STM32 MPUs

Hi,I try to compile my custom system, which is based on the Cubemx in the latest V4.0.0 ecosystem. I have a few problemslttng - how properly disable this package or what shall I add to compile it wellAfter compilation is finished, the *.tsv files are not created - empty folders as following How to properly resolve issues with the licenseWARNING: st-image-weston-1.0-r0 do_st_write_license_create_summary: IMG LIC SUM: File does not exist with open file /home/mw/openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15/build-openstlinuxweston ...I added a meta-layer to reproduce the compilation process if necessary.git@github.com:voloviq/meta-osd32mp1-brk.gitThe configuration process was based on https://wiki.st.com/stm32mpu/wiki/How_to_create_your_own_machineThanks for any hints.Best Regards

Hi,I have tried to migrate from ST Ecosystem 3.1 to 4.0.2 but I am unable to flash the image into the eMMC memory. My issue is similar to this one but I am using the Octavos SiP OSD32MP157c. As suggested by the original poster, the issue is probably the TF-A configuration.The flash process is started but when the u-boot is supposed to be loaded the processor reboots and the process is repeated.The image is created using device trees generated by STM32CubeMX 6.6.1 and I first generated the new trees and then copied changed parts from the generated trees for the DK2 board and my previous configuration.For flashing I use STM32_Programmer_CLI and a micro-USB port. The flash log is attached below:NOTICE: CPU: STM32MP157C?? Rev.Z NOTICE: Model: STMicroelectronics custom STM32CubeMX board - openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15 INFO: PMIC version = 0x21 INFO: Product_below_2v5=0: HSLVEN protected by HW INFO: Reset reason (0x54): INFO: System reset generated by MPU (MPSYSRST) INFO: FC

In our custom board which uses STM32MP153C, right when TF-A is about to launch u-boot it displays PSCI power domain map logs and MPU reboots with MPSYSRST reason. Any idea what is the root cause? We are using the device tree based on the dk2 sample device tree. INFO:   ARM GICv2 driver initializedINFO:   stm32mp HASH1 (9): SecureINFO:   stm32mp RNG1 (15): SecureINFO:   stm32mp USART1 (18): SecureINFO:   ETZPC: SPI6 (4) could be non secureINFO:   ETZPC: CRYP1 (9) could be non secureINFO:   ETZPC: I2C6 (12) could be non secureINFO:   SP_MIN: Initializing runtime servicesINFO:   SP_MIN: Preparing exit to normal worldINFO:   PSCI Power Domain Map:INFO:     Domain Node : Level 1, parent_node -1, State ON (0x0)INFO:     Domain Node : Level 0, parent_node 0, State ON (0x0)INFO:   &

I am asking because I want to remove some of them, how can I find out what are the names and what they do? What is the best way to do this?

We are trying to verify the signature generated by the STM signing tool. From this link, in the table for "On STM32MP15x lines", we need to know what data type is used for each field when the signature hash (sha256) is created so we can replicate it.

Currently our custom board powers up in two cases, one when you press the power on button, and another when you plug in the USB cable. We would like to disable PMIC from powering on when USB cable is plugged in. We see that the STPMIC specification document (section 5.4.2), lists the three conditions which triggers PMIC turn-on and also the register bit VBUSOTG_DET_DIS which needs to be set to 1 to disable this feature.Since I cannot write to PMIC register address 0x40, was wondering if there is any device tree setting which sets this bit? Any information on how to do this would be very much appreciated.Thanks!

I was analyzing stm32mp15 disco and eval board schematics for my project. I noticed that DDR pinning are not the same between them. Disco boards have 512MB and EVAL boards have 2x512MB. My project also going to have two variants. I wanted to know why pinouts of these two schematics are different and can I use same pcb for both of my variants with one 1x512 and other with 2x512. (1x512 will have two locations on the pcb but only one chip. Is it possible?)

Hello we are using STM32mp153 for multiple products, and I'm considering buiding a generic TF-A bootloader with one device tree with all the clocks enabled for all the peripherals. I would like to do this in order to make it more generic across our portfolio.Does this come with any potential problems down the line?

..

Does it have a linux driver and is it supported? How can I use it?

Hello,actually, our processor STM32MP153 (dual core) is completely full with the "normal" memory regions regarding the MCUs (M4) code and data. So now, we decided to add the RETRAM_data form ORIGIN = 0x00001000, LENGTH = 0x0000F000 /*60k*/The linker works as expected but if I start a DMA transfer with variables placed inside this area, the DMA stucks directly with a transfer error. The same code but placed in the RAM1_dataORIGIN = 0x10030000, LENGTH = 0x00010000 /* 64K*/works fine.I know this problem only if the area is not accessible by the DMA as on the H7 but if I have a look at the memory layout and master/slave bus matrix, both DMA should be able to send data from the RetRAM.Any ideas what could lead to this problem? I checked also to place the segment at 0x38000000 but this leads to a linux boot stuck.My actual workaround is to place variables not used by the DMA into the RetRAM section .bss1 with the __attribute__((section (".bss1")))command. This works but, this is my next smal

In stm32mp157c-ed1.dts, the MPU pin setting is written in the stm32mp15-pinctrl.dtsi file, but in CubeMX, a node called @pinctrl is created in the generated .dts file and written in it.Why?Can I delete this part and create a ***.pinctrl.dtsi file?

Hello , Can anyone please point me to the example of SPI communication between stm32(F4/L0/L4) with A7 of STM32MP1?A7 could be a driver or any user application but both should be able to send and receive the data over spi.on A7 I suppose it will be SPI5.please assist.Thanks in advance!Br,,P

I have a custom build on the stm32mp157f dk2.stm TF-A v2.6-r1 or vanilla TF-A v2.7, using sp_min.stm U-boot 2020.10-r2 or vanilla Barebox-2022.06.I've also tried U-boot 2021.10-r1, but that one crashes the kernel even earlier.Kernel: vanilla 5.18/vanilla 5.19.FSBL starts fine.Both v2.6 and v2.7 starts just fine. sp_min too.Both TF-A versions load and executes BL33 (stm u-boot or vanilla barebox) just fine.But when I use stm TF-A v2.6 kernel crashes when started from BL33.Works fine with vanilla TF-A 2.7 all other components being equal.I can see the kernel decompression but stops executing directly after that.This happens regardless if I do the fip manually or if TF-A puts it together.Kernel 5.18 or 5.19 doesn't matter. But when I use vanilla TF-A both kernels boot.TF-A is configured exactly the same using the variable configuration.

When I use cmake,/opt/st/stm32mp1/4.0.1-openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15/sysroots/x86_64-ostl_sdk-linux/usr/share/cmake-3.22/Modules/CMakeTestCXXCompiler.cmake:62: error: The C++ compiler "/opt/st/stm32mp1/4.0.1-openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15/sysroots/x86_64-ostl_sdk-linux/usr/bin/arm-ostl-linux-gnueabi/arm-ostl-linux-gnueabi-g++" is not able to compile a simple test program. It fails with the following output: Change Dir: /home/zxb/qt_proj/build-untitled1-Stm32mp1-Debug/CMakeFiles/CMakeTmp Run Build Command(s):/opt/st/stm32mp1/4.0.1-openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15/sysroots/x86_64-ostl_sdk-linux/usr/bin/ninja cmTC_1f845 && [1/2] Building CXX object CMakeFiles/cmTC_1f845.dir/testCXXCompiler.cxx.o [2/2] Linking CXX executable cmTC_1f845 FAILED: cmTC_1f845 : && /opt/st/stm32mp1/4.0.1-openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15/sysroots/x86_64-ostl_sdk-linux/usr/bin/arm-ostl-linux-gnueabi/arm-ostl-linux-gnueabi-g++ --sysroot=/

Hi,I follow "https://wiki.st.com/stm32mpu/wiki/PMIC_hardware_components#Support_in_Cortex-A7_Secure" can port the ecosystem 3.0 on the kit "stm32mp157c-odyssey" successfully, but fail to port the ecosystem4.0.The error log is :" 14:13:55 : Start Embedded Flashing service 14:13:55 : Memory Programming ... 14:13:55 : Opening and parsing file: tf-a-stm32mp157c-odyssey-usb.stm32 14:13:55 :  File     : tf-a-stm32mp157c-odyssey-usb.stm32 14:13:55 :  Size     : 225.76 KB  14:13:55 :  Partition ID : 0x01  14:13:55 : Download in Progress: 14:13:56 : File download complete 14:13:56 : Time elapsed during download operation: 00:00:01.064 14:13:56 : RUNNING Program ...  14:13:56 :  PartID:   :0x01  14:14:14 : Error: Start operation failed at partition 0x01 14:14:14 : Error: TSV flashing service failed 

I am getting "Error: can't fuse again in the OTP" when I tried registering a key on boards that shouldn't have been fused before. Is there a way to confirm if they are fused? Heres a picture of the output: Am I missing a step? I've been following "STM32 MPU ROM code secure boot" and "How to use U-Boot stm32key command" articles on how to enable secure boot.

Our hardware has an STM32MP151DAx.In order to make use of the Cube Programmer and the boot-via-USB method I need the equivalent of the tf-a-stm32mp157d-ev1-usb.stm32 file but for our hardware.I have worked through the IOC stuff, have DTS files and got the hang of building TF-A (and U-Boot and Linux) according to instructions but how do I build this -USB variant as used in the .TSV files ?Jon

Hi,We are using stm32mp157a-dk1 board.So we want to change the default console serial port(UART4) to anthour serial port(UART 7). I have changed respective .dts files in uboot it is working properly up to uboot.Know i want to change in kernal.So i Refereed St wiki.They said that we need to enable to CONFIG_DEBUG_LL, CONFIG_STM32MP1_DEBUG_UART and CONFIG_EARLY_PRINTK using the Linux Menuconfig.Refer below link for your referencehttps://wiki.st.com/stm32mpu/wiki/Dmesg_and_Linux_kernel_log#earlyprintk I am unable to active those configurations through linux menuconfig. I am using linux-5.13 version. How can i activate CONFIG_DEBUG_LL, CONFIG_STM32MP1_DEBUG_UART and CONFIG_EARLY_PRINTK?I am attaching my screenshots for linux-menuconfig and my uboot log.could you please help us in this regards.

Hi,We are using stm32mp157a processor in my custom board.I have download stm32mp157a-dk1 discovery kit code.The difference between discovery board and my custom board is Debug Serial port Configuration.In Discovery Board The debug uart pins are PG11-->UART4_TXPB2-->UART4_RX. where as in my custom board the pins are different PE0--> UART8_TXPE1-->UART8_RXSo I have changed the pin muxing and nodes in dts files. And debug_uart_init() in Uboot source code. Up to Uboot It is working fine. I have done the same configuartion except debug_uart_init() in kernal source code.why because debug_uart_init() function is not there in linux source code. when can i get kernal log in my custom board. Please help me in this regards. You can see my log below.

When I change the download directory for my yocto project in my local.conf file, this note pops up, what is this and should I do anything about it.NOTE: /home/gencay/STM32MPU_workspace/openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15/layers/meta-st/meta-st-openstlinux/recipes-st/images/st-image-resize-initrd.bb: ENABLE_PARTITIONS_IMAGE not enabled:45

To detach signing from cube programmer I made an open source C signing tool, using openssl. Should be rather easy to use in various build environments.https://github.com/milkylainen/stm32mp1sign

Hello everyone,I am trying to define a machine configuration file for a custom board using the DistributionPackage (I am following this approach from the ST wiki and my directory tree is the same described there).Currently, I managed to successfully boot my custom board using the DeveloperPackage, this makes me believe that the device tree files I generated using the ST MX Cube are okay.From my understanding, it is not possible to generate the filesystem files (e.g. ext4) within the DeveloperPackage (I used the ones from the StarterPackage); can you confirm this?So far, I populated the mx directory under <path of STM32MP1_Distribution_Package>/layers/meta-st/meta-st-stm32mp-addons/with the device tree files I generated using the DeveloperPackage and its directory tree looks as it follows:mx └── es-frank-ghiacciolo    ├── kernel    │   └── stm32mp157c-es-frank-ghiacciolo-mx.dts    ├── optee-os    │   └── stm32mp157c-e

I'm targeting:openstlinux-5.15-yocto-kirkstone-mp1-v22.07.27DISTRO: openstlinux-eglfsMACHINE: stm32mp15-discoImage+SDK build fine. I create the raw file as usual, image SD card, and boot fails with:NOTICE: Model: STMicroelectronics STM32MP157F-DK2 Discovery BoardNOTICE: Board: MB1272 Var4.0 Rev.C-02INFO:   PMIC version = 0x20INFO:   Reset reason (0x15):INFO:     Power-on Reset (rst_por)INFO:   FCONF: Reading TB_FW firmware configuration file from: 0x2ffe2000INFO:   FCONF: Reading firmware configuration information for: stm32mp_ioINFO:   Using SDMMCINFO:     Instance 1INFO:   Boot used partition fsbl1NOTICE: BL2: v2.6-stm32mp1-r1.0(debug):v2.6-dirtyNOTICE: BL2: Built : 13:14:26, Nov 23 2021WARNING: CRC check failed on FWU metadataWARNING: loading of FWU-Metadata failed, using Bkup-FWU-MetadataWARNING: CRC check failed on FWU meta

Hi MPU's Community, A "minor"patch release 4.0.2 is available for OpenSTLinux Distribution package.The corresponding oe-manifest tag is openstlinux-5.15-yocto-kirkstone-mp1-v22.07.27. It contains some fix of issues detected early after DV4.0.0 / openstlinux-5.15-yocto-kirkstone-mp1-v22.06.15 release and you may face. Here is the changelog : V4.0.2 ecosystem release        o layers/meta-st/meta-st-openstlinux from [182b150] to [cbb3a2e]     - [Update] CONF: update distro to Yocto 4.0.2     - [Update] GSTREAMER1.0: update to 1.20.2     - [Update] SYSTEMD: update to 250.5        o layers/meta-st/meta-st-stm32mp from [d00d3a0] to [00c2494]     - [Update] CLASS: FLASHLAYOUT: correct bad exit of loop     - [Update] CONF: correct examples to used according to devicetree package  &n