Re: final cfg build problems

[Richard Henderson <rth at cygnus dot com>]

On Mon, Nov 01, 1999 at 06:35:32AM -0700, Jeffrey A Law wrote:
> One thought for dealing with these various problems is to not optimize the
> cfg, delete dead code, etc etc when we rebuild the cfg.  Just build the
> cfg, notes and quit.

Yes, that's probably the best solution.  I'll check in the
following patch to that effect.


r~



	* toplev.c (rest_of_compilation): Don't optimize the CFG
	when rebuilding, just before dbr.

Index: toplev.c
===================================================================
RCS file: /cvs/gcc/egcs/gcc/toplev.c,v
retrieving revision 1.247
diff -c -p -d -r1.247 toplev.c
*** toplev.c	1999/10/29 21:45:04	1.247
--- toplev.c	1999/11/01 21:01:37
*************** rest_of_compilation (decl)
*** 4323,4331 ****
  	(dbr_sched_time,
  	 {
             /* ??? Keep the CFG up to date after cross-jumping.  */
! 	   find_basic_blocks (insns, max_reg_num (), rtl_dump_file, 1);
  	   count_or_remove_death_notes (NULL, 1);
! 	   life_analysis (insns, max_reg_num (), rtl_dump_file, 1);
  
             dbr_schedule (insns, rtl_dump_file);
  	 });
--- 4323,4331 ----
  	(dbr_sched_time,
  	 {
             /* ??? Keep the CFG up to date after cross-jumping.  */
! 	   find_basic_blocks (insns, max_reg_num (), rtl_dump_file, 0);
  	   count_or_remove_death_notes (NULL, 1);
! 	   life_analysis (insns, max_reg_num (), rtl_dump_file, 0);
  
             dbr_schedule (insns, rtl_dump_file);
  	 });