On Fri, 12 Jul 2019, Casey Schaufler wrote: > I'm not disagreeing with that. What I'm saying is that the capability > check interface is not the right place to pass that information. The > capability check has no use for the object information. I would much > rather see a security_pass_object_data() hook that gets called after > (or before) the security_capable() hook in the places where you want > the extra information. Extending existing security models is a core feature of the LSM framework. The Linux capability code has no use for object metadata by design, but extending that model to MAC (and other models) via LSM hooks is well within scope and of course already happening e.g. mediating Linux capabilities wrt SELinux subject types. Adding object metadata extends the function of the capability hook along these lines, so that more effective MAC policies may be implemented by LSMs. -- James Morris <jmorris@xxxxxxxxx>
