[patches] Re: more robust purge_dead_edges

Jan Hubicka jh@suse.cz
Tue Jul 24 15:29:00 GMT 2001


> 	if (computed_jump_p (insn))
> 	  return;
> or
> 	if (! any_condjump_p (insn)
> 	    && ! simplejump_p (insn)
> 	    && ! returnjump_p (insn))
I would vote for this one.
We don't want to see tablejumps too. I am not sure if they qualify
as computed jumps.

Honza
> 	  return;
> 
> 
> r~



More information about the Gcc-patches mailing list