Re: current status of ixg(4)

[Masanobu SAITOH <msaitoh%execsw.org@localhost>]

On 2015/04/02 18:17, Masanobu SAITOH wrote:
> On 2015/04/02 18:13, Masanobu SAITOH wrote:
> > On 2015/04/02 0:01, Thor Lancelot Simon wrote:
> > > On Tue, Mar 31, 2015 at 03:38:45PM +0200, 6bone%6bone.informatik.uni-leipzig.de@localhost wrote:
> > > > On Fri, 27 Mar 2015, Masanobu SAITOH wrote:
> > > >
> > > > > This change have commited now.
> > > > >
> > > > > New patch:
> > > > >
> > > > >     http://www.netbsd.org/~msaitoh/ixg-20150327-0.dif
> > > >
> > > > I have tested the patch and found no problems.
> >
> >   Thanks. I'll commit that diff.
> >
> > > > My server (HP G5) can handle with the new driver package rates up to 200,000
> > > > packets per second. Then CPU0 is running at 100% with interrupts.
> > > >
> > > > If I have not charged me, it comes with 200,000 packets per second and an
> > > > MTU of 1500 bytes to a maximum of 2.4GB.
> > > >
> > > > Is it possible to optimize some parameters for the interrupt throttling?
> > >
> > > I think for your workload you may actually want interrupt distribution over
> > > multiple CPUs.  But we have some support for that in -current now, right?
> > >
> > > The next thing you're likely to run into is contention for the kernel lock
> > > in ip_input.  But if I recall correctlly how this driver's built, I don't
> > > think you should be seeing that time charged to the interrupt handler, so
> > > I *think* you are not there yet.
> > >
> > > Thor
> >
> >   It's correct what tls@ said.
> >
> >   For ixg(4), it has multiqueue support, but it's not enabled because
> > NetBSD-current has not MSI/MSI-X support yet. The remaining is to merge
> > knakahara@'s MSI/MSI-X support and use the API. The change would be
> > merged in a few weeks.
> >
> >   For layer 2, NET_MPSAFE enables the MP capability.
> >
> >   For layer 3, it's now working in progress.
> >
> >   For the detail about NET_MPSAFE and MSI/MSI-X, see Ryota Ozaki and
> > Kengo Nakahara's paper and presentations in AsiaBSDCon 2015:
> >
> >      http://www.netbsd.org/gallery/presentations/
>
>   There are some graphs from page 60 to 63 in their slides.
>
>
> > Other than NET_MPSAFE and MSI/MSI-X, ixg(4) still have a room to
> > improve the performance. I have not merged all of FreeBSD's change
> > yet. I checked all changes and I knew there were some changes that

s/some changes/some changes which improve performance/

> > I have not merged yet.
> >
> >   Thanks.


--
-----------------------------------------------
                SAITOH Masanobu (msaitoh%execsw.org@localhost
                                 msaitoh%netbsd.org@localhost)