[BUG,REGRESSION?] 3.11.6+,3.12: GbE iface rate drops to few KB/s

Eric Dumazet eric.dumazet at gmail.com
Mon Nov 18 12:13:28 EST 2013


On Mon, 2013-11-18 at 08:54 -0800, Stephen Hemminger wrote:
> On Mon, 18 Nov 2013 11:44:48 +0100
> Simon Guinot <simon.guinot at sequanux.org> wrote:

> > c9eeec26 tcp: TSQ can use a dynamic limit
> > 
> 
> But without that patch there was a performance regression for high speed
> interfaces whihc was caused by TSQ. 10G performance dropped to 8G

Yes, this made sure we could feed more than 2 TSO packets on TX ring.

But decreasing minimal amount of queuing from 128KB to ~1ms of the
current rate did not please NIC which can have a big delay between
ndo_start_xmit() and actual skb freeing (TX completion)

So
http://git.kernel.org/cgit/linux/kernel/git/davem/net.git/commit/?id=98e09386c0ef4dfd48af7ba60ff908f0d525cdee

restored this minimal amount of buffering, and let the bigger amount for
40Gb NICs ;)






More information about the linux-arm-kernel mailing list