Hi Xiang, thanks for your reply and for your work on armv8-m.
Bye, Michael Am Do., 10. Dez. 2020 um 15:40 Uhr schrieb Xiang Xiao < xiaoxiang781...@gmail.com>: > On Wed, Dec 9, 2020 at 10:48 PM Michael Jung <mij...@gmx.net> wrote: > > > Hello, > > > > I recently bought an STM32L52E-DK development kit and I am playing around > > with ST's port of ARM's TrustedFirmware-M (see ST's UM2671). I would > like > > to basically leave this untouched, but replace ST's 'non-secure' > > demo-application with a NuttX image. It would be nice if you could give > me > > some advice before I start coding: > > > > 1.) The STM32L5 is an armv8-m architecture chip, which since recently > > (thanks!) seems to be supported by NuttX. > > 1.1.) Is armv8-m support already mature? Or do you expect considerable > > effort in getting this working? > > > > We use armv8-m on two different chipset, the result is very good so far. > > > > 1.2.) With TrustedFirmware-M (TFM) the complete 'secure' domain is > owned > > by TFM. NuttX will only manage the 'non-secure' domain. Is there anything > > inherent in NuttX's armv8-m support that requires access to the 'secure > > domain'? > > > > No, the arch code only touches the core registers in the unsecure domain. > > > > 2.) To my understanding there currently is no support for STM32L5 devices > > in NuttX. > > 2.1.) Is anybody out there already working on STM32L5? > > 2.2.) Given that STM32L5 is armv8-m based all hardware macros have been > > extended for 'secure domain' support. For example many registers are > > present twice for access control from 'secure' and 'non-secure' > contexts. I > > think this would warrant (yet another) 'arch/arm/src/stm32l5' directory. > Do > > you agree? > > > > Thanks! > > Michael > > >
