Hi!
On the following testcase, match.pd during GENERIC folding
changes the -1U / x < y into __imag__ .MUL_OVERFLOW (x, y),
but unfortunately unlike for normal calls nothing sets TREE_SIDE_EFFECTS on
the call. There is the process_call_operands function that non-internal
call creation calls and it is usable for internal calls too,
e.g. TREE_SIDE_EFFECTS is derived from checking whether the
call has side-effects (non-ECF_{CONST,PURE}; we have those for internal
calls) and from whether any of the arguments has TREE_SIDE_EFFECTS.
Fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux, ok for
trunk?
2020-04-28 Jakub Jelinek <ja...@redhat.com>
PR tree-optimization/94809
* tree.c (build_call_expr_internal_loc_array): Call
process_call_operands.
* gcc.c-torture/execute/pr94809.c: New test.
--- gcc/tree.c.jj 2020-04-27 14:31:09.260018763 +0200
+++ gcc/tree.c 2020-04-28 00:22:07.537891047 +0200
@@ -11523,6 +11523,7 @@ build_call_expr_internal_loc_array (loca
CALL_EXPR_ARG (t, i) = args[i];
SET_EXPR_LOCATION (t, loc);
CALL_EXPR_IFN (t) = ifn;
+ process_call_operands (t);
return t;
}
--- gcc/testsuite/gcc.c-torture/execute/pr94809.c.jj 2020-04-28
00:26:23.647107159 +0200
+++ gcc/testsuite/gcc.c-torture/execute/pr94809.c 2020-04-28
00:25:42.289718195 +0200
@@ -0,0 +1,12 @@
+/* PR tree-optimization/94809 */
+
+int
+main ()
+{
+ int a = 0;
+ unsigned long long one = 1;
+ ((-1ULL / one) < a++, one);
+ if (a != 1)
+ __builtin_abort ();
+ return 0;
+}
Jakub
