Re: [tuples][patch] Convert pass_copy_prop to tuples.

[Diego Novillo <dnovillo at google dot com>]

On 3/25/08 1:08 AM, Bill Maddox wrote:

> @@ -880,31 +869,37 @@ init_copy_prop (void)
>  	     value was loop invariant, it will be hoisted by LICM and
>  	     exposed for copy propagation.  */
>  	  if (stmt_ends_bb_p (stmt))
> -	    DONT_SIMULATE_AGAIN (stmt) = false;
> +            prop_set_simulate_again (stmt, true);
>  	  else if (stmt_may_generate_copy (stmt)
> -		   && loop_depth_of_name (GIMPLE_STMT_OPERAND (stmt, 1)) <= depth)
> -	    DONT_SIMULATE_AGAIN (stmt) = false;
> +                   /* Since we are iterating over the statements in
> +                      BB, not the phi nodes, STMT will always be an
> +                      ssignment.  */

s/ssignment/assignment/

OK with that change.  Any new tests fixed?  ISTR some copy-prop specific 
tests.


Diego.