On 07 Sep 2021 10:40, Simon Marchi wrote:
> On 2021-08-19 5:23 p.m., Mike Frysinger via Gdb-patches wrote:
> > The sim side is already ready to go for cris, so wire it up.
> > ---
> >  gdb/configure.tgt | 1 +
> >  1 file changed, 1 insertion(+)
> > 
> > diff --git a/gdb/configure.tgt b/gdb/configure.tgt
> > index 55ae7541b619..c678027faef4 100644
> > --- a/gdb/configure.tgt
> > +++ b/gdb/configure.tgt
> > @@ -743,6 +743,7 @@ arm*-*-*)		gdb_sim=arm ;;
> >  avr*-*-*)		gdb_sim=avr ;;
> >  bfin-*-*)		gdb_sim=bfin ;;
> >  bpf-*-*)		gdb_sim=bpf ;;
> > +cris-*-*|cris32-*-*)	gdb_sim=cris ;;
> >  frv-*-*)		gdb_sim=frv ;;
> >  ft32-*-*)		gdb_sim=ft32 ;;
> >  h8300*-*-*)		gdb_sim=h8300 ;;
> > 
> 
> Any reason not to use
> 
>   cris*-*-*
> 
> similar to other arches?

the two cris tuples here are what's used throughout the tree.  i don't have
an opinion on it in general, but having gdb match the rest of the tree seems
like the best bet.  i get that earlier in this file it does "cris*".  maybe
the cris maintainers want to rectify this ? :)

$ grep 'cris-[*]' configure* */configure*
configure:  cris-*-* | crisv32-*-*)
configure:  cris-*-* | crisv32-*-*)
configure.ac:  cris-*-* | crisv32-*-*)
configure.ac:  cris-*-* | crisv32-*-*)
gas/configure:      cris-*-*aout*)	emulation="crisaout criself" ;;
gas/configure:      cris-*-*)		emulation="criself crisaout" ;;
gas/configure.ac:      cris-*-*aout*)	emulation="crisaout criself" ;;
gas/configure.ac:      cris-*-*)		emulation="criself crisaout" ;;
gas/configure.tgt:  cris-*-linux-* | crisv32-*-linux-*)
gas/configure.tgt:  cris-*-* | crisv32-*-*)		fmt=multi ;;
ld/configure.tgt:cris-*-*aout*)		targ_emul=crisaout
ld/configure.tgt:cris-*-linux-* | crisv32-*-linux-*)
ld/configure.tgt:cris-*-* | crisv32-*-*)	targ_emul=criself
sim/configure:    all|cris-*-* | crisv32-*-*)
sim/configure.ac:    SIM_TARGET([cris-*-* | crisv32-*-*], [cris])

$ grep 'cris[*]' configure* */configure*
gdb/configure.tgt:cris*)
-mike