Re: purpose of the skb head pool

netdev

Re: purpose of the skb head pool

To:	Robert Olsson <Robert.Olsson@xxxxxxxxxxx>
Subject:	Re: purpose of the skb head pool
From:	Florian Weimer <fw@xxxxxxxxxxxxx>
Date:	Thu, 01 May 2003 12:38:38 +0200
Cc:	Christoph Hellwig <hch@xxxxxx>, davem@xxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to:	<16046.30879.738356.495523@xxxxxxxxxxxx> (Robert Olsson's message of "Tue, 29 Apr 2003 15:05:35 +0200")
References:	<20030429135506.A22411@xxxxxx> <16046.30879.738356.495523@xxxxxxxxxxxx>
Sender:	netdev-bounce@xxxxxxxxxxx
User-agent:	Gnus/5.09002 (Oort Gnus v0.20) Emacs/21.3 (gnu/linux)

Robert Olsson <Robert.Olsson@xxxxxxxxxxx> writes:

>  2.6.66 IP. Forwarding of two input simplex flows. eth0->eth1, eth2->eth3 
>  Fixed affinity CPU0: eth0, eth3. CPU1: eth1, eth2. Which common for routing
>  and should be "worst case" for other use. The test should give a very high 
>  load on the packet memory system. As seen at least we don't see any
>  improvement from skb_head_pool code. 
>
>
>  Vanilla 2.5.66              381 kpps
>  Magazine                    431 kpps
>  Magazine + no skb_head_pool 435 kpps

Can you rerun this test with random source/destination addresses, to
get more realistic (for some configurations) numbers?

[More with this subject...]

<Prev in Thread]	Current Thread	[Next in Thread>
Re: purpose of the skb head pool, Florian Weimer <= Re: purpose of the skb head pool, David S. Miller

Previous by Date:	Re: [PATCH] af_pppox: create module infrastructure for protocol modules, David S. Miller
Next by Date:	Re: purpose of the skb head pool, David S. Miller
Previous by Thread:	Re: [PATCH] af_pppox: create module infrastructure for protocol modules, David S. Miller
Next by Thread:	Re: purpose of the skb head pool, David S. Miller
Indexes:	[Date] [Thread] [Top] [All Lists]