[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YGYq0E/vZe65R7kk@lunn.ch>
Date: Thu, 1 Apr 2021 22:19:28 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Ioana Ciornei <ciorneiioana@...il.com>
Cc: davem@...emloft.net, kuba@...nel.org, netdev@...r.kernel.org,
ruxandra.radulescu@....com, Ioana Ciornei <ioana.ciornei@....com>
Subject: Re: [PATCH net-next 2/3] dpaa2-eth: add rx copybreak support
On Thu, Apr 01, 2021 at 11:13:50PM +0300, Ioana Ciornei wrote:
> On Thu, Apr 01, 2021 at 08:49:43PM +0200, Andrew Lunn wrote:
> > Hi Ioana
> >
> > > +#define DPAA2_ETH_DEFAULT_COPYBREAK 512
> >
> > This is quite big. A quick grep suggest other driver use 256.
> >
> > Do you have some performance figures for this?
> >
>
> Hi Andrew,
>
> Yes, I did some tests which made me end up with this default value.
>
> A bit about the setup - a LS2088A SoC, 8 x Cortex A72 @ 1.8GHz, IPfwd
> zero loss test @ 20Gbit/s throughput. I tested multiple frame sizes to
> get an idea where is the break even point.
>
> Here are 2 sets of results, (1) is the baseline and (2) is just
> allocating a new skb for all frames sizes received (as if the copybreak
> was even to the MTU). All numbers are in Mpps.
>
> 64 128 256 512 640 768 896
>
> (1) 3.23 3.23 3.24 3.21 3.1 2.76 2.71
> (2) 3.95 3.88 3.79 3.62 3.3 3.02 2.65
>
> It seems that even for 512 bytes frame sizes it's comfortably better when
> allocating a new skb. After that, we see diminishing rewards or even worse.
Nice. If you need to respin, consider putting this in patch 0/3.
Andrew
Powered by blists - more mailing lists