Messages in this thread | | | Subject | Re: RFC: fragmentation code experiment | Date | Sun, 24 Oct 1999 01:31:15 +0100 (BST) | From | Alan Cox <> |
| |
> UDP only currently, right? How does NFS product out-of-order fragments on > local networks on a regular basis? Is this all NFS implementations, or > just linux.
Linux sends UDP fragments in reverse order. So yes - most definitely
> > > out-of-order fragment reassembly would have simpler (and > > > better-performing) network code. > > > > Nope, its not on the fast path > > But the code would still be simpler, no?
Probably very little. There are lots of more productive things you could do to speed it up without breaking the behaviour. Things like special casing reverse order packets(thats the easiest for a receiver btw) as you know the packet size for the final frame immediately so can reassembl in place
> This is the part i could use some help understanding. As i understand it, > fragmentation only happens at routing when two links have different MTUs. > Is this covered by what you mean by 'local'?
It also occurs with UDP sending larger frames than the MTU - hence it happens a lot with UDP. In fact thats the major source of fragmented frames nowdays
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |