[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <50532464.5050904@linux.vnet.ibm.com>
Date: Fri, 14 Sep 2012 18:04:44 +0530
From: "Srivatsa S. Bhat" <srivatsa.bhat@...ux.vnet.ibm.com>
To: Peter Zijlstra <peterz@...radead.org>
CC: Fengguang Wu <fengguang.wu@...el.com>, paulmck@...ux.vnet.ibm.com,
jack@...e.cz,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Thomas Gleixner <tglx@...utronix.de>,
"rusty@...tcorp.com.au" <rusty@...tcorp.com.au>,
Tejun Heo <tj@...nel.org>,
Michael Wang <wangyun@...ux.vnet.ibm.com>
Subject: Re: WARNING: at kernel/rcutree.c:1558 rcu_do_batch+0x386/0x3a0(),
during CPU hotplug
On 09/14/2012 05:55 PM, Peter Zijlstra wrote:
> On Fri, 2012-09-14 at 17:48 +0530, Srivatsa S. Bhat wrote:
>> #! /bin/bash
>
> CPUPATH="/sys/devices/system/cpu"
>
>> NUMBER_OF_CPUS=`ls -d /sys/devices/system/cpu/cpu[0-9]* | wc -l`
>
> apply the above
>
>> cd /sys/devices/system/cpu
>
> skip this, so running the script doesn't change PWD
>
>> while [ 1 ]
>
> while :;
>
>> do
>> for ((i=1; i < NUMBER_OF_CPUS; i++))
>> do
>> sleep 1;
>
> Also play with shorter sleeps like: sleep .1, I've found that higher
> hotplug rate triggers some races faster.
>
>> state=`cat cpu$i/online`
>> if [ $state -eq 0 ]
>> then
>> echo 1 > cpu$i/online
>> else
>> echo 0 > cpu$i/online
>
> echo $((state^1)) > $CPUPATH/cpu$i/online
>
>> fi
>> done
>> done
>
>
Thanks a lot Peter for your suggestions!
Fengguang, like I had mentioned some time before, somebody had pointed
me to a hotplug test-suite that I believe is being used in LTP.
I have a copy of that suite, which I have attached with this mail.
I have never actually used it seriously before (because I could trigger
bugs with my simple script!), so I don't exactly know what state it is
in ;-)
Regards,
Srivatsa S. Bhat
Download attachment "lhcs_regression-1.6.tgz" of type "application/x-compressed-tar" (17854 bytes)
Powered by blists - more mailing lists