empty function optimizations

[Joern Rennecke]

> Some conforming programs would break, but these are fairly obscure (e.g.
> programs that compare function pointers).  Debugging might get harder.
> But if you factor out common code, you can get greater savings than the
> compiler can find and your gains are portable.

How about especially marking function references that are function
calls (the SH already does this for linker relaxation) ?
We could then factor out common code if all references to a function
are for a function call.