On 05/21/10 10:42, James Carlson wrote: > Rishi Srivatsavai wrote: >> Thanks David. I will see if there is a RFE filed on this already, >> if not I will file one. The RFE should provide a way to determine >> whether the driver (and device) supported emitting frames >> with a source address different from device MAC address. >> If such a method was available we don't have to limit bridging >> support by media type. > > I don't think that's true. There are at least two issues here: are the > MAC headers (and addresses contained within) and frame formats directly > compatible, and can the driver transmit packets unmolested? > > The potential 802.11 issue is with the latter part of that. But even if > that were solved with some kind of new predicate, the former part would > still be a problem. You wouldn't want to bridge Ethernet onto FDDI. > (Well, if you were a customer you might want to do so, but you'd just be > wrong. ;-}) > > So, I don't think the existing MAC-type test goes away.
Yes agreed; I was incorrect in saying the RFE above will remove the need for the media type test. The RFE will allow an additional check to ensure only drivers that have the capability to send packets without changes are used by bridging (and possibly VNICs). Rishi _______________________________________________ rbridges-dev mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/rbridges-dev
