[Bug c++/111806] g++ generates better code for variant at -Os compared to -O3

[hiraditya at msn dot com via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111806

--- Comment #1 from AK <hiraditya at msn dot com> ---
It seems like we could 'sink' the 4 common instructions (of .L2) at -O3

L2:
        add     rsp, 48
        xor     eax, eax
        pop     rbx
        ret


Or is it due to some kind of tail duplication?