[Bug target/63890] [4.9/5/6 regression] Compiling trivial program with -O -p leads to misaligned stack

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63890

--- Comment #22 from Richard Henderson <rth at gcc dot gnu.org> ---
I believe that Honza is over-thinking this in #c12.
Mike reports that the patch in #c9 works, which makes
sense to me based on where darwin emits its profiler call.

The patch could be improved to note the darwin-specific
nature of the patch.  E.g.

  (TARGET_MACHO && crtl->profile)

but otherwise I think it's fine.