Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

look for IP defragmentation soft ip cores

Status
Not open for further replies.

wannerz

Newbie
Newbie level 3
Joined
May 2, 2022
Messages
4
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
30
Hi, All,
I want to look for a soft ip cores that work for tcp/ip, and they can support ip defgramentation.
 

Do you really want to work on TCP/IP packet structuring at the hardware IP core level?
Normally these are not handled at the DataLink layer. They are done at a much higher level by software.
 

Thanks dpaul, yes, they are mostly handled by high level protocal, but in some cases, the router or other layer 3 device will split an ip packet into two or multiple, so you need to reassemble the packet that to do the filter and later action, in another case, if you split a encrypt ip packet, then you can't decrypt it if you reassemble it.
 

As it is not usually done in hardware, it would be difficult to find such a free IP core version.
 

@dpaul actually, IP fragmentation would likely mostly be done in hardware. Passing it up to software would likely be rather slow. Imagine trying to refragment a 25 or 100G link on a processor - good luck with that.

@wannerz do you really need fragmentation? - I dont know how much it is actually used, and is not even allowed in IPv6.
In FPGAs, you only get a MAC core, so a streaming output would give you the ethernet directly. So I dont think vendors even provide cores to handle much more of the IP stack. I think most assume there is no fragmentation.
 

@TrickyDicky thanks for your reply, the question of whether or not a IP fragmentation is necessary?
If you design a ipsec device, if the path MTU is 1514, and the incoming ip packet is 1514, because we need add another ip header in tunnel mode, so if we don't do the fragmenation, the packet length will be more than MTU, this will encounter an error, os fragmentation is necessary.
Also in other circumstance, if the path MTU changed during transmission, for encrypt ipsec packet, we can't decrypt them if we do the defragmentation.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top