[Bug middle-end/30314] optimize multiply-by-constant overflow (wrap) test

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=30314

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
More like
--- gcc/match.pd.jj     2022-05-27 11:29:49.403465453 +0200
+++ gcc/match.pd        2022-05-31 17:41:04.381972072 +0200
@@ -5969,6 +5969,17 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
        && (!TYPE_UNSIGNED (TREE_TYPE (@2)) || TYPE_UNSIGNED (TREE_TYPE (@0))))
    (ovf @1 @0))))

+/* Optimize __builtin_mul_overflow_p (x, cst, (utype) 0) if all 3 types
+   are unsigned to x > (umax / cst).  */
+(simplify
+ (imagpart (IFN_MUL_OVERFLOW:cs@2 @0 integer_nonzerop@1))
+  (if (INTEGRAL_TYPE_P (TREE_TYPE (@0))
+       && TYPE_UNSIGNED (TREE_TYPE (@0))
+       && TYPE_MAX_VALUE (TREE_TYPE (@0))
+       && types_match (TREE_TYPE (@0), TREE_TYPE (TREE_TYPE (@2)))
+       && int_fits_type_p (@1, TREE_TYPE (@0)))
+   (convert (gt @0 (trunc_div! { TYPE_MAX_VALUE (TREE_TYPE (@0)); } @1)))))
+
 /* Simplification of math builtins.  These rules must all be optimizations
    as well as IL simplifications.  If there is a possibility that the new
    form could be a pessimization, the rule should go in the canonicalization