https://gcc.gnu.org/bugzilla/show_bug.cgi?id=119482
--- Comment #23 from Andrew Pinski <pinskia at gcc dot gnu.org> --- >It uses flatten. With flatten gcc compilation is a lot slower (40+s) vs clang >(6s). So it turns out that is because flatten in GCC and LLVM do 2 different things. GCC flattens all levels of the call while LLVM just does the first level of calls. I found this out on accident: https://discourse.llvm.org/t/rfc-function-level-flatten-depth-attribute-for-depth-limited-inlining/89017 So maybe this is works as expected ....
