Re: [hatari-devel] "next" and "step" commands to Hatari debugger?

[Eero Tamminen <oak@xxxxxxxxxxxxxx>]

Hi,

On sunnuntai 04 marraskuu 2012, Eero Tamminen wrote:
> > > "next" (continue to next instruction, used for passing subroutines)
> > > would be done with current conditional breakpoints, I just need
> > > to add an option to them to silence their output on creation,
> > > hitting and removal of a temporary internal breakpoint...
> > 
> > That could be useful, too!
[...]
> the output from creating, hitting and removal of these breakpoints
> is hugely verbose.

I've added now also the "next" command for CPU & DSP.


It required adding GetNextPC() wrappers  which don't do any output,
to both CPU and DSP disassembly code.  I'm not completely
sure about whether the DSP version is correct though.

Laurent, could you check that this is OK:
  http://hg.tuxfamily.org/mercurialroot/hatari/hatari/rev/97d057207ac2
?

I took code from dsp56k_execute_one_disasm_instruction(),
but there 1 is deducted from the instruction length.  In
normal DSP disassembly function that works fine, but with
the new function, the deduction would give just the current
PC address (when real instruction length was 1).

So, I'm a bit mystified how the original 
dsp56k_execute_one_disasm_instruction() actually can
work OK...


I also made m68k dissembly output flags & column values
configurable in the debugger.  But I wonder should I
expose these settings also in the Hatari configuration
as valid values might change between Hatari versions...

Any comments?


Next I'll update the docs for these debugger updates.


	- Eero