STM32CubeMX (MCUs)

Dear STMicroelectronics Support Team, I hope this message finds you well. I would like to report a clock configuration issue encountered when working with TouchGFX and STM32CubeMX for the STM32F746G-DISCO board. When creating a project using TouchGFX and selecting the STM32F746G-DISCO board, everything works as expected initially. However, after opening the project in STM32CubeMX and enabling the SDMMC interface, the tool prompts to "Resolve clock issues." Upon accepting the proposed solution, CubeMX automatically modifies the clock configuration in a way that alters the LTDC (TFT LCD) pixel clock from its original value of 9.6 MHz to a different value. As a result, the display no longer renders correctly. After investigation, I realized that the auto-resolve feature in CubeMX does not preserve the required clock settings for the display, and instead prioritizes SDMMC configuration. To fix

when I use V6.14.1 everything goes well, bug when I update to V6.15,the same IOC file shows LTDC RGB565 confict with usart2, when I disabled usart2, re-enabled LTDC RGB565,LTDC_R1 pin come out but no R0,R2, it must be a bug here!

I'm developing an app on the demo board STM32U5A9J-DKI've created the project from TouchGFX 4.25.0 project STM32U5A9J-DK FreeRTOSIt perfectly works, compile and debug with both TouchGFX and CubeIDE 1.18.1I then opened the .ioc file of the project with Open CubeMX 6.14.1. It says the project was set up with 6.14.0 and FW_U5 V1.7.0, so I migrated it.I've updated the Software Packs from X-CUBE-FREERTOS 1.0.0 to 1.3.1. CubeMX says no issue found and upgraded.Then I generated the code. There, the new code rewrote the app_freertos.c file deleting the initialization of the TouchGFX task/* Definitions for defaultTask */ osThreadId_t defaultTaskHandle; const osThreadAttr_t defaultTask_attributes = { .name = "defaultTask", .priority = (osPriority_t) osPriorityNormal, .stack_size = 128 * 4 }; /* Definitions for GUI_Task */ osThreadId_t GUI_TaskHandle;

When I use CubeMX, the default language of my genereted code is C. I would like to develop in C++, as I can do in STM32CubeIDE. (I want to use VSCode to program)Is there any option to do this? I've tried just change the main.c to main.cpp, but it didn't work.(I'm a noob. So, if this questions is elementar, please, let me know likewise).

I just upgraded STM32CubeIDE from 1.17 to 1.19 which in turn upgraded MX from 6.13 to 6.15Prior to updating, in MX Clock Configuration I could clock the SPI system on the STM32H533 at the same rate as the CPU clock (250 MHz with no error). Now, to avoid the configuration error, I have to reduce it to 125 MHz.My code ran fine at 250 MHz in the previous version. IS THIS A FIX OR A BUG?Thanks!

In my project for an STM32H7S3, I have a number of GPIO settings.However, every time I open the file, one particular pin has its settings reverted to the defaults.Also, if I save the IOC file without making any changes at all, I can see in the diff that the settings for that pin have been removed:I'm currently using Cube 6.15.0, though this file was created with an earlier version.I'm not willing to share my IOC file here, though I'd be happy to send it privately to someone at ST.(See also

Hello everybody,I am trying to generate a project for the STM32H7S78 using the latest version of STM32CubeMX. However, when I click on "Generate Code", the process stops at "Copying library files" and does not proceed any further.I usually work with IAR Embedded Workbench for ARM (v9.40.2), but I also tried using STM32CubeIDE, and the issue occurs there as well.Do you have any idea how to solve this problem?Thank you very much in advance! 

I cannot enable DCMI for STM32H753ZIT in STM32CubeMX version 6.15.0. I'd assume DCMI can be indeed used and that this is just a bug on CubeMX's part, but I'd like to verify before I do any orders. Related:- https://community.st.com/t5/stm32-mcus-products/minimum-clock-speed-and-enabling-pssi-problem/td-p/709772- 

Operating System: Windows 10STM32CubeMX version: 6.10.0STM32CubeU5 Firmware Package V1.4.0 / 20-October-2023When I attempt to validate my X-CUBE-AI model on target, I receive the following pop-up:Exploring STM32CubeMX.log shows the following warning:[WARN] IP:2119 - SRAM6PowerDownInRunMode not found: check if there is a RefParameter without a condition.The U5A5 doesn't even have an SRAM6, if I am reading the datasheet correctly.  When I look at the configuration options in CubeMX, I can't see anything for SRAM6 in the PWR section.

We have run into an issue with the NetXduo driver in V1.5.0 for the STM32H5 series (and maybe others, we did not check).WhatWhen IGMP is enabled (to allow reception of multicast messages), the reception of broadcast messages gets disabled. This breaks basic network functionality like ARP, making the device unreachable for others on the network.WhyIn the NetX driver (Middlewares/ST/netxduo/common/drivers/ethernet/nx_stm32_eth_driver.c), there is a structure called FilterConfig. Line 1500 sets the member FilterConfig.BroadcastFilter to ENABLE - which will be interpreted by the HAL to configure the broadcast filter in the ETH preipheral (Bit 5/DBF in register ETH_MACPFR on the STM32H573 used here).The structure is not used until a call to nx_igmp_enable() trickles down to _nx_driver_hardware_multicast_join() in the driver, at which point a different bit (to allow multicast reception)

Hi,I'm using a STM32f767ZIT with FreeRTOS and LWIP on a custom board.After running into some weird UDP transmission issues, I discoverd that in the generated ethernetif.c file, the DCache is not cleaned in the low_level_output function.Adding this at line 414 ( after Txbuffer[i].len = q->len; ) in low_level_output() fixed my issues.SCB_CleanDCache_by_Addr(    (uint32_t *)((uint32_t)(Txbuffer[i].buffer) & ~0x1F),    Txbuffer[i].len + ((uint32_t)(Txbuffer[i].buffer) & 0x1F));In the HAL_ETH_RxLinkCallback, the SCB_InvalidateDCache_by_Addr is used, so this seems like an oversight?

arm-none-eabi-gcc -mcpu=cortex-m4 -g3 -DDEBUG -c -x assembler-with-cpp -MMD -MP -MF"Core/Startup/startup_stm32f429igtx.d" -MT"Core/Startup/startup_stm32f429igtx.o" --specs=nano.specs -mfpu=fpv4-sp-d16 -mfloat-abi=hard -mthumb -o "Core/Startup/startup_stm32f429igtx.o" "../Core/Startup/startup_stm32f429igtx.s"arm-none-eabi-gcc "../Core/Src/eth.c" -mcpu=cortex-m4 -std=gnu11 -g3 -DDEBUG -DUSE_HAL_DRIVER -DSTM32F429xx -c -I../Core/Inc -I../Drivers/STM32F4xx_HAL_Driver/Inc -I../Drivers/STM32F4xx_HAL_Driver/Inc/Legacy -I../Drivers/CMSIS/Device/ST/STM32F4xx/Include -I../Drivers/CMSIS/Include -I../LWIP/App -I../LWIP/Target -I../Middlewares/Third_Party/LwIP/src/include -I../Middlewares/Third_Party/LwIP/system -I../Drivers/BSP/Components/lan8742 -I../Middlewares/Third_Party/LwIP/src/include/netif/ppp -I../Middlewares/Third_Party/LwIP/src/include/lwip -I../Middlewares/Third_Party/LwIP/src/include/lwip/apps -I../Middlewares/Third_Party/LwIP/src/include/lwip/priv -I../Middlewares/Third_Part

Missing pre-processor symbol check in file "Middlewares/ST/ST_Cryptolib/Inc/Commonn/types.h"current line 49:#if !defined(_STDINT_H) && !defined(__stdint_h) && !defined(_STDINT) new line 49:#if !defined(_STDINT_H) && !defined(__stdint_h) && !defined(_STDINT) && !defined(_GCC_STDINT_H)To solve this bug add ``&& !defined(_GCC_STDINT_H)`` on line 49. Cheers,Andreas

Hi, I have configured My RTC with LSE of 32.768Khz for STM32G0B1CETxAnd In NVIC Setting RTC & TAMP interrupts are enabled..All these are done using Cube MxAfter HAL_RTC_Init() call from main;  RTC_TAMP_IRQHandler is getting called continuously. Time and alarm is configured after  HAL_RTC_Init(). which is not even called.Any inputs could be helpful.Thank youSuraj Prem

hi everyone, i got a problem when i use MX generate code and open folder，i try to reinstall and didn't solve。it shows not responsible

Hi, we are working on an open-source library OS in Rust for embedded architectures, result of a collaboration between the French research institution INRIA and the Freie Universität Berlin, called Ariel OS. Our OS currently supports a variety of STM32 MCUs and we expect the list to grow with time.To streamline user experience, we would like to provide some ways of automatically generating the clock configuration for the STM32 MCUs. We are envisioning a utility where users simply ask for specific frequency ranges for specific clocks and this utility would figure out the correct dividers/multipliers/multiplexers to use to comply with the users requirements.To achieve this, we would need to be able to recreate the entire clock tree logic. As far as I understand, CubeMX recreates the clock tree in software and is able to calculate frequencies based on the chosen/defaults values of the

Hi, I am using the Nucleo-H7A3ZI-Q Board and was trying to use the DMA to transmit SPI frames. Only creating simple SPI TX Circular DMA request works when generating the code from STM32CubeIDE 1.12.1 (with STM32CubeMX 6.8.1). I get two linker scripts STM32H7A3ZITXQ_FLASH.ld and STM32H7A3ZITXQ_RAM.ld. In the STM32H7A3ZITXQ_FLASH.ld, I get this memory sectionMEMORY { ITCMRAM (xrw) : ORIGIN = 0x00000000, LENGTH = 64K FLASH (rx) : ORIGIN = 0x08000000, LENGTH = 2048K DTCMRAM1 (xrw) : ORIGIN = 0x20000000, LENGTH = 64K DTCMRAM2 (xrw) : ORIGIN = 0x20010000, LENGTH = 64K RAM (xrw) : ORIGIN = 0x24000000, LENGTH = 1024K RAM_CD (xrw) : ORIGIN = 0x30000000, LENGTH = 128K RAM_SRD (xrw) : ORIGIN = 0x38000000, LENGTH = 32K }When I was generating the code from STM32CubeMX (alone) to generate code with the "toolchain/IDE" option set to Makefile, the same code does not work. Diff reveals n

Hi,After STM32CubeMX update to the 6.15.0 version, the HAL_DMA_ConfigChannelAttributes() calls get removed in the generated FSBL and Appli (secure) code while no changes in the *.ioc file have been made.Could you please elaborate on the behavior and consequences? 

Hello, what do I wrong?  I am trying to setup a new stm32f7xx project, with CMSIS_V2 and LWIP.I use only the CubeMX application at the version 6.15.0, I have tried several hours to get this setting but nothing working.If I select CMSIS_V1 the LWIP use the RTOS (WITH_RTOS is Enabled), if a switch to CMSIS_V2 it is disabled.All settings are CubeMX default.

Recently I'm trying to setup CLion and STM32CubeMX as I want to switch to a new IDE (By the way, the fact that STM32CubeIDE still doesn't have features like auto-triggered content asist, and that its poor, buggy GUI has bad compatibility with high DPI screen are just annoying). However, when I successfully generate the project in CubeMX and open it in CLion, CLion refuses to compile the project no matter what. I tried the bundled MinGW, CubeCLT1.18 and the older 1.16, and also CLion 2025.1, 2025.1.2, 2025.1.3. All didn't work. Up till this point I've been using CubeMX 6.15, and so I tried to install and use the older version (6.14.0). Everything suddenly works then. The project compiles without problem. Therefore, I suspect that there are some problems in the newest release of CubeMX (6.15) that are generating faulty CMake files.Here are some screenshots of what happened. I used the same version of CLion and CubeCLT in the same configuration on two projects generated by CubeM

16bit SDRAM with 4 banks, 9bit column address, 13bit row should show as 32MB memory area,but it is shown as 16MB.You have to change column address bits to 10bit and back to 9 bits to set the value correctly (32MB).It is dangerous beacause code generated with wrong 16MB area generate exception, if more memory have been used.

  When I try to log in to my ST account in STM32CubeMX, it sends me an error message that says, 'This feature requires an internet connection.' However, my internet connection is working well, and I can use other browsers to surf the internet normally. How can I solve this problem?

When I assign pinouts for LED1 and LED3 on the NUCLEO-H753ZI in STM32CubeMX with the BSP selected, they are no longer assigned when the project is reloaded.

When I use cubeide, it is OK to select c++ as targeted language. However, I can not find such a option in cubemx. Does cubemx support to generate c++ project with cmake? If not, could it be supported in the future? 

Hello,I'm working with the Nucleo-H7S3L8 board in a macOS development environment, and so far, the experience has been quite frustrating. The documentation is inconsistent, scattered across multiple sources, and often inaccurate. This has made development more difficult than it should be.For context:I’ve previously worked with the Nucleo-F401RE, and the development experience was significantly smoother — configuration was straightforward, and documentation was far more coherent.Here's the issue I'm currently facing:I imported the Template_XIP project into STM32CubeIDE. The project compiled, uploaded to the board successfully, and I was able to debug without any issues. However, after attempting to enable SPI1 using both STM32CubeIDE and STM32CubeMX, and generating code based on those settings, I encountered the following build error in the Template_XIP_Appli project: