https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126964

--- Comment #5 from Jerry DeLisle <jvdelisle at gcc dot gnu.org> ---
(In reply to Josef Melcr from comment #4)
> (In reply to Jerry DeLisle from comment #3)
> > Patch submitted: The attached patch is obsolete
> > 
> > https://gcc.gnu.org/pipermail/fortran/2026-August/064524.html
> 
> I wanted to test your patch to see if it fixes the slowdown, but it won't
> apply cleanly for me. The diff has lines which don't exist in the repo, such
> as
> 
> trans.h:
> @@ -5559,7 +5559,7 @@ gfc_conv_subref_array_arg (gfc_se *se, gfc_expr *
> expr, int g77,
>                          sym_intent intent, bool formal_ptr,
>                          const gfc_symbol *fsym, const char *proc_name,
>                          gfc_symbol *sym, bool check_contiguous,
> -                        bool deep_copy)                               <--
> +                        bool deep_copy, bool span_only)
> 
>  #define GFC_DECL_SPAN(node) (DECL_LANG_SPECIFIC(node)->span)            <--
> +#define GFC_DECL_SPAN_NORMALIZED(node) \
> +  (DECL_LANG_SPECIFIC(node)->span_normalized)
>  
> trans-array.cc:
> +      if (VAR_P (GFC_TYPE_ARRAY_OFFSET (type)))
> +     scale (GFC_TYPE_ARRAY_OFFSET (type));
> +    }
> +
>    /* Load the span once here, like the bounds above, so that element    <--
>       addressing does not reload it from the descriptor.  The descriptor <--
>       itself is not available in an outlined region, such as an OpenMP   <--
> 
> Could you rebase it please?

I see that, too many items in the worktree and got crossed up. After I fix I
will attach here and repost to the list.

Reply via email to