On Friday 26 February 2010, Spencer Oliver wrote:
> I have made a few tweaks that enable the armv7m to use the generic
> arm cmd group.

Mostly looks OK.  I notice you didn't do the "arm reg" bit;
understandable, that'd get kind of messy.  (Though I'd have
left a FIXME there.)

Another nit -- instead of referencing explicitly "ARMv7-M" I'd
have the comments say "Microcontroller Profile". since the same
issues apply to ARMv6-M (Cortex-M0 and Cortex-M1).

I see ARM11 and Cortex-A semihosting will fail politely; good!

One issue I noticed there:  The jim_mcrmrc() routine shouldn't
reject the microcontroller profile ... v7m does include basic
coprocessor support!  (But Cortex-M3 doesn't.)  Better instead
to just check whether that core's ops vector includes mcr() or
mrc() methods (whichever is needed).

That way when some V7-M processor starts using coprocessor
instructions -- I can imagine Cortex-M4 doing its DSP stuff
that way -- we won't need to change as much.

(Random note ... testing/examples/cortex/cm3-ftest.cfg uses a
coprocessor opcode in l_nocp_err to trigger a particular fault
and thus the relevant exception vector catch logic.)




> semihosting and disassemble cmds now use the generic arm cmds.
> 
> rather than posting lots of patches, here is my repo address.
> latest 4 are of interest.
> 
> http://repo.or.cz/w/openocd/ntfreak.git/shortlog/refs/heads/wip
> 


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

Reply via email to