Re: btrfs: hang on boot due to tests

[Chris Mason]

On 06/10/2014 07:14 AM, Ed Tomlinson wrote:
> On Monday 09 June 2014 12:17:50 Sasha Levin wrote:
>> On 06/09/2014 11:59 AM, Chris Mason wrote:
>>> On 06/09/2014 11:16 AM, Sasha Levin wrote:
>>>>> Hi all,
>>>>>
>>>>> It seems that some recent changes to btrfs tests make it hang during boot:
>>>>>
>>>>> [   49.730033] NMI watchdog: BUG: soft lockup - CPU#34 stuck for 23s! [swapper/0:1]
>>>>> [   49.730033] Modules linked in:
>>>>> [   49.730033] hardirqs last enabled at (6389143): restore_args (arch/x86/kernel/entry_64.S:829)
>>>>> [   49.730033] hardirqs last disabled at (6389144): apic_timer_interrupt (arch/x86/kernel/entry_64.S:1021)
>>>>> [   49.730033] softirqs last enabled at (6389142): __do_softirq (./arch/x86/include/asm/preempt.h:22 kernel/softirq.c:296)
>>>>> [   49.730033] softirqs last disabled at (6389139): irq_exit (kernel/softirq.c:346 kernel/softirq.c:387)
>>>>> [   49.730033] CPU: 34 PID: 1 Comm: swapper/0 Not tainted 3.15.0-rc8-next-20140606-sasha-00021-ga9d3a0b-dirty #597
>>>
>>> This is 3.15-rc8 + linux-next?  I'll try to reproduce here, but the
>>> tests were working for me.
>>
>> Yes, it's the latest -next tree available.
>>
>> Also note that it doesn't happen every time, so might be some sort of a race?
> 
> I've noticed that with -rc8 and now .15 btrfs fails to automount (or mount) about 1 in 2 times requiring a reboot to get it to work.
> I have not seen anything in logs.  Might this be related?

It shouldn't be, the lockup Sasha hit is only in the self testing code,
and he's triggering it in the qgroups self tests.

Do you have CONFIG_BTRFS_FS_RUN_SANITY_TESTS=y?

I'd love a sysrq-t or sysrq-w while you're stuck.

-chris
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/