CFG branch merge 8 - SMALL_REGISTER_CLASSES versus profiling

[Jan Hubicka]

> > ! 	      /* Now, NEXT_INSERT_AFTER may be an instruction that uses the
> > ! 	         return value.  However, it could also be something else,
> > ! 	         like a CODE_LABEL, so check that the code is INSN.  */
> > ! 	      if (next_insert_after
> > ! 		  && GET_CODE (next_insert_after) == INSN
> > ! 		  && reg_referenced_p (return_reg,
> > ! 				       PATTERN (next_insert_after)))
> > ! 		insert_after = next_insert_after;
> 
> I think all this should use keep_with_call_p and not be
Aarghh. sure. Of course :) I already done that when
merging in the Janis patch to cfg-branch and obviusly
forgot the commit..

I probably can prepare updated patch at end of weekend.
Honza
> dependent on SMALL_REGISTER_CLASSES.  As can be seen with
> ia64 when you put stuff before the gp reload.
> 
> 
> r~