Re: [PATCH] tcp: splice as many packets as possible at once

[Jarek Poplawski <jarkao2@gmail.com>]

On Mon, Jan 19, 2009 at 07:14:20AM +0100, Willy Tarreau wrote:
> On Sun, Jan 18, 2009 at 07:27:19PM -0800, David Miller wrote:
> > From: Herbert Xu <herbert@gondor.apana.org.au>
> > Date: Mon, 19 Jan 2009 14:08:44 +1100
> > 
> > > On Mon, Jan 19, 2009 at 01:42:06AM +0100, Willy Tarreau wrote:
> > > >
> > > > Just for the record, I've now re-integrated those changes in a test kernel
> > > > that I booted on my 10gig machines. I have updated my user-space code in
> > > > haproxy to run a new series of tests. Eventhough there is a memcpy(), the
> > > > results are EXCELLENT (on a C2D 2.66 GHz using Myricom's Myri10GE NICs) :
> > > > 
> > > >   - 4.8 Gbps at 100% CPU using MTU=1500 without LRO
> > > >     (3.2 Gbps at 100% CPU without splice)
> > > 
> > > One thing to note is that Myricom's driver probably uses page
> > > frags which means that you're not actually triggering the copy.
> 
> So does this mean that the corruption problem should still there for
> such a driver ? I'm asking before testing, because at these speeds,
> validity tests are not that easy ;-)

I guess, David meant the performance: there is not much change because
only a small part could be copied. The most harmed should be jumbo
frames in linear only skbs. But no corruption is expected.

Jarek P.