Quoting James Morris (jmorris@private): > On Tue, 24 Aug 2004, Serge E. Hallyn wrote: > > > Is this the generally accepted view? My hope was that the fact > > that these will not affect networking performance would make them > > more acceptable. > > We had hooks at similar levels initially and they were rejected for poking > too far into the stack (e.g. at the protocol or net device level). I did feel a little dirty with the inet_* hooks, though even that seemed generic enough to be fair. Maybe I should see just how much I can do with just the netdev hooks.
This archive was generated by hypermail 2.1.3 : Wed Aug 25 2004 - 11:21:36 PDT