From: cross@spitfire.i.gajendra.net   
      
   In article ,   
   John Dallman  wrote:   
   >In article <10c6irh$er0$1@reader2.panix.com>,   
   >cross@spitfire.i.gajendra.net (Dan Cross) wrote:   
   >> In article ,   
   >> John Dallman  wrote:   
   >   
   >> >That is changing, reasonably quickly. ARM stopped releasing new   
   >> >cores that could do A32 or T32 in 2023, having been phasing them   
   >> >out since 2021.   
   >   
   >I should have said "ARM stopped releasing new _A-profile_ cores that   
   >could do A32 or T32 in 2023 ..."   
   >   
   >> I wonder if this suggests that they'll introduce a compressed   
   >> instruction set a la Thumb for 64 bit mode; -M profile seems to   
   >> top out at ARMv8.1; and according to the ARMv8-M ARM, only   
   >> supports T32.   
   >   
   >ARM v8-M does not have 64-bit registers or instructions, or virtual   
   >memory. It has an optional, simple, memory protection system.   
      
   Yes.  As I mentioned, ARMv8.1 in M profile is T32 only,   
   according to the ARM (Arch. Ref. Manual in this context---leave   
   it up to ARM to overload the acronym ARM so many different   
   ways).   
      
   There is no M profile for ARMv9 at present.  My question earlier   
   was, if there were, one wonders whether they would use 64-bit   
   registers and introduce a compressed instruction set.   
      
   >The   
   >additions at ARMv8.1M are not the same as the ones in ARM v8.1A.   
   >   
   >> Presumably at some point they'll introduce an ARMv9 core for   
   >> the embedded market and this will become an issue.   
   >>   
   >> Or maybe they won't.  We could be in a world of 32-bit embedded   
   >> cores in that space for a very long time indeed.   
   >   
   >It depends what you're doing, really. Qualcomm cellphone-derived SoCs   
   >with 64-bit Cortex-A cores are already widely used in robotics and   
   >similar kinds of "embedded" uses. But there's no need at all for 64-bit   
   >in tiny microcontrollers.   
      
   Well, then I suppose they'll either split their product line or   
   introduce a 32-bit M profile for V9.   
      
   I'm do not entirely agree with that assessment re: 64-bit in   
   MCUs, however: a lot of work is going into cryptographically   
   signed secure boot stacks and hardware attestation for firmware;   
   64-bit registers can make implementing cryptography primitives   
   with large key sizes much easier.   
      
   E.g., the PSP on AMD EPYC is a Cortex-A5 right now (presumably   
   because they wanted a real MMU, and not just an MPU), but I   
   could imagine that changing over time.   
      
   	- Dan C.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)