STM32CubeProgrammer 2.20: Check out the new MCU features supported and a big welcome to Apple silicon

STM32CubeProgrammer 2.20 is the first version of the ST tool to natively support Apple silicon Macs, to offer Japanese and Chinese localization, and greatly improve incremental update programming. Behind the scenes, our teams applied 184 fixes and dealt with a total of 295 tickets, which is a testament to the massive endeavor that goes behind such a utility. Ensuring more regions can enjoy STM32CubeProgrammer in their native language, supporting more processor architectures, and making new features on STM32 devices accessible is a big part of STM32CubeProgrammer, but it’s not the only one. Closing nearly 300 tickets demonstrates the commitment ST developers bring to ensure all users enjoy a more optimal and smooth workflow.

What’s new in STM32CubeProgrammer 2.20?

Native support of Apple silicon

The new version of STM32CubeProgrammer is an important one for Mac users as it brings native support for Apple silicon. Many in our communities work on legacy systems to protect their work against unreliable environments, which is why we also waited to support the new processor architecture. However, Apple recently announced that macOS Tahoe would be their last OS to support Intel-based Mac computers and that macOS 27 will be the last version to offer Rosetta, the translation process that ensures Apple Silicon devices can run Intel-based apps. Hence, it was essential to bring native Apple silicon support now to ensure that users planning a migration will get the best STM32CubeProgrammer experience on macOS for years to come.

Japanese and Chinese localization

To make STM32CubeProgrammer more inclusive, we are bringing Chinese and Japanese localization. It reflects our desire to make the entire STM32Cube ecosystem more accessible to developers in those regions. The vast majority of users who speak Japanese and Chinese as their first language also speak English fluently and would have no issue using our tool in English. However, given our presence and dedication in those markets, we also know that translating our solutions opens the door for students, enthusiasts, and experts who prefer to work in their native language. That’s why we localized the STM32 MCU and MPU Developer Zone in Chinese and Japanese and why we are now bringing the same devotion to STM32CubeProgrammer.

Improvements to incremental programming

Version 2.20 of STM32CubeProgrammer brings important updates to the incremental programming features ST introduced in version 2.19 (more information below), by bringing new MCU support and performance improvements. In terms of compatibility, developers can now perform incremental programming on STM32WB5 and STM32WB35 devices as well as STM32G0B0, STM32G0B1, and STM32G0C1 MCUs. Additionally, besides bug fixes, the 2.20 update improves performance. In our internal tests, while a 1 MB firmware takes about 23 second to load on an STM32H7, it takes only 5 seconds when using incremental updates to write to two memory sectors, even if they are not contiguous. When developers do this over and over as they write their application, the savings become substantial.

Support for the STM32MP21

As is customary, a new release of STM32CubeProgrammer is synonymous with new STM32 device support, and version 2.20 is no exception. In this instance, we are adding compatibility with the new STM32MP21, which includes a single-core Cortex-A35 running at 1.5 GHz and a Cortex-M33. The MPU also features a camera interface to enable computer vision applications on more cost-effective systems that don’t run a complex GUI requiring a GPU. Concretely, STM32CubeProgrammer now supports programming the STM32MP21 one-time programmable partition, using PKCS#11 when encrypting its firmware, and ensure compatibility with the ST power management IC. Put simply, it puts the new device on par with the other MPUs in our line up.

MCE Support for the STM32N6

While STM32CubeProgrammer has officially supported the STM32N6 since version 2.18, version 2.20 brings support for external memory encryption. The microcontroller is our most powerful STM32 MCU and our first to feature a neural processing unit. Consequently, while it features the most embedded RAM ever on an MCU (4.2 MB), it doesn’t contain any flash, as engineers will require significantly more storage than can be accommodated on the die. However, it means that encryption and decryption operations are trickier. STM32CubeProgrammer can now take advantage of the STM32N6 Memory Cipher Engine (MCE) to encrypt and decrypt information from and to the flash and SRAM on the fly.

Expansion of the feature set for the STM32H5

With this latest version STM32CubeProgrammer, ST brings the ability to use Secure Manager on the external flash of an STM32H573. The STM32H5 represents the new mainstream STM32 MCU thanks to the most powerful Cortex-M33 implementation. STM32CubeProgrammer already supports previous versions of the device. Put simply, ST continues to make all its security features available on our latest devices and with various configurations, ensuring engineers can benefit from them as soon as possible. Moreover, version 2.20 also supports the factory reset option bytes on STM32H5 devices. Put simply, restoring the STM32 device to its factory settings only takes a few clicks, detailed on our community, which makes sharing an MCU with other teams or on another project a lot simpler.

What’s STM32CubeProgrammer? An STM32 flasher and debugger

At its core, STM32CubeProgrammer helps debug and flash STM32 microcontrollers. As a result, it includes features that optimize these two processes. For instance, version 2.6 introduced the ability to dump the entire register map and edit any register on the fly. Previously, changing a register’s value meant changing the source code, recompiling it, and flashing the firmware. Testing new parameters or determining if a value is causing a bug is much simpler today. Similarly, engineers can use STM32CubeProgrammer to flash all external memories simultaneously. Traditionally, flashing the external embedded storage and an SD card demanded developers launch each process separately. STM32CubeProgrammer can do it in one step.

Another challenge for developers is parsing the massive amount of information passing through STM32CubeProgrammer. Anyone who flashes firmware knows how difficult it is to track all logs. Hence, we brought custom traces that allow developers to assign a color to a particular function. It ensures developers can rapidly distinguish a specific output from the rest of the log. Debugging thus becomes a lot more straightforward and intuitive. Additionally, it can help developers coordinate their color scheme with STM32CubeIDE, another member of our unique ecosystem designed to empower creators.

What are some of its key features?

New MCU support

Most new versions of STM32CubeProgrammer support a slew of new MCUs. For instance, version 2.16 brought compatibility with the 256 KB version of the STM32U0s. The device was the new ultra-low power flagship model for entry-level applications thanks to a static power consumption of only 16 nA in standby. STM32CubeProgrammer 2.16 also brought support for the 512 KB version of the STM32H5, and the STM32H7R and STM32H7S, which come with less Flash so integrators that must use external memory anyway can reduce their costs. Put simply, ST strives to update STM32CubeProgrammer as rapidly as possible to ensure our community can take advantage of our newest platforms rapidly and efficiently.

SEGGER J-Link probe support

To help developers optimize workflow, we’ve worked with SEGGER to support the J-Link probe fully. This means that the hardware flasher has access to features that were previously only available on an ST-LINK module. For instance, the SEGGER system can program internal and external memory or tweak the read protection level (RDP). Furthermore, using the J-Link with STM32CubeProgrammer means developers can view and modify registers. And since version 2.17, we added the ability to generate serial numbers and automatically increment them within STM32CubeProgrammer, thus hastening the process of flashing multiple STM32s in one batch.

We know that many STM32 customers use the SEGGER probe because it enables them to work with more MCUs, it is fast, or they’ve adopted software by SEGGER. Hence, STM32CubeProgrammer made the J-Link vastly more useful, so developers can do more without leaving the ST software. In fact, version 2.18 added the ability to securely install the Bluetooth stack on an STM32WB via a J-Link probe. Hence, developers can use their SEGGER tool for more use cases, making these features more widespread.

Exporting options

Other quality-of-life improvements aim to make STM32CubeProgrammer more intuitive. For instance, it is now possible to export an STM32’s option bytes. Very simply, they are a way to store configuration options, such as read-out protection levels, watchdog settings, power modes, and more. The MCU loads them early in the boot process, and they are stored in a specific part of the memory that’s only accessible by debugging tools or the bootloader.

By offering the ability to export and import option bytes, STM32CubeProgrammer enables developers to configure MCUs much more easily. Since version 2.17, developers can now edit memory fields in ASCII to make certain section a lot more readable. And version 2.18 brought a synthetic option byte view to see and edit multiple option bytes on a single row instead of having to scroll through detailed lists. For expert users who know exactly what they want to do, this synthetic view makes changing an option byte a lot quicker.

We also introduced improvements to the user experience, such as a project mode that allows users to save and restore configuration and connection settings, option byte values, firmware lists, external flash loaders, security firmware updates (root security system extension binaries), stack install settings for the STM32WB, and automatic mode parameters. In a nutshell, we want developers to collaborate more efficiently by importing and exporting major project elements so they can focus on their code rather than ticking boxes and applying the same settings repeatedly.

Automating the developer’s life

STM32CubeProgrammer 2.19 brought a series of automations that simplified operations. For instance, the utility now automatically downloads the X-CUBE-RSSe file, which contains the latest version of RSS extension binaries, option byte templates, and the personalization data files for the STM32HSM-V2 hardware security module. Users will see a message on the SFI/SFIx tab on the secure programming section of STM32CubeProgrammer telling them that a new version is available for download and a tab informing them of the main changes. Given the highly sensitive nature of these security libraries and features, it ensures developers keep their software patched and free of known issues.

This latest software revision also enhances serial numbering by enabling users to choose whether they prefer a big- or little-endian format. Big-endianness mimics a left-to-right approach, as is usual in most languages using an alphabet, and is most common when dealing with network addresses. However, little-endianness is very prominent on embedded systems because storing the least significant byte (LSB) in the lowest memory address makes reading the address more efficient on resource-constraint systems. For instance, an operation that only needs the LSB will have faster access to this information. That’s why most MCUs default to little-endianness, but STM32CubeProgrammer gives the option to use both.

Similarly, STM32CubeProgrammer 2.19 brought incremental programming when using the command line interface by automatically updating only the sectors in the flash where changes have been made. Instead of flashing the firmware onto the entire flash, the tool figures out which memory sectors contain updates and changes only those sections of the flash. Most STM32 devices support it already, and we will continue to add compatibility with more MCU series and memory configurations over time. Finally, ST brings debug authentication with multiple target boards. Put simply, users can choose one target board and seamlessly switch between several ones.

Automating the installation of a Bluetooth LE stack

Until now, developers updating their Bluetooth LE wireless stack had to figure out the address of the first memory block to use, which varied based on the STM32WB and the type of stack used. For instance, installing the basic stack on the STM32WB5x would start at address 0x080D1000, whereas a full stack on the same device would start at 0x080C7000, and the same package starts at 0x0805A000 on the STM32WB3x with 512 KB of memory. Developers often had to find the start address in STM32CubeWB/Projects/STM32WB_Copro_Wireless_Binaries. The new version of STM32CubeProgrammer comes with an algorithm that determines the right start address based on the current wireless stack version, the device, and the stack to install.

A portal to security on STM32

Readers of the ST Blog know STM32CubeProgrammer as a central piece of the security solutions present in the STM32Cube Ecosystem. The utility comes with Trusted Package Creator, which enables developers to upload an OEM key to a hardware secure module and to encrypt their firmware using this same key. OEMs then use STM32CubeProgrammer to securely install the firmware onto the STM32 SFI microcontroller. Developers can even use an I²C or SPI interface, which gives them greater flexibility. Additionally, the STM32H73x, STM32H7Bx, STM32L5, STM32U5, and STM32H5 also support external secure firmware install (SFIx), meaning that OEMs can flash the encrypted binary on memory modules outside the microcontroller.

X-CUBE-RSSE

To facilitate updates to RSSe (Root Security System binaries extensions) binaries, STM32HSM-V2 personalization files, and option bytes templates, these elements are now delivered separately in the X-CUBE-RSSe expansion package supported by both STM32CubeProgrammer and Trusted Package Creator tools. Consequently, these elements are no longer part of the lastest version of CubeProgrammer and should be downloaded separately.

Secure Manager

Secure Manager is officially supported since STM32CubeProgrammer 2.14 and STM32CubeMX 1.13. Currently, the feature is exclusive to our new high-performance MCU, the STM32H573, which supports a secure ST firmware installation (SSFI) without requiring a hardware secure module (HSM). In a nutshell, it provides a straightforward way to manage the entire security ecosystem on an STM32 MCU thanks to binaries, libraries, code implementations, documentation, and more. Consequently, developers enjoy turnkey solutions in STM32CubeMX while flashing and debugging them with STM32CubeProgrammer. It is thus an example of how STM32H5 hardware and Secure Manager software come together to create something greater than the sum of its parts.

Other security features for the STM32H5

STM32CubeProgrammer enables many other security features on the STM32H5. For instance, the MCU now supports secure firmware installation on internal memory (SFI) and an external memory module (SFIx), which allows OEMs to flash encrypted firmware with the help of a hardware secure module (HSM). Similarly, it supports certificate generation on the new MCU when using Trusted Package Creator and an HSM. Finally, the utility adds SFI and SFIx support on STM32U5s with 2 MB and 4 MB of flash.

Making SFI more accessible

Since version 2.11, STM32CubeProgrammer has received significant improvements to its secure firmware install (SFI) capabilities. For instance, in version 2.15, ST added support for the STM32WBA5. Additionally, we added a graphical user interface highlighting addresses and HSM information. The GUI for Trusted Package Creator also received a new layout under the SFI and SFIx tabs to expose the information needed when setting up a secure firmware install. The Trusted package creator also got a graphical representation of the various option bytes to facilitate their configuration.

Secure secret provisioning for STM32MPx

Since 2.12, STM32CubeProgrammer has a new graphical user interface to help developers set up parameters for the secure secret provisioning available on STM32MPx microprocessors. The mechanism has similarities with the secure firmware install available on STM32 microcontrollers. It uses a hardware secure module to store encryption keys and uses secure communication between the flasher and the device. However, the nature of a microprocessor means more parameters to configure. STM32CubeProgrammers’ GUI now exposes those settings previously available in the CLI version of the utility to expedite workflows.

Double authentication

Since version 2.9, the STM32CubeProgrammer supports a double authentication system when provisioning encryption keys via JTAG or a Boot Loader for the Bluetooth stack on the STM32WB. Put simply, the feature enables makers to protect their Bluetooth stack against updates from end-users. Indeed, developers can update the Bluetooth stack with ST’s secure firmware if they know what they are doing. However, a manufacturer may offer a particular environment and, therefore, may wish to protect it. As a result, the double authentication system prevents access to the update mechanism by the end user. ST published the application note AN5185 to offer more details.

PKCS#11 support

Since version 2.9, STM32CubeProgrammer supports PKCS#11 when encrypting firmware for the STM32MP1. The Public-Key Cryptography Standards (PKCS) 11, also called Cryptoki, is a standard that governs cryptographic processes at a low level. It is gaining popularity as APIs help embedded system developers exploit its mechanisms. On an STM32MP1, PKCS#11 allows engineers to segregate the storage of the private key and the encryption process for the secure secret provisioning (SSP).

SSP is the equivalent of a Secure Firmware Install for MPUs. Before sending their code to OEMs, developers encrypt their firmware with a private-public key system with STM32CubeProgrammer. The IP is thus unreadable by third parties. During assembly, OEMs use the provided hardware secure module (HSM) containing a protected encryption key to load the firmware that the MPU will decrypt internally. However, until now, developers encrypting the MPU’s code had access to the private key. The problem is that some organizations must limit access to such critical information. Thanks to the new STM32CubeProgrammer and PKCS#11, the private key remains hidden in an HSM, even during the encryption process by the developers.

Supporting new STM32 MCUs

STM32WL3, STM32N6, STM32C0, STM32H7R3/7S3/7R7/7S7

While nearly every version of STM32CubeProgrammer comes with new MCU support, 2.18 was particularly noteworthy for the number of added devices. Users can now work with the STM32WL3 announced just a few weeks ago, the STM32N6 launched a few days ago, the new STM32C0 devices with 64 KB and 256 KB of flash.

STM32CubeProgrammer also brings additional feature support for the STM32H7R3/7S3/7R7/7S7, all STM32 MPUs, and the STM32U5. For instance, the STM32H7R/S MCUs can now perform Secure Firmware Installation, while the STM32MP25 gets a GUI to manage PMIC registers and export settings to a binary file, which makes porting them to another project a breeze. And the STM32U5 can now restore its option byte configuration to factory settings if developers make an error that gets them stuck.

STM32C0, STM32MP25, and STM32WB05/6/7

Since version 2.17, STM32CubeProgrammer supports STM32C0s with 128 KB of flash. It also recognizes the STM32MP25, which includes a 1.35-TOPS NPU, and all the STM32WB0s, including the STM32WB05, STM32WB05xN, STM32WB06, and STM32WB07. In the latter case, we brought support only a few weeks after their launch, thus showing that STM32CubeProgrammer keeps up with the latest releases to ensure developers can flash and debug their code on the newest STM32s as soon as possible.

Access to the STM32MP13’s bare metal

Microcontrollers demand real-time operating systems because of their limited resources, and event-driven paradigms often require a high level of determinism when executing tasks. Conversely, microprocessors have a lot more resources and can manage parallel tasks better, so they use a multitasking operating system, like OpenSTLinux, our Embedded Linux distribution. However, many customers familiar with the STM32 MCU world have been asking for a way to run an RTOS on our MPUs as an alternative. In a nutshell, they want to enjoy the familiar ecosystem of an RTOS and the optimizations that come from running bare metal code while enjoying the resources of a microprocessor.

Consequently, we are releasing today STM32CubeMP13, which comes with the tools to run a real-time operating system on our MPU. We go into more detail about what’s in the package in our STM32MP13 blog post. Additionally, to make this initiative possible, ST updated its STM32Cube utilities, such as STM32CubeProgrammer. For instance, we had to ensure that developers could flash the NOR memory. Similarly, STM32CubeProgrammer enables the use of an RTOS on the STM32MP13 by supporting a one-time programmable (OTP) partition.

Traditionally, MPUs can use a bootloader, like U-Boot, to load the Linux kernel securely and efficiently. It thus serves as the ultimate first step in the boot process, which starts by reading the OTP partition. Hence, as developers move from a multitasking OS to an RTOS, it was essential that STM32CubeProgrammer enable them to program the OTP partition to ensure that they could load their operating system. The new STM32CubeProgrammer version also demonstrates how the ST ecosystem works together to release new features.

STM32WB and STM32WBA support

Since version 2.12, STM32CubeProgrammer has brought numerous improvements to the STM32WB series, which is increasingly popular in machine learning applications, as we saw at electronica 2022. Specifically, the ST software brings new graphical tools and an updated wireless stack to assist developers. For instance, the tool has more explicit guidelines when encountering errors, such as when developers try to update a wireless stack with the anti-rollback activated but forget to load the previous stack. Similarly, new messages will ensure users know if a stack version is incompatible with a firmware update. Finally, STM32CubeProgrammer provides new links to download STM32WB patches and get new tips and tricks so developers don’t have to hunt for them.

Similarly, STM32CubeProgrammer supports the new STM32WBA, the first wireless Cortex-M33. Made official a few months ago, the MCU opens the way for a Bluetooth Low Energy 5.3 and SESIP Level 3 certification. The MCU also has a more powerful RF that can reach up to +10 dBm output power to create a more robust signal.

STM32H5 and STM32U5

The support for STM32H5 began with STM32CubeProgrammer 2.13, which added compatibility with MCUs, including anything from 128 KB up to 2 MB of flash. Initially, the utility brought security features like debug authentication and authentication key provisioning, which are critical when using the new life management system. The utility also supported key and certificate generation, firmware encryption, and signature. Over time, ST added support for the new STM32U535 and STM32U545 with 512 KB and 4 MB of flash. The MCUs benefit from RDP regression with a password to facilitate developments and SFI secure programming.

Additionally, STM32CubeProgrammer includes an interface for read-out protection (RDP) regression with a password for STM32U5xx. Developers can define a password and move from level 2, which turns off all debug features, to level 1, which protects the flash against certain reading or dumping operations, or to level 0, which has no protections. It will thus make prototyping vastly simpler.

STLINK-V3PWR

In many instances, developers use an STLINK probe with STM32CubeProgrammer to flash or debug their device. Hence, we quickly added support for our latest STLINK-PWR probe, the most extensive source measurement unit and programmer/debugger for STM32 devices. If users want to see energy profiles and visualize the current draw, they must use STM32CubeMonitor-Power. However, STM32CubeProgrammer will serve as an interface for all debug features. It can also work with all the probe’s interfaces, such as SPI, UART, I2C, and CAN.

Script mode

The software includes a command-line interface (CLI) to enable the creation of scripts. Since the script manager is part of the application, it doesn’t depend on the operating system or its shell environment. As a result, scripts are highly sharable. Another advantage is that the script manager can maintain connections to the target. Consequently, STM32CubeProgrammer CLI can keep a connection live throughout a session without reconnecting after every command. It can also handle local variables and even supports arithmetic or logic operations on these variables. Developers can thus create powerful macros to automate complex processes. To make STM32CubeProgrammer CLI even more powerful, the script manager also supports loops and conditional statements.

A unifying experience

STM32CubeProgrammer aims to unify the user experience. ST brought all the features of utilities like the ST-LINK Utility, DFUs, and others to STM32CubeProgrammer, which became a one-stop shop for developers working on embedded systems. We also designed it to work on all major operating systems and even embedded OpenJDK8-Liberica to facilitate its installation. Consequently, users do not need to install Java themselves and struggle with compatibility issues before experiencing STM32CubeProgrammer.

Qt 6 support

Since STM32CubeProgrammer 2.16, the ST utility uses Qt 6, the framework’s latest version. Consequently, STM32CubeProgrammer no longer runs on Windows 7 and Ubuntu 18.04. However, Qt 6 patches security vulnerabilities, brings bug fixes, and comes with significant quality-of-life improvements.

• MOOC – Programming STM32 MCUs using STM32CubeProgrammer
• MOOC – STM32Cube Tools in practice
• Download STM32CubeProgrammer