[dpdk-users] issues with dpdk-pdump

Rami Rosen roszenrami at gmail.com
Wed Oct 24 07:11:44 CEST 2018


Hi Srinivas,

This works for me on dpdk-18.05, with an I40E nic, running testpmd
with the same parameters
that you use, " -c ff -n 4 -r 1 -m 800",  launching dpdk-pdump as you
did, namely with
/dpdk-pdump  --  --pdump 'port=0,queue=*,rx-dev=/tmp
/rc.pcapng,tx-dev=/tmp/tc.pcapng'

Did you ran testpmd in your successful test I suggested with the same
parameters that you are
using now for your app, namely "-c ff -n 4 -r 1 -m 800"? if not,
I suggest doing so.


Also,  to be on the safe side, I suggest running DPDK- 18.05 on a
clean branch. I see that in the log
you posted there is the message:

rte_eth_promiscuous_enable: Function not supported
Girish rte_eth_dev_confugure called 5 0 1
rte_eth_promiscuous_disable: Function not supported

And it seems as if you added logging "Girish rte_eth_dev_confugure called 5 0 1"
so it could be that by mistake you changed something in the DPDK core.

Regards,
Rami Rosen

On Mon, 22 Oct 2018 at 23:49, Srinivas Kotamarti
<srinivas_kotamarti at affirmednetworks.com> wrote:
>
> Hi Rami,
> Thanks for replying to my message.
>
> I followed the procedure like you said and here is what I get when I run the dpdk-pdump after running the testpmd.
>
> .(8.4.0.0-490.REL)root at an3000-26-5-ssm:/opt/thoroughbred/dpdk/tools# ./dpdk-pdump --  --pdump 'port=0,queue=*,rx-dev=/tmp/rc.pcapng,tx-dev=/tmp/tc.pcapng'
> EAL: Detected 7 lcore(s)
> EAL: Detected 1 NUMA nodes
>
> EAL: Multi-process socket /var/run/dpdk/rte/mp_socket_10550_2ce0b99fb94
> EAL: PCI device 0000:03:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 15ad:7b0 net_vmxnet3
> EAL: PCI device 0000:04:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 15ad:7b0 net_vmxnet3
> EAL: PCI device 0000:0b:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 15ad:7b0 net_vmxnet3
> EAL: PCI device 0000:0c:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 15ad:7b0 net_vmxnet3
> EAL: PCI device 0000:13:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 15ad:7b0 net_vmxnet3
> EAL: PCI device 0000:1b:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 15ad:7b0 net_vmxnet3
> rte_eth_promiscuous_disable: Function not supported
> rte_eth_allmulticast_disable: Function not supported
> Port 4 MAC: 00 00 00 01 02 03
> rte_eth_promiscuous_enable: Function not supported
> Girish rte_eth_dev_confugure called 5 0 1
> rte_eth_promiscuous_disable: Function not supported
> rte_eth_allmulticast_disable: Function not supported
> Port 5 MAC: 00 00 00 01 02 03
> rte_eth_promiscuous_enable: Function not supported
>
>
> I actually past the crash (because of rte_memconfig) that I reported earlier.  It seems to be because of the --no-shconf that we are passing for rte_eal_init. Once I removed it I got past the error but now I get the following error
>
> dpdk-pump[25734]: EAL: Probing VFIO support...
> dpdk-p  DPDPK_PDUMP rte_eal_init internal_config.process_type: 1 RTE_PROC_PRIMARY:0EAL: Probing VFIO support...
> dpdk-p EAL: Cannot initialize tailq: RTE_ACL
> dpdk-pump[25734]: EAL: Cannot initialize tailq: RTE_ACL
> dpdk-pump[25734]: EAL: Cannot init tail queues for objects#012
> dpdk-pump[25734]: PANIC in main():
> dpdk-pump[25734]: Cannot init EAL
> dpdk-p Tailq 0: qname:<VFIO_RESOURCE_LIST>, tqh_first:(nil), tqh_last:0x7f2b9d85f07c
> dpdk-p Tailq 1: qname:<UIO_RESOURCE_LIST>, tqh_first:0x7f27fe1f3e40, tqh_last:0x7f27fe1b8bc0
> dpdk-p Tailq 2: qname:<RTE_HASH>, tqh_first:0x7f27e2a8b300, tqh_last:0x7f27e2c05140
> dpdk-p Tailq 3: qname:<RTE_FBK_HASH>, tqh_first:(nil), tqh_last:0x7f2b9d85f10c
> dpdk-p Tailq 4: qname:<RTE_MEMPOOL>, tqh_first:0x7f27fe1b7ec0, tqh_last:0x7f27fe1b7ec0
> dpdk-p Tailq 5: qname:<RTE_RING>, tqh_first:0x7f27fe0f6d00, tqh_last:0x7f27fde03f40
> dpdk-p Tailq 6: qname:<RTE_LPM>, tqh_first:(nil), tqh_last:0x7f2b9d85f19c
> dpdk-p Tailq 7: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 8: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 9: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 10: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 11: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 12: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 13: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 14: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 15: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 16: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 17: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 18: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 19: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 20: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 21: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 22: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 23: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 24: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 25: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 26: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 27: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 28: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 29: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 30: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p Tailq 31: qname:<>, tqh_first:(nil), tqh_last:(nil)
> dpdk-p EAL: FATAL: Cannot init tail queues for objects
>
>
> These are the arguments now that we pass to rte_eal_init ' -c ff -n 4 -r 1 -m 800'.
>
> And, here is how I start the dpdk-pdump
> ./dpdk-pdump --  --pdump 'port=0, queue=*, rx-dev=/tmp/rx.pcap, tx-dev=/tmp/tx.pcap'
>
>
> And I am running with DPDK 18.05.
>
> Any help is greatly appreciated.
>
> Thanks.
>
> --Srinivas
>
>
> -----Original Message-----
> From: Rami Rosen [mailto:roszenrami at gmail.com]
> Sent: Saturday, October 20, 2018 8:55 AM
> To: Srinivas Kotamarti <srinivas_kotamarti at affirmednetworks.com>
> Cc: users at dpdk.org
> Subject: Re: [dpdk-users] issues with dpdk-pdump
>
> Hi Srinivas,
> Can you do a simple test to make sure your setup is ok ?
> The suggested test is: instead od running you app, simply do the following:
> First, bind a port to DPDK.
> Then, run testpmd app (build/app/testpmd; as testpmd calls
> rte_pdump_init() if RTE_LIBRTE_PDUMP is defined).
> Then run again the dpdk-pdump, and post the result.
> For me this works with dpdk-18.08, I am aware that you are using an older version, but it should work also for you I believe.
> Please note that I had to remove the while spaces in the command, but anyhow these whitespaces did not cause the error about rte_eal_config_attach() you had, but a parsing error; after I removed these whitespaces, everything was fine, no errors; what I actually ran is; ./build/app/dpdk-pdump --  --pdump 'port=0,queue=*,rx-dev=/work/rc.pcapng,tx-dev=/work/tc.pcapng'
>
> and the log shows:
>
> EAL: Detected 4 lcore(s)
> EAL: Detected 1 NUMA nodes
> EAL: Multi-process socket /var/run/dpdk/rte/mp_socket_15023_7a49f061eb1
> EAL: Probing VFIO support...
> EAL: PCI device 0000:01:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 8086:1584 net_i40e
> EAL: PCI device 0000:02:00.0 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 8086:1528 net_ixgbe
> EAL: PCI device 0000:02:00.1 on NUMA socket -1
> EAL:   Invalid NUMA socket, default to 0
> EAL:   probe driver: 8086:1528 net_ixgbe
> Port 2 MAC: 00 00 00 01 02 03
> Port 3 MAC: 00 00 00 01 02 03
>
>
> Regards,
> Rami Rosen
> http://ramirose.wixsite.com/ramirosen


More information about the users mailing list