[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20140930074259.GC11709@breakpoint.cc>
Date: Tue, 30 Sep 2014 09:42:59 +0200
From: Florian Westphal <fw@...len.de>
To: Tom Herbert <therbert@...gle.com>
Cc: Florian Westphal <fw@...len.de>,
Linux Netdev List <netdev@...r.kernel.org>,
Francois Romieu <romieu@...zoreil.com>,
Hayes Wang <hayeswang@...ltek.com>
Subject: Re: [PATCH next] r8169: add support for Byte Queue Limits
Tom Herbert <therbert@...gle.com> wrote:
> Watch inflight and limit in the byte_queue_limits for the queue.
> inflight must always go back to zero when link goes idle.
Yes, inflight goes to 0 when link is idle.
Output of
while true; do
for n in inflight limit; do
echo -n $n\ ; cat $n;
done; sleep 1;
done
during netperf run, 100mbit peer:
inflight 0
limit 3028
inflight 6056
limit 4542
[ no changes during test ]
limit 4542
inflight 3028
limit 6122
inflight 0
limit 6122
[ changed cable to 1gbit peer, restart netperf ]
inflight 37850
limit 36336
inflight 33308
limit 31794
inflight 33308
limit 31794
inflight 27252
limit 25738
[ no changes during test ]
inflight 27252
limit 25738
inflight 0
limit 28766
[ change cable to 100mbit peer, restart netperf ]
limit 28766
inflight 27370
limit 28766
inflight 4542
limit 5990
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 6056
limit 4542
inflight 0
[ end of test ]
I think thats what its supposed to look like :-)
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists