[Bug middle-end/91146] [9/10 Regression] -Werror=array-bounds if compile with -fsanitize=address

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91146

--- Comment #9 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Seems even with that patch
3677              if (stmt_may_clobber_ref_p_1 (def_stmt, ref, tbaa_p))
in tree-ssa-sccvn.c thinks the
  .ASAN_MARK (UNPOISON, &D.48886, 4);
call may clobber
MEM[(struct small_vector_template_common *)&v].D.48455.BeginX
With the following completely untested patch we optimize like withouf
-fsanitize=address and don't warn.
--- gcc/internal-fn.def.jj      2020-01-18 13:47:09.517357706 +0100
+++ gcc/internal-fn.def 2020-03-17 16:32:41.210800195 +0100
@@ -309,7 +309,7 @@ DEF_INTERNAL_FN (UBSAN_OBJECT_SIZE, ECF_
 DEF_INTERNAL_FN (ABNORMAL_DISPATCHER, ECF_NORETURN, NULL)
 DEF_INTERNAL_FN (BUILTIN_EXPECT, ECF_CONST | ECF_LEAF | ECF_NOTHROW, NULL)
 DEF_INTERNAL_FN (ASAN_CHECK, ECF_TM_PURE | ECF_LEAF | ECF_NOTHROW, "..R..")
-DEF_INTERNAL_FN (ASAN_MARK, ECF_LEAF | ECF_NOTHROW, NULL)
+DEF_INTERNAL_FN (ASAN_MARK, ECF_LEAF | ECF_NOTHROW, "..W.")
 DEF_INTERNAL_FN (ASAN_POISON, ECF_LEAF | ECF_NOTHROW | ECF_NOVOPS, NULL)
 DEF_INTERNAL_FN (ASAN_POISON_USE, ECF_LEAF | ECF_NOTHROW | ECF_NOVOPS, NULL)
 DEF_INTERNAL_FN (ADD_OVERFLOW, ECF_CONST | ECF_LEAF | ECF_NOTHROW, NULL)
--- gcc/tree-ssa-alias.c.jj     2020-03-03 11:04:46.367821907 +0100
+++ gcc/tree-ssa-alias.c        2020-03-17 17:40:22.295385869 +0100
@@ -3032,6 +3032,14 @@ call_may_clobber_ref_p_1 (gcall *call, a
        default:
          /* Fallthru to general call handling.  */;
       }
+  else if (gimple_call_internal_p (call)
+          && gimple_call_internal_fn (call) == IFN_ASAN_MARK)
+    {
+      ao_ref dref;
+      tree size = gimple_call_arg (call, 2);
+      ao_ref_init_from_ptr_and_size (&dref, gimple_call_arg (call, 1), size);
+      return refs_may_alias_p_1 (&dref, ref, false);
+    }

   /* Check if base is a global static variable that is not written
      by the function.  */