When considering EIM bursts (performance, burst length, etc) we should take into account, that some parameters (maximum burst length, which defines length in clocks of back-to-back EIM access) depend on master, which performs the access (EIM always service as slave). The EIM can split a master burst in order to meet own settings, but the EIM cannot join two master accesses. So, the maximum  burst length is defined by master.  Usually three options are applied for EIM burst accesses : - ARM block copy instructions (LDM / STM) ; - ARM NEON copy instructions (VLDM / VSTM) ; - i.MX SDMA.   Below are some details regarding these options.   1. ARM. ARM provide recommendations below about the fastest way to copy memory on a Cortex-A8. http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.faqs/ka13544.html   According to section 8.1.2 (Supported AXI transfers) of ARM Cortex-A9 Technical Reference Manual, it is possible to get maximum 64 bytes (16 beats x 4 bytes) burst for read and 8 bytes burst for write. “INCR N (N:1-16) 32-bit read transfers INCR N (N:1-2) for 32-bit write transfers” http://infocenter.arm.com/help/topic/com.arm.doc.ddi0388i/DDI0388I_cortex_a9_r4p1_trm.pdf 2. SDMA.   According to section 55.4.3.1 (Burst DMA Unit) of the i.MX6 DQ RM : “Perform up to 8-beat read and write bursts to the ARM platform memory, which optimizes throughput when accessing SDRAM-type devices because of an internal, 36-byte FIFO”. This means, that burst length of the SDMA cannot be greater than 32 bytes (8 beat x 4 bytes). As for performance and implementation of SDMA approach, please look at the following : “Measure SDMA Memory To Memory Copy Performance on i.MX6Q” https://community.freescale.com/docs/DOC-103127

Important: If you have any questions or would like to report any issues with the DDR tools or supporting documents please create a support ticket in the i.MX community. Please note that any private messages or direct emails are not monitored and will not receive a response.   This is the detailed programming aid for the registers associated with DRAM initialization (DDR3 and LPDDR2) of the MX6UL/ULL/ULZ (consolidated RPA). The last work sheet tab in the tool formats the register settings for use with the ARM DS5/RealView debugger. It can be manually converted by the user to a DCD file format used by uboot or other bootloaders (note the removal of debugger specific commands in this tab). The programming aids were developed for internal NXP validation and development boards.   This tool serves as an aid to assist with programming the DDR interface of the MX6UL/ULL/ULZ and is based on the DDR initialization scripts developed by the R&D team and no guarantees are made by this tool.   The following are some general notes regarding this tool: Refer to the "How To Use" tab in the tool as a starting point to use this tool. Note that in the "DStream .ds file" tab there are DS5 debugger specific commands that should be commented out or removed when using the DRAM initialization for non-debugger specific applications (like when porting to bootloaders). This tool may be updated on an as-needed basis for bug fixes or future improvements.  There is no schedule for aforementioned maintenance.  

The i.MX27 PDK, with Smart Speed™ technology, is a completely integrated hardware and software solution designed to simplify product development so you can focus on the critical differentiation needed for market success. Reduce development time and design products that have power to spare, even when running multiple applications simultaneously. Receive stellar Ethernet and video performance in a system design that dramatically reduces power consumption. Features i.MX27 Applications Processor - ARM9™ 128 MB DDR SDRAM 256 MB NAND FLASH Power Management (PMIC MC13783) + Power Circuitry Audio HS USB PHY Touch Controller 10/100 Ethernet port Accelerometer MMA7450L (Freescale) User I/O Connectivity (FM, 802.11, Bluetooth, USB OTG, USB HS) Button 2.7" TFT Display 2MP Camera Module SD card, ATA HDD External Connectors (dock, headphones, TV out, GPS) Microphone Speaker Debug Ethernet Port Debug Serial Port JTAG Reset, Interrupt, Boot Switches Debug LEDs CodeTest Interface Power Source Current/Power Monitoring

As the Voice over IP (VoIP) market grows, the next evolution of the market is Video and Voice over IP (V2IP). iWave’s iW-RainboW-G15S is a Pico-ITX Single Board Computer (SBC) which has a Freescale’s i.MX6 DualLite ARM Cortex-A9 core based CPU which can operate up to 800MHz speed/core with 1GB (expandable) DDR3 RAM. iWave has provided V2IP on iWave’s Pico ITX i.MX6 which gives more quality in video and audio streaming. iWave has expertise in HD video streaming over V2IP. The SIP(Session Initiation Protocol) protocol for connection between i.MX6 PICO ITX board and host PC which is used in the application Linphone through network (Ethernet). Linphone is a comprehensive solution consisting of an extensive set of algorithms and codecs designed for Digital Voice and Video applications. Above design can be utilized for peer to peer communication between i.MX6 PICO ITX single board computer and PC with Android OS. The i.MX6 single board computer and PC are connected to an Ethernet. The Audio CODEC used on the iMX6 PICO ITX dev board is ALC5610 from Realtek with inbuilt Headphone amplifier and MIC which is used to provide a complete audio solution for portable products. The video pixel rates are typically from 25 MHz up to 297 MHz, but HDMI can support higher rates up to 340 MHz’s i.MX6 PICO ITX board supports LVDS connector to connect different LVDS LCDs. It also supports backlight connector with 15V 300mA output for LCD backlight. iWave's i.MX 6 PICO ITX SBC supports 8/10bit CMOS Camera Interface. External clock for camera is provided using on board Oscillator of frequency 26MHz. i.MX6 CPU supports MIPI CSI interface. The V2IP systems use some existing standard video codec and audio codec to reduce the program material to a bit stream and then use an Internet Protocol (IP) network to carry that bit stream encapsulated in stream of IP packets. This is typically accomplished using some variant of the RTP protocol. Freeescale i.MX6 multimedia applications processor provides sufficient power to provide high-quality audio through wideband audio, in-call audio and video recording on flash. Image: V2IP on iWave’s Android i.MX6 Pico ITX Single Board Computer Video Streaming Platform feature: Pico ITX board with i.MX6 dual lite CPU 320X240p camera OS: Android jelly-bean (4.3) MIC HDMI with 1920X1080p display Android NDK and SDK H264 video codec Freescale’s VPU     For further information or enquiries please write to [email protected] or visit www.iwavesystems.com

MX6UL_DDR3_调校_应用手册_V3_20160511.doc

i.MX8M_DDR3L_register_programming_aid is created for DDR3L validation board.

Q: Looking at the datasheet, why a 14.7KOhm pulldown and a 4.7KOhm pullup must be used? When useing 4.7KOhm resistors in series from an MCU to control the boot pins, pull up is ok, but pull down is not. The EIM boot pins only see 4.7KOhm and not 14.7KOhm as recommended in the design guide. How did we determine those resistor values? at boot there is a pull-up of 100KOhm in the pad if I understand well, 4.7KOhm would do the job as well as 14.7K. right? A: 4.7k will work fine for pull-up or pull-down. 10k pull-downs is to limit current drain while allowing use of simple SPST switches to configure boot-up. There are on-chip 100k pull resistors, so this limits max resistor size. This document was generated from the following discussion: EIM boot resistor value

[中文翻译版] 见附件   原文链接： https://community.nxp.com/docs/DOC-345359 

Play MPEG4 Video only gst-launch filesrc location=test.mpeg ! mfw_vpudecoder codec-type=std_mpeg4 ! mfw_v4lsink H.264 Video only gst-launch filesrc location=test.avi ! mfw_avidemuxer ! mfw_vpudecoder codec-type=std_avc ! mfw_v4lsink AVI(H264+MP3) gst-launch filesrc location=test.avi ! mfw_avidemuxer name=demux demux. !  mfw_vpudecoder \     codec-type=std_avc ! mfw_v4lsink demux. ! queue max-size-buffers=0 ! <mp3_decoder_plugin> ! alsasink MP4(H264+MP3) gst-launch filesrc location=test.mp4 ! mfw_mp4demuxer name=demux demux. ! \    mfw_vpudecoder codec-type=std_mpeg4 ! mfw_v4lsink demux. ! \    queue max-size-buffers=0 ! <mp3_decoder_plugin> ! alsasink <mp3_decoder_plugin> can be replaced by mad RAW Video Test gst-launch videotestsrc ! video/x-raw-yuv,format=\(fourcc\)I420 !  mfw_v4lsink

Sorry I can't find a place to share this input file. It is to reproduce the VPU JPEG decoder issue I reported in Corrupted MJPG decoding result with FSL 3.10.17 BSP on i.MX6Q VPU​

In order to improve the speed of compilation, VMWare Player 14.0 is installed on local hard disk, and Ubuntu 18.04 LTS is installed on a SSD with at least 500GB size and USB3.1 specification. When installing ubuntu 18.04 LTS to SSD, it should be allocated at least 350GB of disk space, because compiling this version of android requires a larger disk space. The following are detailed compilation steps: Part l Configuring Ubuntu 18.04 LTS 1. Installing Ubuntu 18.04 on VMplayer 14.0 After installation is done, root user should be set at first. # sudo passwd root Then follow these steps to configuration ubuntu 18.04 for environment of compiliation --Changing sources of ubuntu 18.04 mirror If you are Chinese users, you can do the step, which can improve your system performance. # sudo cp /etc/apt/sources.list /etc/apt/sources.list.bak # sudo geit /etc/apt/source.list Comment I.MX customers outside China do not need to modify Ubuntu source list, or can modify it to local mirror site of Ubuntu 18.04, which can improve the speed of software upgrade.    Delet all sources and copy following lines here, Then save it and exit Changing ubuntu source deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-updates main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-backports main restricted universe multiverse deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-security main restricted universe multiverse # deb https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse # deb-src https://mirrors.tuna.tsinghua.edu.cn/ubuntu/ bionic-proposed main restricted universe multiverse    Then running these 2 commands to update sources and packages    # sudo apt-get update    # sudo apt-get upgrade 2. Installing packages for compiliation Packages for compiliation # sudo apt-get install flex bison gperf build-essential zlib1g-dev lib32ncurses5-dev x11proto-core-dev libx11-dev lib32z1-dev libgl1-mesa-dev tofrodos python-markdown libxml2-utils xsltproc # sudo apt-get install uuid-dev:i386 liblzo2-dev:i386 gcc-multilib g++-multilib subversion openssh-server openssh-client uuid uuid-dev zlib1g-dev liblz-dev lzop liblzo2-2 liblzo2-dev git-core curl # sudo apt-get install u-boot-tools mtd-utils android-tools-fsutils openjdk-8-jdk device-tree-compiler aptitude libcurl4-openssl-dev nss-updatedb # sudo apt-get install chrpath texinfo gawk cpio diffstat gdisk m4 libz-dev libssl-dev Part 2 Compiling Android Q10.0.0_2.1.0 BSP 1. Downloading NXP source code for Android Q10.0.0_2.1.0    File name is imx-android-10.0.0_2.1.0.tar.gz.    Copy the file to ~/, and decompress it.    # cd ~/    # tar zxvf ./imx-android-10.0.0_2.1.0.tar.gz    Then “imx-android-10.0.0_2.1.0” directory is created at ~/, now run the command to download android source code. # source ./imx-android-10.0.0_2.1.0/imx_android_setup.sh Comment imx_android_setup.sh is a script file, which includes all steps needed by the environment of android Q10.0.0_2.1.0 BSP. If network environment is enough good, several hours later, it will be done. 2. Compiling Android Q10.0.0_2.1.0 Referring to steps in Android_User's_Guide.pdf, We summaries steps for compilation here: (1) Preparing cross-compile tool chain    In Android_User's_Guide.pdf, 2 kinds of tool chain are recommended for users. --- gcc-arm-8.3-2019.03-x86_64-aarch64-elf.tar.xz --- gcc-arm-8.3-2019.03-x86_64-aarch64-linux-gnu.tar.xz    Users can select one of them, and then decompress it to /opt/ directory. On how to download them or more details, see Android_User's_Guide.pdf, page 3. Here we will use gcc-arm-8.3-2019.03-x86_64-aarch64-linux-gnu.tar.xz as tool chain. (2) Beginning to Compile Android Q10.0.0_2.1.0 BSP    Since this version of Android BSP requires high memory capacity when compiling, if the memory configuration of the virtual machine is incorrect, it is very likely to cause the compilation to fail. The following is a list of variable tests for user reference: # cd android_build # export AARCH64_GCC_CROSS_COMPILE=/opt/gcc-arm-8.3-2019.03-x86_64-aarch64-linux-gnu/bin/aarch64-linux-gnu- # source build/envsetup.sh # lunch evk_8mp-userdebug # ./imx-make.sh -j2 2>&1 | tee build-log.txt Part 3 Errors During Compilation 1. Allocating 8GB Memory For VMware Player # ./imx-make.sh -j1 2>&1 | tee build-log.txt 2. Allocating 12GB Memory For VMware Player # ./imx-make.sh -j4 2>&1 | tee build-log.txt # ./imx-make.sh -j4 2>&1 | tee build-log.txt (Run it again) # ./imx-make.sh -j4 2>&1 | tee build-log.txt (Run it again)       So if we use 4 thread to compile BSP, command for compilation will have to be run for 3 times. NXP TIC Team Weidong Sun 2020/4/30

imx7 supports TRULY-WVGA-TFT3P5581E display, the connector is J30, which is a connector on the imx7d board, normally, one can find the part number from BOM file, but imx7 Sabre-SDB board  BOM file missed this information. got connector information from expert team as below: The connector manufacture’s part number is: DF40C-24DS-0.4V(51)   For more detailed information about sabre board information information is available from nxp.com in the Design files for i.MX7Dual   For data sheet of this, which is under NDA, if you need it, pls create a salesforce case to get this, I couldn’t post this on public community.

U-boot expects uImage format for the kernel image. In order to LTIB generate a uImage file: $ export SYSCFG_KTARG=uImage $ ./ltib -p kernel Setup in U-Boot the kernel bootargs: u-boot> setenv bootargs noinitrd console=ttymxc0,115200 init=/linuxrc root=/dev/nfs nfsroot=10.29.244.27:/tftpboot/rootfs ip=dhcp Change 10.29.244.27 to your host IP. The procedure above is needed when default bootloader used by ltib was redboot. In some ltib releases (before 2010) default bootloader is u-boot. In this case, ltib will create uImage by default

The i.MX6 Multi-Mode DDR Controller (MMDC) has profiling capabilities to monitor the operation of the controller. The profiling capability counts certain events related to a specified AXI-ID during a profiling period. The events that can be counted are: The number of read accesses during the profiling period (MMDCx_MADPSR2[RD_ACC_COUNT] register field) The number of write accesses during the profiling period (MMDCx_MADPSR3[WR_ACC_COUNT] register field) The number of bytes read during the profiling period (MMDCx_MADPSR4[RD_BYTES_COUNT] register field) The number of bytes written during the profiling period (MMDCx_MADPSR5[WR_BYTES_COUNT] register field) The number of MMDC clock cycles during which the MMDC state machine is busy (MMDCx_MADPSR1[BUSY_COUNT] register field) BUSY_COUNT is the number of MMDC clock cycles during the profiling period in which the MMDC state machine is not idle. So this is the time the MMDC spends doing any activity, not just read or write data transfers. The MMDC state machine is active whenever there are any read or write requests in the read and write FIFOs. The MMDC is active during many operations that are not reading or writing data such as arbitration of requests, control cycles, bank open/close, etc. So BUSY_COUNT represents the number of cycles when the controller is busy, not just the number of cycles when the external bus is busy. The number of bytes read and bytes written can be used to determine data throughput and the BUSY_COUNT can be used to determine what part of the time the controller is active/idle. Together these can be used to determine the controller efficiency for a particular application. For detailed information, see the "MMDC profiling" section of the MMDC chapter in the reference manual for the SoC being used.

Q: In mfgtool release 1.0.0 I, "Support only Cortex-A5 as primary boot core (Cortex-M4 boot is not supported on this release)." When will a version supporting M4 boot first be available? A: Mfg tool is based on it's counterpart, the ROM code. Rom code is for the A5 and not for the M4 as a matter of fact if you boot with the M4 and the device doesn't enumerate on the USB port at all then you know why and the MfgTool won't be able to do anything about it. Currently there no plans to provide a MfgTool version that supports boot from M4 core. Customers can modify the exisiting code to make it work with M4 as primary core. Booting from USB with the M4 core should work. The image that is booted must be for execution on M4 (thumb mode). MfgTool uses u-boot and a micro linux kernel to download images and program memories. This Linux can only run on the A5 core. To make it work with the M4 core as primary, it should be sufficient to add some code to the exisiting A5 images. Code will execute on M4 and start the A5 at the proper address as if it is booted from the A5.

Encode From YUV to H.264 gst-launch-0.10 filesrc location=file_in.yuv blocksize=w*h*1.5 ! \ mfw_vpuencoder codec type=std_avc framerate=fr ! filesink location=file_out.mpg Where: file_in.yuv: is the input file, a raw file. w*h*1.5: is the blocksize, it's calculated from input file dimensions: width * height * 1.5 mfw_vpuencoder: is the encoder with hardware acceleration for iMX27 std_avc: chooses the codec type for output file fr: indicates the framerate in that input file was created file_out.mpg: is the output file encoded in H.264 From Camera to H.264 gst-launch-0.10 mfw_v4lsrc ! mfw_vpuencoder codec-type=std_avc \ width=176 height=144 framerate=25 ! filesink location=test.video

When flashing a Linux System on a SD card using the script mk_mx28_sd on a Ubuntu 12.04 host, one needs to modify it so partitions are created correctly.  Just follow these steps on the console: $ cd $SDK/L2.6.35_10.12.01_SDK_scripts $ cat > first_partition_sector.patch << EOF diff -Naur a/mk_mx28_sd b/mk_mx28_sd --- a/mk_mx28_sd        2010-10-06 09:47:42.000000000 -0500 +++ b/mk_mx28_sd        2012-11-30 13:38:34.508199154 -0600 @@ -178,7 +178,7 @@ n p 1 -1 +2048 +32M t b EOF $ patch -p1 < first_partition_sector.patch then, you can run the mk_mx28_sd command again with the device as parameter                $ cd $LIB $ export PATH=$PATH:$SDK/L2.6.35_10.12.01_SDK_scripts $ mk_mx28_sd /dev/$SDX

The native Yocto not support chinese, so need to add chinese font . The first step should download Chinese font  files such as  MSYHMONO.ttf and DroidSansFallback.ttf.Then copy fonts files to /usr/share/fonts/truetype and run command  fc-cache .update font cache. use fc-list :lang=zh can see which Chinese fonts are installed:          /usr/share/fonts/truetype/MSYHMONO.ttf: Microsoft YaHei Mono:style=Regular          /usr/share/fonts/truetype/DroidSansFallbackFull.ttf: Droid Sans Fallback:style=Regular The secend step is to change default fonts. In weston-terminal.c we can see:           3119   weston_config_section_get_string(s, "font", &option_font, "mono");           3120   weston_config_section_get_int(s, "font-size", &option_font_size, 14);  weston-terminal default use mono font.Just to modify /etc/fonts/fonts.conf           <edit name="family" mode="assign" binding="same">                       <string>monospace</string> to   <string>Microsoft YaHei Mono</string> or <string>Droid Sans Fallback</string> could display Chinese character   But it have some problem such like as :   the left use Microsoft YaHei Mono font the Chinese character is overlapping ,the right use Droid Sans Fallback font is en-font could not display. If you have a better way to welcome a message！

The SNVS LDO output (VDD_SNVS_CAP) requires an external capacitor. Freescale's updated recommendation is that this should be a single 0.22 uF capacitor. Freescale is working to get documents in alignment. As of Feb 2013, some documents (such as schematics or user guides) show a single 0.22 uF capacitor, others do not.