Re: [quicwg/base-drafts] Client's initial source CID is unauthenticated (#1479)

[Kazuho Oku <[e-mail]>]

@chris-wood 
>> ... client's CID is never allowed to change during the handshake.
> 
> Can you please point me to where this is written?

My understanding is that the following paragraphs effectively requires that.

> During the handshake, an endpoint might receive multiple packets with the long header, and thus be given multiple opportunities to update the Destination Connection ID it sends. A client MUST only change the value it sends in the Destination Connection ID in response to the first packet of each type it receives from the server (Retry or Initial); a server MUST set its value based on the Initial packet. Any additional changes are not permitted; if subsequent packets of those types include a different Source Connection ID, they MUST be discarded. This avoids problems that might arise from stateless processing of multiple Initial packets producing different connection IDs.
> [4.7. Connection ID Encoding](https://quicwg.org/base-drafts/draft-ietf-quic-transport.html#connection-id-encoding)

> If the assumption is that changing the source CID means the handshake continues indefinitely, then putting it in the TP seems to actually enforce that requirement.

To be precise, having client's initial source CID in the TP allows the server to check if the value was consistent during the handshake. Clients naturally check that, by comparing the DCID field of the Initial / Handshake packets that they receive.

That's what is different from having client's initial destination CID in TP; without that, neither the client nor the server can check if the field was tampered by a middlebox.

> In general, I'd like the TP to authenticate as many things in initial packets as possible.

That's a fair point. Though I might argue that we do not necessarily need to include something that is a constant during the handshake. As @martinthomson points out, we do not authenticate IP addresses. To me it seems that the security properties of client's source CID is equivalent to that.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/1479#issuecomment-400534297