DM365 Leopard SDK Old 2011Q2 Getting Started Guide
This guide includes instructions and tips users should follow to get their DM365 Leopard Board to work with the RR Evaluation SDK. For this guide different commands need to be executed on the host machine and the target. Commands to be run on the Ubuntu host have a yellow background. Commands to be run on the Ubuntu target have an aqua background.
- 1 Installation
- 1.1 Getting the RR LeopardBoard DM365 EVAL SDK 2011Q2
- 1.2 CodeSourcery Toolchain Installation
- 1.3 DVSDK 4 Installation
- 1.4 Install other required packages
- 1.5 Installing the SDK
- 1.6 Build SDK
- 1.7 Install bootloader, kernel, and target file system on LeopardBoard 365
- 1.8 Updating your SDK
- 2 SDK Hints
- 2.1 SD boot recover
- 2.2 Setting up video components
- 2.3 Activation of video capture modules
- 2.4 Video initialization
- 2.5 Output volume modification
- 2.6 Using ipiped and AEW library
- 2.7 AEW Commands
- 2.8 Using GStreamer with the LeopardBoard DM365 EVAL SDK 2011Q2
- 3 SDK common errors
Getting the RR LeopardBoard DM365 EVAL SDK 2011Q2
If you want to get the LeopardBoard DM365 EVAL SDK 2011Q2 you need to go to our download center page and select the LeopardBoard DM365 Evaluation SDK. You will be asked for your email address and some information. Once you have submitted the required information you will receive an email with the following information:
- A link to download the LeopardBoard DM365 SDK 2011Q2 Installer file for GNU/Linux based operative systems.
- A username and password to download the installer.
- A link to the installation instructions.
CodeSourcery Toolchain Installation
New DM365 SDK 2011Q2 uses the Codesourcery toolchain for cross compilation. You need to download and install the toolchain before you can build the SDK.
In order to download and install the CodeSourcery's toolchain for the DM365 SDK 2011Q2 please see Code Sourcery ARM toolchain 2009q1-203.
DVSDK 4 Installation
In addition to the toolchain, it is necessary to install the TI's Digital Video Software Development Kit (DVSDK) package before installing the SDK. The DM365 SDK 2011Q2 uses the new DVSDK 4. The download and installation instructions are detailed below:
1. The dvsdk_dm365-evm_4_02_00_06_setuplinux binary file can be downloaded from TI's site.
2. Install it by executing the binary file with the following commands:
sudo chmod a+wrx dvsdk_dm365-evm_4_02_00_06_setuplinux ./dvsdk_dm365-evm_4_02_00_06_setuplinux
3. You will be asked for the Toolchain's path, you need to specify the path to the binaries' directory of the toolchain as is shown in the following image:
4. Install the DVSDK with the default configuration.
Install other required packages
Usually the SDK will ask you to install any missing package it requires that is not installed on your host machine. You can install them using the following command:
sudo apt-get install package_name
The following is the list of the required packages by the SDK:
It is possible you need to install some additional packages when some features require them, the SDK will ask you for this packages in the form of a error message as is shown below:
Checking for host packages required by the SDK build system: buildsystem (build-essential) is installed make (make) is installed ncurses (libncurses5-dev) is installed texinfo (texinfo) is installed >> minicom (minicom) is NOT installed perl (perl) is installed quilt (quilt) is installed dc (dc) is installed subversion (subversion) is installed curl (curl) is installed git (git-core) is installed autoconf (autoconf) is installed crypt-ssleay (libcrypt-ssleay-perl) is installed quilt (quilt) is installed libusb (libusb-dev) is installed perl-expect (libexpect-perl) is installed mono (mono-devel) is installed There are missing packages, please install them: apt-get install minicom make: *** [.oscheck] Error 1
Installing the SDK
Once you have downloaded the SDK installer you can execute it by running the following commands:
sudo chmod a+x LeopardBoardDM365sdk-EVAL2011Q2-Linux-x86-Install.bin ./LeopardBoardDM365sdk-EVAL2011Q2-Linux-x86-Install.bin
Once you have executed it you can follow the next steps for its installation.
1. Click Yes in the start up window.
2. Click Next in the main window.
3. Accept the license terms in order to use the EVAL SDK and click Next to continue.
4. Choose the destination directory to where the EVAL SDK will be installed. Note: The installation directory must be a folder that doesn't exist since the installer will create it in the process.
5. Once you have selected the installation directory a confirmation window will be displayed. Click Next to start the installation or Back to change any data. Note: Be sure to have an Internet connection before click Next since the installer needs to download some packages
6. Wait while the EVAL SDK is installed in your computer.
7. Once the EVAL SDK has been successfully installed the next window will be displayed. Click Finish to close the installer and start using the EVAL SDK.
I put all my development directories in the work directory in my home folder. You can have more than one development directory installed, just remember to have the DEVDIR shell variable set properly.
Several packages are needed in order to compile the SDK. The SDK build process will let you know what packages are needed. While compiling you will be prompted to install them using sudo apt-get <packages>.
You can get a description of all the available commands for the SDK by running the following command into your SDK's installation directory:
It will show a list as the shown bellow:
RidgeRun Integrated Linux Software Developer's Kit Make targets: build - build kernel, target fs, bootloader kernel - build kernel fs - build target filesystem cmdline - build cmdline for target kernel (fs target does the same as well) bootloader - build bootloader config - allow SDK to be configured config_batch - attempt to configure without user interaction clean - delete all derived files update - Checks for repository updates for the SDK show_updates - Checks for available updates for the SDK's release; use revision=<number> to specify a particular revision or range (same syntax as svn log) or showdirs=1 to enable showing change details in the log displayed. env - Displays commands to run to setup shell environment coreconfig - Select toolchain, bootloader, and kernel (updates bsp/mach/Make.conf) doc - Generate the SDK API documentation into the documentation folder prelink - prelink the root filesystem install - Install all out of date images to target installbootloader - Install bootloader to target installmtd - Install uBoot MTD partition table installkernel - Install kernel to target installfs - Install file system to target installcmdline - Install uBoot kernel command line installbootcmd - Install uBoot boot command svnup - recursively run the 'svn up' command. Make parameters: VERBOSE - set to 1 to output executing commands LOGFILE - set to built output filename Make update parameters: FORCE_UP - set to force the application/removal of patches DEBUG_UP - set to output the executing commands U-boot Installer Parameters ($DEVDIR/installer/u-boot-installer) PORT - Host serial port to use for install HOST_IP - Host IP address to use for TFTP install FORMAT - Set to 1 to force SD card to be formatted FORCEINSTALL - Set to yes to force performing the requested install target ERASECHIP - Set to yes if you want to scrub the FLASH chip during bootloader installation Examples: make # same as 'make build' make VERBOSE=1 make installbootloader PORT=/dev/ttyUSB1
Configure the features you want to have in your board using:
Set the PATH to include the toolchain and aslo set the DEVDIR shell variable
$(make env) # this is the same as using grave accents to execute the output from the command - `make env`
Compile the SDK:
Install bootloader, kernel, and target file system on LeopardBoard 365
Verify bootloader is active
First a minicom (or any other terminal emulator used) session must be opened in order to stop the booting process at the bootloader prompt:
Reboot the board and stop the autoboot when
Hit any key to stop autoboot: ...
is displayed. Then close your minicom session.
Save bootloader to target hardware
The next step is to install the bootloader on the target. Typically you only install the SDK bootloader once.
Update target hardware with new images
Power cycle your target hardware, and install uboot environment / kernel / FS on the target (per your SDK configuration choices):
Similar to the bootloader installation, to install the devdir on the target the booting process should be stopped at the bootloader prompt.
Updating your SDK
Since the SDK is in a SVN respository RidgeRun can periodically make updates to the contents. To have an up-to-date SDK the following command must be executed on the SDK root directory
If, for any reason, modifications to any files that are included in a patch were made, the make update process won't be able to finish successfully. In order to override this the following command must be executed
make update FORCE_UP=1
Note that all changes previously made to these files will be lost.
Also, a debug functionality is implemented where all the executed commands will be displayed on the standard error (usually the same as standard output). To activate it use
make update DEBUG_UP=1
Of course, a combination of both options can be used
make update FORCE_UP=1 DEBUG_UP=1
A brief help menu explaining these features can be displayed using
A list of all current updates available is shown when running the following command
SD boot recover
If you brick your board, this can be recovered from an SD card. In order to make a bootable SD card you first need to have a successfully compiled SDK.
SD card preparation
1. Insert the SD card to the host port
2. Find out the device mounting point (e.g. /dev/sdd)
3. On your devdir root directory type:
make sd_boot_recover device=/dev/sdX
where sdX is your device mounting point.
Install Boot recovery files on SD card
1. Remove SD card from host and reinsert the SD card into host. This will cause your host to mount the VFAT file system on the SD card.
make install sd_boot_recover
Boot from SD to recover NAND image
1. Set switch 2 of DIPSW1 to ON, 1 and 3 to OFF. This is SD card boot mode.
2. Power up the board, it will automatically erase the NAND flash, and burn UBL, u-boot, kernel and file system. The SD card recovery LeopardBoard 365 serial output will provide progress information.
Verify NAND recovery
1. Set all pos of DIPSW1 to OFF for NAND boot mode.
2. Connect serial cable to PC and power cycle the board (it should boot up). Insert the SD card to the board.
Your LeopardBoard 365 now contains the UBL, uboot, kernel, and file system that was originally shipped on LeopardBoard 365 boards.
Having this done the bootloader and devdir must be installed (check Installation for instructions).
Setting up video components
The output and mode used for video playback/capture can be set during the SDK configuration process. The first step is to enter the configuration script:
Inside Architecture configurations select the options you desire on Video Output and Video Default Resolution.
Finally compile your changes and install
make make install
When you use the SDK to set your video output and resolution, the SDK build process creates $DEVDIR/images/cmdline with different kernel command line parameter settings. The table below shows the various video output settings for three of the common settings
|Video Output||DVI||Component||Composite||sysfs file|
Activation of video capture modules
While configuring the SDK go to Kernel configuration->Device drivers->Multimedia support->Video capture adapters and activate the devices you desire. For more details see Hardware setup - LeopardBoard DM365.
LeopardBoard 365 GStreamer Pipelines - SDK 2011Q2 contains instructions showing examples to demonstrate the hardware is working.
Before running a capture pipeline you must be sure you have chained the video previewer, that could be done in two ways:
- Currently the v4l2src element has a build in property called chain-ipipe which by default is set to true.
- If you are not using v4l2src or you have disabled its chain-ipipe property you can chain it by using the ipiped application.
When you boot your board the RidgeRun logo will be displayed, in order to display video the following commands must be executed:
Output volume modification
To increase/decrease the volume for audio playback (or other audio options) type
and modify as desired.
Using ipiped and AEW library
Ipiped is a D-Bus server for controlling and configuring the camera sensor, the dm365 video processor and the aew library. It usually works together with the auto white balance and auto exposure (aew) library in order to give a better control of the quality of the image captured.
You can find a Demo version of the AEW library in our EVAL SDK for DM365 2011Q2. This version is fully functional but has some demo features:
- The library will darken you current image periodically.
- After a while the AEW algorithm will stop working and you will get the capture with the last values configured by it, so you will need to restart your AEW algorithm in order to test it again.
You can find more about Ipiped and the AEW library at Auto exposure and auto white balance library
In order to use Ridgerun's AEW demo a series of commands needs to be entered at the Leopard's console:
640x480 Video Playback
First you need to remove RidgeRun logo image from the screen
The screen will be plain white after the "fbset -disable" command
In order to use the Ridgerun's AEW Demo, first a config script for the 5Mp video sensor mt9p031 needs to be run:
ipipe-client run-config-script dm365_mt9p031_config
This will chain ipiped, set the bayer-pattern and others.
Next, configure auto exposure:
ipipe-client init-aew G EC S C 150000 15 50 640 480 50
This will configure the auto exposure as:
White balance: Gray world algorithm Auto exposure: Electronic centric Gain type: the sensor gain Exposure metering method: center weighted metering Time between algorithm adjustments: 200000 uS Minimum frame rate: 15 Frame segmentation factor: 50 Width: 640 Height: 480 Center percentage: 50%
For more information about RidgeRun's AEW options:
ipipe-client help init-aew
Other options are available, for example white balance can use retinex algorithm, Exposure can use: partial metering, average metering or segmented metering.
Recommended pipeline for video playback:
gst-launch -e v4l2src chain-ipipe=false always-copy=false ! dmaiaccel ! video/x-raw-yuv,format=\(fourcc\)NV12, width=640, height=480, framerate=\(fraction\)30/1 ! dmaiperf \ ! TIDmaiVideoSink sync=false accelFrameCopy=false videoOutput=composite videoStd=D1_NTSC&
640x480 Video Recording
Same as 640x480 video playback, but use the following pipeline:
gst-launch -e -v v4l2src always-copy=FALSE chain-ipipe=false num-buffers=300 ! \ 'video/x-raw-yuv,format=(fourcc)NV12,width=640,height=480' ! dmaiaccel ! dmaienc_mpeg2 ! \ dmaiperf ! mpegpsmux ! filesink location= test8003.mpeg
A file named: test8003.mpeg will be created, no video will be displayed in the composite output.
1280x720 Video Recording
Similar to 640x480 video playback, but instead of using:
ipipe-client init-aew G EC S C 150000 15 50 640 480 50
ipipe-client init-aew G EC S C 150000 15 50 1280 720 50
Use the following pipeline:
gst-launch -e -v v4l2src always-copy=FALSE chain-ipipe=false num-buffers=300 \ ! 'video/x-raw-yuv,format=(fourcc)NV12,width=1280,height=720 \ ' ! dmaiaccel ! dmaienc_mpeg2 ! dmaiperf ! mpegpsmux ! filesink location= test8003.mpeg
A file named: test8003.mpeg will be created, no video will be displayed in the composite output.
Using GStreamer with the LeopardBoard DM365 EVAL SDK 2011Q2
You can combine the SDK and the LeopardBoard DM365 with the GStreamer's API in order to get powerful multimedia applications in an easy and faster way. Some examples of use of GStreamer to implement basic multimedia pipelines can be found at LeopardBoard 365 GStreamer Pipelines - SDK 2011Q2.
SDK common errors
Bash is not the default shell
bash is not your default shell Please run: sudo rm /bin/sh sudo ln -s /bin/bash /bin/sh make: *** [.oscheck] Error 255
Solution: Set bash as your default shell. Run the following commands.
sudo rm /bin/sh sudo ln -s /bin/bash /bin/sh
Missing packages in host computer
If you are running the "make config" command for the first time, you may need to install some packages in order to compile the SDK. You might receive a message as follows:
work@work-laptop:leopard-dm365$ make config The system is ubuntu-10.04 running in a 32 bit architecture SYSID: ubuntu-10.04_32 Checking for host packages required by the SDK build system: buildsystem (build-essential) is installed >> libtool (libtool) is NOT installed make (make) is installed >> ncurses (libncurses5-dev) is NOT installed texinfo (texinfo) is installed >> minicom (minicom) is NOT installed perl (perl) is installed quilt (quilt) is installed dc (dc) is installed subversion (subversion) is installed >> curl (curl) is NOT installed quilt (quilt) is installed >> libusb (libusb-dev) is NOT installed >> perl-expect (libexpect-perl) is NOT installed >> mono (mono-devel) is NOT installed There are missing packages, please install them: apt-get install libtool libncurses5-dev minicom curl libusb-dev libexpect-perl mono-devel make: *** [.oscheck] Error 1
Use the apt-get command to install the missing packages, for example, if you get the following message:
>> curl (curl) is NOT installed
Then you should install that package as follows:
sudo apt-get install curl
arm-linux-gnueabi-gcc: command not found
If you have not installed the toolchain yet, you will get the following erros:
kernel/bounds.c /bin/sh: arm-linux-gnueabi-gcc: command not found make: *** [kernel/bounds.s] Error 127 make: *** [prepare0] Error 2 make: Leaving directory `/home/work/leopard-dm365/kernel/linux-2.6.32' make: *** [build] Error 2 make: Leaving directory `/home/work/leopard-dm365/kernel' make: *** [kernel_build] Error 2
You need to install the toolchain, follow the instructions Install here.
Unable to find the dvsdk
echo "Unable to find the file $DVSDK_DIR/Rules.make. Aborting" ; exit 1 ; \ fi \ fi You need to install dvsdk_3_10_00_19 provided by Texas Instruments You can abort now and install the dvsdk_3_10_00_19 binary package on the default path at /home/work/dvsdk/dvsdk_3_10_00_19, so it will be automatically detected next time you try to build. Otherwise you can enter the location where you installed the package below. Please provide the path to dvsdk_3_10_00_19 installation location or press ctrl-c to abort
You need to download and install the dvsdk in your home directory.
Autotools and automake not found
checking for automake-1.6 >= 1.7 ... not found. checking for automake-1.5 >= 1.7 ... not found. . . . checking for autoconf-2.53 >= 2.52 ... not found. checking for autoconf-2.52 >= 2.52 ... not found.
You need to install the package autoconf by running the following command:
sudo apt-get install autoconf2.59