Re: [PATCH] PR/44970, fix broken fwprop incremental dataflow

[Jack Howarth <howarth at bromo dot med dot uc dot edu>]

On Tue, Oct 19, 2010 at 04:08:09PM +0200, Paolo Bonzini wrote:
> On 10/19/2010 03:28 PM, Jack Howarth wrote:
> > 
> > On x86_64-apple-darwin10, this causes the regression of...
> > 
> > FAIL: gcc.c-torture/execute/arith-rand-ll.c compilation,  -Os  (internal compiler error)
> > UNRESOLVED: gcc.c-torture/execute/arith-rand-ll.c execution,  -Os
> > 
> > at -m32.
> 
> Thanks, consider this squashed in:

Paolo,
   I can confirm that this fixes the issue on x86_64-apple-darwin10.
           Jack

> 
> Index: gcc/df-scan.c
> ===================================================================
> --- gcc/df-scan.c (revision 163855)
> +++ gcc/df-scan.c (working copy)
> @@ -2841,7 +2864,8 @@ df_ref_record (enum df_ref_class cl,
>        /*  If this is a multiword hardreg, we create some extra
>           datastructures that will enable us to easily build REG_DEAD
>           and REG_UNUSED notes.  */
> -      if ((endregno != regno + 1) && insn_info)
> +      if (collection_rec
> +         && (endregno != regno + 1) && insn_info)
>         {
>           /* Sets to a subreg of a multiword register are partial.
>              Sets to a non-subreg of a multiword register are not.  */
> 
> Paolo