News for hackers. Stuff that matters.
Re:STM32MP2: ST’s first Linux capable 64-bit MPU with NPU, G... by user_7832 on Saturday August 26, 2023 @01:16PM (#37274751)
I’m curious if any hardware dev can chime in on this - would it be possible to use such an chip with an RTOS? I’ve seen rtos’es which run on esp32 style mmu-less chips, and it seems that anything with an a core runs an (non-real-time) OS. But could the A chip in this be used to run an RTOS without a host system, for repeatability and real-time operation?
Re:STM32MP2: ST’s first Linux capable 64-bit MPU with NPU, G... by 5436436347 on Saturday August 26, 2023 @01:41PM (#37275013)
There are a number of RTOS’s out there now, both proprietary (e.g. VxWorks) and OSS (Zephyr) that support A-series chips with an MMU, and can operate with a single (kernel) context. They still provide bounded latencies compared to full-fat OS’s, but don’t have as much determinism as microcontrollers without page tables and caches. The cores themselves (maybe not the A35?) don’t generally work without the MMU enabled, because most ARM cores also disable the cache if the MMU is off.
Re:STM32MP2: ST’s first Linux capable 64-bit MPU with NPU, G... by user_7832 on Sunday September 03, 2023 @04:09AM (#37368638)
Thank your for your reply! I didn’t realise that Zephyr supports A series chips already.
Re:STM32MP2: ST’s first Linux capable 64-bit MPU with NPU, G... by jpcfl on Saturday August 26, 2023 @01:29PM (#37274897)
Yes.