Date: Tue, 5 Dec 1995 12:26:30 -0700 (MST) From: Terry Lambert <terry@lambert.org> To: wollman@lcs.mit.edu (Garrett A. Wollman) Cc: terry@lambert.org, wollman@lcs.mit.edu, hackers@FreeBSD.org Subject: Re: ipx on 802.3 Message-ID: <199512051926.MAA02226@phaeton.artisoft.com> In-Reply-To: <9512051838.AA02904@halloran-eldar.lcs.mit.edu> from "Garrett A. Wollman" at Dec 5, 95 01:38:59 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> > Once in, another cleanup could be done. The hard part on the 802.3 LLC > > is codifying the state table. I've seen companies buy code from Microsoft > > and hack COFF objects from MSVC 2.x into something usable by GCC to get > > an 802.3 before. > > Which is just fine, since we already have 99% of what's needed > anyway. The original question, which you seem to have never grasped, > is: > > How can you tell an Ethernet interface to use 802.3 > encapsulation rather than Ethernet v2? > > Or, to spell it out more explicitly: > > How can you tell an Ethernet interface to send an 802.3-style > length field and 802.3-style LLC header rather than the > two-byte type field specified in Ethernet v2? By linking the encapsulation with the address family of the socket used to do the ioctl(). If you get a LINK1 on an AF_INET socket, it's a physical media select. If you get one on an AF_IPX socket, it's an LLC select. Means you can't do 802.3 with IP until the thing is fixed correctly, but since the 802.3 is being added for IPX, IP can wait. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199512051926.MAA02226>