From mboxrd@z Thu Jan 1 00:00:00 1970 From: P@draigBrady.com Subject: Re: e1000 w/ NAPI + SMP = 99% CPU utilization Date: Wed, 09 Jun 2004 10:01:02 +0100 Sender: netdev-bounce@oss.sgi.com Message-ID: <40C6D1CE.9050202@draigBrady.com> References: <16582.49526.374109.580312@robur.slu.se> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: quoted-printable Cc: Chris Carpinello , netdev@oss.sgi.com Return-path: To: Robert Olsson In-Reply-To: <16582.49526.374109.580312@robur.slu.se> Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org Robert Olsson wrote: > Chris Carpinello writes: >=20 > Hello! >=20 > Is seems like your network load @ ~202 Mbps gets you system into=20 > continuing polling as we see very few interrupts on your eth3. > This means that rx_softirq reschedules itself do_softirq() kicks > ksoftird to prevent the rx_softirq from monopolize the system. > So now all the work gets accounted in ksoftird And by design=20 > ->poll is strictly serialized per device to guarantee ordering and=20 > avoid cache bouncing we only see one ksoftirq used as use only have=20 > one input device. >=20 > P=E1draig suggest binding to separate CPU's. This is normally a good=20 > thing but as you only have one input device it will not help. agreed. All traffic is on eth3 so you can't share it over CPUs > And didn't we just see a fix for ifconfig down oops? yep, seems like it: http://marc.theaimsgroup.com/?l=3Dlinux-netdev&m=3D108631346103966&w=3D2 P=E1draig.