Unkillable process in STOP state

[Johan Schuijt-Li]

Yeah, the only difference we had that for us the status was 'Ds' rather then 'STOP'. So the status is most likely irrelevant, at least the traces are exactly the same! :)

- Johan


> On 12 Nov 2015, at 11:45, Slawa Olhovchenkov <slw at zxy.spb.ru> wrote:
> 
> On Thu, Nov 12, 2015 at 07:05:46AM +0100, Johan Schuijt-Li wrote:
> 
>> This seems like the exact same problem as that we’ve had, more details can be found in the following PR:
>> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=200992 <https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=200992>
> 
> May be. Other time I am see "sleep without queue" status
> 
>> The patch attached there solved all our problems.
> 
> I am try this patch on next update, thanks.
> 
>> - Johan
>> 
>> 
>>> On 12 Nov 2015, at 01:12, Slawa Olhovchenkov <slw at zxy.spb.ru> wrote:
>>> 
>>> I have amd64, STABLE r288167.
>>> 
>>> root at edge09:/home/admin # procstat -k -k 627
>>> PID    TID COMM             TDNAME           KSTACK                       
>>> 627 100167 tcpkali.new      -                mi_switch+0xe1 thread_suspend_switch+0x170 thread_single+0x4e5 exit1+0xbe sys_sys_exit+0xe amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 100172 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 100173 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 100174 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 100175 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 100178 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 100180 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 102207 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 102208 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 102209 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> 627 102211 tcpkali.new      -                mi_switch+0xe1 sleepq_timedwait_sig+0x8b _sleep+0x238 kern_kevent+0x401 sys_kevent+0x12a amd64_syscall+0x357 Xfast_syscall+0xfb 
>>> root at edge09:/home/admin # procstat -t 627
>>> PID    TID COMM             TDNAME           CPU  PRI STATE   WCHAN    
>>> 627 100167 tcpkali.new      -                 10  120 stop    -         
>>> 627 100172 tcpkali.new      -                  3  152 stop    -         
>>> 627 100173 tcpkali.new      -                  5  152 stop    -         
>>> 627 100174 tcpkali.new      -                  3  152 stop    -         
>>> 627 100175 tcpkali.new      -                  2  149 stop    -         
>>> 627 100178 tcpkali.new      -                  5  152 stop    -         
>>> 627 100180 tcpkali.new      -                  2  132 stop    -         
>>> 627 102207 tcpkali.new      -                  5  136 stop    -         
>>> 627 102208 tcpkali.new      -                  3  152 stop    -         
>>> 627 102209 tcpkali.new      -                  5  139 stop    -         
>>> 627 102211 tcpkali.new      -                  1  120 stop    -         
>>> 
>>> kill -STOP don't have effect.
>>> gdb can't be attached.
>>> _______________________________________________
>>> freebsd-stable at freebsd.org mailing list
>>> https://lists.freebsd.org/mailman/listinfo/freebsd-stable
>>> To unsubscribe, send any mail to "freebsd-stable-unsubscribe at freebsd.org"
>>