[patch] Workaround ada (gnat-4.3.0) on x86_64

[Joel Brobecker]

> FYI: not sure if it should be applied upstream as it is just a workaround of
> a GCC bug.

I don't mind installing guards against bogus debugging info, but
I'm not sure in this case that we're doing it in the right place.
But the problem is that I don't get the entire picture, so it's hard
for me to judge.

> procedure AdaFF is
>   type rec is
>     record
>       null;
>     end record;
>   X : rec;
> begin
>   null;
> end AdaFF;

Could you send me the executable so that I can inspect it?

Thanks,
-- 
Joel