================
@@ -34,21 +34,22 @@ void test_invalid_call_1(int s) {
int some_func2(int a, int b);
void test_invalid_call_2() {
- // CHECK: -RecoveryExpr {{.*}} 'int' contains-errors
+ // CHECK: -RecoveryExpr {{.*}} '<dependent type>' contains-errors
----------------
mizvekov wrote:
If calling ActOnCallExpr produces a better type, we can still call it inside a
SFINAETrap, in order to avoid repeated diagnostics, right?
I think we still assert that we can't have a dependent type when we know we can
evaluate a constraint, that happens before going into the evaluator.
https://github.com/llvm/llvm-project/pull/136295
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
