g77.f-torture/execute/980520-1.f

Steven Bosscher s.bosscher@student.tudelft.nl
Thu Jul 17 07:54:00 GMT 2003


Op do 17-07-2003, om 03:10 schreef Jan Hubicka:
> > >>>>> Geoff Keating writes:
> > 
> > Geoff> I couldn't find any patch from Toon, but I did find this patch from
> > Geoff> David Edelsohn, which Toon liked, and I hope that's what you are
> > Geoff> referring to:
> > 
> > Geoff> This patch is OK, please commit it David!
> > 
> > 	Zack didn't like the patch and suggested something else that
> > Steven Bosscher offered to implement.  Honza also was going to work on
> > teaching the Fortran front-end to create the equivalent of "static
> > inline".  I think those other approaches are better.  I would like to know
> Creating equivalent of static inlining does not help as without inlining
> the function is output anyway.  Apparently teaching frontend to deffer
> the function is very dificult due to funny design of whole thing, so I
> think your patch probably best sollution to the problem.

No, the solution is that flag_inline_trees must die.  It's not even a
flag, toplev shouldn't know about it because it has nothing to do with
trees, etc.

I'll work on this RSN.

Gr.
Steven



More information about the Gcc-patches mailing list