call_usermodehelper() returns -513 when ocfs2 umounting filesystems

call_usermodehelper() returns -513 when ocfs2 umounting filesystems

[Xue jiufei]

Hi, 
We have encountered an error when umounting ocfs2 filesystems.
Function ocfs2_leave_group() calls call_usermodehelper() to stop
heartbeat thread, but it returns -513(ERESTARTNOINTR) in one test.
And after that error, every times umounting the filesystem, it
returns the same error.
And at the same time, there's another kworker thread which is
sending messages to other nodes always return return errno
-512(ERESTARTSYS). So I think these two threads may have
pending signals remain.
This error can not reproduced any longer. Has any ideas?

The log is as follows:
[58463.684504] (kworker/u:1,17332,0):o2net_send_tcp_msg:1332 ERROR: sendmsg returned -512 instead of 104
[58915.213299] ocfs2: Error -513 running user helper "/usr/sbin/ocfs2_hb_ctl -K -u A5FD0ED8733D4F9C98D23B326AD7DE10"
[59406.443615] ocfs2: Error -513 running user helper "/usr/sbin/ocfs2_hb_ctl -K -u A5FD0ED8733D4F9C98D23B326AD7DE10"

Thanks.
		--joyce

Re: call_usermodehelper() returns -513 when ocfs2 umounting filesystems

[Xue jiufei]

On 2013/7/30 15:30, Xue jiufei wrote:
> Hi, 
> We have encountered an error when umounting ocfs2 filesystems.
> Function ocfs2_leave_group() calls call_usermodehelper() to stop
> heartbeat thread, but it returns -513(ERESTARTNOINTR) in one test.
> And after that error, every times umounting the filesystem, it
> returns the same error.
> And at the same time, there's another kworker thread which is
> sending messages to other nodes always return return errno
> -512(ERESTARTSYS). So I think these two threads may have
> pending signals remain.
> This error can not reproduced any longer. Has any ideas?
> 
> The log is as follows:
> [58463.684504] (kworker/u:1,17332,0):o2net_send_tcp_msg:1332 ERROR: sendmsg returned -512 instead of 104
> [58915.213299] ocfs2: Error -513 running user helper "/usr/sbin/ocfs2_hb_ctl -K -u A5FD0ED8733D4F9C98D23B326AD7DE10"
> [59406.443615] ocfs2: Error -513 running user helper "/usr/sbin/ocfs2_hb_ctl -K -u A5FD0ED8733D4F9C98D23B326AD7DE10"
> 
> Thanks.
> 		--joyce
The kernel version is 3.0.58.
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
> 

Re: call_usermodehelper() returns -513 when ocfs2 umounting filesystems

[Li Zefan]

On 2013/7/30 16:51, Xue jiufei wrote:
> On 2013/7/30 15:30, Xue jiufei wrote:
>> Hi, 
>> We have encountered an error when umounting ocfs2 filesystems.
>> Function ocfs2_leave_group() calls call_usermodehelper() to stop
>> heartbeat thread, but it returns -513(ERESTARTNOINTR) in one test.
>> And after that error, every times umounting the filesystem, it
>> returns the same error.
>> And at the same time, there's another kworker thread which is
>> sending messages to other nodes always return return errno
>> -512(ERESTARTSYS). So I think these two threads may have
>> pending signals remain.
>> This error can not reproduced any longer. Has any ideas?

You should have made the question more clear, and that is:

How is it possible that a workqueue kernel thread has pending signals?

>>
>> The log is as follows:
>> [58463.684504] (kworker/u:1,17332,0):o2net_send_tcp_msg:1332 ERROR: sendmsg returned -512 instead of 104
>> [58915.213299] ocfs2: Error -513 running user helper "/usr/sbin/ocfs2_hb_ctl -K -u A5FD0ED8733D4F9C98D23B326AD7DE10"
>> [59406.443615] ocfs2: Error -513 running user helper "/usr/sbin/ocfs2_hb_ctl -K -u A5FD0ED8733D4F9C98D23B326AD7DE10"
>>

Re: call_usermodehelper() returns -513 when ocfs2 umounting filesystems

[Tejun Heo]

On Tue, Jul 30, 2013 at 04:51:09PM +0800, Xue jiufei wrote:
> The kernel version is 3.0.58.

That's way too old for upstream to do any meaningful debugging without
more specific information.  Kernel threads usually don't deal with
signals but there are a handful weird places in kernel which make use
of actual signals to affect the control flow.  ISTR modifying some so
that it's not too crazy but I could be misremembering....

Thanks.

-- 
tejun