On Thu, 26 Feb 2015, Andy Lutomirski wrote:
I'm still extremely nervous about allowing nonempty pA to propagate to
setuid or nonzero fP programs. It's less obviously dangerous if pA is
never a superset of pP, but it could still cause problems with setuid
programs that execute intentionally deprivileged helpers.
Well but the intend of the ambient caps is that all processes spawned have
those caps. So they should not be dropped implicitly by othe mechanisms
because they could spawn scripts etc that need these privs again.