Re: [patch] cse.c: gcc_assertify an "if" statement whose conditionis always false.

[Mark Mitchell <mark at codesourcery dot com>]

Jakub Jelinek wrote:
> On Wed, Jan 26, 2005 at 03:21:30PM -0500, Kazu Hirata wrote:
>
> > 2005-01-26  Kazu Hirata  <kazu@cs.umass.edu>
> >
> > 	* cse.c (fold_rtx): gcc_assertify an "if" statement whose
> > 	condition is always false.
> >
> > Index: cse.c
> > ===================================================================
> > RCS file: /cvs/gcc/gcc/gcc/cse.c,v
> > retrieving revision 1.332
> > diff -c -d -p -r1.332 cse.c
> > *** cse.c	26 Jan 2005 12:44:58 -0000	1.332
> > --- cse.c	26 Jan 2005 12:49:24 -0000
> > *************** fold_rtx (rtx x, rtx insn)
> > *** 3239,3246 ****
> >       extends.  */
> >    enum machine_mode mode_arg0;
> >  
> > !   if (x == 0)
> > !     return x;
> >  
> >    mode = GET_MODE (x);
> >    code = GET_CODE (x);
> > --- 3239,3247 ----
> >       extends.  */
> >    enum machine_mode mode_arg0;
> >  
> > ! #ifdef ENABLE_CHECKING
> > !   gcc_assert (x != NULL_RTX);
> > ! #endif
> >  
> >    mode = GET_MODE (x);
> >    code = GET_CODE (x);
>
>
> If x is never NULL, I think deleting it is equally good.
> If x is NULL, you'll segfault on GET_MODE (x), which is as good as
> assertion failure.

I dunno, I like asserts -- they document the assumption.  But, there's 
no point in guarding ggc_assert with ENABLE_CHECKING, as ggc_assert 
already is conditionalized on that macro.  OK with that change.

--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com
(916) 791-8304