One more small optimization:  Where possible, for loops that do a search and then try to break out of the loop early, eliminate the attempt to break out of the loop after the last search.  And with that, I'm declaring __dynamic_cast done.  Though if anyone sees any problems, has suggestions for improvements, or wants to contribute some test cases, that is certainly welcome feedback.

git-svn-id: https://llvm.org/svn/llvm-project/libcxxabi/trunk@148246 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/test/dynamic_cast_stress.cpp b/test/dynamic_cast_stress.cpp
index a6fce73..2106ab7 100644
--- a/test/dynamic_cast_stress.cpp
+++ b/test/dynamic_cast_stress.cpp
@@ -73,9 +73,9 @@
 Timing results I'm seeing (median of 3 microseconds):
 
                           libc++abi    gcc's dynamic_cast
-B<Width/2, Depth> -O3      50.694         93.190           libc++abi 84% faster
-B<Width/2, Depth> -Os      55.235         94.103           libc++abi 70% faster
-A<Width, Depth>   -O3      14.895         33.134           libc++abi 122% faster
-A<Width, Depth>   -Os      16.515         31.553           libc++abi 91% faster
+B<Width/2, Depth> -O3      48.334         93.190           libc++abi 93% faster
+B<Width/2, Depth> -Os      58.535         94.103           libc++abi 61% faster
+A<Width, Depth>   -O3      11.515         33.134           libc++abi 188% faster
+A<Width, Depth>   -Os      12.631         31.553           libc++abi 150% faster
 
 */