Fix lto-symtab ICE during Ada LTO bootstrap

[Arnaud Charlet]

> > Will it also fix
> > 
> > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61954
> 
> Yes, probably, as well as the m68k issue, if that's really doable.

Right, that's also why I think it's a more promising approach.
Pretending that system.address is just an unsigned integer is bound to cause
troubles, although handling system.address as a void* in gigi/gcc
is also bound to cause some other troubles at this stage, but I suspect
we'll have to bite the bullet at some point.

Arno