On Thu, Oct 8, 2026 at 9:58 PM Samuel Thibault <[email protected]> wrote: > Digging this old thread back, because I happened to think about the issue > for an unrelated microkernel-related security question. > I had kept this in my inbox without having the time to really look at > it.
Hey, it only took 3.5 years! :D I know the feeling, I'm drowning in my inbox too. > Now that I look at it, it looks completely reasonable: we have > auth_user/server_authenticate that allows to safely pass a port from > server to user. Having another one that allows to safely pass a port > from user to server just looks like the logical dual operation, so with > this actual user of it, we should just implement it :) Great! And it sounds like an exciting thing for me to hack on; I'll try to get to it next week then. Thanks, Sergey
