On Thu, Dec 22, 2011 at 09:23:54PM -0500, Mouse wrote: > ...making them appear asynchronously with respect to userland > userspace's calls to userland kernel would, yes, be difficult. (Or > anything else that lets the ptracer regain control; syscalls are just > the most convenient such thing.) > > But is that important? Do we have cases now where new VM appears in a > userspace's VM asynchronously with respect to syscalls it makes? If > so, what can trigger the creation of such a thing?
There are none; I don't see that there's a problem... -- David A. Holland dholl...@netbsd.org