Hi Andrew
Andrew Gallatin wrote:
> Francesco DiMambro wrote:
>
> > By the way I measured my driver last night against Linux which uses
> TOE,
> > get slow Rx out of the equation, I can get 1.2G from the Niagara system
> > with a single strand pegged. Nothing fancy just netperf -H ..... --
> -S1M
> > -s1M.
>
> Have you compared to LSO? This is a bit of an apples and oranges
> comparison, but I see 2.2Gb/s from a single netperf using my GLDv3
> LSO driver on a similar system (T2000, 1000MHz cpu).
Just my own, and it wasn't that much in it, MDT was ahead, what
was more interesting was it allows UDP to be accelerated as well.
Would you be prepared to share the options you used to get the
result, and the version of netperf? It's good to know there's more
to be had on that platform thanks.
>
> Also, can you point to a definition of the MDT interface?
Yes there is docs in the PSARC case, I had to leave it behind
at Sun so I no longer have the info. Hopefully folks on the alias
will give you the PSARC number, or just give you the docs.
You may want to look at the 'multidata.h' header file for the
definition of the multidata_t, and all the helper functions.
> Are there
> any consumers of it already in the tree so I can see how it works?
No, just in the stack implementation.
> All this talk of it has made me quite curious.
Good to hear, I hope that guys at Sun share with you the info
you need, maybe even the 'Cassini' implementation.
Frank
>
> Thanks,
>
> Drew
_______________________________________________
networking-discuss mailing list
[email protected]
