[dpdk-dev] [Bug 301] DPDK driver is crashing for Mellanox-5 NIC card

[dpdk-dev] [Bug 301] DPDK driver is crashing for Mellanox-5 NIC card

[bugzilla]

https://bugs.dpdk.org/show_bug.cgi?id=301

            Bug ID: 301
           Summary: DPDK driver is crashing for Mellanox-5 NIC card
           Product: DPDK
           Version: 19.05
          Hardware: x86
                OS: Linux
            Status: CONFIRMED
          Severity: critical
          Priority: Normal
         Component: ethdev
          Assignee: dev@dpdk.org
          Reporter: ullas-d.bhat@hpe.com
  Target Milestone: ---

Created attachment 44
  --> https://bugs.dpdk.org/attachment.cgi?id=44&action=edit
Back trace for core dump file

Hardware Details:

Server : ProLiant DL380 Gen10 (Processor - Intel(R) Xeon(R) Gold 6126 CPU @
2.60GHz)

OS Version : Red Hat Enterprise Linux Server release 7.6 (Maipo)

Kernel version : 3.10.0-957.el7.x86_64

# lspci |grep -i ether
12:00.0 Ethernet controller: Mellanox Technologies MT27710 Family [ConnectX-4
Lx]
12:00.1 Ethernet controller: Mellanox Technologies MT27710 Family [ConnectX-4
Lx]

# lspci | grep Mellanox | awk '{print $1}' | xargs -i -r mstvpd {}
ID:      HPE Eth 10/25Gb 2p 640SFP28 Adptr
PN:      817751-001
EC:      G-5903
SN:      ACA91800ML
V0:      PCIe GEN3 x8 10/25Gb 15W
V2:      5918
V4:      B883038282D0
V5:      0G
VA:      HP:V2=MFG:V3=FW_VER:V4=MAC:V5=PCAR
VB:      HPE ConnectX-4 Lx SFP28

# ethtool -i ens1f1
driver: mlx5_core
version: 4.6-1.0.1
firmware-version: 14.23.8036 (HP_2420110034)
expansion-rom-version:
bus-info: 0000:12:00.1
supports-statistics: yes
supports-test: yes
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: yes

Problem Description:
_________________________

We are running our DPDK app with latest Mellanox NIC card which has 10GBPS of
VLAN capacity, while app running in PMD mode for receiving and transmitting RTP
packets.After sometime our app crashes and generates core dump file.
We have attached a file with back trace for core dump. We also see same crash
with dpdk-18.11 version.

This issue is not seen when run with Intel I350 NIC Card


Crash back trace snippet 

Program terminated with signal 11, Segmentation fault.
#0  rte_atomic16_read (v=0x6666666666666678) at
/home/dpdk-stable-18.08.1/x86_64-native-linuxapp-gcc/include/generic/rte_atomic.h:258
258             return v->cnt;
Missing separate debuginfos, use: debuginfo-install glibc-2.17-260.el7.x86_64
libgcc-4.8.5-36.el7.x86_64 libibverbs-46mlnx1-1.46101.x86_64
libmnl-1.0.3-7.el7.x86_64 libnl3-3.2.28-4.el7.x86_64
libpcap-1.5.3-11.el7.x86_64 libstdc++-4.8.5-36.el7.x86_64
numactl-libs-2.0.9-7.el7.x86_64
(gdb) bt
#0  rte_atomic16_read (v=0x6666666666666678) at
/home/dpdk-stable-18.08.1/x86_64-native-linuxapp-gcc/include/generic/rte_atomic.h:258
#1  rte_mbuf_refcnt_read (m=0x6666666666666666) at
/home/dpdk-stable-18.08.1/x86_64-native-linuxapp-gcc/include/rte_mbuf.h:820
#2  rte_pktmbuf_prefree_seg (m=0x6666666666666666) at
/home/dpdk-stable-18.08.1/x86_64-native-linuxapp-gcc/include/rte_mbuf.h:1638
#3  mlx5_tx_complete (txq=<optimized out>) at
/home/dpdk-stable-18.08.1/drivers/net/mlx5/mlx5_rxtx.h:577
#4  mlx5_tx_burst (dpdk_txq=<optimized out>, pkts=0x7ffd96e5cce0, pkts_n=1) at
/home/dpdk-stable-18.08.1/drivers/net/mlx5/mlx5_rxtx.c:505
#5  0x000000000047d04a in rte_eth_tx_burst (queue_id=0, nb_pkts=<optimized
out>, tx_pkts=0x7ffd96e5cce0, port_id=<optimized out>) at
/usr/local/include/dpdk/rte_ethdev.h:4101
#6  RteSendPacket (m=0x7f31ec5eab40, port=<optimized out>) at
/home/dpdk-stable-18.08.1/examples/ocmp-dpdk/rte-iflib.c:105
#7  PMD () at /home/dpdk-stable-18.08.1/examples/ocmp-dpdk/rte-iflib.c:612
#8  0x00000000004d4e73 in rte_eal_mp_remote_launch (f=f@entry=0x47a3a0
<DISPATCHER>, arg=arg@entry=0x0, call_master=call_master@entry=CALL_MASTER)
    at /home/dpdk-stable-18.08.1/lib/librte_eal/common/eal_common_launch.c:62
#9  0x000000000047e3a4 in RteMainLoop () at
/home/dpdk-stable-18.08.1/examples/ocmp-dpdk/rte-iflib.c:630
#10 0x000000000046f95e in main (argc=<optimized out>, argv=<optimized out>) at
/home/dpdk-stable-18.08.1/examples/ocmp-dpdk/dpdk-app.cc:377


Occurrence frequency:
________________
Always

-- 
You are receiving this mail because:
You are the assignee for the bug.