Comment # 4 on bug 845245 from Bernhard Voelker

As per discussions like [1] and [2], it is problematic to skip the
chmod(3) syscall because e.g. the ACLs on an NFS file system may
change even if the UNIX file permission bits do not.
Thus said, the only instance knowing if the ctime has to be changed
is the kernel.  Trying to workaround that in chmod(1) seems to be
quirky.

[1] http://lists.gnu.org/archive/html/bug-coreutils/2010-01/msg00303.html
[2] https://lists.gnu.org/archive/html/coreutils/2013-10/msg00030.html