Re: kirkwood mv643xx_eth wrong tcp checksums
Lennert Buytenhek wrote on 23.11.2009 14:42:
> No matter which MTU the user/admin sets, the hardware/driver should
> just work. I.e. packets should always go out with correct checksums.
I agree, that's right. Without this new knowledge it was hard to follow
why tcp communication did not work anymore.
> You can probably figure out which method iperf is using by
> strace'ing it, and seeing which of the above system calls it uses
> to put data in the socket's transmit queue.
It seems as if iperf does not support zero copy, only a few newer svn
trunks seem to support it. Of course there are some better tools to test
tcp performance. I just used it because this was the first tool that
came to my mind.
I will post some new results after having made some tests this evening.
so far, thanks very much for your help.