[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20120613124900.GA1741@localhost>
Date: Wed, 13 Jun 2012 20:49:00 +0800
From: Fengguang Wu <fengguang.wu@...el.com>
To: Steven Rostedt <srostedt@...hat.com>
Cc: "Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>,
Dave Jones <davej@...hat.com>,
LKML <linux-kernel@...r.kernel.org>,
Lai Jiangshan <laijs@...fujitsu.com>
Subject: Re: rcu-torture boot hang
Hi Steven,
On Tue, Jun 12, 2012 at 08:11:22PM +0800, Fengguang Wu wrote:
> Hi Paul,
>
> In kernel boot tests with the attached config, I find the 3.5-rc2+
> kernels all hang here:
>
> [ 6.522546] rcu-torture:--- Start of test: nreaders=2
> nfakewriters=4 stat_interval=0 verbose=0 test_no_idle_hz=0
> shuffle_interval=3 stutter=5 irqreader=1 fqs_duration=0 fqs_holdoff=0
> fqs_stutter=3 test_boost=1/0 test_boost_interval=7
> test_boost_duration=4 shutdown_secs=0 onoff_interval=0 onoff_holdoff=0
It turns out that commit 5963e317b1e9d2a4511503916d8fd664bb8fa8fb
("ftrace/x86: Do not change stacks in DEBUG when calling lockdep") is
the root cause of this boot hang. The commit is reported by git bisect
and confirmed to fix the boot hang by reverting in on 3.5-rc2.
[ rcu torture tests ]
...It would hang here before reverting the commit...
[ 1.611901] Testing tracer function: PASSED
Attached is my kconfig. I'm running the kernel tests in kvm:
kvm -cpu kvm64 -enable-kvm \
-kernel ./gpxelinux.lkrn \
-initrd ./kvm.gpxe \
-m 384M \
-net nic,vlan=0,macaddr=00:00:00:00:00:00,model=e1000 \
-net user,vlan=0,hostfwd=tcp::$port-:22 \
-boot order=nc \
-no-reboot \
-watchdog i6300esb \
$(qemu_drives) \
-pidfile $pidfile \
-serial file:$dmesg \
Thanks,
Fengguang
View attachment ".config" of type "text/plain" (88737 bytes)
Powered by blists - more mailing lists