On Friday 17 July 2009, Spencer Oliver wrote:
> 
> > I guess I'll have an update coming along in a while ...
> 
> I have been thinking about the changes to the cortex reg's.
> And do not see any benefit, we were already getting the info from reg20 but
> just putting it in 4 virtual regs.

The original code is problematic in at least three ways:

 - Treats CM3-specific DCRDR selectors as architectural.
   (Patch at least *commented* this flaw, which won't
   really hurt till the next V7-M part comes along.  The
   Cortex-M0 and M1 seem be V6-M.)

 - Hard-wired array sizes and indices.

 - Uses undocumented chip-internal hacks.  (Which you
   say were once documented, then retracted by ARM for
   R1 parts, reasons unknown.)

If what you say about the R1 docs is correct, then a
partial reversion of this patch will be appropriate.

I plan to look at this over the weekend.

- Dave

_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to