[ovs-discuss] Bug version 2.1 - ovs-vswitchd: pthread_mutex_lock failed (Invalid argument)

Ben Pfaff blp at nicira.com
Mon Mar 17 22:53:39 UTC 2014


Well, I assume ovs-vswitchd dumping core.  If not, you need to make it
dump core.

Once you have a core dump, invoke GDB on the ovs-vswitchd binary and the core
dump, then type "thread apply all bt" and paste the result into an email.

On Mon, Mar 17, 2014 at 3:45 PM, André Martins <aanm at ua.pt> wrote:
> Sure, how?
>
>
> On 03/17/2014 10:39 PM, Ben Pfaff wrote:
>>
>> On Mon, Mar 10, 2014 at 08:47:38PM +0000, Andr? Martins wrote:
>>>
>>> Hello,
>>> after I add groups and flows via ovs-ofctl, for example:
>>> ovs-ofctl --protocols=OpenFlow13 add-flow
>>> unix:/usr/local/var/run/openvswitch/s4.mgmt 'in_port=2
>>> actions=group:1'
>>> When I exit mininet, it halts when it tries to stop the open vSwitch
>>> switches.
>>> This is the last 5 lines when I run sudo ovs-vswitchd --pidfile
>>> 2014-03-10T20:40:59Z|00114|rconn|INFO|s5<->tcp:127.0.0.1:6633:
>>> connection failed (Protocol error)
>>> 2014-03-10T20:41:21Z|00115|ofproto|INFO|s4: 3 flow_mods 10 s ago (3 adds)
>>> 2014-03-10T20:41:21Z|00116|ofproto|INFO|s3: 3 flow_mods 10 s ago (3 adds)
>>> 2014-03-10T20:41:21Z|00117|ofproto|INFO|s5: 2 flow_mods in the 4 s
>>> starting 10 s ago (2 adds)
>>> ovs-vswitchd: pthread_mutex_lock failed (Invalid argument) <----This
>>> only happens when I add groups and flows it doesn't happen if I only
>>> run and close mininet
>>
>> Can you get a backtrace (on all threads) from ovs-vswitchd?
>>
>
>



-- 
"I don't normally do acked-by's.  I think it's my way of avoiding
getting blamed when it all blows up."               Andrew Morton



More information about the discuss mailing list