Messages in this thread | | | Subject | Re: zero-copy TCP fileserving | Date | Fri, 24 Sep 1999 19:46:48 +0100 | From | Philip Blundell <> |
| |
>Latency? zero-copy TCP is not going to do anything to the latency, >since a NIC that does TCP checksumming on TX will have to do it as >'store and forward' since the checksum is in the header. The world is >about bulk data transfers ;-)
True enough. But if you are transmitting bulk data (ie back to back packets), which is where zero copy is likely to be especially appealing, you can hide the latency by loading (and checksumming) one packet as the previous one is still going out on the wire.
>I am still looking forward to the day when we will have zero-copy TX >support, it should give us even more performance on the Linux boxes >running our tape silos.
Yeah. I have some half-done patches somewhere to add iovec support to skb's for transmit purposes, which I cooked up a while ago with this sort of thing in mind. As it happens one of the projects I am working on at the moment is one where zero copy networking in general (though not TCP specifically) would be a bonus; I'll try to dig them out again.
p.
- 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/
| |