https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124358
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jakub at gcc dot gnu.org
--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
The problem is that gimple_fold_indirect_ref uses STRIP_NOPS and then has
/* *(foo *)fooarrptr => (*fooarrptr)[0] */
if (TREE_CODE (TREE_TYPE (subtype)) == ARRAY_TYPE
&& TREE_CODE (TYPE_SIZE (TREE_TYPE (TREE_TYPE (subtype)))) == INTEGER_CST
&& useless_type_conversion_p (type, TREE_TYPE (TREE_TYPE (subtype))))
{
tree type_domain;
tree min_val = size_zero_node;
tree osub = sub;
sub = gimple_fold_indirect_ref (sub);
if (! sub)
sub = build1 (INDIRECT_REF, TREE_TYPE (subtype), osub);
type_domain = TYPE_DOMAIN (TREE_TYPE (sub));
if (type_domain && TYPE_MIN_VALUE (type_domain))
min_val = TYPE_MIN_VALUE (type_domain);
if (TREE_CODE (min_val) == INTEGER_CST)
return build4 (ARRAY_REF, type, sub, min_val, NULL_TREE, NULL_TREE);
}
Without the
#if GENERIC
(simplify
(pointer_plus (convert:s (pointer_plus:s @0 @1)) @3)
(convert:type (pointer_plus @0 (plus @1 @3))))
#endif
rule, this doesn't trigger as ARRAY_TYPE is just the type of the inner
POINTER_PLUS and there is the outer one which has int * type.
So, wonder whether I shouldn't change the rule to
(pointer_plus (convert:type @0) (plus @1 #3))))
instead.
