storage media
=> ums 0 mmc 1
This command will expose the mmc device as storage device (/dev/sdX) to Ubuntu PC, further user can mount and partition, format and copy the files to device.
Booting Secure DRA7xx Devices SD/eMMC
The Processor SDK Linux Automotive u-boot supports booting on secure (HS) devices. In order to boot the secure DRA7xx devices, at minimum, the MLO binary must be signed and formatted properly with the necessary keys and certificates. Signing is done using the IFT tool (contact local sales representative to get M-Shield-Dk/IFT tools). This process can be used with test secure (HS) devices with the TI dummey key. The sequence of authentication of images for HS devices as follows. The initial authentication is done by ROM code and is required. The last 2 steps are optional and constitute extended secure boot authentication.
• ROM -> PKC – ROM code authenticates Public Key Certificate against Root eFuse key
• ROM -> PPA – ROM code authenticates Primary Protected Application and loads to secure RAM and executes
• ROM -> ISW – ROM code loads and authenticates initial software (equivalent to MLO/u-boot-spl for non-secure
devices)
• MLO -> u-boot – The MLO authenticate u-boot
• u-boot->kernel – The u-boot boot kernel
In any stage if authentication fails, then device will enter a deadloop.
Install Mshield-DK
Install Mshield-DK tool and set the path to “TI_SECURE_DEV_PKG” environment variable corresponding to the version you installed (LITE, STD, for FULL). for example
#export TI_SECURE_DEV_PKG="${HOME}/mshield-dk_std_4_5_3"
The SDK has been validated with M-Shield DK Lite/Standard version 4.5.3.
Building secure signed u-boot & MLO image
The defconfig “dra7xx_hs_evm_config” is used to compile the u-boot. Use below command to compile u-boot binaries from within the u-boot source tree
#make distclean
#make dra7xx_hs_evm_defconfig
You can also change the value of DEFAULT_UBOOT_CONFIG in the SDK Rules.mak file to be
‘dra7xx_hs_evm_defconfig’ and use the SDK top-level Makefile to rebuild and install.
#make u-boot_clean
#make u-boot
The signed binaries u-boot-spl_HS_MLO (the signed MLO file, also copied to as a file named ‘MLO’ in the same path) and u-boot.img (the main u-boot image) are created.
Copy the signed images to the SD card’s boot partition.
#cp MLO u-boot.img /media/boot/
or use the ‘u-boot_install’ make target.
Booting the HS device
Insert the SD card to secure dra7xx(HS) EVM and reset the board. If the board is not booting then check the u-boot build log for any warnings or errors. The M-Shield packages are prebuilt to work on secure devices programmed with TI dummy (development) keys. If attempting to boot a custom keyed device (that is, not a device with TI dummy keys), the M-Shield package that is being pointed to by the “TI_SECURE_DEV_PKG” environment variable should be rebuilt with the right keys/signatures for the HS device in use.
USB DRD (Dual Role Device) Support
The USB ports has capable of switching between device mode or host mode dynamically without removal of the
USB cable.
• make sure dr_mode property in DTB is set to “otg” for DRD support.
• mount the debugfs node
# mount -t debugfs debugfs /mnt
• load the gadget module.
# modprobe <gadget.ko>
• To switch between host mode for USB1 and USB2 port.
# echo "host" > /mnt/48890000.usb/mode
# echo "host" > /mnt/488d0000.usb/mode
• To switch between host mode for USB1 and USB2 port.
# echo "device" > /mnt/48890000.usb/mode
# echo "device" > /mnt/48890000.usb/mode
Note: The role switching can be done dynamically based on board USB-ID (SW1 switch on TI-EVM).
# If host-type (A-type) usb cable is connected, the port switched to host mode.
# If device-type (B-type) usb cable is connected, the port switched to device mode.
Note:
• Refer DRD support [7] for more information.
• Refer drive-vbus-using-gpio [8] in order to drive the vbus using gpio.
Device Firmware Upgrade (DFU)
The Device Firmware Upgrade (DFU) feature is used to program or flash the firmware to memory devices such as eMMC/MMCSD/QSPI/NOR/NAND/RAM devices. The u-boot has DFU support for eMMC, MMC/SD, QSPI and RAM devices.
DFU Supported Devices in u-boot
• eMMC
• MMC/SD
• RAM
• QSPI
SPL-DFU support (USB Peripheral boot mode)
This method is used to flash the binary images from ubuntu Host PC (using dfu-utils tool) to EVM over usb interface using Device Firmware Update (DFU) feature. This is used to flash the eMMC, or QSPI to fresh/factory boards.
• Use default “dra7xx_evm_defconfig” to build spl/u-boot-spl.bin, u-boot.img.
# make dra7xx_evm_defconfig
• select SPL->DFU support from menuconfig->”Boot Images”->”Enable SPL with DFU to load binares to memory device”
• Build spl/u-boot-spl.bin and u-boot.img
# make
• Set SYSBOOT SW2 switch to USB Peripheral boot mode
SW2[7..0] = 0000000
• Connect EVM Superspeed port (USB1 port) to PC (Ubuntu) through USB cable.
• From Ubuntu PC, load spl/u-boot-spl.bin to EVM. Issue below command and reset the board.
$ sudo ./usbboot -S spl/u-boot-spl.bin
• Load the u-boot.img to RAM.
# sudo dfu-util -l
Found DFU: [0451:d022] devnum=0, cfg=1, intf=0, alt=0, name="kernel"
Found DFU: [0451:d022] devnum=0, cfg=1, intf=0, alt=1, name="fdt"
Found DFU: [0451:d022] devnum=0, cfg=1, intf=0, alt=2, name="ramdisk"
$ sudo dfu-util c 1 -i 0 -a 0 -D "u-boot.img" -R
• Now EVM will boot to u-boot prompt.
DFU Usage from u-boot
Use default “dra7xx_evm_defconfig” to build MLO and u-boot.img.
make dra7xx_evm_defconfig
make
Connect the EVM with Ubuntu Host through micro USB device cable (USB1 port on EVM).
• From target u-boot prompt
Reset the default environment variable
# env default -a
set the dfu_alt_info environment variable to emmc dfu settings
# setenv dfu_alt_info ${dfu_alt_info_emmc}
• Run DFU command chosing eMMC as memory device
dfu
# dfu 0 mmc 1
set_config: high speed config #1: usb_dnload
• From PC host side (Ubuntu)
To list the available partition to load the file.
# sudo dfu-util -l
dfu-util 0.5
(C) 2005-2008 by Weston Schmidt, Harald Welte and OpenMoko Inc.
(C) 2010-2011 Tormod Volden (DfuSe support)
This program is Free Software and has ABSOLUTELY NO WARRANTY
dfu-util does currently only support DFU version 1.0
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=0, name="rawemmc"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=1, name="boot"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=2, name="rootfs"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=3, name="MLO"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=4, name="MLO.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=5, name="u-boot.img.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=6, name="spl-os-args.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=7, name="spl-os-image.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=8, name="spl-os-args"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=9, name="spl-os-image"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=10, name="u-boot.img"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=11, name="uEnv.txt
The above list shows list of dfu partitions or file name.
Flashing rawemmc partition
The “rawemmc” partition is used to update the disk image to raw eMMC sector 0. This is used to flash the eMMC to fresh boards. Please refer to section Using USB peripheral boot mode
Update existing MLO and u-boot.img of eMMC boot partition (vfat)
This method is used to update the MLO/u-boot.img files in boot(vfat) partion of eMMC
• To update the MLO file
#sudo dfu-util -D MLO -c 1 -i 0 -a 3
• To update the u-boot.img file
#sudo dfu-util -D u-boot.img -c 1 -i 0 -a 10
Update existing boot and rootfs partition (ext4) of eMMC
This method is used to update the existing boot partition (vfat) or rootfs partition (ext4) images in eMMC
• To update the boot partition disk image
#sudo dfu-util -D boot.img -c 1 -i 0 -a 1
• To update the rootfs partition disk image
#sudo dfu-util -D rootfs.img -c 1 -i 0 -a 2
Steps to create boot (vfat) and rootfs (ext4) raw disk image
In order to create raw boot and rootfs disk image along with partition structure follow the steps mentioned below.
1. Insert the mmc/sd card to Ubuntu PC, the size of sdcard shall be less than or equal to size of eMMC populated in EVM.
2. Prepare mmc/sd card mentioned in step4 of Processor SDK Linux Automotive User Guide [9]. By default the mksdboot.sh script creates two partitions, 64MB of boot partition with vfat filesystem and remaining size of sd card with rootfs partition with ext4 filesystem.
3. After sd card is created with boot and rootfs partition, mount the sdcard to update uenv.txt and overwrite uenv.txt by uenv-emmc.txt in “boot” partition. Copy any additional files required to boot/rootfs partition.
#mount /dev/sdY1 /media/boot
#mount /dev/sdY2 /media/rootfs
#cp /media/boot/uenv-emmc.c /media/boot/uenv.txt
#sync
Create the vfat raw boot disk image along with partition structure
#dd if=/dev/<sdY1> of=boot.img bs=1M &
Create the ext4 raw rootfs disk image along with partition structure
#dd if=/dev/<sdY2> of=rootfs.img bs=1M &
Flashing/Upgrading custom file to eMMC vfat/ext4 partition
- If eMMC device is not partitoned, then first the eMMC device must be partitioned to vfat or ext4 filesystem. ReferFlashing_binaries_to_the_factory_boards_using_Device_Firmware_Upgrade_(DFU).pdf
- Once the eMMC device has been partitioned to vfat/ext4, user can write into existing file or partition. For example as shown in above list of dfu interfaces, user can upgrade the MLO or u-boot.img file from ubuntu PC. The alternate interface number 3 need to selected for upgrading MLO, and 10 need to selected for u-boot.img.
#sudo dfu-util -D MLO -c 1 -i 0 -a 3
#sudo dfu-util -D u-boot.img -c 1 -i 0 -a 10
Adding a custom file to DFU configuration
- Add a custom/new file to the root directory of existing rootfs(ext4) or boot(vfat) partition of sdcard and flash the filesystem raw disk images to eMMC. For example consider the file zImage, dra7-evm-lcd10.dtb is added to root directory of rootfs(ext4) and README file in boot(vfat) partition. Make sure file read/write permission is set.
- From u-boot source repository, edit the file include/configs/dra7xx_evm.h, add a new entry to DFU_ALT_INFO_EMMC macro.
#define DFU_ALT_INFO_EMMC \
"dfu_alt_info_emmc=" \
"rawemmc raw 0 3751936;" \
"boot part 1 1;" \
"rootfs part 1 2;" \
"MLO fat 1 1;" \
"ZImage ext4 1 2;" \
"dra7-evm-lcd10.dtb ext4 1 2;" \
"README fat 1 1;" \
"MLO.raw raw 0x100 0x100;" \
Processor SDK Linux Automotive Software Developers Guide 39
"u-boot.img.raw raw 0x200 0x400;" \
"spl-os-args.raw raw 0x80 0x80;" \
"spl-os-image.raw raw 0x900 0x2000;" \
"spl-os-args fat 1 1;" \
"spl-os-image fat 1 1;" \
"u-boot.img fat 1 1;" \
"uEnv.txt fat 1 1\0"
- Compile the u-boot to create MLO, u-boot.img and load MLO/u-boot.img to EVM
- Execute “dfu-util -l” from PC host, the new interfaces added will be listed
#sudo dfu-util -l
dfu-util 0.5
(C) 2005-2008 by Weston Schmidt, Harald Welte and OpenMoko Inc.
(C) 2010-2011 Tormod Volden (DfuSe support)
This program is Free Software and has ABSOLUTELY NO WARRANTY
dfu-util does currently only support DFU version 1.0
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=0, name="rawemmc"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=1, name="boot"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=2, name="rootfs"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=3, name="MLO"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=4, name="zImage"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=5, name="dra7-evm-lcd10.dtb"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=6, name="README"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=7, name="MLO.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=8, name="u-boot.img.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=9, name="spl-os-args.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=10, name="spl-os-image.raw"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=11, name="spl-os-args"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=12, name="spl-os-image"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=13, name="u-boot.img"
Found DFU: [0403:bd00] devnum=0, cfg=1, intf=0, alt=14, name="uEnv.txt"
- Now you can upgrade the new added files by chosing the respective alternate interface number (alt=
)
Upgrade the zImage
#sudo dfu-util -c 1 -i 0 -a 4 -D zImage
Upgrade the dra7-evm-lcd10.dtb
#sudo dfu-util -c 1 -i 0 -a 4 -D dra7-evm-lcd10.dtb
Using dfu-utils from Windows PC Host
Follow the below steps to install the dfu-tools.
1. Download the dfu-utils dfu-util-0.6-win32.zip.bz2 [11] and unzip to some folder (say c:\dfu)
2. Download the zdaig application [12] and execute zdaig application, install winUSB driver.
3. Go through DFU section mentioned above for setup the EVM for DFU.
Note: For flashing fresh boards, refer to peripheral usb boot mode section.
The initial bootloader (u-boot-spl.bin) shall be loaded to EVM using usbboot tool running from Ubuntu Host PC.
For more details refer Flashing_binaries_to_the_factory_boards_using_Device_Firmware_Upgrade_(DFU).pdf
- Connect the EVM (running DFU) to windows PC through superspeed usb cable. Windows host will detect the EVM as “download gadget”.
- Copy required binary files to flash c:\dfu folder. Open command prompt and execute dfu-util.exe.
c:\dfu> dfu-util –l
This will list all dfu interfaces
- Follow DFU section mentioned above for flashing the binaries to eMMC.
c:\dfu> dfu-util -D <file-name> -c 1 -i 0 -a <alt number>
IPC
The table below shows the remote cores and their corresponding definitions in the kernel dtsi files (${INSTALL_DIR}/board-support/linux/arch/arm/boot/dts/dra7.dtsi, and dra74x.dtsi), as well as the argument to be used in the loading/unloading commands.
| Remote Core | Definition in dtsi file | Argument in loading/unloading |
|---|---|---|
| IPU1 | ipu@58820000 | 58820000.ipu |
| IPU2 | ipu@55020000 | 55020000.ipu |
| DSP1 | dsp@40800000 | 40800000.dsp |
| DSP2 | dsp@41000000 | 41000000.dsp |
For example, the argument of 55020000.ipu corresponds to IPU2 as can be seen from dra7.dtsi.
ipu2: ipu@55020000 {
compatible = "ti,dra7-rproc-ipu";
In the sections below, 55020000.ipu will be used as the example. For a specific use case, please select the corresponding argument which is applicable.
Unloading and loading remotecores at runtime
It is possible to unload and reload a remotecore at runtime from Linux using the sysfs interface.
target cd /sys/bus/platform/drivers/omap-rproc/
target echo 55020000.ipu > unbind
target $ echo 55020000.ipu > bind
The echo 55020000.ipu > unbind command tears down the communication channels between the A15 and the remotecore and unloads the remotecore. Any application level shutdown that needs to be performed needs to be handled by the system integrator.
The echo 55020000.ipu > bind loads the appropriate firmware binary onto the remotecore.
Changing the remotecore binary at runtime
To change the remotecore binary at runtime
1. Unload the remotecore using unbind.
2. Change the remotecore binary in the firmware folder. Default location is /lib/firmware on the target
filesystem.
3. Load the remotecore using bind.
target cd /sys/bus/platform/drivers/omap-rproc/
target echo 55020000.ipu > unbind
target cp /home/root/new-binary.xem4 /lib/firmware/dra7-ipu2-fw.xem4
target echo 55020000.ipu > bind
If it is desirable to avoid overwriting the existing remote binaries, the method of symbolic links can be used instead of direct copy. For example, Processor SDK provides two types of DSP remotecore binaries: one for DSPDCE (dra7-dsp1-fw.xe66.dspdce-fw) and another one for OpenCL (dra7-dsp1-fw.xe66.opencl-monitor). dra7-dsp1-fw.xe66 is created as a symbolic link by default pointing to the OpenCL binary. When it is needed to switch to DSPDCE, the symbolic link of dra7-dsp1-fw.xe66 can be updated pointing to dra7-dsp1-fw.xe66.dspdce-fw.
target cd /sys/bus/platform/drivers/omap-rproc/
target echo 40800000.dsp > unbind
target rm /lib/firmware/dra7-dsp1-fw.xe66
target ln -s /lib/firmware/dra7-dsp1-fw.xe66.dspdce-fw /lib/firmware/dra7-dsp1-fw.xe66
target $ echo 40800000.dsp > bind
After the switch, copycodectest application can be run to verify that DSPDCE firmware is loaded. This application fills the input buffer with a number entered as the argument and after process the output buffer is tested for the same pattern.
usage: copycodectest pattern.
Example:
target # copycodectest 123
Sample console output:
root@am57xx-evm:~# copycodectest 123
0x22070: Opening Engine..
Created dsp_universalCopy
Processor SDK Linux Automotive Software Developers Guide 42
Fill input buffer with pattern 123
Verifing the UniversalCopy algorithm
copycodectest executed successfully
Loading firmware during initial boot without using udev
During the default boot, firmware is supplied to the kernel by udev. Starting the udev service on boot causes a few seconds increase in boot time. In cases where a quick boot is required, the user may not start the udev service in boot. In such cases, firmware can be supplied to the kernel using the sysfs interface. An example script is shown below.
FW_NAMES="dra7-dsp1-fw.xe66 dra7-dsp2-fw.xe66 dra7-ipu1-fw.xem4 dra7-ipu2-fw.xem4"
for FW in FW_NAMES ; do
echo 1>/sys/class/firmware/FW/loading
cat /lib/firmware/FW>/sys/class/firmware/FW/data
echo 0 > /sys/class/firmware/$FW/loading
done
Handling the IPU bitbanding region
The ARM Cortex-M4 memory map includes a bit-banding region of memory from 0x4000:0000 to 0x400F:FFFF and 0x4200:0000 to 0x43FF:FFFF. Here is a Cortex-M4 memory map picture from ARM:
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0553a/CHDBIJJE.html
Many Vayu components running on the IPUs, including IPC, must access peripherals physically located in this bit-banding region. As a result, these accesses must be performed indirectly using a virtual memory address, mapped using the IPU’s AMMU.
Many of the components aligned on mapping this memory using one Large AMMU page that maps 512M of physical memory beginning at 0x4000:0000 to virtual memory beginning at 0x6000:0000. Then the components (by default) access the peripherals using the 0x6XXX:XXXX address space.
IPC specifics
IPC follows the convention of, by default, accessing memory physically located at 0x4XXX:XXXX using virtual memory at 0x6XXX:XXXX. You can see an example of this here – note the mailbox addresses configured here are in the 06XXX:XXXX range:
http://git.ti.com/cgit/cgit.cgi/ipc/ipcdev.git/tree/packages/ti/sdo/ipc/family/vayu/InterruptIpu.xs
In that same file, you can see that these addresses are configurable, and the default 0x6XXX:XXXX addresses are only used if other addresses haven’t already been configured by the system integrator (e.g. in a .cfg script). Users can override these default mailbox addresses using the ti.sdo.ipc.family.vayu.InterruptIpu module’s mailboxBaseAddr[] array, documented here:
http://software-dl.ti.com/dsps/dsps_public_sw/sdo_sb/targetcontent/ipc/latest/docs/cdoc/ti/sdo/ipc/family/vayu/InterruptIpu.html#metamailbox.Base.Addr
Using the Late attach functionality
To satisfy the startup time requirements of specific use cases, one may need the boot loader to boot a remote core before booting the A15 with the linux kernel. The kernel then attaches to the already booted remote core for further communication. We refer to this feature as the “Early Boot – Late Attach” functionality. The “Early Boot” functionality is provided by the boot loader. The “Late Attach” functionality is a feature of the Linux Kernel. This functionality was added to Processor SDK – Linux Automotive from v3.01.
Below are the steps required to test early boot late attach functionality.
Configuring U-Boot
Please enable the configuration option CONFIG_LATE_ATTACH in the file u-boot/configs/dra7xx_evm_defconfig.h. Change the lines
CONFIG_LATE_ATTACH=n
to
CONFIG_LATE_ATTACH=y
and rebuilt u-boot using the configuration dra7xx_evm_config.
Building the Linux kernel device tree
Linux kernel contains the “Late Attach” feature builtin. The feature is enabled or disabled on a per remote core basis at boot time through device tree node attributes. To enable “Late Attach” for a remote core, 3 attributes need to be set on the remote core and each of the timer, mmu nodes used by the remotecore. These three attributes are
1. ti,late-attach
2. ti,no-idle-on-init
3. ti,no-reset-on-init.
These three attributes together signal to the kernel that
1. Late attach feature is in use for the remote core.
2. The remotecore and other nodes have been configured and are in use before the kernel boot. These should not be reset or idled during kernel boot.
The Linux kernel tree delivered with Processor SDK Linux Automotive includes a dts file that shows how to enable late attach feature for IPU2 on a DRA7xx evm with 10″ 1280×800 LCD. This file was built in the following manner.
The timers used by IPU2 are found from $PSDKLA/board-support/linux/arch/arm/boot/dts/dra7-evm.dts
&ipu2 {
status = "okay";
memory-region = <&ipu2_cma_pool>;
mboxes = <&mailbox6 &mbox_ipu2_legacy>;
timers = <&timer3>;
watchdog-timers = <&timer4>, <&timer9>;
};
The MMU used by IPU2 can be found from $PSDKLA/board-support/linux/arch/arm/boot/dts/dra7.dtsi.
ipu2: ipu@55020000 {
compatible = "ti,dra7-rproc-ipu";
reg = <0x55020000 0x10000>;
reg-names = "l2ram";
ti,hwmods = "ipu2";
iommus = <&mmu_ipu2>;
ti,rproc-standby-info = <0x4a008920>;
status = "disabled";
};
IPU2 uses timer3 to supply the OS tick and timer4 and timer9 as watchdog timers. IPU2 also uses an MMU referred to as mmu_ipu2 in the device tree. All of these nodes including the IPU2 node need to have the three attributes described above set.
We also reserve a small region of memory(1 MB) for U-Boot to store the pagetables of the MMU’s for the DSP’s and the IPU’s. The memory reservation prevents the kernel from reusing the memory.
&reserved_mem {
latea_pagetbl: late_pgtbl@bfc00000 {
reg = <0x0 0xbfc00000 0x0 0x100000>;
no-map;
status = "okay";
};
};
The file
$PSDKLA/board-support/linux/arch/arm/boot/dts/dra7-evm-lcd-lg-late-attach.dts
has the full example. This file is tructured such that it includes an existing configuration first
#include "dra7-evm-lcd-lg.dts"
and sets only the atrributes required for enabling late attach functionality.
Late attach can be enabled for other remotecores by setting the late attach atributes on the corresponding device tree nodes. Late attach can be enabled for other hardware configurations by choosing a different base .dts file. e.g. If the EVM has an OSD lcd instead of 10″ LCD, you can change the include to “dra7-evm-lcd-osd.dts” instead of “dra7-evm-lcd10.dts”.
Build the .dtb file and use it for linux boot instead of the regular dtb file.
Boot setup
- Configure the EVM for boot from SD card.
- Use the MLO and u-boot.img built above with the early boot functionality enabled.
- Use the dtb file built with the late attach functionality enabled.
- Copy the IPU2 binary into the boot partition of the SD card. The IPU2 binary can be found in
/lib/firmware/dra7-ipu2-fw.xem4 on the target filesystem.
Boot the board and test the IPU2 functionality as you would normally do. You will not notice any difference in the functionality except that the IPU2 was already loaded and running before the Linux Kernel was booted.
Confirming Late attach functionality
There is no default indication of whether late attach feature is in use or not. This is by design as use of late attach should be invisible to the user. For debug purposes, enable the DEBUG flag on drivers/remoteproc/omap_remoteproc.c during kernel compilation. If late attach functionality is in use, additional traces will now show up in the dmesg logs.
root@dra7xx-evm:~# dmesg | grep late-attach
[ 2.609644] omap-rproc 55020000.ipu: device will be late-attached.
Customizing Early Boot for a Usecase
The Early boot code in U-Boot does the necessary configuration to bring up a remotecore. This includes the timers and the MMUs. It does not configure any other peripherals by default. Some usecases may require additional peripheral configuration before running the remotecore. U-Boot includes placeholder functions that can be populated for this purpose. These can be found in the file
$GLSDK/board-support/u-boot/board/ti/dra7xx/evm.c.
/*
* If the remotecore binary expects any peripherals to be setup before it has
* booted, configure them here.
*
* These functions are left empty by default as their operation is usecase
* specific.
*/
u32 ipu1_config_peripherals(u32 core_id, struct rproc *cfg)
{
return 0;
}
u32 ipu2_config_peripherals(u32 core_id, struct rproc *cfg)
{
return 0;
}
u32 dsp1_config_peripherals(u32 core_id, struct rproc *cfg)
{
return 0;
}
u32 dsp2_config_peripherals(u32 core_id, struct rproc *cfg)
{
return 0;
}
Validation and Limitations
Late attach functionality has been validated for all the remotecores IPU2,IPU1,DSP1 and DSP2. However there is the following limitation.
• U-Boot has only been updated to load remotecore binaries from the boot partition of the SD card or eMMC. If you desire to load remotecore binaries from other media (e.g. QSPI), the code to load the binaries needs to be added to U-Boot.
Debug guide
A more detailed user guide of Early Boot and Late Attach functionality can be found at Early Boot and Late Attach in Linux
##Changing MPU Frequency (DVFS)
DRA7xx supports running MPU (CPU0/1) at more than one frequency.
User may chose to run at different frequencies based on performance requirements for the application/use-case targeted.
The Data Manual contains all the frequencies supported by the Hardware and the Processor SDK Linux_Automotive Datasheet available with current release contains details of supported frequencies in the Software.
Please refer to the following wiki for details on how to change CPU frequency and voltage DVFS UserGuide
Testing OpenCL
OpenCL is supported in the SDK from the 3.02 release. To test OpenCL, please follow the below steps.
1. OpenCL requires reserving areas of memory to allow transfer of data between A15 and DSP. To reserve memory, apply the below kernel patch, rebuild the device tree file and boot with the updated device tree.
http://review.omapzoom.org/38240 dra7xx: dts: reserve memory for opencl
2. Ensure that the OpenCL firmware is loaded to the DSP. In the default PSDKLA filesystem, OpenCL firmware is not loaded by default. Follow the below steps to select the right firmware and reboot.
root@dra7xx-evm:~# cd /lib/firmware/
root@dra7xx-evm:/lib/firmware# ls -la *dsp*-fw.xe66
lrwxrwxrwx 1 1000 pulse 55 Feb 13 2017
dra7-dsp1-fw.xe66 ->
ipc/ti_platforms_evmDRA7XX_dsp1/test_omx_dsp1_vayu.xe66
lrwxrwxrwx 1 1000 pulse 55 Feb 13 2017
dra7-dsp2-fw.xe66 ->
ipc/ti_platforms_evmDRA7XX_dsp2/test_omx_dsp2_vayu.xe66
root@dra7xx-evm:/lib/firmware# rm dra7-dsp1-fw.xe66 dra7-dsp2-fw.xe66
root@dra7xx-evm:/lib/firmware# ln -s dra7-dsp1-fw.xe66.opencl-monitor
dra7-dsp1-fw.xe66
root@dra7xx-evm:/lib/firmware# ln -s dra7-dsp2-fw.xe66.opencl-monitor
dra7-dsp2-fw.xe66
root@dra7xx-evm:/lib/firmware# ls -la *dsp*-fw.xe66
lrwxrwxrwx 1 root root 32 Feb 24 2017
Processor SDK Linux Automotive Software Developers Guide 47
dra7-dsp1-fw.xe66 -> dra7-dsp1-fw.xe66.opencl-monitor
lrwxrwxrwx 1 root root 32 Feb 24 2017
dra7-dsp2-fw.xe66 -> dra7-dsp2-fw.xe66.opencl-monitor
root@dra7xx-evm:/lib/firmware# sync
- Ensure that the cmemk module is inserted in the kernel.
root@dra7xx-evm:~# lsmod | grep cmemk
cmemk 35110 2
- Ensure that CMEM memory reservations are in effect. Link [14]
root@dra7xx-evm:~# cat /proc/iomem | grep -i cmem
40500000-405fffff : CMEM
a0000000-abffffff : CMEM
- OpenCL examples are present in the filesystem in /usr/share/ti/examples/opencl. The examples are prebuilt and can be executed on the target.
root@dra7xx-evm:/usr/share/ti/examples/opencl# ls
Makefile dgemm float_compute monte_carlo
offline_embed sgemm vecadd_openmp
buffer dsplib_fft make.inc null
ooo_callback simple vecadd_openmp_t
ccode edmamgr matmpy offline
platforms vecadd
root@dra7xx-evm:/usr/share/ti/examples/opencl# cd platforms/
root@dra7xx-evm:/usr/share/ti/examples/opencl/platforms# ls
Makefile Makefile.rtos main.cpp main.o platforms
root@dra7xx-evm:/usr/share/ti/examples/opencl/platforms# ./platforms
PLATFORM: TI AM57x
Version: OpenCL 1.1 TI product version 01.01.11.2 (Feb 19 2017
21:41:17)
Vendor : Texas Instruments, Inc.
Profile: FULL_PROFILE
DEVICE: TI Multicore C66 DSP
Type : ACCELERATOR
CompUnits : 2
Frequency : 0.6 GHz
Glb Mem : 163840 KB
GlbExt1 Mem: 0 KB
GlbExt2 Mem: 0 KB
Msmc Mem : 1024 KB
Loc Mem : 128 KB
Max Alloc : 147456 KB
root@dra7xx-evm:/usr/share/ti/examples/opencl/platforms# cd ../vecadd
Processor SDK Linux Automotive Software Developers Guide 48
root@dra7xx-evm:/usr/share/ti/examples/opencl/vecadd# ./vecadd
DEVICE: TI Multicore C66 DSP
Offloading vector addition of 8192K elements...
Kernel Exec : Queue to Submit: 13 us
Kernel Exec : Submit to Start : 120 us
Kernel Exec : Start to End : 33010 us
Success!
Testing OpenCV
The 3.02 release also supports OpenCV. OpenCV is accelerated by using the C66x DSP’s through OpenCL.
1. Download OpenCV test data from https://github.com/Itseez/opencv_extra/archive/master.zip unzip this file in /usr/share/OpenCV. This will create a directory opencv_extra-master with the below contents.
root@dra7xx-evm # ls /usr/share/OpenCV/opencv_extra-master/
3d android_manager_internal_docs classifiers gpu_demos_pack
learning_opencv_v2 README.md testdata
The testsuite expects the testdata folder in /usr/share/OpenCV. Create a symbolic link to the testdata folder in /usr/share/OpenCV or move the testdata to the required location.
root@dra7xx-evm:/usr/share/OpenCV# ln -s opencv_extra-master/testdata .
- Run the tests with the following commands
root@dra7xx-evm:/usr/share/OpenCV# cd titestsuite/
root@dra7xx-evm:/usr/share/OpenCV/titestsuite# ls
runtests setupEnv.sh test_core_ocl test_imgproc_ocl
root@dra7xx-evm:/usr/share/OpenCV/titestsuite# source setupEnv.sh
root@dra7xx-evm:/usr/share/OpenCV/titestsuite# ifconfig
The tests run for ~70 mins on DRA75x platform which has 2 DSP’s. Processing is offloaded transparently to the DSP \here possible.
For more information on OpenCV and accelerating OpenCV using the C66x DSP’s via OpenCL, please refer to the below links.
Processor SDK Linux Automotive Software Developers Guide 49
1. TI Processors Wiki OpenCV page [15]
2. How to add a new DSP Kernel [16]
3. Creating OpenCL C Kernel optimized for C66 core [17]
OpenCV can be rebuilt and installed via the yocto pacakge opencv.
SOC Performance monitoring tools
Introduction
The SOC Performance monitoring tools are a set of tools that are included in the default filesystem that allow the user to visualize various SOC parameters real-time on the screen.
Currently, there are two tools and a suite of scripts and utilities to use them.
1. soc-performance-monitor
2. soc-ddr-bw-visualize
Both these applications are Wayland applications and need to be invoked after running Weston.
These tools bring in the capability to visualize the following:
1. DDR BW Utilization
1. Overall DDR BW Usage
2. Split of the traffic between the two EMIF’s
3. A real time “top” like functionality that depicts the list of “Top 6” initiators generating the traffic.
2. Voltage of the various rails
3. Frequency of the various cores
4. Temperature (read from on die temperature sensors)
5. CPU Load information of the various processor cores including the GPU and DSP.
6. Boot time results (requires rebuild of u-boot and kernel), refer instructions below.
7. Power plot (Will be available soon. Note that this requires board modification on the EVM)
Getting started
• Prepare the card with root filesystem provided with the installer.
• Boot up
• Start weston
target # /etc/init.d/weston start
• Copy the required scripts into a temporary folder (this is to allow you to experiment with the settings later)
target # mkdir temp
target # cd temp
target # cp /etc/glsdkstatcoll/* .
target # cp /etc/visualization_scripts/* .
• You should see the following file in the directory after the above operation.
target # ls -al
drwxr-xr-x 2 root root 4096 Mar 22 18:01 .
drwxr-xr-x 3 root root 4096 Mar 22 18:01 ..
-rw-r--r-- 1 root root 114 Mar 22 18:01 config.ini
Processor SDK Linux Automotive Software Developers Guide 50
-rw-r--r-- 1 root root 265 Mar 22 18:01 dummy_boot_time_results.sh
-rw-r--r-- 1 root root 419 Mar 22 18:01 dummy_cpu_load.sh
-rw-r--r-- 1 root root 899 Mar 22 18:01 getFrequency.sh
-rw-r--r-- 1 root root 2293 Mar 22 18:01 getTemp.sh
-rw-r--r-- 1 root root 371 Mar 22 18:01 getVoltage.sh
-rw-r--r-- 1 root root 254 Mar 22 18:01 initiators.cfg
-rw-r--r-- 1 root root 143 Mar 22 18:01 list-boot-times.sh
-rw-r--r-- 1 root root 367 Mar 22 18:01 send_boot_times_to_monitor.sh
-rw-r--r-- 1 root root 496 Mar 22 18:01 soc_performance_monitor.cfg
-rw-r--r-- 1 root root 133 Mar 22 18:01 start_visualization_test.sh
• Running the soc-performance-monitor, this tool has two pre-requisites.
1. The name of the fifo configured in the file soc_performance_monitor.cfg needs to be created
2. The file soc_performance_monitor.cfg should be present in the current directory. This should be done in the
above steps.
• Creating the fifo (mentioned in the soc_performance_monitor.cfg)
target # mkfifo /tmp/socfifo
• Run the tool for various performance metrics
target # soc-performance-monitor &
• Run the tool for DDR BW Visualization
target # mkfifo /tmp/statcollfifo
target # soc-ddr-bw-visualizer &
The following sections will talk about the how to populate the data into tools and further controls that are possible.
Quick guide to available plugins
Plugins are the entities (scripts/native binaries) that can be used to send commands to the SOC Performance Monitoring tools.
The main intent of this is to separate the visualization engine from the data collection part and allow full configuration of the application.
When the application (soc-performance-monitor) is invoked, it starts up with the default data which is set to zero. To populate the real values, the user can use the scripts provided in the prebuilt filesystem.
Temperature data
The temperature data is read from the on-die temperature registers and sent to the visualization tool. The file system comes with a script that does this functionality.
target # sh getTemp.sh
Invoking the above command will populate the temperature table with the current temperature.
Voltage data
The voltage data is read from the omapconf utility and then parsing out the required information to be later sent to the visualization tool. The file system comes with a script that does this functionality.
target # sh getVoltage.sh
Invoking the above command will populate the Temperature table with the configured voltage for the various rails.
Frequency data
The frequency data is read from the omapconf utility and then parsing out the required information to be later sent to the visualization tool. The file system comes with a script that does this functionality.
target # sh getFrequency.sh
Invoking the above command will populate the Frequency table with the configured frequency for the various cores.
CPU Load information
The CPU load information need individual plugin modules for each of the cores. This is envisioned to be different for different systems. The default filesystem contains the plugins required for reading the MPU(A15) and the GPU(SGX544 MP2). Other plugins for measuring the loads for the IPU1, IPU2, DSP1 and DSP2 will be available at a later time.
Measuring the MPU load
The filesystem is populated with a binary which is called “mpuload” that reads the /proc/stat interface and derives the load. The user can run the utility in the background with the
target # mpuload FIFO <time in seconds>
Example usage:
target # mpuload /tmp/socfifo 1000 &
After running this binary the MPU load in the Bar Graph of the CPU load will be updated dynamically at an interval of 1 second.
Measuring the GPU load
The filesystem is populated with a binary called as “pvrscope” that reads the SGX registers via a library called libPVRScopeDeveloper.a This utility invokes the APIs provided by IMG as part of the Imagination PowerVR SDK and then populates the required FIFO.
Usage instructions:
target # pvrscope <option> <time_seconds>
options:
-f write into the FIFO (/tmp/socfifo)
-c output to console
time:
1-n specified in seconds
0 run forever
After running this utility, the GPU load in the BAR Graph of the CPU load area will be updated at an interval of 1 second.
Boot time measurement
This feature will be provided at future release.
Order of execution
The performance visualization tools have to be executed in the following order.
• Launch weston
• Create required FIFOs
• Configure the .cfg file to suit the required settings
• Run the soc-performance-monitor and/or soc-ddr-bw-visualizer
• Run the plugins to populate data
Config file format
The config file has the following format.
There are 3 different kinds of sections that can be defined, please refer to the particular section for more details.
The generic format is:
[SECTION_NAME]
VALUE_1
VALUE_2
..
..
VALUE_N
SPECIAL VALUE
<blank line>
Types of sections
1. GLOBAL
2. TABLE
3. BAR GRAPH
GLOBAL section:
The SECTION_NAME is specified as GLOBAL followed by a sequence of key value pairs.
[GLOBAL]
KEY_1=VALUE_1
KEY_2=VALUE_2
..
..
KEY_n=VALUE_n
<blank>
Global configurations
The list of recognized global values are:
• REFRESH_TIME_USECS
• FIFO
• MAX_HEIGHT
Processor SDK Linux Automotive Software Developers Guide 53
• MAX_WIDTH
• X_POS
• Y_POS
REFRESH_TIME_USECS:
• This will dictate the interval at which the utility is going to run.
• The value is specified in micro seconds
• This value decides a major trade-off, lower rate will increase the CPU load and GPU load.
• The ideal value is about 100000 usecs
FIFO:
• The value of this field is the named pipe or fifo that can be used to communicate with the application.
• User would need to create a fifo (application will prompt if it doesn’t exist)
MAX_HEIGHT, MAX_WIDTH:
• The width and height of the application.
• This can be adjusted based on the number of tables and bar graph entities.
X_POS, Y_POS:
• Decide the starting offset of the application.
• Note that there are commands to move the application (Refer commands section).
TABLE section:
The section name can be one of the following:
• BOOT_TIME
• TEMPERATURE
• VOLTAGE
• FREQUENCY
[TABLE_NAME]
VALUE_1
VALUE_2
..
..
VALUE_N
TITLE="TABLE TITLE",UNIT="unit to be displayed"
<blank line>
NOTE: The TITLE=list is a list of comma separated values and TITLE and UNIT are the only supported values.
BAR GRAPH section:
This section is the simplest section and does not allow much configuration other than the names and the title.
It follows the following format:
[GRAPH_NAME]
VALUE_1
VALUE_2
..
..
VALUE_N
TITLE OF THE GRAPH
<blank line>
Commands:
The FIFO can be used to communicate with the soc-performance-monitor application and pass data from the command line or from other applications.
There are a few commands that have been implemented to aid in modifying the running application via the FIFO.
The commands in general have the following format:
"INSTRUCTION: DATA_1 ... DATA_N"
and they can be sent to the soc-performance-monitor by simply doing an echo:
echo "INSTRUCTION: DATA_1 ... DATA_N" > FIFO
The currently supported list of supported commands are:
1. TABLE
2. CPULOAD
NOTE: To execute a sequence of commands in a sequence, it is advised that a delay of REFRESH_TIME_USECS be inserted between two commands.
TABLE command
The format of the TABLE command is:
"TABLE: ROW_NAME value unit"
When this command is issued, the tool will find a table entry with the ROW_NAME in Column 0 and then update the Column 1 of the table with “value unit”
If the ROW_NAME is not found, then this command will have no effect. Please note that this brings in a restriction that all the tables rows will need to have a unique name. In order to ensure this, the soc_performance_monitor.cfg file will have to be reviewed to ensure unique names.
Example: To update the FREQUENCY table for MPU, the user can send the following command:
echo "TABLE: FREQ_MPU 1500 MHz" > /tmp/socfifo
CPULOAD command
The format of the CPULOAD command is:
"CPULOAD: CORE_NAME value" > FIFO
CORE_NAME has to be one of the names specified in the soc_performance_monitor.cfg.
value is in the range 0 to 100
Usually, the CPULOAD command is invoked through an application monitors the load of a specific core.
In each system, the mechanism to retrieve the CPULOAD of a particular core can vary and it is for this reason that several plugins have been provided and serve as an example for further extension.
Example: To update the CPULOAD table for GPU, the user can send the following command:
echo "CPULOAD: GPU 87" > /tmp/socfifo
Executing in debug mode
To launch the application in debug mode for very verbose data on the internal working of the tool, launch the tool with the following option:
# soc-performance-monitor 1
Build instructions
The full source of the tool is available and the required recipes have been updated as part of the recipes and upstreamed to meta-arago.
Essentially, if the user builds the Yocto filesystem as documented in the SDG, the tool will get recompiled as part of it.
Configuration of the soc-ddr-bw-visualizer
Refer to #Using_the_statistics_collector_.28bandwidth_application.29
• The total time that the tool runs is configured using config.ini.
• To allow finer granularity of control to choose the initiators of interest, the user will have to modify the initiators.cfg.
The tool will have to relaunched for the new settings to take effect.
Additional Procedures
Build Environment Setup
NOTE: From this release, each component i.e kernel, u-boot or any userspace application should be cross-compiled
Cross Compiler setup
The cross compiler setup for the Rules.make is done as part of the setup.sh script in the SDK installation folder. The script for cross compiler ensures that the Linaro cross compiler toolchain is installed in the ${HOME} (or user specified) folder of the host machine.
Note:Please ensure that the PATH variable is set in your machine to point to the cross compiler setup.
To compile the code (if not using the top level Makefile to build kernel and u-boot), please ensure the environment variables ARCH and CROSS_COMPILE is set to arm and to the linaro cross compiler path respectively.
Rebuilding the SDK components
The installer contains a top level Makefile to allow the re-building of the various components.
Rebuild the SDK components by first entering the SDK directory using:
host cd{INSTALL_DIR}
The installer Makefile has a number of build targets which allows you to rebuild the various components. For a complete list execute:
host $ make help
After that, each of the build targets listed by ‘make help’ can then be executed using:
host make <target>_clean
host make <target>
host $ make <target>_install
For example, to compile the Linux Kernel, you can use the following commands.
host make linux_clean
host make linux
host $ make linux_install
In order to install the resulting binaries on your target, execute one of the “install” targets. Where the binaries are copied is controlled by the EXEC_DIR variable in ${INSTALL_DIR}/Rules.make. By default, this variable is set up to point to your NFS mounted target file system when you execute the installer setup (setup.sh) script, but can be manually changed to fit your needs.
You can remove all components generated files at any time using:
host $ make clean
And you can rebuild all components using:
host $ make all
You can then install all the resulting target files using:
host $ make install
Creating your own Linux kernel image
The pre-built Linux kernel image (zImage) provided with in the SDK installation folder is compiled with a default configuration. You may want to change this configuration for your application, or even alter the kernel source itself.
This section shows you how to recompile the Linux kernel provided with the SDK, and shows you how to boot it instead of the default Linux kernel image.
1. If you haven’t already done so, follow the instructions in #Starting_your_software_development to setup your build environment.
2. Recompile the kernel provided with the SDK by executing the following:
host cd{INSTALL_DIR}
host make linux_clean
host make linux
host $ make linux_install
- You will need a way for the boot loader (u-boot) to be able to reach your new zImage. Copy the new zImage that is generated in arch/arm/boot/ directory to the rootfs/boot folder.
- Copy the exported Linux kernel modules from the EXEC_DIR to the /lib/modules directory to the root file system
Using the statistics collector (bandwidth application)
The default filesystem includes a utility to determine the bandwidth statistics of each initiator in the J6 SoC.
This section will give an overview of how the user can use this effectively.
Target side
The tool is called glsdkstatcoll and is present under the /usr/bin folder. This tool requires a config file as an input which is also part of the filesystem Copy the sample config file into your working directory:
target # cp /etc/glsdkstatcoll/* .
target # glsdkstatcoll -h
There are two levels of configuration that the tool allows:
1. Configuration of the interval and the total time can be controlled via config.ini
2. The list of initiators can be configured using the initiators.cfg file.
Note that both these files have to be present under the current directory.
After configuration, the user can launch the tool as below:
target # glsdkstatcoll -f config.ini
The tool will run for the specified duration in seconds “TOTAL_TIME” in config.ini and will sample every “INTERVSAL_US” microseconds.
The results from this tool will be available in statcollector.csv
target # ls -al statcollector.csv
Host side
To visualize this data, first install matplotlib:
host # sudo apt-get install python-matplotlib
host # git clone git://git.ti.com/glsdk/example-applications.git
host # cd example-applications/bandwidth-tool/host
Since the analysis of the data will be done over many iterations, the user has been provided with a config file in which some basic configuration can be done.
host # vi configstat.ini # Set the IP Address of the target and also the directory where the statcollector.csv was generated
Once the above setup is completed, the data can be visualized as below:
host # python statcoll_plot.py
This will tool will do the following:
1. Fetch the file from the target
2. Figure out all the initiators that did not have any traffic and exclude them from the plot.
3. Plot the total traffic from EMIF1_SYS and EMIF2_SYS ports
4. Display the peak, average and average(active) traffic on the various ports.
5. Note that if the INTERVAL_US is other than 30000 microseconds, user will have to edit the statcoll_plot.py
If the target does not have the Ethernet capabilities, then the file can be fetched and the same tool can be run like below:
host # python statcoll_plot.py -f <name of the file>
There are some further enhancements planned, they will be posted in the git place holder.
Setting up Tera Term
Tera Term is a commonly used terminal program on Windows. If you prefer to use it instead of Minicom, you can follow these steps to set it up.
1. Download Tera Term from this location [18], and start the application.
2. In the menu select Setup->General… and set:
Default port: COM1
- In the menu select Setup->Serial Port… and set the following:
Port: COM1
Baud rate: 115200
Data: 8 bits
Parity: none
Stop: 1 bit
Flow control: none
References
[1] http://processors.wiki.ti.com/index.php/Category:Processor_SDK_Linux_Automotive
[2] http://releases.ubuntu.com/14.04/
[3] http://processors.wiki.ti.com/index.php/How_to_Build_a_Ubuntu_Linux_host_under_VirtualBox
[4] http://processors.wiki.ti.com/index.php/Download_CCS#Code_Composer_Studio_Version_6_Downloads
[5] http://software-dl.ti.com/infotainment/esd/jacinto6/processor-sdk-linux-automotive/latest/index_FDS.html
[6] http://processors.wiki.ti.com/index.php/RPMsg_Kernel_Client_Application
[7] http://processors.wiki.ti.com/index.php/Template:GLSDK_USB_DRD#Introduction
[8] http://processors.wiki.ti.com/index.php/Template:GLSDK_USB_DRD#Driving_VBUS_using_GPIO
[9] http://processors.wiki.ti.com/index.php/DRA7xx_Processor_SDK_Linux_Automotive_Software_Developers_Guide#Starting_your_software_development
[10] http://processors.wiki.ti.com/index.php/File:Flashing_binariesto_the_factory_boards_using_Device_Firmware_Upgrade(DFU).pdf
[11] http://dfu-util.sourceforge.net/releases/
[12] https://sourceforge.net/projects/libwdi/files/zadig/
[13] http://processors.wiki.ti.com/index.php/GLSDK_DRA7xx_PM_DVFS_User_Guide
[14] http://downloads.ti.com/mctools/esd/docs/opencl/memory/ddr-partition.html
[15] http://processors.wiki.ti.com/index.php/OpenCV
[16] http://processors.wiki.ti.com/index.php/OpenCV#OpenCV_OpenCL_related_framework_details:_how_to_add_new_DSP_kernel
[17] http://processors.wiki.ti.com/index.php/OpenCV#Creating_OpenCL_C_kernel_optimized_for_C66_core
[18] http://logmett.com/index.php?/download/tera-term-477-freeware.html
