> Receivers can tell the process ID of the sender, so they can do security checks. Does this run into security corner cases around pid reuse? (Or race conditio

> Receivers can tell the process ID of the sender, so they can do security checks. Does this run into security corner cases around pid reuse? (Or race conditions like a process dropping it's privileges after sending?) I remember the kdbus authors talking about making a lot of security metadata attach to the message itself, rather than indirecting through a pid, maybe for these reasons?