netdev
[Top] [All Lists]

Re: [Fwd: Re: possible bug x86 2.4.2 SMP in IP receive stack]

To: Pekka Pietikainen <pp@xxxxxxxxxxxxxx>
Subject: Re: [Fwd: Re: possible bug x86 2.4.2 SMP in IP receive stack]
From: Andrew Morton <andrewm@xxxxxxxxxx>
Date: Wed, 07 Mar 2001 21:15:17 +1100
Cc: Bob Felderman <feldy@xxxxxxxx>, kuznet@xxxxxxxxxxxxx, netdev@xxxxxxxxxxx
References: <200103061854.KAA03475@xxxxxxxx>, <200103061854.KAA03475@xxxxxxxx> <20010306211530.A14203@xxxxxxxxx>
Sender: owner-netdev@xxxxxxxxxxx
Pekka Pietikainen wrote:
> 
> On Tue, Mar 06, 2001 at 10:54:22AM -0800, Bob Felderman wrote:
> > a 1.5Gigabit/sec udp stream. It might be possible to reproduce this
> > using 100Mbit ethernet, but it might require 1gbit ethernet with jumbo
> > frames. I'm currently using a 9000 bytes MTU. Our interrupt routine
> > will deliver only a single ethernet packet to the higher levels
> > for each interrupt, so maybe that also stresses the IP fragmentation
> > code.
> Verified on Alteon GigE (1500 byte frames seemed to be ok, 9000 killed the
> receiver) using the script&/proc settings. I'd have to do some walking to
> get the Oops so I'll do that tomorrow :)

So you're saying that you can reproduce this crash with acenic?

And that it's specific to jumbo frames, SMP and the /proc settings
which Bob is using?

Bob is using a non-zerocopy kernel.  What were you using?

<Prev in Thread] Current Thread [Next in Thread>