Hi Matus,
I get some bt details below:



vpp#
Thread 5 "vpp_wk_0" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7ffe33b1d700 (LWP 8348)]
0x00007fffb3d5d3c8 in rte_atomic16_add_return (v=0x7fffb3f9a5e8 
<bond_ethdev_tx_burst_8023ad+854>, inc=-1)
    at 
/home/vpp/vpp/build-root/install-vpp_debug-native/dpdk/include/dpdk/generic/rte_atomic.h:323
323return __sync_add_and_fetch(&v->cnt, inc);
(gdb) bt
#0  0x00007fffb3d5d3c8 in rte_atomic16_add_return (v=0x7fffb3f9a5e8 
<bond_ethdev_tx_burst_8023ad+854>, inc=-1)
    at 
/home/vpp/vpp/build-root/install-vpp_debug-native/dpdk/include/dpdk/generic/rte_atomic.h:323
#1  0x00007fffb3d632a5 in __rte_mbuf_refcnt_update (m=0x7fffb3f9a5d6 
<bond_ethdev_tx_burst_8023ad+836>, value=-1)
    at 
/home/vpp/vpp/build-root/install-vpp_debug-native/dpdk/include/dpdk/rte_mbuf.h:758
#2  0x00007fffb3d63c96 in rte_pktmbuf_prefree_seg (m=0x7fffb3f9a5d6 
<bond_ethdev_tx_burst_8023ad+836>)
    at 
/home/vpp/vpp/build-root/install-vpp_debug-native/dpdk/include/dpdk/rte_mbuf.h:1408
#3  rte_pktmbuf_free_seg (m=0x7fffb3f9a5d6 <bond_ethdev_tx_burst_8023ad+836>) 
at 
/home/vpp/vpp/build-root/install-vpp_debug-native/dpdk/include/dpdk/rte_mbuf.h:1444
#4  rte_pktmbuf_free (m=0x7fffb3f9a5d6 <bond_ethdev_tx_burst_8023ad+836>) at 
/home/vpp/vpp/build-root/install-vpp_debug-native/dpdk/include/dpdk/rte_mbuf.h:1467
#5  0x00007fffb3d66dfd in dpdk_interface_tx (vm=0x7fffb6828080, 
node=0x7fffb65296c0, f=0x7fffb6778f40) at 
/home/vpp/vpp/build-data/../src/plugins/dpdk/device/device.c:505
#6  0x00007ffff7953d16 in dispatch_node (last_time_stamp=11677216614527570, 
frame=0x7fffb6778f40, dispatch_state=VLIB_NODE_STATE_POLLING, 
type=VLIB_NODE_TYPE_INTERNAL,
    node=0x7fffb65296c0, vm=0x7fffb6828080) at 
/home/vpp/vpp/build-data/../src/vlib/main.c:988
#7  dispatch_pending_node (vm=vm@entry=0x7fffb6828080, 
pending_frame_index=pending_frame_index@entry=17, 
last_time_stamp=11677216614527570)
    at /home/vpp/vpp/build-data/../src/vlib/main.c:1138
#8  0x00007ffff7954c27 in vlib_main_or_worker_loop (is_main=0, 
vm=0x7fffb6828080) at /home/vpp/vpp/build-data/../src/vlib/main.c:1616
#9  vlib_worker_loop (vm=0x7fffb6828080) at 
/home/vpp/vpp/build-data/../src/vlib/main.c:1641
#10 0x00007ffff689ab98 in clib_calljmp () at 
/home/vpp/vpp/build-data/../src/vppinfra/longjmp.S:110
#11 0x00007ffe33b1cdd0 in ?? ()
#12 0x00007fffb3e281b7 in eal_thread_loop (arg=0x0) at 
/home/vpp/vpp/build-root/build-vpp_debug-native/dpdk/dpdk-stable-18.02.1/lib/librte_eal/linuxapp/eal/eal_thread.c:153
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(gdb)


and my startup.conf is :

/etc/vpp# grep -v "#" startup.conf

 

unix {

  log /var/log/vpp/vpp.log

  full-coredump

  interactive

  cli-listen localhost:5002

  gid vpp

  exec  /etc/vpp/setvppinterface

}

 

api-trace {

  on

}

 

api-segment {

  gid vpp

}

 

cpu {

          main-core 1

          corelist-workers 4

}

dpdk {

          dev 0000:42:00.0

        dev 0000:42:00.1

        dev 0000:04:00.0

        dev 0000:04:00.1

 

        vdev eth_bond0,mode=4,slave=0000:04:00.0,slave=0000:04:00.1

        vdev eth_bond1,mode=4,slave=0000:42:00.0,slave=0000:42:00.1

 

          uio-driver vfio-pci

        huge-dir /dev/hugepages

          socket-mem 2048,2048

}




Regards,

George




On 7/9/2018 13:23,Matus Fabian -X (matfabia - PANTHEON TECHNOLOGIES at 
Cisco)<[email protected]> wrote:

Hi George,

 

I don’t have any issue with “show nat44 sessions” under heavy load (it just 
print number of session for each NAT user/internal network IP, no sessions 
details). Without core file analysis hard to say where is the issue.

 

Regards,

Matus

 

 

From:[email protected] <[email protected]> On Behalf Of Geoge
Sent: Saturday, July 7, 2018 4:59 AM
To: vpp-dev <[email protected]>
Subject: [vpp-dev] The command "show nat44 sessions" occurs "worker thread 
deadlock".

 

 

Hello everyone,

 

When I use "show nat44 sessions" ,with the VM sends packtes to vpp do nat 
translate, the vpp will crash. 

 

The vpp shows below:

 

vpp#     show nat44 sessions

Connection closed by foreign host.

 

vpp# vlib_worker_thread_barrier_sync_int: worker thread deadlock

Aborted

 

My vpp version is :vpp v18.04.

 

In my opinion, I think that the "CLI command" can not be disposed under a large 
number of packets in processing.

 

Please give me some advice, thank you very much!

 

Regards,

Geoge

 
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#9809): https://lists.fd.io/g/vpp-dev/message/9809
Mute This Topic: https://lists.fd.io/mt/23181323/21656
Group Owner: [email protected]
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub  [[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
  • [... Geoge
    • ... Matus Fabian -X (matfabia - PANTHEON TECHNOLOGIES@Cisco) via Lists.Fd.Io
      • ... George
    • ... Geoge

Reply via email to