Is it possible to use reassembly on non-split packets?
I need to dissect multiple packets, but they are not necessarily related (ie. non-split packets).
Are you saying that you've got 2 fragments of your upper-layer protocol in a single IP packet/Ethernet frame and you want to be able to reassemble them?
I believe that, yes, this will work. I'm pretty sure I've seen this happen with SCTP. It was bizarre (though apparently legal--IIRC) that SCTP could fragment and put the 2 fragments in a single IP packet.
How this would work I'm not really sure; the SCTP dissector might serve as a useful reference.
By "multiple fragments in a single packet" do you mean that, in a single link-layer packet, you have multiple sub-packets that are not, by themselves, complete packets, and must then be reassembled into a complete packet?
Or do you mean that, in a single link-layer packet, you have multiple complete packets?
Multiple complete packets, which I don't know how to dissect in a single dissector.
So are you saying, for example, that, in your other question, frames 39 through 44 each have multiple MASP packets in them, and that all of each of those packets is contained within one frame?
Asked: 2017-12-14 16:06:57 +0000
Seen: 431 times
Last updated: Dec 14 '17
Can you expand your question, over what protocol are these packets running, are they sequential, fragmented, anything else you can think of to explain the issue.
These packets are running over UDP. I'm still fairly new to wireshark, so I'm still not familiar with some terms, like "sequential" or "fragmented". When I was reading the developer's guide section 9.4, I found that if a packet is too large, it can be split, and the data is spread between multiple packets. To do my dissection, I need to reassemble these split packets. What I want to know, is that is it possible to use the reassembly API to assemble packets that aren't split?
I'm missing something here, if you have the complete packet because it isn't split, why do you need to reassemble?
Note that this is quite an advanced topic for dissectors and there isn't much documentation apart from in the reassembly code itself, and the existing dissectors that use it.
It's for a project, if it isn't possible to apply reassembly to packets that aren't split, then is there another way to dissect multiple packets in a single dissector?
So do you have multiple packets for your protocol inside a single link-layer frame, a single packet for your protocol that is made up from multiple link-layer frames, or both?