Here are some debug methods for kernel performance requirements or related issues. It includes all the common methods such as oops/panic issues, memory issues, and so on. Please check it in the attachments for details. OS and System analysis Oops/Panic case addr2line objdump gdb Pstore Kdump Memory debugging SLAB KASAN Kmemleak Performance Perf Ftrace eBPF/bcc

Symptoms   When configure a gpio pin for a driver in the dts/dtsi file like below example,   e.g.   a-switch {            compatible = "a-switch-driver";            pinctrl-names = "default";            pinctrl-0 = <&pinctrl_switch>;            gpios = <&lsio_gpio1 1 GPIO_ACTIVE_HIGH>;            status = "okay"; };   pinctrl_switch: switch_gpio {     fsl,pins = < IMX8QXP_SPI2_SDO_LSIO_GPIO1_IO01    0x21 >; };   then you may get the error when request the gpio in the driver during the kernel boot up.   Error message like this: a-switch: failed to request gpio a-switch: probe of a-switch failed with error -22   Linux version: L5.4.x   Diagnosis   Because the gpio_mxc_init function run before the function imx_scu_driver_init. The pm_domains for gpio is not ready before running mxc_gpio_probe, so gpio request will be failed.     Solution   There are two ways to resolve this issue 1. Build the driver as a module. i.e. select the driver in kernel’s menuconfig as “M”. Then , run “insmod” to load the driver after the kernel boot up.   OR   2. Apply below patch, let gpio driver init after scu driver. diff --git a/drivers/gpio/gpio-mxc.c b/drivers/gpio/gpio-mxc.c index 1dfe513f8fcf..52b5799040b3 100644 --- a/drivers/gpio/gpio-mxc.c +++ b/drivers/gpio/gpio-mxc.c @@ -892,7 +892,7 @@ static int __init gpio_mxc_init(void) return platform_driver_register(&mxc_gpio_driver); } -subsys_initcall(gpio_mxc_init); +device_initcall(gpio_mxc_init);  

adv7180 is the 8 bits parallel CSI interface TVin to iMX8QXP validation board. Its weaving mode de-interlace can be supported on both iMX8QXP B0 and C0 chips, but blending mode de-interlace can only work on iMX8QXP C0 chips.   ISL79987 is the 4 virtual channel TVin chip which can input 4 CVBS cameras to iMX8QXP with MIPI CSI2 inteface, it can only work with iMX8QXP C0 chips. The iMX8QXP B0 chips have MIPI CSI2 virtual channel errata.   To test the capture to file: $ /unit_tests/V4L2/mx8_v4l2_cap_drm.out -cam 1 -num 300 -fmt YUYV -of   To test the preview on screen: $ killall weston $ /unit_tests/V4L2/mx8_v4l2_cap_drm.out -cam 1 -fmt RGBP -num 30000   Note: 1. For weaving mode de-interlace, when the ISI is doing de-interlace, it can't do CSC at the same time, so preview will get color issue, because the real output video is always YUYV format. 2. For blending mode de-interlace, it must use ISI0, so for ISL79987, only one camera can use blending mode, the other three cameras are still using weaving mode. The preview color is OK for such use case. 3. The patch is for L4.19.35 BSP.     2019-11-14 update: Add the test application "mx8_v4l2_cap_drm.tar.gz" to support YUYV render to display. Test command to render 4 weaving mode cameras:    ./mx8_v4l2_cap_drm.out -cam 0xF -fmt YUYV -num 30000     2020-04-29 update: Add "0006-isl7998x-fix-the-mipi_bps-overwrite-issue-from-set_f.patch", it fixed the issue that MIPI bps information in isl7998x_data->format.reserved[0] had been overwritten by isl7998x_set_fmt().   2021-06-11 update: Added the patches based on L5.4.70_2.3.0 GA BSP.

1 - Introduction: The Ultra Secured Digital Host Controller (uSDHC) provides the interface between the host processor and the SD/SDIO/MMC cards. Most recent versions provides the ability to automatically select a quantized delay (in fractions of the clock period) regardless of on-chip variations such as process, voltage, and temperature (PVT). The auto tuning is performed during runtime at hardware level, no software enablement is needed to drive this feature. 2 - Failure description: SDIO cards can implement an optional feature that uses DATA[1] to signal the card's interrupt to the i.MX device, this feature can be enabled by the SDIO card device and does not depends on i.MX uSDHC driver configuration. NXP Linux BSP is enabling the auto tuning for high SDIO frequencies (SDR104 and SDR50). Out of reset uSDHC_VEND_SPEC2 register is configured to use DATA[3:0] for calibration, this setup can conflict with the SDIO interrupt as DATA[1] signal can be asserted asynchronously. SDIO failures can be observed when running SDIO applications that requires high usage of the SDIO interface (e.g Download of large files), SDIO controller cannot return an accurate DLL causing failures such as "CMD53 read error". Failure can be observed on i.MX8MM EVK and i.MX8MN EVK boards, both devices are running 88w8987 Wi-Fi chipset at 208Mhz (SDR104). Users can observe an SDIO crash followed by error message below at Linux Kernel level. [ 401.945627] cmd53 read error=-84 [ 401.974677] moal_read_data_sync: read registers failed 3 - Impacted devices: The following devices are impacted by this limitation. - i.MX6 Family:   i.MX6SL, i.MX6SLL, i.MX6SX, i.MX6UL, i.MX6ULZ and i.MX6ULL. - All i.MX7 and i.MX7ULP family:   i.MX7D, i.MX7S and i.MX7ULP. - All i.MX8M Family:   i.MX8MQuad, i.MX8M Mini, i.MX8M Nano, i.MX8M Nano UL and i.MX8M Plus. - All i.MX8/8X Family:   i.MX8DQXP, i.MX8DX and i.MX8QM. NXP Linux BSP is enabling the auto tuning for SDR104 and SDR50 modes. Other operation modes are not impacted by this limitation. Users can poll uSDHCx_CLK_TUNE_CTRL_STATUS register when running SDIO applications to confirm. TAP_SEL_PRE field is updated automatically during run time and constant variations can point to an incorrect delay cell calculated by the uSDHC controller.   All NXP Wi-Fi chipsets are enabling SDIO interrupt during firmware load, failures can be observed with any Wi-Fi vendor enabling SDIO asynchronous interrupt. 4 - Software changes: Recommendation is to enable auto tuning for DATA[0] and CMD signals only, DATA[1] should not be used for auto calibration to avoid a possible conflict with SDIO interrupt. This setup can only be used if SDIO interface length are well matched. Software patches can be found at codeaurora.org. Fix is already included in L5.10.52-2.1.0 BSP, users can add fsl,sdio-interrupt-enabled property to uSDHC device tree node to enable SW workaround. https://source.codeaurora.org/external/imx/linux-imx/commit/?h=lf-5.10.y&id=3b3d6dec05277f7786d813592a31ea4a1ce60a74 https://source.codeaurora.org/external/imx/linux-imx/commit/?h=lf-5.10.y&id=b9b5a43df1d709809b2b654ad8f8181b00a4ee55 https://source.codeaurora.org/external/imx/linux-imx/commit/?h=lf-5.10.y&id=95a846af9f82dc6ea60064d9d12d5d2378e23941  

This is based on L5.10.35 BSP where you have to install QT static build: Qt 5.15 static build: Assuming your sysroot is at "/sysroot-cross" and your toolchain is at "/Toolchain" your qt-source is at /Qt-5.15 PATH=/sysroot-cross/bin:/sysroot-cross/sbin:/Toolchain/bin mkdir /Qt-5.15/mkspecs/qws/linux-imx6-g++ create in this dir the textfile "qmake.conf" with this content: ####################### snip qmake.conf ############################## include(../../common/linux.conf) include(../../common/qws.conf) # modifications to g++.conf QMAKE_CC                = arm-linux-gnueabi-gcc QMAKE_CFLAGS            = -pipe -isystem /sysroot-cross/include -isystem /sysroot-cross/usr/include QMAKE_CXX               = arm-linux-gnueabi-g++ QMAKE_CXXFLAGS          = -pipe -isystem /sysroot-cross/include -isystem /sysroot-cross/usr/include QMAKE_INCDIR            = /sysroot-cross/include /sysroot-cross/usr/include QMAKE_LIBDIR            = /sysroot-cross/lib /sysroot-cross/usr/lib QMAKE_LINK              = arm-linux-gnueabi-g++ QMAKE_LINK_SHLIB        = arm-linux-gnueabi-g++ QMAKE_LFLAGS            = -L/sysroot-cross/lib -L/sysroot-cross/usr/lib -Wl,-rpath-link -Wl,/sysroot-cross/lib QMAKE_LFLAGS           += -Wl,-rpath-link -Wl,/sysroot-cross/usr/lib #Opengl QMAKE_INCDIR_OPENGL = /Vivante/include QMAKE_INCDIR_OPENGL += /Vivante/include/GL QMAKE_INCDIR_OPENGL += /Vivante/include/EGL QMAKE_INCDIR_OPENGL += /Vivante/include/GLES2 QMAKE_LIBDIR_OPENGL = /Vivante/lib QMAKE_INCDIR_OPENGL_ES1 = $$QMAKE_INCDIR_OPENGL QMAKE_LIBDIR_OPENGL_ES1 = $$QMAKE_LIBDIR_OPENGL QMAKE_INCDIR_OPENGL_ES1CL = $$QMAKE_INCDIR_OPENGL QMAKE_LIBDIR_OPENGL_ES1CL = $$QMAKE_LIBDIR_OPENGL QMAKE_INCDIR_OPENGL_ES2 = /Vivante/include QMAKE_INCDIR_OPENGL_ES2 += /Vivante/include/EGL QMAKE_INCDIR_OPENGL_ES2 += /Vivante/include/GLES2 QMAKE_LIBDIR_OPENGL_ES2 = $$QMAKE_LIBDIR_OPENGL QMAKE_INCDIR_EGL = $$QMAKE_INCDIR_OPENGL_ES2 QMAKE_LIBDIR_EGL = $$QMAKE_LIBDIR_OPENGL QMAKE_LIBS_EGL = -lEGL -lGAL -lGLESv2 -lGLES_CM QMAKE_LIBS_OPENGL_ES2 = -lEGL -lGAL -lGLESv2 -lGLES_CM QMAKE_LIBS_OPENGL = -lEGL -lGAL -lGLESv2 -lGLES_CM QMAKE_LIBS_OPENGL_QT = -lEGL -lGAL -lGLESv2 -lGLES_CM QMAKE_LIBS_OPENGL_ES1 = QMAKE_LIBS_OPENGL_ES1CL = # modifications to linux.conf QMAKE_AR                = arm-linux-gnueabi-ar cqs QMAKE_OBJCOPY           = arm-linux-gnueabi-objcopy QMAKE_STRIP             = arm-linux-gnueabi-strip QMAKE_CFLAGS_RELEASE   = -pipe -isystem /sysroot-cross/include -isystem /sysroot-cross/usr/include load(qt_config) ####################### snip qmake.conf ############################## create in the same dir the text file "qplatformdefs.h" ####################### snip qplatformdefs.h ############################## #include "../../linux-g++/qplatformdefs.h" ####################### snip qplatformdefs.h ############################## now goto dir /Qt-5.15 cd /Qt-5.15 call configure with ./configure -opensource -confirm-license -release -no-rpath -no-fast \     -no-sql-ibase -no-sql-mysql -no-sql-odbc -no-sql-psql -no-sql-sqlite2 \     -no-qt3support -no-mmx -no-3dnow -no-sse -no-sse2 -no-sse3 -no-ssse3 \     -no-sse4.1 -no-sse4.2 -no-avx -no-optimized-qmake -no-nis -no-cups -pch \     -reduce-relocations -force-pkg-config -prefix /usr -no-armfpa -make libs \     -nomake docs -little-endian -embedded armv6 -qt-decoration-styled \     -depths all -xplatform qws/linux-imx6-g++ -iconv -largefile -qt-gfx-linuxfb \     -qt-gfx-multiscreen -qt-mouse-pc -qt-mouse-linuxinput -qt-libpng \     -plugin-gfx-directfb -system-zlib -no-accessibility -no-gfx-transformed \     -no-gfx-qvfb -no-gfx-vnc -no-kbd-tty -no-kbd-linuxinput -no-kbd-qvfb \     -no-mouse-linuxtp -no-mouse-tslib -no-mouse-qvfb -no-libmng -no-libtiff \     -no-gif -no-libjpeg -no-freetype -no-stl -no-glib -no-openssl -no-egl \     -no-xmlpatterns -no-exceptions -no-multimedia -no-audio-backend -no-phonon \     -no-phonon-backend -no-webkit -no-script -no-scripttools -no-svg -no-script \     -no-declarative -no-sql-sqlite -no-qdbus -no-opengl -static -nomake tools \     -nomake examples -nomake demos when configuring is finished call make after a looong time, when everything goes right, we have a staticly compiled Qt. DO NOT call "make install". We will install manually: copy from /Qt-5.15/bin the files moc, uic, rcc and qmake to somewhere in PATH, eg. /sysroot-cross/bin copy the contents of dir /Qt-5.15/mkspecs to /sysroot-cross/usr/mkspec copy the contents of dir /Qt-5.15/plugins to /sysroot-cross/usr/plugins copy the contents of dir /Qt-5.15/include to /sysroot-cross/usr/include copy the contents of dir /Qt-5.15/lib to /sysroot-cross/usr/lib Test application camtest: if you don't have/want directfb plugin remove from camtest.pro the lines LIBS += -L/sysroot-cross/usr/plugins/gfxdrivers QTPLUGIN += QDirectFBScreen and the lines from main.cpp #include <QtPlugin> Q_IMPORT_PLUGIN(qdirectfbscreen) generate makefile by typing /sysroot-cross/bin/qmake -spec /sysroot-cross/usr/mkspecs/qws/linux-imx6-g++ camtest.pro then make you should set and activate your framebuffers with this script ################# snip ################################ fbset -fb /dev/fb0 -g 1024 768 1024 2304 16 echo -n 0 > /sys/class/graphics/fb0/blank fbset -fb /dev/fb1 -g 1024 768 1024 1536 32 echo -n 0 > /sys/class/graphics/fb1/blank modprobe galcore modprobe uvcvideo modprobe mxc_v4l2_capture ################# snip ################################ if you use directfb then your /etc/directfbrc file should look like this: ######################## snip /etc/directfbrc ############# system=fbdev fbdev=/dev/fb1 mode=1024x768 depth=32 pixelformat=ARGB no-cursor window-surface-policy=systemonly ######################## snip /etc/directfbrc ############# to start the application with directfb: ./camtest -qws -display directfb without directfb using linuxfb: ./camtest -qws -display linuxfb:/dev/fb1 Notes about application: 1. The application shows 2 webcams in background-framebuffer (BG-FB). The foreground-framebuffer (FG-FB) shows the qt-gui. FG-FB is configured to be fully opaque and uses color-keying. On the BG-FB one cam is overlayed on the other cam using IPU. Optimization possibilities: the app copies the frames from the cams with memcpy. This wouldn't be necessary, when the kernel usb-webcam interface (uvc) would support V4L2_MEMORY_USERPTR method. through this way, you could pass the mapped IPU mmapped inbufs directly to v4l2 output buffers. If you get errors like NOSPC (-28) from uvc, this is a limitation of USB. My board is a MX6QSabre, where the two webcams are connected to the same usb-controller. With both webcams I had to limit the frame size to 320x250 and 160x120 at 25Hz. You might try higher res if you have other type of webcams (not usb). Have fun  

The i.MX 8QuadXPlus Multisensory Enablement Kit (MEK) is a NXP development platform based on Cortex A-35 + Cortex-M4 cores. Built with high-level integration to support graphics, video, image processing, audio, and voice functions, the i.MX 8X processor family is ideal for safety-certifiable and efficient performance requirements. This tutorial shows how to enable the Cortex-M4 using the MCUXpresso SDK package and loading the binary from the network. NOTE: It is also possible to load the Cortex-M4 image from the SCFW using the imx-mkimage utility. But now we are going to focus on MCUXpresso. Setting up the machine   Install cmake on the host machine: $ sudo apt-get install cmake Download the armgcc toolchain and export the location as ARMGCC_DIR: $ export ARMGCC_DIR=<your_path_to_arm_gcc>/gcc-arm-none-eabi-9-2020q2/ NOTE: The ARMGCC_DIR variable needs to be exported on the terminal used for compilation. To setup the TFTP server on the host machine: Configuring your Host PC for TFTPPermalink   The first step is to install all the prerequisite packages for TFTP: $ sudo apt-get install xinetd tftpd tftp Create a TFTP folder in your desired location with root owner and the “rwx” permission for all users: $ sudo mkdir /tftpboot $ sudo chmod –R 777 /tftpboot $ sudo chown –R root /tftpboot Create a configuration file for the TFTP with the following content. (The server_args parameter must match with the folder created above) $ cat /etc/xinetd.d/tftp service tftp { protocol = udp port = 69 socket_type = dgram wait = yes user = root server = /usr/sbin/in.tftpd server_args = -s /tftpboot disable = no } Restart the xinetd service: $ sudo /etc/init.d/xinetd restart You can place any file at the TFTP folder and load it through U-Boot, you can also create symbolic links from your building directory avoiding to copy and paste your zImage and dtb files every time. Configuring your Host PC for NFSPermalink   Install all the needed packages for NFS: $ sudo apt-get install nfs-kernel-server Create a folder for placing your rootfs: $ mkdir /tftpboot/rfs Add the following line in the end of your /etc/exports file: /tftpboot/rfs *(rw,no_root_squash,no_subtree_check) Restart the NFS service: $ sudo service nfs-kernel-server restart Place your rootfs or create a symbolic link for the NFS folder.    Downloading the SDK Download the MCUXpresso following these steps: Click on “Select Development Board”; Select MEK-MIMX8QX under “Select a Device, Board, or Kit” and click on “Build MCUXpresso SDK” on the right; Select “Host OS” as Linux and “Toolchain/IDE” as GCC ARM Embedded; Add “FreeRTOS” and all the wanted Middleware and hit “Request Build”; Wait for the SDK to build and download the package. Building the image All demos and code examples available on the SDK package are located in the directory <<SDK_dir>>/boards/mekmimx8qx/. This tutorial shows how to build and flash the hello_world demo but similar procedures can be applied for any example (demo, driver, multicore, etc) on the SDK. To build the demo, enter the armgcc folder under the demo directory and make sure that the ARMGCC_DIR variable is set correctly. $ cd ~/SDK_2.3.0_MEK-MIMX8QX/boards/mekmimx8qx/demo_apps/hello_world/armgcc $ export ARMGCC_DIR=<your_path_to_arm_gcc>/gcc-arm-none-eabi-9-2020q2/ Run the build_release.sh script to build the code. $ ./build_release.sh NOTE: If needed, give the script execution permission by running chmod +x build_release.sh. This generates the M4 binary (hello_world.bin) under the release folder. Copy this image to the /tftpboot/ directory on the host PC. NOTE: This procedure shows how to build the M4 image that runs on TCM. To run the image from DDR, use the build_ddr_release.sh script to build the binary under the ddr_release folder. Flashing the image Open two serial consoles, one for /dev/ttyUSB0 for Cortex-A35 to boot Linux, and one for /dev/ttyUSB1 for Cortex-M4 to boot the SDK image. On the A35 console, with a SD Card with U-Boot, stop the booting process and enter the following commands to load the M4 binary to TCM: => dhcp => setenv serverip <ip_from_host_pc> => tftp 0x88000000 hello_world.bin => dcache flush => bootaux 0x88000000 Then the M4 core will load the image to the /dev/ttyUSB1 console.    

SW Environment Setup: 1. Prepare L5.10.35 Yocto and build Image  The prebuilt image also is available and useable. 2. Flash image to the SD card  Refer to the Yocto User Guide. 3. Compile flash.bin without M4 and flash it to sdcard (flash.bin as attachment)  make SOC=iMX8QM flash sudo dd if=flash.bin of=/dev/sde bs=1k seek=32 conv=fsync HW Environment Setup: Prepare the imx8qm MEK CPU board and base board and DB9 male cable, connect to CAN0 and CAN1 female connector on base board. (Pin to Pin connection) User Case: 1. Power on board and configure specify dtb file in uboot  setenv fdt_file imx8qm_mek.dtb 2. Boot up and config bitrate for can0 and can1 in kernel root@imx8qmmek:~# ip link set can0 up type can bitrate 500000 root@imx8qmmek:~# ip link set can1 up type can bitrate 500000 3. Check CAN0 and CAN1 devices root@imx8qmmek:~# ifconfig can0: flags=193<UP,RUNNING,NOARP> mtu 16 unspec 00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00 txqueuelen 10 (UNSPEC) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 0 bytes 0 (0.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 85 can1: flags=193<UP,RUNNING,NOARP> mtu 16 unspec 00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00 txqueuelen 10 (UNSPEC) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 0 bytes 0 (0.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 device interrupt 86 3. Run candump for CAN1 root@imx8qmmek:~# candump can1 & [1] 1215 [ 65.624580] can: controller area network core [ 65.630225] NET: Registered protocol family 29 [ 65.641158] can: raw protocol 4. Run cansend for CAN0 root@imx8qmmek:~# cansend can0 5A1#11.2233.44556677.88 can1 5A1 [8] 11 22 33 44 55 66 77 88   The above red is output result from CAN1.

This note show how to use the open source gstreamer1.0-rtsp-server package on i.MX6QDS and i.MX8x to stream video files and camera using RTP protocol.  The i.MX 6ULL and i.MX 7 doesn't have Video Processing Unit (VPU). Real Time protocol is a very common network protocol for delivering media over IP networks. On the board, you will need a GStreamer pipeline that encodes the raw video, adds the RTP payload, and sends over a network sink. A generic pipeline would look as follows: video source ! video encoder ! RTP payload ! network sink Video source: often it is a camera, but it can be a video from a file or a test pattern, for example. Video encoder: a video encoder as H.264, H.265, VP8, JPEG and others. RTP payload: an RTP payload that matches the video encoder. Network sink: a video sync that streams over the network, often via UDP.   Prerequisites: MX6x o MX8x board with the L5.10.35 BSP installed. A host PC with either Gstreamer or VLC player installed. Receiving h.264/h.265 Encoded RTP Video Stream on a Host Machine Using GStreamer GStreamer is a low-latency method for receiving RTP video. On your host machine, install Gstreamer and send the following command: $ gst-launch-1.0 -v udpsrc port=5000 caps = "application/x-rtp, media=(string)video, clock-rate=(int)90000, encoding-name=(string)H264, payload=(int)96" ! rtph264depay ! decodebin ! videoconvert ! autovideosink sync=false   Using Host PC: VLC Player Optionally, you can use VLC player to receive RTP video on a PC. First, in your PC, create a sdp file with the following content:  stream.sdpv=0m=video 5000 RTP/AVP 96c=IN IP4 127.0.0.1a=rtpmap:96 H264/90000 After this, with the GStreamer pipepline on the device running, open this .sdp file with VLC Player on the host PC. Sending h.264 and h.265 Encoded RTP Video Stream GStreamer provides an h.264 encoding element by software named x264enc. Use this plugin if your board does not support h.264 encoding by hardware or if you want to use the same pipeline on different modules. Note that the video performance will be lower compared with the plugins with encoding accelerated by hardware. # gst-launch-1.0 videotestsrc ! videoconvert ! x264enc ! rtph264pay config-interval=1 pt=96 ! udpsink host=<host-machine-ip> port=5000 Note: Replace <host-machine-ip> by the IP of the host machine. In all examples you can replace videotestsrc by v4l2src element to collect a stream from a camera   i.MX8X # gst-launch-1.0 videotestsrc ! videoconvert ! v4l2h264enc ! rtph264pay config-interval=1 pt=96 ! udpsink host=<host-machine-ip> port=5000   i.MX 8M Mini Quad/ 8M Plus # gst-launch-1.0 videotestsrc ! videoconvert ! vpuenc_h264 ! rtph264pay config-interval=1 pt=96 ! udpsink host=<host-machine-ip> port=5000 i.MX6X The i.MX6QDS does not support h.265 so the h.264 can work: # gst-launch-1.0 videotestsrc ! videoconvert ! vpuenc_h264 ! rtph264pay config-interval=1 pt=96 ! udpsink host=<host-machine-ip> port=5000   Using Other Video Encoders While examples of streaming video with other encoders are not provided, you may try it yourself. Use the gst-inspect tool to find available encoders and RTP payloaders on the board: # gst-inspect-1.0 | grep -e "encoder"# gst-inspect-1.0 | grep -e "rtp" -e " payloader" Then browse the results and replace the elements in the original pipelines. On the receiving end, you will have to use a corresponding payload. Inspect the payloader element to find the corresponding values. For example: # gst-inspect-1.0 rtph264pay   Install rtp in your yocto different form L5.10.35 BSP, to install gstreamer1.0-rtsp-server in any Yocto Project image, please follow the steps below: Enable meta-multimedia layer: Add the following on your build/conf/bblayers.conf: BBLAYERS += "$"${BSPDIR}/sources/meta-openembedded/meta-multimedia" Include gstreamer1.0-rtsp-server into the image: Add the following on your build/conf/local.conf: IMAGE_INSTALL_append += "gstreamer1.0-rtsp-server" Run bitbake and mount your sdcard. Copy the binaries: Access the gstreamer1.0-rtsp-server examples folder: $ cd /build/tmp/work/cortexa9hf-vfp-neon-poky-linux-gnueabi/gstreamer1.0-rtsp-server/$version/build/examples/.libs Copy the test-uri and test-launch to the rootfs /usr/bin folder. $ sudo cp test-uri test-launch /media/USER/ROOTFS_PATH/usr/bin Be sure that the IPs are correctly set: SERVER: => ifconfig eth0 $SERVERIP CLIENT: => ifconfig eth0 $CLIENTIP Video file example SERVER: => test-uri file:///home/root/video_file.mp4 CLIENT: => gst-launch-1.0 playbin uri=rtsp://$SERVERIP:8554/test You can try to improve the framerate performance using manual pipelines in the CLIENT with the rtspsrc plugin instead of playbin. Follow an example: => gst-launch-1.0 rtspsrc location=rtsp://$SERVERIP:8554/test caps = 'application/x-rtp'  ! queue max-size-buffers=0 ! rtpjitterbuffer latency=100 ! queue max-size-buffers=0 ! rtph264depay ! queue max-size-buffers=0 ! decodebin ! queue max-size-buffers=0 ! imxv4l2sink sync=false   Camera example SERVER: => test-launch "( imxv4l2src device=/dev/video0 ! capsfilter caps='video/x-raw, width=1280, height=720, framerate=30/1, mapping=/test' ! vpuenc_h264 ! rtph264pay name=pay0 pt=96 )" CLIENT: => gst-launch-1.0 rtspsrc location=rtsp://$SERVERIP:8554/test ! decodebin ! autovideosink sync=false The rtspsrc has two properties very useful for RTSP streaming: Latency: Useful for low-latency RTSP stream playback (default 200 ms); Buffer-mode: Used to control buffer mode. The slave mode is recommended for low-latency communications. Using these properties, the example below gets 29 FPS without a sync=false property in the sink plugin. The key achievement here is the fact that there is no dropped frame: => gst-launch-1.0 rtspsrc location=rtsp://$SERVERIP:8554/test latency=100 buffer-mode=slave ! queue max-size-buffers=0 ! rtph264depay ! vpudec ! imxv4l2sink      

Use case: iMX8QXP system can be a video input source to another system.   Hardware Pins： LCDIF_D00 ~ LCDIF_D07 LCDIF_CLK LCDIF_VSYNC LCDIF_HSYNC LCDIF_EN   Reference patch： It is based on L5.4.70_2.3.0 GA BSP.  File: L5.4.70_2.3.0-iMX8QXP-LCDIF-add-YUV422-8-bits-output.patch Customer can change the timing parameters in file "panel-lcdif-yuv422.c" as needed, the default timing is a 1280x720 P30 mode: static const struct display_timing yuv422_lcd_timing = {     .pixelclock = { 74250000, 74250000, 74250000 },     .hactive = { 1280, 1280, 1280 },     .hfront_porch = { 220, 220, 220 },     .hback_porch = { 110, 110, 110 },     .hsync_len = { 40, 40, 40 },     .vactive = { 720, 720, 720 },     .vfront_porch = { 20, 20, 20 },     .vback_porch = { 5, 5, 5 },     .vsync_len = { 5, 5, 5 },     .flags = DISPLAY_FLAGS_DE_HIGH, };   Test application: drm_test_yuv.zip: it can set framebuffer to UYVY mode, in this case, no CSC is needed, the data in framebuffer memory will be same as output on display data interface. drm_test_rgb.zip: it can set framebuffer to RGBA mode, in this case, RGB to YUV CSC is needed, application can draw RGB data into framebuffer as normal, the LCDIF will convert it to YUV422 format on the fly, then output the YUV data to display interface.    

This article will describe one suggestion for one issue that UART continuously generate RX interrupts and receive 0xFF even when Rx line is continuously high in some cases on imx6 series. Below I will explain with imx6DL. Some settings are just to make it easier to reproduce. BSP version: L5.4.70-2.3.0 Board HW: MCIMX6DL-SDB When issue happen Config imx6DL UART3 as the serial port to 1200 baud, 8-N-1 format. Keep the RX Line high. Make the RX line low and keep it for a short time (360 usec-370 usec).   At this condition, you will find that the UART will continuously generate RX interrupts and show receiving 0xFF even you make the RX line return to be high. Why issue happen The low time is not in the correct range and out of our spec. In the imx6DL AEC document, there is one chapter named UART Receiver like blow   If using 1200 band, that means one valid bit time is 833 usec. And there is a definition that “tolerate 1/(16 x Fbaud_rate) tolerance in each bit”. That’s means in the case of 1200 baud. A range of valid bit is 781 to 885 usec. But is reproducing, the Low level time is 360 usec. This time out of range will make UART state machine to be confused. How to fix Actually, the best way is following our spec. If there is such an unknown situation in the customer’s environment, then the following method could be regarded as a suggestion to fix the issue meet by the customer. The interrupt handler will check USR1[AWAKE]. 2    If AWAKE is asserted, clear it and proceed as usual (assume we have valid data), else, check if USR1[AGTIM] is asserted. 3    if AGTIM is asserted, clear it and proceed as usual, else perform software reset (assume we have invalid data). Checking AGTIM is for one race condition when the RX fifo has some characters (less than RXTL) but no more data is coming in. When following this procedure, the UART will perform a software reset when a block interrupt occurs. Notes: From customer report, error could be cleared if a valid start-bit is detected on the RX line. This needs to be verified by the customer themselves. Test code has been included in attachment.   Besst Regards

The A53 Debug Console Changing consists in several major updates like: RDC settings, Pinmux, Clocks and Ecosystem Updates.

Some case need configure the GPIO as power off button. One solution is to use “gpio-keys” to send the “KEY_POWER” event to the system. Co-work with systemd, system gets power off.  

 

     

Steps to replace the Wi-Fi/Bluetooth firmware on the i.MX 8M series on Linux    Applicable to versions L5.4.47, L5.4.70, L5.10.9   1. Download the newest firmware. you can download the attachment in this thread and unzip it. 2. Copy it to the EVK board. 3. Copy the firmware to /lib/firmware/nxp   root@imx8mmevk: cp pcieuart8997_combo_v4.bin sdiouart8987_combo_v0.bin  /lib/firmware/nxp If the Linux version is L5.4.3，Then the step3 is to copy firmware to lib/firmware/mrvl/ root@imx8mmevk: cp pcieuart8997_combo_v4.bin sdiouart8987_combo_v0.bin  /lib/firmware/mrvl    

Purpose This is early communication to notify i.MX 8M Dual/8M QuadLite/8M Quad customers of a potential incorrect PCIe power supply configuration on certain NXP BSP Linux and Android versions. Description The PCIE_VPH power supply is selectable in software  between 1.8V and 3.3V. When the PCIE_VPH supply is configured to operate at 3.3V, the 1.8V internal regulator (disabled by default) must be enabled to prevent overstress conditions on the PCIe PHY. If the 1.8V internal regulator is left disabled when the PCIE_VPH supply is configured to operate at 3.3V, it could potentially impact the product lifetime of the device.   Impact •i.MX 8M Dual/8M QuadLite/8M Quad (other i.MX processors are not impacted) •Only Impacts Linux/Android kernel versions earlier than L5.4.70_2.3.2 or Linux 5.10.9_1.0.0 releases MITIGATION •When the PCIE_VPH supply is configured to operate at 3.3V users need to enable the internal regulator by setting the IOMUXC_GPR_GPR14 and IOMUXC_GPR_GPR16 registers - PCIE1_VREG_BYPASS and PCIE2_VREG_BYPASS bit to 0. •There are 3 software patches for each release. Software patch details in the Code Aurora Forum (CAF): •For L5.4.70_2.3.2 patch release, the git log references are: •MLK-25349-3 PCI: imx: clear vreg bypass when pcie vph voltage is 3v3 •MLK-25349-2 arm64: dts: imx8mq-evk: add one regulator used to power up pcie phy •MLK-25349-1 dt-bindings: imx6q-pcie: add one regulator used to power up pcie phy • •The L5.4.70_2.3.2, LF_5.10 Q2 and later BSP releases correctly configure and enable the internal regulator by setting the IOMUXC_GPR_GPR14 and IOMUXC_GPR_GPR16 registers The Patch MLK-25349 which correctly enables the internal regulator is already included in the L5.4.70_2.3.2 patch release and release versions after it. MITIGATION •The following branches of Linux/Android BSP releases contain the MLK-25349 patch. The patch is attached below for each respective release.   •Other branches which are not listed should try to apply the nearest Patch version patch. If a user encounters any conflicts in applying, they should back porting from below nearest patch release version below. imx_4.9.51_ga, imx_4.9.y_android_imx8m_ga_v2                           - Patch attached  imx_4.9.88_ga, imx_4.9.y_android_2.0.0_ga                                   - Patch attached  imx_4.14.y and imx_4.14.98_2.3.0, imx_4.14.98_2.3.0_android     - Patch attached  imx_4.19.y and imx_4.19.35_1.1.0, imx_4.19.35_1.1.0_android     - Patch attached  imx_5.4.y, imx_5.4.3_2.0.0, imx_5.4.3_2.0.0_android                     - Patch attached Documentation Change Description – 1 of 3 for Datasheet Updated Datasheets and Reference Manual will be published to nxp.com. Updated Hardware Design guide and Schematics have already been published on nxp.com.  Updated the descriptions of PCIE_VPH in the Datasheet Table 8, "Operating ranges"     Documentation Change Description – 2 of 3 for Reference Manual (RM) Updated the description of field 12 "PCIE1_VREG_BYPASS" in 8.2.4.15 GPR14 General Purpose Register (IOMUXC_GPR_GPR14)           Documentation Change Description – 3 of 3 for RM Updated the description of field 12 "PCIE2_VREG_BYPASS" in 8.2.4.17 GPR16 General Purpose Register (IOMUXC_GPR_GPR16)   REFERENCES •i.MX 8M Dual / 8M QuadLite / 8M Quad Product Lifetime Usage  •i.MX 8M Dual / 8M QuadLite / 8M Quad Applications Processors Data Sheet for Industrial Products •i.MX 8M Dual / 8M QuadLite / 8M Quad Applications Processors Data Sheet for Consumer Products •i.MX 8MDQLQ Hardware Developer’s Guide  •i.MX 8M Dual/8M QuadLite/8M Quad Applications Processors Reference Manual  

Software environment: L5.4.47_2.2.0 Hardware i.MX8QXPC0 EVK board In the uuu script we can see the bootloader imx-boot-imx8qxpc0mek-sd.bin-flash is necessary. The default BSP build generate in the yocto project is with the spl, some customers are confused about the how to build the imx-boot-imx8qxpc0mek-sd.bin-flash. Here I give the manually compile way and generate it in yocto. In the yocto generate it is more convenient than the manually compile way. Hope this can do help for you.

i.MX8DXL DDR3L EVK board, nor flash using is MT25QU512ABB8ESF-0SIT. This doc will show reference of FlexSPI configuration parameters to make booting from MT25Q flash, with QUAD pad and DDR mode. HW: i.MX8DXL DDR3L EVK board SW: Linux 5.4.70 BSP From RM 5.9.3.2 FlexSPI serial flash BOOT operation, the FlexSPI boot flow as :   FlexSPI configuration parameters,  could be think as two kind group: parameter for FlexSPI controller,  parameter related to the operation on nor flash.   Full parameter table check check i.MX8DXL RM Table 5-20. FlexSPI Configuration block. Let us check MT25Q data sheet for its feature, note our target is DDR mode(80MHZ) and QUAD pad:     Now let us change the FlexSPI configuration parameters: 1>readSampleClkSrc , set as 2 , that is loop back from SCK pad; this filed default set as 0, as found default value booting will met failure in this use case, so change to 2. 2>deviceModeCfgEnable set 1, deviceModeSeq.seqNum set 1 , deviceModeSeq. seqId set to 4; deviceModeArg set 0x5f. i.MX8DXL will send some cmd to flash to make MT25Q enter DDR mode and QUAD mode, so deviceModeCfgEnable =1. For seqNum=1, seqId =4; means index 4 of LUT table will store this sequence, and cost one LUT entry. We will explain how to change LUT entry later. For deviceModeArg=0x5f, check MT25Q data sheet, its enhanced volatile register could be write to configure the flash working mode:  3>controllerMiscOption as 0x40, this parameter only for FlexSPI controller itself, means as” External device works using DDR commands”. 4>deviceType=1(Serial Nor),  sflashPadType=4 (QUAD pad),  serialClkFreq=4(80MHZ CLK), these parameter also only for FlexSPI controller. 5>sflashA1Size fill actual size, in terms of bytes 6>LUT entry changes, check 8DXL RM Table 5-21: So LUT entry 0 is sequence for Read command, entry 1 is for Read Status sequence, entry 3 is for Write Enable sequence,  entry 15 is for Dummy command sequence. Other index LUT entry(for example 2,4,6,7,8,10,12,13,14) is could be used for store your sequence for some cmd your flash device neede. We store sequence of writing MT25Q enhance volatile register as LUT entry 4. Check 8DXL RM,  Figure 15-6. LUT and sequence structure:   Each LUT entry (sequence) will using 16 byte,  one sequence consists of up to 8 instructions, each instruction will using 16bit. Each instruction  format as opcode—num_pads—operand. Check RM 15.2.4.8 Programmable Sequence Engine, for supported instructions:   Actually the Write enable sequence is run first before the other sequence, as we will write Mt25Q volatile register, before that need issue Write enable sequence. Check MT25Q data sheet: For this sequence only need one instruction, that is 0x0406, at this time still using is SDR and one pad mode:  Opcode (CMD_SDR),  one pad (0), operand (6).   LUT entry 1, Read status sequence, it is READ STATUS REGISTER (05h) of MT25Q , check data sheet: It use two instructions: 0x0405: opcode(CMD_SDR), pad (one pad), operand (0x5, READ STATUS REGISTER) 0x2404: opcode(READ_SDR), pad (one pad), operand (0x4 , byte number)   LUT entry 4, that is for make MT25Q enter DDR mode and quad pad: From MT25Q data sheet: It will use two instructions, that is 0x0461: opcode (CMD_SDR),  one pad (0), operand (0x61 WRITE ENHANCED VOLATILE CONFIGURATION REGISTER) 0x2001: opcode (WRITE_SDR 08), one pad(0), operand (1 byte data size) The 0x5f will be send out as data.   Next check LUT entry read , at this time MT25Q had enter QUAD pad and DDR mode: LUT entry 0, Read sequence, it is fast read data from MT25Q, from data sheet: will use four instructions , that is : 86ED, opcode (CMD_DDR ), pad ( four pad), operand (0xEDh fast read) 8a18, opcode (RADDR_DDR), pad (four pad), operand (0x18 , three byte address) B210, opcode(DUMMY_ADDR), pad (four pad), operand(0x10, dummy cycle) A604, opcode (READ_DDR), pad (four pad) , operand (0x4, data byte)   Reference: 1.i.MX8DXL Reference Manual 2.MT25Q data sheet              

1.  Introduction 1.1.        Purpose This application note introduces a procedure of how to port AVB/TSN stack and run referring feature demos on i.MX8DXL board. This can help users who want to run AVB/TSN demos to quickly understand and customized their own codes. Since many of the standards are only for TSN switch/bridges and i.MX8DXL is design to be a TSN/AVB endpoint, the demos did not implement a full stack or full standards. They only demonstrated the basic end-to-end point (talker to listener) A/V streaming without bridge or switch. The software used for example in this documentation are based on the opensource such as gstreamer and alsa utils.   1.2.        Overview 1.2.1.     AVB/TSN AVB (Audio Video Bridging) is a common name for the set of technical standards which provide improved synchronization, low-latency, and reliability for switched ethernet network. AVB was initially developed by the IEEE Audio Video Bridging task group of the IEEE 802.1 standards committee. In November 2012, AVB group was renamed to TSN (Time-Sensitive Networking) task group to reflect the expanded scope of its work, which is to provide the specifications that will allow time-synchronized low latency streaming services through IEEE 802 networks. The referring standards shows as follows:     TSN protocol additions QoS components supported in HW TSN MAC + SW driver Managed Object components expose i/f to allow support of standardized network config protocols (local & remote) Transport API to allow other transport layer to use TSN QoS Stack extensions to map traffic priority to application task scheduling Real Time, gPTP based, Best Effort 1.2.2.     Demo introduction   The two streams are defined as below to grantee time sensitive (sub-microsecond synchronization), low latency and bandwidth on the ethernet: Stream A: SR class A, AVTP Audio Format, PCM 16-bit sample, 48 kHz, stereo, 12 frames per AVTPDU. Stream B: SR class B, AVTP Compressed Video Format, H.264 profile High, 1920x1080, 30 fps. The two TSN streams would be allocated into different TC (traffic control) class for egress. Different TC class would be mapped to different hardware queues with specific DMA channel which supported by ENET_OoS IP. The demos were built by follow blocks:   Linux Traffic Control: streams egress control Linux ptp: clock sync in network Libavtp: Time Sensitive Applications AV Transport protocol Gstreamer: avtp plugin uses the libavtp to transmit and receive AVTP audio/video (audio pcm, video h264).   1.2.3.     Traffic control Multiply queue qdiscs + CBS: The CBS class is actually handled by hardware IP to select which queue for transmitting.   CBS parameters come straight from the IEEE 802.1Q-2018 specification. They are the following: idleSlope: rate credits are accumulated when queue isn’t transmitting; sendSlope: rate credits are spent when queue is transmitting; hiCredit: maximum amount of credits the queue is allowed to have; loCredit: minimum amount of credits the queue is allowed to have;     2.  Build demo 2.1.        Build yocto $ DISTRO=fsl-imx-xwayland MACHINE=imx8dxlevk source imx-setup-release.sh -b ./xwayland $ bitbake imx-image-full Prepare a SD card and burn it with the built out images. 2.2.        Rebuild kernel Rebuild the kernel after applying the 0001-qenet-add-queue-avoid-panic.patch, and overwrite the Image and imx8dxl-evk.dtb on the boot partition of the SD card. 2.3.        Install the toolchain $ bitbake -f fsl-image-validation-imx -c populate_sdk $ sh tmp/deploy/sdk/fsl-imx-xwayland-glibc-x86_64-fsl-image-validation-imx-aarch64-imx8dxlevk-toolchain-5.4-zeus.sh The toolchain would be installed into /opt/fsl-imx-xwayland/5.4-zeus   2.4.        Create a install folder $ mkdir <your install folder> Create a folder to install all of the shared libraries, binaries and configure files which built out manually in this doc. After built done, you should copy all of the contents in this folder to target board root.   2.5.        Build libavtp $ source /opt/fsl-imx-xwayland/5.4-zeus/environment-setup-aarch64-poky-linux $ git clone https://github.com/Avnu/libavtp.git $ cd libavtp $ meson build --prefix=<your install folder>/usr $ ninja -C build Copy the built out .so and .pc into the toolchain rootfs: $ sudo cp build/libavtp.so* /opt/fsl-imx-xwayland/5.4-zeus/sysroots/aarch64-poky-linux/usr/lib $ sudo cp build/meson-private/*.pc /opt/fsl-imx-xwayland/5.4-zeus/sysroots/aarch64-poky-linux/usr/lib/pkgconfig/ Copy the .so into the install folder: $ cp build/libavtp.so* <install folder>/usr/lib/ To make sure you have avtp package installed correctly:     $ pkg-config --list-all | grep avtp   2.6.        Build ALSA aaf plugin $ cd <yocto build>/tmp/work/aarch64-poky-linux/alsa-plugins/1.1.9-r0/alsa-plugins-1.1.9 $ ./configure --build=x86_64-linux --host=aarch64-poky-linux --target=aarch64-poky-linux --prefix=<install folder>/usr --disable-silent-rules --disable-dependency-tracking --with-libtool-sysroot=<yocto build>/xwayland/tmp/work/aarch64-poky-linux/alsa-plugins/1.1.9-r0/recipe-sysroot --disable-static --enable-aaf --disable-jack --disable-libav --disable-maemo-plugin --disable-maemo-resource-manager --enable-pulseaudio --enable-samplerate --with-speex=lib $ make $ make install   2.7.        Build Gstreamer AVTP plugins （1.17.x） 2.7.1.     Build Gstreamer core $ git clone https://gitlab.freedesktop.org/gstreamer/gstreamer.git $ patch -p1 < gstreamer-1.0-pass-build.patch $ meson build --prefix=<install folder>/usr $ ninja -C build $ sudo ninja -C build install After Gstreamer is installed into <your install folder>, please fix the “prefix” path in the .pc files by, and copy to the toolchain folders: $ cd <your install folder> $ grep -lR <your install folder> ./lib/pkgconfig/ | xargs sed -i 's/<your install folder>/\/usr/g' $ cp -rf ./usr/* /opt/fsl-imx-xwayland/5.4-zeus/sysroots/aarch64-poky-linux/usr/ 2.7.2.     Build gst-plugins-base $ git clone https://gitlab.freedesktop.org/gstreamer/gst-plugins-base.git $ cd gst-plugins-base $ patch -p1 < gst-plugins-base-pass-build.patch $ meson build --prefix=<your install folder>/usr $ ninja -C build $ sudo ninja -C build install   2.7.3.     Build gst-plugins-bad $ git clone https://gitlab.freedesktop.org/gstreamer/gst-plugins-bad.git $ cd gst-plugins-bad $ meson build --prefix=<your install folder>/usr $ ninja -C build $ sudo ninja -C build install   After gst-plugins-base and gst-plugins-bad installed into <your install folder>, please fix the “prefix” path in the .pc files and copy them into the toolchain folders: $ cd <your install folder> $ grep -lR <your install folder> ./lib/pkgconfig/ | xargs sed -i 's/<your install folder>/\/usr/g' $ cp -rf ./usr/* /opt/fsl-imx-xwayland/5.4-zeus/sysroots/aarch64-poky-linux/usr/   2.8.        Build H.264 SW plugins 2.8.1.     Build x264 As the yocto actually has the x264 recipes, but not included in our bblayers, we need to copy the x264 source into our bblayers path under <yocto>/source to build: $ cp -rf ./poky/meta/recipes-multimedia/x264 ./meta-openembedded/meta-multimedia/recipes-multimedia/ $ vi ./meta-openembedded/meta-multimedia/recipes-multimedia/x264_git.bb Remove the LICENSE_FLAGS line $ bitbake -f x264 -c do_install $ sudo cp -rf tmp/work/aarch64-poky-linux/x264/r2917+gitAUTOINC+72db437770-r0/image/usr/* /opt/fsl-imx-xwayland/5.4-zeus/sysroots/aarch64-poky-linux/usr/ 2.8.2.     Build gst-plugins-ugly $ git clone https://gitlab.freedesktop.org/gstreamer/gst-plugins-ugly.git $ cd gst-plugins-ugly $ meson build --prefix=<your install folder>/usr $ ninja -C build $ sudo ninja -C build install   2.8.3.     Build libav $ cp -rf poky/meta/recipes-multimedia/gstreamer/gstreamer1.0-libav meta-openembedded/meta-multimedia/recipes-multimedia/gstreamer-1.0/ Remove the LICENSE_FLAGS line $ vim ./poky/meta/recipes-multimedia/gstreamer/gstreamer1.0-libav_1.16.2.bb $ bitbake -f gstreamer1.0-libav -c do_install $ cp /opt/samba/nxf39444/imx-yocto-bsp-i.mx8dxl/xwayland/tmp/work/aarch64-poky-linux/gstreamer1.0-libav/1.16.2-r0/image/usr/lib/gstreamer-1.0/libgstlibav.so <your install folder>/usr/lib/gstreamer-1.0   2.8.4.     Install binaries Final step is to copy all of your built out files from <your install folder> into your board / root, and boot up the board. $ export GST_PLUGIN_PATH=/usr/lib/gstreamer-1.0/ $ gst-inspect-1.0 To check if the above Gstreamer plugins we built out can be found by gst-instpect.   3.  System Setup 3.1.        VLAN The ENTE_QoS is assigned to eth0 instance. So create eth0.5 for vlan id 5: $ ip link add link eth0 name eth0.5 type vlan id 5 egress-qos-map 2:2 3:3 $ ip link set eth0.5 up   3.2.        Qdiscs The TSN control plane is implemented through the TC (Traffic Control) system. The transmission algorithms specified in the FQTSS (Forwarding and Queuing for Time-Sensitive Streams) chapter of IEEE 802.1Q-2018 are supported via TC Qdiscs (Queuing Discipline). 3.2.1.     MQPRIO qdisc $ tc qdisc add dev eth0 parent root handle 100 mqprio num_tc 3 map 0 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 queues 1@0 1@1 1@2 hw 1 3.2.2.     CBS qdisc Q1 CBS for audio, Q2 CBS for video: $ tc qdisc replace dev eth0 parent 100:3 handle 888 cbs idleslope 3648 sendslope -996352 hicredit 12 locredit -113 offload 1 $ tc qdisc replace dev eth0 parent 100:2 handle 777 cbs idleslope 98688 sendslope -901312 hicredit 153 locredit -1389 offload 1 3.2.3.     TimeSync Run the ptp4l and phc2sys in background, and use check_clocks to check the ptp sync works. $ ptp4l -i eth0 -f ./gPTP.cfg --step_threshold=1 & $ pmc -u -b 0 -t 1 "SET GRANDMASTER_SETTINGS_NP clockClass 248 clockAccuracy 0xfe offsetScaledLogVariance 0xffff currentUtcOffset 37 leap61 0 leap59 0 currentUtcOffsetValid 1 ptpTimescale 1 timeTraceable 1 frequencyTraceable 0 timeSource 0xa0" $ ./check_clocks -d eth0 4.  Run demo 4.1.        ALSA AAF audio To run the alsa AAF demo, please add aaf0 and converter0 plugin device into /etc/asound.conf: pcm.aaf0 {    type aaf    ifname eth0.5    addr 01:AA:AA:AA:AA:AA    prio 2    streamid AA:BB:CC:DD:EE:FF:000B    mtt 50000    time_uncertainty 1000    frames_per_pdu 12    ptime_tolerance 100 } pcm.converter0 {    type linear    slave {                  pcm "hw:0,0"                  format S16_LE    } } The “aaf0” plugin device defines the ethernet interface which AAF runs on, the socket priority which mapping to Traffic Class in kernel TC, the stream-id for the aaf streaming. The “converter0” plugin device is used for convert the S16_BE format to S16_LE for the wm8960 PCM audio.   Select one device as AVB talker, and run: $ speaker-test -p 25000 -F S16_BE -c 2 -r 48000 -D aaf0   Select one device as AVB listener, and run: $ arecord -F 25000 -t raw -f S16_BE -c 2 -r 48000 -D aaf0 | aplay -F 25000 -t raw -f S16_BE -c 2 -r 48000 -D converter0   You can hear the sound on the listener device.   You can also check which qdisc queue is used for AAF by: $ tc -s qdisc   4.2.        Gstreamer AAF audio Select one device as AVB talker, and run: $ gst-launch-1.0 clockselect. \( clock-id=realtime audiotestsrc samplesperbuffer=12 is-live=true ! audio/x-raw,format=S16BE,channels=2,rate=48000 ! avtpaafpay mtt=50000000 tu=1000000 streamid=0xAABBCCDDEEFF000B processing-deadline=0 ! avtpsink ifname=eth0.5 address=06:98:c0:22:df:35 priority=3 processing-deadline=0 \)   Select one device as AVB listener, and run: $ gst-launch-1.0 clockselect. \( clock-id=realtime avtpsrc ifname=eth0.5 ! avtpaafdepay streamid=0xAABBCCDDEEFF000B ! queue max-size-bytes=0 max-size-buffers=0 max-size-time=0 ! audioconvert ! audioresample !  alsasink device="hw:0,0" \)   5.  Packet sniffer Use tcpdump on board to dump the L2 ethernet packet: $ tcpdump -i eth0 ether proto 0x22f0 -w dump.pcap The AVTP ether protocol code is 0x22f0 embedded inside the ether frame, or you can use "vlan 5" VLAN id for tcpdump parameters to dump. Then open this dump.pcap in the windows/Linux PC by the wireshark tool, it will automatically show the protocol inside the package, it can also parser the IEEE1722 (AVTP) CVF/AFF package header as below:   To measure the package latency from transmit port (talker) to receive port (listener), you can use the tcpdump on both end-points. And compare the Epoch Time the packet dumped: "Epoch Time: 1596252905.688243000 seconds". The delta of the epoch time of the same packet is around 100us~500us. This latency actually includes the AF_PACKET clone cost in kernel netfilter, also the tcpdump application schedule latency.   6.  Revision history summarizes the changes done to this document since the initial release. Table2. Revision history Revision number Date Substantive changes 1 5/2021 Initial release    

This doc share one OpenGL ES sample code, it is running on i.MX8 MEK board with QNX SDP7.1. HW: i.MX8 MEK board, HDMI display SW: QNX SDP7.1, i.MX8 MEK board BSP, and this sample code   This sample code will draw 3D object model, and with some animation. Reference: https://www.nxp.com/products/processors-and-microcontrollers/arm-processors/i-mx-applications-processors/i-mx-8-processors/i-mx-8-family-arm-cortex-a53-cortex-a72-virtualization-vision-3d-graphics-4k-video:i.MX8 https://github.com/NXPmicro/gtec-demo-framework https://github.com/syoyo/tinyobjloader-c https://github.com/nothings/stb https://3dhaupt.com/futuristic-car-game-ready-download/ https://wallpapersafari.com/w/Y5JZNh https://www.pngwing.com/en/free-png-ysaus https://www.shadertoy.com/view/Ms2SWW#