netdev
[Top] [All Lists]

Re: Network buffer hang was Re: [PATCH] 2.6 workaround for Athlon/Optero

To: dada1 <dada1@xxxxxxxxxxxxx>
Subject: Re: Network buffer hang was Re: [PATCH] 2.6 workaround for Athlon/Opteron prefetch errata
From: Ben Greear <greearb@xxxxxxxxxxxxxxx>
Date: Thu, 11 Sep 2003 18:46:24 -0700
Cc: "Nakajima, Jun" <jun.nakajima@xxxxxxxxx>, Andi Kleen <ak@xxxxxx>, linux-kernel@xxxxxxxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to: <0b2901c37867$1db399a0$890010ac@edumazet>
Organization: Candela Technologies
References: <uqD5.3BI.3@xxxxxxxxxxxxxxxx> <m3iso0arlx.fsf@xxxxxxxxxxxxxxxxxxxxxx> <0a5801c37821$54eb8180$890010ac@edumazet> <20030911051121.GA7751@xxxxxxxxxxxxx> <0a7701c37829$c4bdef40$890010ac@edumazet> <20030911120956.GB7751@xxxxxxxxxxxxx> <0b2901c37867$1db399a0$890010ac@edumazet>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5b) Gecko/20030827
dada1 wrote:
This is not a kernel crash. But total freeze as all memory is used by
network buffers, in no more than 10 seconds.

I believe I have hit this as well, by starting 500kbps streams on 500
vlans at once.  My machine locks so hard that I have not been able to
get any info out of it...but the realtek driver shows and error message
about not being able to allocate any buffers if I ping it....

I was using kernel 2.4.21 or so, and had tcp buffers turned up
quite high.  I have seen this on P-IV and AMD (single Athlon) systems,
as well as a VIA C3.


--
Ben Greear <greearb@xxxxxxxxxxxxxxx>
Candela Technologies Inc  http://www.candelatech.com



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