(usagi-users 03710) Re: shim6 implementation

On Wed, 13 Sep 2006, Sebastien Barre wrote:

> Well, you are now several people to suggest me a (partial) move to 
> userland, as well as to have a look at HIP implementation (which I 
> didn't learn about before). For HIP impl, I'll get informed, and 
> regarding the move to userland, I think this requires a careful 
> architecture re-design. So i will think about this re-design, and 
> propose it here before to dive into the source code.

We were suggested earlier to move to userspace by kernel maintainers:

http://www.linux.sgi.com/archives/netdev/2005-03/msg00899.html

And so we did. In retrospect, moving from kernel to userland 
helped us to reduce the development and maintenance time a lot.

However, I am not sure if the same constraints apply for shim6 
implementation as it there is e.g. no asymmetric crypto involved in shim6. 
There are also other relatively big protocols in the kernel, such as SCTP 
and DCCP. Basically, the thing is that linux kernel is still monolithic 
despite of some efforts to move the weight towards userspace 
(nfs-user-server, udev).

-- 
Miika Komu                                       http://www.iki.fi/miika/