STM32 MPUs Embedded software and solutions

I'm trying to figure out how to add OpenCV to my STM32MP157-DK2 Discovery Board. I am not new to ST Microcontrollers, but I am new to the MP1 and Cortex A processors. I'm also not new to Linux but am completely new to embedded Linux and Yocto.I can see that there is a recipe for OpenCV included in the Ecosystem V1.2 release distribution package, but when I build the release and flash, OpenCV is not present on the DK2 board. I was able build OpenCV specifically with the bitbake command, but again, it is not present in the image. I have seen a few references to installing OpenCV in the wiki but not enough details to get to the next step.I am wondering what the correct or preferred method for getting OpenCV included in the image. Preferably, enough information or pointers to documentation that I can understand the bigger picture and know how to do this in general and not simply how just to add OpenCV in this case.

For measuring the USB signal quality, it is necessary to print out test pattern on the USB interface. How this can be done on a customer board with STM32MP15x controller. Is a test software or source code available?

Hi, I installed node.js with apt-get on my STM32MP157C-DK2, which worked great.Now I want to install another package but it failsapt-get install nodejs-npm... Need to get 7286 kB of archives. After this operation, 0 B of additional disk space will be used. E: You don't have enough free space in /var/cache/apt/archives/.But I think there is still enough free space on the SD Card:root@stm32mp1:~# df -h Filesystem Size Used Avail Use% Mounted on /dev/root 715M 682M 0 100% / devtmpfs 150M 0 150M 0% /dev tmpfs 214M 0 214M 0% /dev/shm tmpfs 214M 20M 195M 10% /run tmpfs 214M 0 214M 0% /sys/fs/cgroup tmpfs 214M 4.0K 214M 1% /tmp /dev/mmcblk0p4 58M 8.5M 46M 16% /boot tmpfs 214M 80K 214M 1% /var/volatile /dev/mmcblk0p7 14G 74M 13G 1% /usr/local /dev/mmcblk0p5 15M 8.4M 5.1M 63% /vendorhope someone can help. Otherwise I have to cancel my work with node.jsregardsLukas

What's the process to add a native gcc to openstlinux-weston so I can compile from the command line on my DK2 dev board? I was able to rebuild using bitbake and started poking around to find which layer to enable to get a native gcc running on the STM32MP1 dev board (DK2) but I can't figure it out. Or should I use a different approach, is there a pre-built image for the DK2 with native gcc?Thanks,Andrew

I'm following the OpenAMP_TTY_echo example.To do a Tx it basically uses two functions: VIRT_UART_Init and VIRT_UART_TransmitAfter I use VIRT_UART_Init, I can see in the Linux terminal a log message something that suggests the initialization is working.However, VIRT_UART_Transmit returns error.I was debugging and found that rpmsg_send_offchannel_raw function is returning because "dst == RPMSG_ADDR_ANY"However, VIRT_UART_Init, is initializing destiny as RPMSG_ADDR_ANYstatus = OPENAMP_create_endpoint(&huart->ept, RPMSG_SERVICE_NAME, RPMSG_ADDR_ANY, VIRT_UART_read_cb, NULL);These are automatic generated functions by STM32CubeMX.Am I missing something? Should I set the dest address somehow?Note this is code generated with STM32CubeMX, I only added the call to VIRT_UART functions.

Hi , iam trying to use spi1 pheripheral for the cortex M4 and iam using Pz0 as CLK, pz1as MISO, pz2 as MOSI, pz3 as NSS , but i tried several ways to do it by changing the kernel DTS and TF-A as well but no result , could anyone please help me how can i configure a spi pheripheral to cortex M4 with above given pins Thanks & regardsAkshay.

Hello. I am currently working on a custom stm32mp157a Linux based board with a KSZ9896 as network switch. At this point everything works on the board expect the RGMII connection to the switch. The RGMII is currently running in 100MBit mode for easier debugging, all signals seem to be received correctly. The stm32mp1 s not receiving any data over RGMII. Even when enabling the loopback on the interface in RGMII no data is seen using tcpdump (-Qin -i eth0). This is interesting since running ethtool -S eth0 shows the rx and tx counters going up together as packages are sent and looped back. Both connections are 100Mbit with a 125 MHz clock coming from the switch and flow control has been disabled.Could anyone point me in the right direction where to start looking?

When booting off SPI NOR I would expect the device (an ST32MP151D) to possibly check the device ID then read a block starting from 0. Where is this fully documented ?   What I actually see is two pairs of negative pulses on nCS then it floats open circuit. It's all within some tens of microseconds. The device is loaded with two copies of my TF-A and one U-Boot at the addresses in the TSV which are 0, 256k and 512k.   Where is the documentation or application note on what exactly it does ?   There is nothing in UART4 (PC10 and PC11).

I'm working on a project using the Seeed Studio STM32MP157C-Odyssey-SoM (it is a module with the STM32MP1, DDR, eMMC and STPMIC). I am making a new PCB into which the module is plugged.   When using the module on the original host board from Seeed, both the eMMC (SDMMC2 controller) and the SD-Card (SDMMC1 controller) boot work. On my board only the eMMC boot works. When using the SWD debugger, I found that the CPU does not even enter the FSBL, but remains stuck in the boot ROM. By checking the SD-Card connection, I found that the CPU does not even attempt a write (and the SDMMC1 controller has only the reset values in its registers).   I also determined that the boot mode is properly detected in the SYSCFG_BOOTR register.   Is there a way to debug the boot ROM?   I understand the code is not available, but I hope to be able to determine what is happening by reading peripheral registers, which ones should I check?   Thanks for your help

Hello,I am working with an STM32MP157 -DK2The Distribution Package currently offers a Yocto version that is aligned with Dunfell (in particular meta-st-linux)Is this repository maintained somewhere in a way that it is aligned with current Yocto releases such as hardknott?Regards

I have the following pinctrl entry in my 5.10 device tree to use the USART1 and the serial port fails, due to an incorrect pin configuration for PZ7 AF7.usart1_pins_c: usart1-2 { pins1 { pinmux = <STM32_PINMUX('Z', 7, AF7)>, /* USART1_TX */ <STM32_PINMUX('A', 12, AF7)>; /* USART1_RTS */ bias-disable; drive-push-pull; slew-rate = <0>; }; pins2 { pinmux = <STM32_PINMUX('Z', 1, AF7)>, /* USART1_RX */ <STM32_PINMUX('A', 11, AF7)>; /* USART1_CTS_NSS */ bias-pull-up; }; };Does anybody know why this is happening?Does the PZ7 require any specific setup to be used as USART1_TX?Best regards,--Mike

Hi everyone, I am using stm32mp157f-dk2 with ecosystem 3.1.I want to configure the usb type C port(CN7) as serial port, and communicate with the host PC.This article helps me a lot. https://wiki.st.com/stm32mpu/wiki/USB_overview#How_to_configure_USB_Gadget_through_configfsI modified the usbotg-gadget-config.bb, the stm32_usbotg_eth_config.sh and create my own service, as attached.After that, I can successfully set up the device. I can see ttyGS0 show up in my device.However, when I connect it to the host, although the host is able to detect my device, it can't recognize my device.What am I missing?I suspect that it is the stlink causing the issue, as it occupy the ttyACM0 already and the host cannot create another tty corresponding to the GS0. I am not sure how will the host OS deal with this situation.Regards, Leung10/3 update: I found that my PC can't boot up when the device is connected. It scared me. I am still investigating why it is failing, but I think the stlink is not the root

I am using starter image "en.st-android-10.0.0-2020-02-21-eval-sd-starter.tar.gz".CubeProgrammer gave me "Error: an error occurred while uploading data from the virtual partition 0xF1" . Below is full log:16:56:19 : USB speed : High Speed (480MBit/s)16:56:19 : Manuf. ID : STMicroelectronics16:56:19 : Product ID : DFU in HS Mode @Device ID /0x500, @Revision ID /0x000016:56:19 : SN : 0046003E333851063438333016:56:19 : FW version : 0x011016:56:19 : Device ID : 0x050016:56:23 : Start Embedded Flashing service16:56:23 : Memory Programming ...16:56:23 : Opening and parsing file: fsbl-programmer.img16:56:23 : File : fsbl-programmer.img16:56:23 : Size : 245360 Bytes16:56:23 : Partition ID : 0x01 16:56:23 : Download in Progress:16:56:24 : File download complete16:56:24 : Time elapsed during download operation: 00:00:00.43116:56:24 : RUNNING Program ... 16:56:24 : PartID: :0x01 16:58:14 : Error: an error occured while uploading data from the virtual partition 0xF116:58:1

..

Hi all,I have been trying to get DSI interface on a custom board using STM32MP1.The device tree exactly done as given here : LInk.I do have DK2 kit , which i probe I am able to see the signals at DSI PHY even if display is not connected.In case of my board , I am unable to get it up. See the device tree attached for more details.The reference design is taken from DK2.The probing of driver i believe fails silently since neither I do any failure logs related to DSI in kernel boot logs. modetest -M : Also does not give positive result. I can see encoder field is empty and unpopulated.Any leads are appreciated.

I'm working on a driver present in the Linux 5.4 kernel. In RS485 communication I receive everything I send, without being connected to another device to return what I sent.I use the command echo -e "Test" > /dev/ttyLP0 for the submission.I know that by default in RS485 communication I receive everything I send, but I want to disable the return (echo).BUG:My setup is:PLC - Programmable Logic ControllerProcessor: LS1027 (It's not a ST processor)Linux Kernel 5.4 (tty driver present in this kernel version, who owns the stm32-usart.c)I use a USB logic analyzer to analyze the signal sent through the echo -e "Test" > /dev/ttyLP0. With rs485 serial pins: D+, D- and GND.Hardware:Best regardsEduardo Gerhardt

I am trying to configure the TIM2_CH4 to capture mode on EV1 board so I can sense a fan speed from a PWM fan but I keep getting cat: read error: Connection timed outI have configured the timer as follows: &timers2{ pinctrl-names = "default", "sleep"; pinctrl-0 = <&tim2_pins_mx>; pinctrl-1 = <&tim2_sleep_pins_mx>; status = "okay";   /* USER CODE BEGIN timers2 */ dma-names = "ch3"; pwm { /* enable PWM on TIM2 */ status = "okay"; }; /* USER CODE END timers2 */ }; tim2_pins_mx: tim2_mx-0 { pins { pinmux = <STM32_PINMUX('A', 3, AF1)>; /* TIM2_CH4 */ bias-disable; drive-push-pull; slew-rate = <0>; }; };   tim2_sleep_pins_mx: tim2_sleep_mx-0 { pins { pinmux = <STM32_PINMUX('A', 3, ANALOG)>; /* TIM2_CH4 */ }; };I am using TIM12_CH1 to generate a pulse and I have shorted the two pins on the board.What am I doing wrong? Is there a better way to detect the sense on a PWM fan?

I am configuring the CPU to wake up from UART reception and/or GPIO interrupt:echo enabled > /sys/devices/platform/soc/4000f000.serial/tty/ttySTM1/power/wakeupecho enabled /sys/devices/platform/soc/4000f000.serial/power/wakeupAnd it goes to sleep with:echo "mem" > /sys/power/stateAfter 1 hour in sleep mode, the time is 20 minutes slow. This is not happening in Run mode.Additional Information:Linux 5.4.313.0V RTC battery backup

Hi!I ordered now a STM32H747I-DISCO board. Have looked in the source tree from U-Boot (ST fork and mainline). I know there is a support for STM32H743I-EVAL board. Wondering why there isn't a support for the STM32H747I-DISCO.Generic question:Has somebody interests when I add U-Boot support for the STM32H747I-DISCO Board?(Maybe I could also add Linux support. Okay, 32MiB is maybe not enough RAM...)It's maybe only a gimmick...

Hi,I'm using OpenST-linux version 4.9 on a STM32MP157c-ev1 and I wrote a simple script to run my firmware for Core M4. I would like to know how I can make the script run automatically after system boot.Here's my script if could of any help for answers, I wrote it following this guide, I tested it launching manually and it works, but now I would like to it runs automatically after system finishes booting. How can I do?RegardsMarco #! /bin/sh   ################### CONFIGURATION VARIABLES ###################################################   FIRMAWARE_PATH="/lib/firmware/"   M4_FIRMWARE_NAME="xmi_ccu-test_CM4.elf"   REMOTE_PROCESSOR="remoteproc0"   ########################### SCRIPT BODY ######################################################   # Add firmware path if different then /lib/firmware if [ $FIRMAWARE_PATH != "/lib/firmware/" ]; then   echo -n $FIRMAWARE_PATH$FIRMAWARE_PATH > "/sys/module/firmware_class/parameters/path" fi   # Load firm

SAI2 Block A configured as a MasterTX and works. Fs configured for 48KHz, bit clock configured as 64*Fs and MCLK as 256*MCLK. I've verified these signals are sent to the audio codec which is configured as a slave (receives MCLK, SCLK, LRCLK (Fs)). The audio codec provides a transmit line and receive line for data. I'm successful with my asynchronous block A in sending audio to the audio codec but when I configure block B as synchronous to block A, the FIFO never fills, and I've verified this inside the HAL_SAI_Receive function. It just times out. I've verified that the audio data is on the SDB line as well via an oscilloscope.Might be related, but on the development boards GPIO connector which exposes all SAI2 pins except MCLK, all signals read as 0v even though reading off of the audio codecs lines, I get the correct signals that I'm generating on those lines. Just an extra problem. I don't have this issue with the I2C that I connect to the audio codec with.Heres the configuration of

Hi dears I wish you a good day ,I am working on my bachelor graduation prohect which is ( an overcurrent protection relay design ) and I using an STM32MP1 to have all the functions and required controlling inside it. And I want to add a communication interface ports with my device to share real time arc Fault data with other devices using local networks and I want to add at least 2 10/100/1000 Mbps Ethernet (RGMII) and 2 Optical Transceivers SFF 1000BASE. I want to add this ETHERNET and SFF module to my project and after a lot of research I noticed that the MPUs only can connect with only one ethernet for and I also found the ETHERNET switches, but I analyzed the marketing, and the experts says that ethernet switches are not good in real time data transferring and the best way to implement the multiple ethernet and SFF transceivers with the project is using the FPGA. And I checked some devices that using for communication, and I noticed that they used the FPGA for that I started t

Hi, I know that the ecosystem is already at 3.1.0 or higher, but we started our project in 2019 and working with image based on that release (1.0.0) and don't plan to upgrade.I wanted to send a link to the SDK from your website to our customers, but the link to 1.0.0 doesn't workI am using:https://www.st.com/content/st_com/en/products/embedded-software/mcu-mpu-embedded-software/stm32-embedded-software/stm32-mpu-openstlinux-distribution/stm32mp1dev.html#If I attempt to download any other SDK (1.1.0 for example) it worksOnly the 1.0.0 option is broken

/* USER CODE BEGIN PD */ #define ADC_BUF_LEN 4096 /* USER CODE END PD */   ...   /* USER CODE BEGIN PV */ uint16_t adc_buf[ADC_BUF_LEN];   ... /* USER CODE BEGIN 2 */ HAL_ADC_Start_DMA(&hadc2, (uint32_t*)adc_buf, ADC_BUF_LEN);I've got the above code added to my project with the attached IOC, but when I hit a debug breakpoint after a second or so there are no values in the adc_buf.In addition, when I probe the ANA1 (Arduino Pin A3) with nothing plugged in, I get a sine wave that sits offset at ~3v and has a peak to peak voltage of about 800mV and 128MHz frequency. See pic below. It's also notable that there is no "DMA Continuous Requests" option in Parameter Settings which I understand I need to set to Enabled.

I'm building an elf file, SCPing it to the controller and using the fw_cortex_m4.sh to load it on the m4, but no matter what I set the APB2DIV scaler to the number of times per second TIM16 is called doesn't change. The only thing that does seem to affect the frequency is the prescaler on the TIM16 pin, but I've already got it set to 0 and I only get about 1.4kH.