https://gcc.gnu.org/bugzilla/show_bug.cgi?id=115172
--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
That said,
--- gcc/ubsan.cc.jj 2024-03-22 09:23:37.695296775 +0100
+++ gcc/ubsan.cc 2024-05-21 12:10:24.261454107 +0200
@@ -1776,13 +1776,17 @@ instrument_bool_enum_load (gimple_stmt_i
|| TREE_CODE (gimple_assign_lhs (stmt)) != SSA_NAME)
return;
+ addr_space_t as = TYPE_ADDR_SPACE (TREE_TYPE (rhs));
+ if (as != TYPE_ADDR_SPACE (utype))
+ utype = build_qualified_type (utype, TYPE_QUALS (utype)
+ | ENCODE_QUAL_ADDR_SPACE (as));
bool ends_bb = stmt_ends_bb_p (stmt);
location_t loc = gimple_location (stmt);
tree lhs = gimple_assign_lhs (stmt);
tree ptype = build_pointer_type (TREE_TYPE (rhs));
tree atype = reference_alias_ptr_type (rhs);
gimple *g = gimple_build_assign (make_ssa_name (ptype),
- build_fold_addr_expr (rhs));
+ build_fold_addr_expr (rhs));
gimple_set_location (g, loc);
gsi_insert_before (gsi, g, GSI_SAME_STMT);
tree mem = build2 (MEM_REF, utype, gimple_assign_lhs (g),
seems to fix that. But wonder how many of the other ~180 spots that create
MEM_REFs don't need similar fixes.
