On 18 Jan 2023 22:53, Mark Wielaard wrote:
> On Sun, Jan 15, 2023 at 04:47:44PM -0500, Mike Frysinger wrote:
> > On 15 Jan 2023 21:20, Mark Wielaard wrote:
> > > --- a/sim/mn10300/sim-main.h
> > > +++ b/sim/mn10300/sim-main.h
> > > @@ -38,7 +38,14 @@ mn10300_core_signal ((SD), (CPU), (CIA), (MAP), (NR_BYTES), (ADDR), (TRANSFER),
> > >  #include "sim-fpu.h"
> > >  #include "sim-signal.h"
> > >  
> > > +/* These datastructures and defines are only used by the igen generated
> > > +   support.c, semantics.c, idecode.c and engine.c files.  */
> > > +#if defined(SUPPORT_C) \
> > > +    || defined(SEMANTICS_C) \
> > > +    || defined(IDECODE_C) \
> > > +    || defined(ENGINE_C)
> > >  #include "mn10300-sim.h"
> > > +#endif
> > 
> > a nifty hack.  can you move the sim-fpu.h & sim-signal.h inside this
> > block ?
> 
> Sure. That does mean sim/mn10300/{dv-mn103cpu.c,interp.c,op_utils.c}
> need to also explicitly include one or both of these files. But I
> guess that these explicit includes are the intention.

yes, that's the direction we want to go

> See attached. OK to push like that?

lgtm, thanks
-mike