Linux Ethernet Bridging - Mar 2015 - [Bridge] [RFC PATCH v2] bridge: make it possible for packets to traverse the bridge without hitting netfilter

On 02/26/15 17:34, David Miller wrote:
> From: Imre Palik <imrep at amazon.de>
> Date: Thu, 26 Feb 2015 11:19:25 +0100
> 
>> If you are looking for peculiarities in my setup then here they are:
>> I am on 4k pages, and perf is not working :-(
>> (I am trying to fix those too, but that is far from being a low hanging
fruit.)
>> So my guess would be that the packet pipeline doesn't fit in the
cache/tlb
> 
> Pure specualtion until you can actually use perf to measure these
> things.
> 
> And I don't want to apply patches which were designed based upon
> pure speculation.
> 
I did performance measurements in the following way:

Removed those pieces of the packet pipeline that I don't necessarily need
one-by-one.  Then measured their effect on small packet performance.

This was the only part that produced considerable effect.

The pure speculation was about why the effect is more than 15% increase in
packet throughput, although the code path avoided contains way less code than
15% of the packet pipeline.  It seems, Felix Fietkau profiled similar changes,
and found my guess well founded.

Now could anybody explain me what else is wrong with my patch?  I run out of
ideas what to improve.

Thanks

Imre

On Fri, Mar 06, 2015 at 11:34:29AM +0100, Imre Palik
wrote:
> On 02/26/15 17:34, David Miller wrote:
> > From: Imre Palik <imrep at amazon.de>
> > Date: Thu, 26 Feb 2015 11:19:25 +0100
> > 
> >> If you are looking for peculiarities in my setup then here they
are:
> >> I am on 4k pages, and perf is not working :-(
> >> (I am trying to fix those too, but that is far from being a low
hanging fruit.)
> >> So my guess would be that the packet pipeline doesn't fit in
the cache/tlb
> > 
> > Pure specualtion until you can actually use perf to measure these
> > things.
> > 
> > And I don't want to apply patches which were designed based upon
> > pure speculation.
> > 
> 
> I did performance measurements in the following way:
> 
> Removed those pieces of the packet pipeline that I don't necessarily
> need one-by-one.  Then measured their effect on small packet
> performance.
> 
> This was the only part that produced considerable effect.
> 
> The pure speculation was about why the effect is more than 15%
> increase in packet throughput, although the code path avoided
> contains way less code than 15% of the packet pipeline.  It seems,
> Felix Fietkau profiled similar changes, and found my guess well
> founded.
> 
> Now could anybody explain me what else is wrong with my patch?
We have to come up with a more generic solution for this.

These sysfs tweaks you're proposing look to me like an obscure way to
tune this.

From: Imre Palik <imrep.amz at gmail.com>
Date: Fri, 06 Mar 2015 11:34:29 +0100
> Removed those pieces of the packet pipeline that I don't necessarily
> need one-by-one.  Then measured their effect on small packet
> performance.
> 
> This was the only part that produced considerable effect.
> 
> The pure speculation was about why the effect is more than 15%
> increase in packet throughput, although the code path avoided
> contains way less code than 15% of the packet pipeline.  It seems,
> Felix Fietkau profiled similar changes, and found my guess well
> founded.
Yes and that's the part being left out, the "why".

That's part of what I expect in the justification.

Look, we're not doing things this way.  We need to find a clean
and generic way to make the netfilter hooks as cheap as possible
when netfilter rules are not in use.