Re: [tree-ssa] Fix function body estimation

[Jan Hubicka <hubicka at ucw dot cz>]

> Op za 18-10-2003, om 22:43 schreef Jan Hubicka:
> > > On Sat, Oct 18, 2003 at 10:26:55PM +0200, Jan Hubicka wrote:
> > > > I see, why do we genericize C trees by gimplifying it?  Should not we
> > > > keep C nodes as they are and gimplify them when we are asked to
> > > > gimplify?
> > > 
> > > I assume you're talking about the call to c_gimplify in finish_function.
> > > The tree inliner only works on GENERIC.  It doesn't work on C trees.
> > > So we need conversion to GENERIC before inlining can happen.
> 
> Thank God, no!  The C trees are in c-common.def.

Good point :)
OK, so all I need is to walk through tree.def and write code size
estimated for that and it will work for all frontends now, right?

Honza
> 
> The way I thought things were until recently is:
> - Front ends put language-specific trees in lang-tree.def.
> - Anything in tree.def is GENERIC
> - Anything that is GENERIC and conforms to the GIMPLE grammar is GIMPLE.
> 
> Correct me if I'm wrong :)
> 
> It would be really nice be clarified and documented.  I posted the
> beginnings of GIMPLE documentation months ago, but at the time its
> grammar was still changing too often.  Perhaps now is a better time...
> 
> Gr.
> Steven