it's the output, not ack coalescing (Re: TSO and FreeBSD vs Linux)

Adrian Chadd adrian at freebsd.org
Fri Aug 16 17:59:31 UTC 2013


On 16 August 2013 10:06, Luigi Rizzo <rizzo at iet.unipi.it> wrote:

> the usenix paper and talk
>
> https://www.usenix.org/conference/usenixfederatedconferencesweek/netmap-novel-framework-fast-packet-io
> should give a short description of what i did.
>
> Basically i manually insert returns (and free mbufs and resources)
> in some place in the path and then compute the time consumed by
> the code as the inverse of the average pps over 0.1 .. 1s intervals.
> By moving the point where i place the return, i can estimate the
> cost of the various sections of code involved.
>
>
Cool. I think I've seen this stuff.

I may have to spin up netmap soon, if only to look at how much more
efficient (and how!) you're driving the CPU and IO bus.



-adrian


More information about the freebsd-net mailing list