[clang] [llvm] [Clang] Make exact dynamic_cast optimization AS aware (PR #213927)

[Jan Patrick Lehr via cfe-commits]

================
@@ -0,0 +1,27 @@
+// RUN: %clang_cc1 %s -triple amdgpu-amd-amdhsa -emit-llvm -std=c++11 \
+// RUN:   -O1 -disable-llvm-passes -fvisibility=hidden -o - | FileCheck %s \
+// RUN:   --implicit-check-not='call {{.*}} @__dynamic_cast'
+
+struct __attribute__((type_visibility("default"))) A {
----------------
jplehr wrote:

Good point. Let me see that I can strip this from any special knobs as much as possible

https://github.com/llvm/llvm-project/pull/213927