In the i.MX51 default WINCE6  release, the eCSPI doesn't support multiple bursts mode and set the wait states. Attached was the document and code for how to enable the multiple bursts mode and how to set the wait states between two burst.

We have a ATK tool which can program image, also it can burn fuse for i.MX51. Since fuse is one time program, so please take care the fuse can't be turn back after programmed.

Very often, customers need to disable the framebuffer auto blank. There are several ways to do so. Modify the code in drivers/tty/vt/vt.c or remove the  “CONFIG_VT_CONSOLE” from the kernel configuration. All these work, but inconvenient. Review the code following: static int vesa_blank_mode; /* 0:none 1:suspendV 2:suspendH 3:powerdown */ static int vesa_off_interval; static int blankinterval = 10*60; core_param(consoleblank, blankinterval, int, 0444); The blank is controlled by blankinterval, which can be set with the name consoleblank. And the consoleblank is a "core_param". core_param in linux can be recognized by kernel. Also can be passed to kernel command line from uboot with bootargs. we could add this to the bootargs that the framebuffer will not go blank: consoleblank=0 Example(verified with imx6 L3.0.35_4.1.0_130816): setenv bootargs_mmc 'setenv bootargs ${bootargs} root=/dev/mmcblk1p1 rootfstype=ext4 rootwait video=mxcfb0:dev=hdmi,1920x1080M@60,if=RGB24,bpp=32 fbmem=28M consoleblank=0'

View Webinar Recording

Brief introduction on the aarch64 linux kernel memory mapping layout and basic management stuffs.  Contents include: Kernel's virtual memory layout and mapping after running i.MX8QM/QXP kernel reserved memory layout Kernel memory allocation method and technology (Buddy, cma, ION...) DMA buffer management, SWIOTLB, IOMMU GPU memory management How to customize the memory for different use cases How to avoid using CMA for a better stability and performance

These questions and answers are about interrupt generation at a dedicated (configurable) video output port. The i.MX6D manual (Rev. 0) Image Processing Unit (IPU) chapter mentions: Every DI has 10 timing generator counters. The IPU Interrupt Generator has 10 DI0 counters (1...10) and just 2 DI1 counters (3 & 😎 as interrupt sources. The Interrupt Control Register lists 11 DI0 counters (0...10) Q1. Are the DI timing-generator counters linked to the counters in the interrupt controller, or are they different counters? A1. Yes, the DI timing generator counters are linked to the counters in the interrupt controller. Q2. Why are there 11 counters listed in the interrupt controller, but just 10 counters in the timing generator? A2. There is disp_clk_en_pre in the interrupt controller. Thus the 11 counters: 10 timing generator counters and 1 disp clock generator counter. Q3. Is configurable timing feasible for DI0 by using the timing generator counters? A3. Yes, using the 10 internal timing counters you can generate various timing relationships. In addition, you can detect any of the interrupt counters. For example, if you use counter 8, then you can detect the interrupt associated with counter 8. Q4. Explain the impact of the DI1 counter access of only channels 3 and 8. A4. DI1 also has 10 timing generator counters and 1 disp clock generator counter, which you can use to generate desired waveforms. This is similar to DI0. The difference is only 2 of the 10 counters (plus another disp_clk) are connected to the interrupt controller for DI1. Therefore, there is a restriction for detection. If you use counter 7, read out the counter 7 interrupt of DI1 is not possible. However, 2 channels should be sufficient. These interrupts are usually used to indicate a frame start or a frame end. We usually use counter 3 to represent Vsync. So normally we only use counter 3 interrupt. DI1 has only 3 accesses because this covers the anticipated use case and the desire was to restrict register size. The extra counters facilitate flexible DI1 timing generation.

This doc show how to use i.MX8QXP Display Controller GammaCor unit to tune gamma. HW: i.MX8QXP MEK board, HDMI monitor SW: i.MX Linux 4.14.98_2.2.0 BSP release, patch in this doc 1.Introduce gamma The gamma, gamma correction, gamma encoding, gamma compression , these words all related one kind operation , see wiki page of it: The device used for image capture/print/display follow this power-law. For example the camera captured image , to view this image on display device as good as original captured image : gamma encoding when camera saved sensor data to image file,  and  gamma decoding when that image file display on your PC LCD monitor. That is : 2. i.MX8QXP Display Controller Gamma Correction Unit The Gamma Correction unit position is located between Frame Gen unit and TCon unit.   More detail see below contents from i.MX8QXP RM: So GammaCor unit could be used as adjust display gamma , or brightness or contrast. To used it, need follow the steps at RM 15.9.2.4.4.8.3.   Something need to note: You need program 33 sample point value into the register, these sample point value range is from 0 to 1023. Note, first write is start sample point value , then the other is delta value: current sample point minus previous sample point value. You can use GammaCor unit on any channel of R/G/B. If you use normalized function f(x), the following formula should be used to clut[i = 0..32] = round( f(i * 32 / 1023) * 1023) 3. i.MX8QXP Linux device driver patch and test code Apply attached  patch 8qxp_dpu_gammacor_4.14.98_2.2.0.diff on Linux kernel. In the kernel patch, function dpu_gammacor_update, I choose not calculate delta value between each sample pint , let user space application calculate delta value and passed to kernel. Apply 8qxp-dpu-gammacor-modetst.diff on libdrm-imx, to get test application which is based on modetest.  Test app will read one greyscale image file 720P.rgb, put it under same folder of test application , calculate sample point value by pow function  , and calling drmModeCrtcSetGamma to pass related value to kernel,  next loop will change sample point value, and will see that greyscale image will changed on HDMI monitor. After system boot up, run below cmd to check result of test application systemctl stop weston ./gamma_show_rgba.out -P 29@32:1280x720@AB24 Reference: a>https://www.nxp.com/webapp/Download?colCode=IMX8DQXPRM b>https://www.nxp.com/webapp/Download?colCode=L4.14.98_2.2.0_MX8QXP&appType=license c> https://source.codeaurora.org/external/imx/libdrm-imx/ d> https://en.wikipedia.org/wiki/Gamma_correction

Debugging with Eclipse and GDB on Linux user space This is a good open-source choice to debug i.MX processors. The integration of popular tools like Eclipse and GDB offers a good and stable connection between host and target. The first step is to install the tools on host. Click here to get instructions of how to install the tools. Let's debug a ready hello world program into the ltib. To extract the hello world package, type on ltib directory: $./ltib -m prep -p helloworld Change the code of hello.c to: #include <stdio.h> int main(int argc, char** argv) {     int i;     for (i = 0; i < 100; ++ i)     {         printf(“Welcome to GDB ! %d /n”, i);     }     return 0; } Change the Makefile to add debug symbols Change the two following lines from: CFLAGS = -Wall CXXFLAGS = -Wall To: CFLAGS = -Wall -g CXXFLAGS = -Wall -g Build and deploy the new source-code: $./ltib –p helloworld –m scbuild $./ltib –p helloworld –m scdeploy Configuring the Target On target, gdbserver needs to be run to perform debug. The gdbserver command has the following structure: gdbserver ip_host:port /full/path/app/app_name If gdbserver is not installed on target, select gdb package on ltib configuration. In this example our host has the 192.168.16.35 IP address and our HelloWorld application is located at /usr/bin/hello on the target board. Execute the gdbserver: gdbserver 192.168.16.35:10000 /usr/bin/hello You can use other port number as long as you use the same number when configuring the Eclipse. Setting a GDB Debug Session on Eclipse Now we will configure Eclipse C/C++ to start a GDB session with our remote i.MX board. We will need to know which is the target board’s IP address. To get your target’s IP address: /sbin/ifconfig In our example the target board has the 192.168.16.36 IP address. Open-up Eclipse and choose the C/C++ perspective. We will import the HelloWorld executable built by LTIB. Go to the menu File -> Import You will see the “Import” screen. Select “C/C++ Executable” option. Hit the “Next” button. Eclipse automaticaly creates a new project when whe use the “Import” option. In the next screen, select the “Search Directory” option and hit the “Browse” button. This session is incomplete and is being edited...

Test digital zoom with ipu for camera preview.   Board :sarbre-sd (imx6dq) BSP   : android 13.4ga In the above flow, one frame buffer is processed in four steps at camera preview. Add the step to change the frame buffer before step 4 , the added step which  zoom one preview frame.   The figure below shows the crop function of ipu lib, we use this function scale the frame.   Test result: preview zoom levle 0:   preview zoom level max:     When taking pictures with 5M pixels and the zoom is over level 1, the picture size is not 2592x1944 but 2016x1512. The underlying reason for it is that ipu crop function only supports the 2048x2048 maximum output .   Thumbnails of test result :  

Installation, patching and building the SDK The i.MX 6 Series Platform SDK v1.1.0 does not enable neither the MMU nor L1/L2 Caches (depending on the benchmark you are running, enabling these yield much better numbers), so there is a need to patch the code to enable these. Please download the SDK from the Freescale portal and the patch attached on this document, then follow these instructions: $ tar zxvf imx6_platform_sdk_v1.1.0.tgz $ cd iMX6_Platform_SDK $ patch -p1 < 0001-add-L2-cache-enable-to-mx6-SDK-1.1.0.patch $ export PATH=$PATH:<toolchain_install_path>/bin $ ./tools/build_sdk For more help, please look at the README.pdf and documents inside the doc folder.

[中文翻译版] 见附件   原文链接： https://community.nxp.com/docs/DOC-345644 

Q: How to setup camera under Android? helping a customer (RTX) debug some issues with camera. They're using i.MX 6Solo and Android 13.4-GA on custom hardware. They added a new camera driver which seems to work when using small console capture program (no gui, no preview), so the route from camera to /dev/video0 seems to work. However, when they try to use camera from Android, entire system freezes. They located the crash to following line in ipu_common.c which basically enables camera CSI0: ipu_cm_write(ipu, reg | IPU_CONF_CSI0_EN, IPU_CONF); While investigating IPU setup, we noticed that CPMEM setup for IDMAC channel 0 is "off": ch 0 word 0 - 00000000 25800000 00000000 E0000000 00077C4F ch 0 word 1 - 01B086B0 00394EC0 0087C000 00009FC0 0000027F As seen from above, EBA0 points to 0x0D843580 and EBA1 to 0x0E53B000, which is in EIM memory space, not DDR memory space, which probably causes issues. We're not sure what could be causing this as camera driver doesn't provide any such address and mostly just handles communication to camera chip. But something gets off in Android framework and we could use any hints about what to look for. A: You can reference to the "i.MXAndroidR13.4GAAdvancedUserGuide.html" "3 Camera&Video Recorder customization" for how to change the camera in Android. Another thing needs be checked is the camera sensor driver, you can reference to "kernel_imx\drivers\media\video\mxc\capture\ov5642.c", "static struct v4l2_int_ioctl_desc ov5642_ioctl_desc", did you implemented the same v4l2_int_ioctl_desc functions? Customer HW is set up so that DDR memory space starts at 0x80000000. This causes problems in myandroid/hardware/imx/mx6/libcamera/CamerHAL.cpp when obtaining buffer addresses where camera should store data. GPU will only return OFFSET into 2GB memory space it can address, so it will return an address below 0x80000000. This needs to be adjusted before passing onto V4L2 when starting capturing.

Overview The first in the series, the i.MX53 Quick Start board is a low cost development platform. Integrated with an ARM® CortexTM-A8 1 GHz processor, the Quick Start board includes a display controller, hardware-accelerated graphics, 1080p video decode and 720p encode as well as numerous connectivity options ideally suited for applications such as human machine interface in embedded consumer, industrial and medical markets. Go to http://www.freescale.com/iMXQuickStart and visit the official Quick Start Board page Hardware Features Processor • i.MX53 1 GHz ARM Cortex-A8 Processor • Power management IC • 1 GB DDR3 memory Display • LVDS connector • VGA connector • Parallel LCD add-on card (via expansion connector) • HDMI add-on card (via expansion connector) Audio • SPDIF output via HDMI add-on card • Freescale SGTL5000 audio codec • Microphone jack • Headphone jack Expansion Connector • Enables parallel LCD or HDMI output • Camera CSI port signals • I2C, SSI, SPI signals Connectivity • Full-size SD/MMC card slot • microSD card slot • 7-pin SATA data connector • 10/100 Base-T Ethernet port • Two High-Speed USB host ports • Micro USB device port Debug • JTAG connector • DB-9 UART port Miscellaneous • 3” x 3” 8-layer PCB • 3-axis Freescale accelerometer (MMA8450QT) • 2A, 5V power supply Tutorials, Training Materials and Documentation Android i.MX 53 QSB Enable WiFi Android i.MX 53 QSB Android Recovery Mode Linux i.MX 53 QSB Board Get Started i.MX 53 QSB Ubuntu Dual Display Running Dual Display on i.MX53QSB   

In order to run the QT5 demos on i.MX6 you should follow the instructions on this link: Building QT for i.MX6 Some of the demos on the release such as  /examples/opengl/hellogl_es2,  consist of a group of multiple widgets appearing on the screen. Normally these demos should work OK in a windowed environment such as Wayland or X11. In the case of Linux only environment, the plugin that draws to the screen is called EGLFS. This plugin has the restriction that it only supports one single widget at a time on the screen surface. Then demos such as hellogl_es2 are *not intended* to work along with this plugin, and it will never work. The errors found when using EGLFS consist on: These issues can be seen in the Qt OpenGL examples.  "hellogl_es2" and "2dpaint" seem to display one rendered frame and then break --   "hellogl_es2" shows the QT word and bubbles, and the GUI is hidden, while  "2dpaint" just shows the openGL version without label. It seems that when including  a QGLWidget on a form, the QGLWidget would work OK, but the rest of the form would not appear. I couldn't click any buttons or do anything.   Along with these problems I would also see one or more of these error messages in the output:   * This plugin does not support setParent!   * This plugin does not support propagateSizeHints()   * QOpenGLContext::swapBuffers() called with non-opengl surface However other demos such as hellowindow work well with EGLFS because they are single widget.  Also all demos created with qtquick will work OK since all visual QML items are rendered as a single widget using the scene graph, a low-level, high-performance rendering stack, closely tied to OpenGL. This is better explained here: Qt5 QPainter vs. QML &amp; Scene Graph.

[中文翻译版] 见附件   原文链接： https://community.nxp.com/docs/DOC-343761 

This ppt provides a tutorial about how to add 24bit LVDS support in Android for iMX6QD.

<analytics uacct="UA-5520491-1" /> How to enable WIFI support for i.MX53 QSB Android After applying every QSB patch, enable WiFi support according to your hardware. Android R4 can be downloaded from Adeneo´s website. AR6102  Change file device/fsl/imx53_loco/BoardConfig.mk -BOARD_WLAN_CHIP_AR6102  := false +BOARD_WLAN_CHIP_AR6102  := true AR6003  Change file device/fsl/imx53_loco/BoardConfig.mk -BOARD_WLAN_CHIP_AR6003  := false +BOARD_WLAN_CHIP_AR6003  := true After complete build ar6000.ko will be created under /system/etc/modules To turn WIFI on, go to Settings > Wireless & network s > Wi-Fi Error message case  In case logcat shows the following error message: E/WifiHW  ( 2086): Cannot access "/data/misc/wifi/wpa_supplicant.conf":Permission denied Reconfigure nfs server file /etc/default/nfs-kernel-server delete this line:   RPCMOUNTDOPT=--manage-gids

i.MX6X Core Board HW User Guide.pdf

This post describes the setup detail for installing Ubuntu based distro in any i.Mx6x NXP Boards. Details are described on: 1. Select your board, Setting the host, Download and compile uboot , dtb and and the Kernel version on your board. 2. Installing the Ubuntu core, Lubuntu graphics desktop version and/or Build your own Ubuntu rootfs with debootstrap. 3. Modify rootfs and Installing needed packages 4. Setting with SD image. 5. Setting Ubuntu on target 6. Adding GPU acceleration 1: Select your board, Setting the host, Download and compile uboot , dtb and and the Kernel version on your board. Supported NXP HW boards: i.MX 6QuadPlus SABRE-SD Board and Platform i.MX 6Quad SABRE-SD Board and Platform i.MX 6DualLite SABRE-SD Board i.MX 6Quad SABRE-AI Board i.MX 6DualLite SABRE-AI Board i.MX 6SoloX SABRE-SD Board i.MX 6SoloX SABRE-AI Board Install host dependences (version tested 14.04): $ sudo apt-get install gparted git build-essential libncurses5 wget u-boot-tools zlib1g-dev ncurses-dev \ cmake libc-dev-armhf-cross pkg-config-arm-linux-gnueabihf build-essential checkinstall cmake \ pkg-config lzop libc6 libstdc++6 debootstrap qemu-user-static binfmt-support Download the compiler toolchain and extract it: $ cd ~/ $ wget -c https://releases.linaro.org/14.09/components/toolchain/binaries/gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux.tar.xz $ tar xf gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux.tar.xz Create general variable environments: $ export target=mx6q (e.g. processor: mx6sx, mx6d, mx6dl,etc) $ export board=sabresd (e.g. sabresd, sabreauto) $ export ARCH=arm $ export CROSS_COMPILE=../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- $ unset LDFLAGS Download u-boot At the release of this document, latest uboot version was imx_3.14.52, it should work with other version as well, so please check the proper version for your board: $ cd ~/ $ wget –c http://git.freescale.com/git/cgit.cgi/imx/uboot-imx.git/snapshot/uboot-imx-rel_imx_3.14.52_1.1.0_ga.tar.gz $ tar -xf uboot-imx-rel_imx_3.14.52_1.1.0_ga.tar.gz $ cd uboot-imx-rel_imx_3.14.52_1.1.0_ga $ make $targetboard_config    # e.g. mx6qsabresd_config $ make Linux Kernel, Firmware, headers, modules and DTS files $ cd ~/ $ wget –c http://git.freescale.com/git/cgit.cgi/imx/linux-2.6-imx.git/snapshot/linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga.tar.gz $ tar xf linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga.tar.gz $ cd linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga $ make imx_v7_defconfig $ make menuconfig $ make -j4 zImage modules dtbs $ cd ~/ move your image to binary folder: $ sudo cp –v uboot-imx-rel_imx_3.14.52_1.1.0_ga/u-boot.imx binary/ $ sudo cp –v linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga/arch/arm/boot/zImage binary/ $ sudo cp –v linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga/arch/arm/boot/dts/i$target-$board.dtb binary/ Now you have the bootloader, device tree and kernel image of your board ready, let’s create the rootfs. 2: Installing the Ubuntu core, Lubuntu graphics desktop version and/or Build your own Ubuntu rootfs with debootstrap. Installing ubuntu core: $ cd ~/ $ sudo mkdir –p core /media/rootfs /media/kernel $ wget –c http://cdimage.ubuntu.com/ubuntu-core/releases/14.04/release/ubuntu-core-14.04.4-core-armhf.tar.gz $ sudo tar –xf ubuntu-core-14.04.4-core-armhf.tar.gz –C core $ sudo cp -vr core/* /media/rootfs $ cd linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga $ sudo make modules_install firmware_install INSTALL_MOD_PATH=/media/rootfs/ ARCH=arm CROSS_COMPILE=../../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- $ sudo make ARCH=arm CROSS_COMPILE=../../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- headers_install INSTALL_HDR_PATH=/media/rootfs/usr Now you should have your ubuntu rootfs on /media/rootfs folder. and you can pass to part 3 of this post. Installing ubuntu Linaro LXDE: $ cd ~/ $ sudo mkdir –p core /media/rootfs /media/kernel $ wget https://releases.linaro.org/14.10/ubuntu/trusty-images/alip/linaro-trusty-alip-20141024-684.tar.gz $ sudo tar -xf linaro-trusty-alip-20141024-684.tar.gz –C core $ sudo mv core/binary/* core/ $ sudo rm –rf core/binary $ sudo cp -vr core/* /media/rootfs $ cd linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga $ sudo make modules_install firmware_install INSTALL_MOD_PATH=/media/rootfs/ ARCH=arm CROSS_COMPILE=../../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- $ sudo make ARCH=arm CROSS_COMPILE=../../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- headers_install INSTALL_HDR_PATH=/media/rootfs/usr Now you should have your ubuntu rootfs on /media/rootfs folder. and you can pass to part 3 of this post. Installing with debootstrap $ cd ~/ $ target=rootfs $ distro=trusty $ sudo debootstrap --arch=armhf --foreign --include=ubuntu-keyring,apt-transport-https,ca-certificates,openssl $distro "$target" http://ports.ubuntu.com $ sudo cp /usr/bin/qemu-arm-static $target/usr/bin $ sudo cp /etc/resolv.conf $target/etc Now have a minimal Ubuntu rootfs - chroot to it and perform the 2nd stage install: $ sudo chroot $target  //Now we are in chroot # distro=trusty # export LC_ALL=C LANGUAGE=C LANG=C # /debootstrap/debootstrap --second-stage Edit the sources.list repositories # cat <<EOT > /etc/apt/sources.list deb http://ports.ubuntu.com/ubuntu-ports/ $distro main restricted universe multiverse deb http://ports.ubuntu.com/ubuntu-ports/ $distro-updates main restricted universe multiverse deb http://ports.ubuntu.com/ubuntu-ports/ $distro-security main restricted universe multiverse EOT # apt-key adv --recv-keys --keyserver keyserver.ubuntu.com 40976EAF437D05B5 # apt-key adv --recv-keys --keyserver keyserver.ubuntu.com 3B4FE6ACC0B21F32 # apt-get update # apt -y -f install # apt-get upgrade # apt-get install nano Now you should be able to login without password, then use passwd command to set one. If you like to add custom users: # passwd root # adduser <myuser> # usermod -a -G tty myuser # usermod -a -G dialout, adm, sudo, dip, plugdev myuser # visudo Under the line that looks like: root ALL=(ALL:ALL) ALL add the following (change user with your actual username) <myuser> ALL=(ALL) ALL your rootfs is ready, exit chroot # exit $ sudo rm $target/etc/resolv.conf $ sudo rm $target/usr/bin/qemu-arm-static $ sudo mv rootfs/* /media/rootfs $ cd linux-2.6-imx-rel_imx_3.14.52_1.1.0_ga $ sudo make modules_install firmware_install INSTALL_MOD_PATH=/media/rootfs/ ARCH=arm CROSS_COMPILE=../../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- $ sudo make ARCH=arm CROSS_COMPILE=../../gcc-linaro-arm-linux-gnueabihf-4.9-2014.09_linux/bin/arm-linux-gnueabihf- headers_install INSTALL_HDR_PATH=/media/rootfs/usr Now you should have your ubuntu rootfs on /media/root. 3: Modify Rootfs and Install needed packages Edit and verify the sources.list repositories $ cd /media/rootfs $ sudo cat <<EOT > etc/apt/sources.list deb http://ports.ubuntu.com/ubuntu-ports/ trusty main restricted universe multiverse deb http://ports.ubuntu.com/ubuntu-ports/ trusty-updates main restricted universe multiverse deb http://ports.ubuntu.com/ubuntu-ports/ trusty-security main restricted universe multiverse EOT Edit networks interfaces and append in the existing file: $ sudo nano etc/network/interfaces auto lo iface lo inet loopback auto eth0 iface eth0 inet dhcp If you require Serial Console, remove and include an additional line at the end of the file for  ttymxc0 output as below, $ sudo nano etc/init/tty1.conf exec /sbin/getty -8 38400 tty1 exec /sbin/getty -L 115200 ttymxc0 If you like to change the localhostname: $ sudo nano etc/hostname and change to “your name” e.g. imx6Q. Set the date and time clock and update $ sudo nano /etc/rc.local  Add this: if [ `date +"%Y"` -eq "1970" ]; then                     date --set="2016-04-01" fi exit 0 (optional for Linaro rootfs) Edit passwd and remove the x in root and linaro lines $ sudo nano etc/passwd root:x:0:0:root:/root:/bin/bash linaro:x:0:0.. and change like this:                                   root::0:0:root:/root:/bin/bash linaro:::0:.. Now you are ready to program your sd image. 4: Setup microSD/SD card For these instructions, we are assuming: DISK=/dev/sdg on your HOST, cat /proc/partitions is very useful for determining the device id. $ cd ~/ $ export DISK=/dev/sdg Erase microSD/SD card: $ sudo dd if=/dev/zero of=${DISK} bs=1M count=10 Install Bootloader $ cd binary/ $ sudo dd if=u-boot.imx of=${DISK} bs=512 seek=2 $ sync Create Partition layout: $ cd ~/ $ sudo fdisk ${DISK} steps:        d ///delete all partitions currently on sd n // create new partition p // Primary partition 1 // partition number 1 2048 //default +1G // n // created 2d parition p 2 default default 1 // firts B // to be fat32 W // write partiotions $ sudo mkfs.vfat ${DISK}1 $ sudo mkfs.ext3  ${DISK}2 Mount ext3 SD partition to /media/rootfs: $ sudo mount ${DISK1} /media/kernel_target $ sudo mount ${DISK}2 /media/rootfs_target Copy Files on the SD. $ cd ~/ $ sudo cp –v binary/ i$target-$board.dtb /media/kernel_target $ sudo cp –v binary/zImage /media/kernel_target $ sudo mv /media/rootfs/* /media/rootfs_target Remove SD: $ sync $ sudo umount /media/kernel_target $ sudo umount /media/rootfs_target Boot the target, in console you should be login as root. root@imx6QSabreSD:~# 5: Setting Ubuntu on target Note: If you have issues with sudo on user UID, need to logout and log as root: imx6Q login: root Welcome to Ubuntu 14.04.4 LTS (GNU/Linux 3.14.52 armv7l) root@imx6Q:~# chown root:root /usr/bin/sudo root@imx6Q:~# chmod 4755 /usr/bin/sudo root@imx6Q:~# exit Login with <user $> or root # # apt-get update # apt-get –f install # apt-get install locales dialog wget # dpkg-reconfigure locales # apt-get upgrade Optional – install some useful packages: # apt-get install openssh-server can-utils usbutils build-essential automake autoconf libtool Get and Install the BSP packages (EULA required) # cd /home/user # mkdir –p vpu_pack # cd vpu_pack # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//firmware-imx-5.3.bin # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//imx-vpu-5.4.32.bin # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//libfslcodec-4.0.8.bin # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//imx-lib-5.1.tar.gz # chmod +x * # ./firmware-imx-5.3.bin --auto-accept --force # mkdir –p /lib/firmware/vpu # cp -ravf firmware-imx-5.3/firmware/* /lib/firmware/ # ./imx-vpu-5.4.32.bin --auto-accept --force # cd imx-vpu-* # make PLATFORM=IMX6Q all # make install # tar -xf imx-lib-5.1.tar.gz # cd  imx-lib-5.1/ # make -j1 PLATFORM="IMX6Q" # make PLATFORM="IMX6Q" install # cd .. # ./libfslcodec-4.0.8 --auto-accept –force # cd libfslcodec-* # ./autogent.sh --prefix=/usr --enable-fhw --enable-vpu # make # make install # mv /usr/lib/imx-mm/video-codec/* /usr/lib # mv /usr/lib/imx-mm/audio-codec/* /usr/lib # rm –rf /usr/lib/imx-mm/ # cd .. # mkdir –p gpu_pack # cd gpu_pack # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//imx-gpu-viv-5.0.11.p7.4-hfp.bin # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//xserver-xorg-video-imx-viv-5.0.11.p7.4.tar.gz # chmod +x * # ./imx-gpu-viv-5.0.11.p7.4-hfp –-auto-accept -–force # cd imx-gpu* # cp g2d/usr/include/* /usr/include/ # cp -d g2d/usr/lib/* /usr/lib/ # cp -Pr gpu-core/usr/* /usr # optional: install demos # cp -r gpu-demos/opt / # optional: install gpu tools # cp -axr gpu-tools/gmem-info/usr/bin/* /usr/bin/ # cd .. Installing gstreamer-imx, IPU, VPU and GPU support: Install build deps, gstreamer1.x, this step could take some time (~350MB): # apt-get install python pkg-config git gstreamer1.0-x gstreamer1.0-tools gstreamer1.0-plugins-good gstreamer1.0-plugins-bad gstreamer1.0-alsa libgstreamer1.0-dev libgstreamer-plugins-base1.0-dev libgstreamer-plugins-good1.0-dev g++-multilib # git clone git://github.com/Freescale/gstreamer-imx.git # cd gstreamer-imx # ln –s /usr/lib/arm-linux-gnueabihf/gstreamer-1.0/ /usr/lib/gstreamer-1.0 # ./waf configure --prefix=/usr --kernel-headers=/include # ./waf # ./waf install # cd ../../ (optional) Install libimxvpuapi library: This library provides a community based open-source API to the NXP imx-vpu library (the low-level IMX6 VPU interface). # git clone git://github.com/Freescale/libimxvpuapi.git # cd libimxvpu* # ./waf configure –-prefix=/usr # ./waf # ./waf install # cd .. note './waf install' installs artifacts to its prefix + /lib/gstreamer-1.0 but they need to be installed to /usr/lib/arm-linux-gnueabihf/gstreamer-1.0 which is why we created a symlink above before installing note g2d lib required to build G2D note that x11 library is required to build EGL sink with Vivante direct textures (only needed for X11 support) note that libfslaudiocodec is required to build audio plugins Now you are ready to test gstreamer 6: Add GPU HW Acceleration for X11 NOTE: The original version of these build instructions can be found in the Gateworks wiki . Many thanks to them for writing this! IMX6 IPU, VPU, and GPU support via GStreamer and Gstreamer-imx plugins. Many of the pieces needed (firmware and source-code) are from NXP and not freely redistributable thus must be downloaded from their mirror and extracted from a shell script that forces you to read and agree to their End User License Agreement (EULA). The following instructions can be used on top of the debootstrap and should work on other sources of Ubuntu or other Linux distributions root filesystems as well You can easily add X11 support to a base image created with the debootstrap instructions above by adding a few package groups. You will need the following: X11 server - ie Xorg Display Manager - this controls the login to the X session Window Manager - manages window position, re-sizing, decorations, etc for X clients If in any case you have installed the Linaro LXDE rootfs, it includes the Xorg X11 server, the lxdm Display Manager, the openbox Window Manager, and others useful user applications including the Chromium browser, if you do not install linaro lxde and want to install it please do: this step could take some time (~650MB)   # apt-get install xinit lxde lxterminal lxappearance lxrandr lxshortcut lxinput xinit  xserver-xorg-dev mesa-utils mesa-utils-extra Notes: you will need to add a non-root user with adduser for Chromium browser to work. You may choose to set up auto-login for that user by editing /etc/lxdm/default.conf and setting the autologin property in the base section at the beginning of the config file. /etc/xdg/lubuntu/lxdm/lxdm.conf This document takes as based kernel version 3.14.52v, vivante 5.0.11p7.4 correspond to the kernel version used. you should check the BSP release notes in order to know which xserver and Vivante GPU files need to be downloaded from the NXP repos. $ sudo nano /etc/lxdm/default.conf    [base]    autologin=user To add hardware GPU acceleration to X11 you need to add some libraries and drivers provided by Freescale from the imx-gpu-viv package. This requires signing Freescales End User License Agreement (EULA). This package provides the following: libg2d - a documented low-level API to the GPU (used by things like libimxvpuapi for gstreamer-imx and the gpu-core drivers) gpu-core - provides all the various OpenGL libs (libGL, libGLESv1_CM, libGLESv1_CL, libGLESv2, libGLSLC, libCLC, libEGL, libGAL, libOpenCL, ls libOpenVG) typically provided by the mesa project. Note that several versions of libEGL/libGAL/libGLESv2/libVIVANTE are provided for different backend rendering systems: dfb, fb, wl, x11. # cd gpu_pack #cd imx-gpu-* # cp gpu-core/usr/lib/dri/vivante_dri.so /usr/lib/xorg/modules/drivers/ # chmod 644 /usr/lib/xorg/modules/drivers/vivante_dri.so # rm /usr/lib/arm-linux-gnueabihf/mesa/libGL.so* # rm /usr/lib/arm-linux-gnueabihf/mesa-egl/libEGL.so* # rm /usr/lib/arm-linux-gnueabihf/mesa-egl/libGLESv2.so* # rm /usr/lib/arm-linux-gnueabihf/mesa-egl/libOpenVG.so* # cd ../../ # cd gpu-pack # wget http://www.nxp.com/lgfiles/NMG/MAD/YOCTO//xserver-xorg-video-imx-viv-5.0.11.p7.4.tar.gz # tar –xf xserver* # cd xserver-org-video-imx* #looks lik have to made #git init # ./fastbuild.sh  BUILD_HARD_VFP=1 XSERVER_GREATER_THAN_13=1 # cd.. # cd kernel-modu* # make Switch to gpu-core x11 backend: # backend=x11 # ln -sf libEGL-${backend}.so /usr/lib/libEGL.so # ln -sf libEGL-${backend}.so /usr/lib/libEGL.so.1 # ln -sf libEGL-${backend}.so /usr/lib/libEGL.so.1.0 # ln -sf libGAL-${backend}.so /usr/lib/libGAL.so # ln -sf libGLESv2-${backend}.so /usr/lib/libGLESv2.so # ln -sf libGLESv2-${backend}.so /usr/lib/libGLESv2.so.2 # ln -sf libGLESv2-${backend}.so /usr/lib/libGLESv2.so.2.0.0 # ln -sf libVIVANTE-${backend}.so /usr/lib/libVIVANTE.so # ln -sf libGAL_egl.dri.so /usr/lib/libGAL_egl.so # for i in egl glesv1_cm glesv2 vg; do         cp /usr/lib/pkgconfig/${i}_${backend}.pc/usr/lib/pkgconfig/${i}.pc     done #rm /usr/lib/*-dfb.so /usr/lib/*-fb.so /usr/lib/*-wl.so (Optional in case you deploy your kernel version with GPU as module) make vivante kernel module (GPU kernel driver) load on boot: # echo vivante >> /etc/modules # nano /etc/udev/rules.d/10-imx.rules KERNEL=="galcore",  MODE="0660", GROUP="video" KERNEL=="mxc_asrc",  MODE="0666" Create an xorg.conf configured for the Vivante fbdev driver: # nano /etc/X11/xorg.conf Section "Device"     Identifier "i.MX Accelerated Framebuffer Device"     Driver "vivante"     Option "fbdev" "/dev/fb0"     Option "vivante_fbdev" "/dev/fb0"     Option "HWcursor" "false" EndSection Section "ServerFlags"     Option "BlankTime"  "0"     Option "StandbyTime"  "0"     Option "SuspendTime"  "0"     Option "OffTime"  "0" EndSection # cd .. Make sure the files copied into the correct places. If all compiled and copied, you should now see a bunch of new libraries in /usr/lib! Congratulations! After you finish you can reboot your system and start playing. Testing Gstreamer examples: show gstreamer-imx plugins: # gst-inspect-1.0 | grep imx imxvpu:  imxvpuenc_mjpeg: Freescale VPU motion JPEG video encoder imxvpu:  imxvpuenc_mpeg4: Freescale VPU MPEG-4 video encoder imxvpu:  imxvpuenc_h264: Freescale VPU h.264 video encoder imxvpu:  imxvpuenc_h263: Freescale VPU h.263 video encoder imxvpu:  imxvpudec: Freescale VPU video decoder imxv4l2videosrc:  imxv4l2videosrc: V4L2 CSI Video Source imxg2d:  imxg2dcompositor: Freescale G2D video compositor imxg2d:  imxg2dvideotransform: Freescale G2D video transform imxg2d:  imxg2dvideosink: Freescale G2D video sink imxipu:  imxipucompositor: Freescale IPU video compositor imxipu:  imxipuvideosink: Freescale IPU video sink imxipu:  imxipuvideotransform: Freescale IPU video transform imxpxp:  imxpxpvideotransform: Freescale PxP video transform imxpxp:  imxpxpvideosink: Freescale PxP video sink imxipuvideosink: # gst-launch-1.0 videotestsrc ! imxipuvideosink imxg2dvideosink: # gst-launch-1.0 videotestsrc ! imxg2dvideosink The imxeglvivsink allows hardware accelerated display to a window on the X11 host # export DISPLAY=:0.0 # gst-launch-1.0 videotestsrc ! imxeglvivsink To test if you have graphics support you can run any glmark2 and/or mesa-utils or can run example of the next route: # cd /opt/viv_samples/vdk/ # ./tutorial1                                                                      //any example root@imx6Q:~# glxgears -info GL_RENDERER   = Vivante GC2000 GL_VERSION    = 2.1 2.0.1 GL_VENDOR     = Vivante Corporation GL_EXTENSIONS = WGL_ARB_extensions_string WGL_EXT_extensions_string WGL_EXT_swap_control GL_EXT_texture_env_add GL_ARB_multitexture GL_ARB_multisample GL_ARB_texture_env_add GL_ARB_texture_compression GL_ARB_texture_env_combine GL_ARB_depth_texture GL_ARB_window_pos …. 1606 frames in 5.0 seconds = 321.130 FPS 1650 frames in 5.0 seconds = 329.834 FPS L_RENDERER   = Vivante GC2000 GL_VERSION    = 2.1 2.0.1 GL_VENDOR     = Vivante Corporation1629 frames in 5.0 seconds = 325.644 FPS 1621 frames in 5.0 seconds = 324.072 FPS 1650 frames in 5.0 seconds = 329.806 FPS 1651 frames in 5.0 seconds = 330.079 FPS