Hi,

 

@'dekelp@nvidia.com'@'Matan Azrad' Can you kindly suggest?

Thank you.

 

 

 

Hi,

 

dpdk version in use: 21.11.2

 

Mlx5 driver crashes in rxq_cq_decompress_v in traffic test after several minutes.

Stack trace:

(gdb) bt

#0  0x00007ffff58612bc in _mm_storeu_si128 (__B=..., __P=<optimized out>)

    at /usr/lib/gcc/x86_64-redhat-linux/12/include/emmintrin.h:739

#1  rxq_cq_decompress_v (rxq=rxq@entry=0x2abe5592f40, cq=cq@entry=0x2abe54fdb00, elts=elts@entry=0x2abe5594638)

    at ../dpdk-21.11/drivers/net/mlx5/mlx5_rxtx_vec_sse.h:142

#2  0x00007ffff5862c84 in rxq_burst_v (no_cq=<synthetic pointer>, err=0x7fffffffb848, pkts_n=4, pkts=<optimized out>,

    rxq=0x2abe5592f40) at ../dpdk-21.11/drivers/net/mlx5/mlx5_rxtx_vec.c:349

#3  mlx5_rx_burst_vec (dpdk_rxq=0x2abe5592f40, pkts=0x7fffffffbf80, pkts_n=32) at ../dpdk-21.11/drivers/net/mlx5/mlx5_rxtx_vec.c:393

#4  0x00005555556a0f41 in rte_eth_rx_burst (nb_pkts=32, rx_pkts=0x7fffffffbf80, queue_id=0, port_id=1)

    at /usr/include/rte_ethdev.h:5721

…

Attached is the error log “Unexpected CQE error syndrome…” and dump file

 

I found there was a similar bug here: https://bugs.dpdk.org/show_bug.cgi?id=334

But the fix (88c0733535d6 extend Rx completion with error handling) should already been included, as I’m using 21.11.2

Also below commit (fix to 88c0733535d6) is already included in my dpdk version.

commit 60b254e3923d007bcadbb8d410f95ad89a2f13fa

Author: Matan Azrad matan@nvidia.com

Date:   Thu Aug 11 19:51:55 2022 +0300

 

    net/mlx5: fix Rx queue recovery mechanism

 

Any suggestion?

Thank you.

 

Br, Xiaoping