Re: [PATCH] S/390 port modernization 3/4

[Jim Blandy <jimb at redhat dot com>]

Andrew Cagney <cagney@gnu.org> writes:

> Ulrich,
> 
> two FYIs
> > ! 	else
> > ! 	  abort ();
> >         }
> 
> GDB doesn't call abort (well ok, there's one call :-) Instead for
> cases like this GDB should call internal_error.
> 
> > ! 	case RETURN_VALUE_STRUCT_CONVENTION:
> > ! 	  error ("Cannot set function return value.");
> > ! 	  break;
> 
> If this happens, GDB's shot itself in the foot so might as well call
> internal_error here as well.
> 
> Jim, since you're handling this, can I assume that you'll clean up any
> nuences that the ARI identifies (as I did for the last round of s390
> patches).  It will save Ulrich the need of respining what are basicly
> correct patches.

In general, sure.  But in this case it seems like Ulrich is already
respinning things pretty quickly.