[dpdk-dev] [dpdk-users] DPDK issue with Marvell NIC QLogic Corp. FastLinQ QL41000

Varghese, Vipin vipin.varghese at intel.com
Wed May 26 11:57:06 CEST 2021
Previous message (by thread): [dpdk-dev] [dpdk-users] DPDK issue with Marvell NIC QLogic Corp. FastLinQ QL41000
Next message (by thread): [dpdk-dev] [PATCH] net/vmxnet3: add interrupt support in MSI-X
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Based on the logs shared from `Turing Team`, the observation is as follows

1. 1GB huge pages are used in the application

```logs
12:46:11.792 §EAL: No free hugepages reported in hugepages-2048kB
12:46:11.792 §EAL: No available hugepages reported in hugepages-2048kB
```

2. Issue comes right at `device porbe to initialize`

```logs
12:46:46.382 §EAL: Probe PCI driver: net_qede (1077:8070) device: 0000:65:00.0 (socket 0)
12:46:46.403 §[QEDE PMD: ()]ecore_load_mcp_offsets:The address of the MCP scratch-pad is not configured
12:46:46.403 §[QEDE PMD: ()]ecore_mcp_cmd_init:MCP is not initialized
12:46:46.404 §[QEDE PMD: ()]ecore_mcp_cmd_and_union:MFW is not initialized!
12:46:46.404 §[QEDE PMD: ()]ecore_hw_get_nvm_info:Shared memory not initialized
12:46:46.404 §[QEDE PMD: ()]ecore_hw_prepare_single:Failed to get HW information
12:46:46.404 §[qed_probe:74(65:00.0:dpdk-port-0)]hw prepare failed
12:46:46.404 §[qede_common_dev_init:2566(65:00.0:dpdk-port-> 0)]qede probe failed rc -3
```

Hence in my opinion this is related to NIC in use and how it is getting initialized. So please try to debug driver 

> -----Original Message-----
> From: dev <dev-bounces at dpdk.org> On Behalf Of Truring Team
> Sent: Wednesday, May 26, 2021 10:12 AM
> To: Nishant Verma <vnish11 at gmail.com>
> Cc: dev at dpdk.org; USERS <users at dpdk.org>
> Subject: Re: [dpdk-dev] [dpdk-users] DPDK issue with Marvell NIC QLogic
> Corp. FastLinQ QL41000
> 
> Yes , I have configured it via grub.
> 
> Grub:
> "default_hugepagesz=1G hugepagesz=1G hugepages=2"
> grub2-mkconfig -o /boot/grub2/grub.cfg
> 
> Manual:
> mkdir /dev/hugepages1G
> mount -t hugetlbfs -o pagesize=1G none /dev/hugepages1G echo 2 >
> /sys/devices/system/node/node1/hugepages/hugepages-
> 1048576kB/nr_hugepages
> 
> Regards
> Puneet
> 
> On Wed, 26 May 2021 at 10:05, Nishant Verma <vnish11 at gmail.com> wrote:
> 
> > meminfo states: Hugepagesize:    1048576 kB
> > But  in your initial mail you state Hugepage size is 1G.
> >
> > have you configured HugePage via grub.... ?
> >
> >
> > Thanks.
> >
> > Regards,
> > NV
> >
> > On Wed, May 26, 2021 at 12:21 AM Truring Team <truring12 at gmail.com>
> wrote:
> >
> >> Hi Nishant,
> >>
> >> cat /proc/meminfo
> >> 11:59:29.200 §MemTotal:       23101672 kB
> >> 11:59:29.200 §MemFree:        15254764 kB
> >> 11:59:29.200 §MemAvailable:   19243184 kB
> >> 11:59:29.200 §Buffers:            6792 kB
> >> 11:59:29.201 §Cached:          4084844 kB
> >> 11:59:29.201 §SwapCached:            0 kB
> >> 11:59:29.201 §Active:          3609016 kB
> >> 11:59:29.201 §Inactive:        1175160 kB
> >> 11:59:29.201 §Active(anon):     695644 kB
> >> 11:59:29.202 §Inactive(anon):    19716 kB
> >> 11:59:29.202 §Active(file):    2913372 kB
> >> 11:59:29.202 §Inactive(file):  1155444 kB
> >> 11:59:29.202 §Unevictable:           0 kB
> >> 11:59:29.202 §Mlocked:               0 kB
> >> 11:59:29.203 §SwapTotal:      11665404 kB
> >> 11:59:29.204 §SwapFree:       11665404 kB
> >> 11:59:29.204 §Dirty:                32 kB
> >> 11:59:29.204 §Writeback:             0 kB
> >> 11:59:29.204 §AnonPages:        694764 kB
> >> 11:59:29.205 §Mapped:           140220 kB
> >> 11:59:29.205 §Shmem:             22800 kB
> >> 11:59:29.205 §Slab:             445016 kB
> >> 11:59:29.205 §SReclaimable:     289968 kB
> >> 11:59:29.205 §SUnreclaim:       155048 kB
> >> 11:59:29.206 §KernelStack:       16624 kB
> >> 11:59:29.206 §PageTables:        35892 kB
> >> 11:59:29.206 §NFS_Unstable:          0 kB
> >> 11:59:29.206 §Bounce:                0 kB
> >> 11:59:29.206 §WritebackTmp:          0 kB
> >> 11:59:29.207 §CommitLimit:    22167664 kB
> >> 11:59:29.207 §Committed_AS:    4342668 kB
> >> 11:59:29.207 §VmallocTotal:   34359738367 kB
> >> 11:59:29.207 §VmallocUsed:      463188 kB
> >> 11:59:29.207 §VmallocChunk:   34358835196 kB
> >> 11:59:29.208 §HardwareCorrupted:     0 kB
> >> 11:59:29.208 §AnonHugePages:    329728 kB
> >> 11:59:29.208 §CmaTotal:              0 kB
> >> 11:59:29.209 §CmaFree:               0 kB
> >> 11:59:29.209 §HugePages_Total:    1024
> >> 11:59:29.209 §HugePages_Free:     1019
> >> 11:59:29.209 §HugePages_Rsvd:        0
> >> 11:59:29.209 §HugePages_Surp:        0
> >> 11:59:29.209 §Hugepagesize:       2048 kB
> >> 11:59:29.209 §DirectMap4k:      381760 kB
> >> 11:59:29.209 §DirectMap2M:    13973504 kB
> >> 11:59:29.215 §DirectMap1G:    11534336 kB
> >>
> >> Regards
> >>
> >>
> >> On Wed, 26 May 2021 at 09:43, Nishant Verma <vnish11 at gmail.com>
> wrote:
> >>
> >>> Can you paste output of /proc/meminfo ?
> >>>
> >>>
> >>>
> >>> Regards,
> >>> NV
> >>>
> >>> On Tue, May 25, 2021 at 9:52 PM Truring Team <truring12 at gmail.com>
> >>> wrote:
> >>>
> >>>> Hi Team,
> >>>>
> >>>> I am trying to run the dpdk test-pmd application (DPDK 20.11 and
> >>>> DPDK
> >>>> 21.05) with NIC QLogic Corp. FastLinQ QL41000 Series 10/25/40/50GbE
> >>>> Controller but facing following issues:
> >>>>
> >>>> If someone can look into the following logs and help me figure out
> >>>> the issue?
> >>>>
> >>>>
> >>>>
> >>>> 1. DPDK 20.11 ( Using 1 GB hugepages)
> >>>>
> >>>>    gdb ./dpdk-testpmd
> >>>>
> >>>> 12:45:59.347 §GNU gdb (GDB) Red Hat Enterprise Linux 7.6.1-114.el7
> >>>>
> >>>> 12:45:59.347 §Copyright (C) 2013 Free Software Foundation, Inc.
> >>>>
> >>>> 12:45:59.348 §License GPLv3+: GNU GPL version 3 or later
> >>>> http://gnu.org/licenses/gpl.html
> >>>>
> >>>> 12:45:59.349 §This is free software: you are free to change and
> >>>> redistribute it.
> >>>>
> >>>> 12:45:59.349 §There is NO WARRANTY, to the extent permitted by law.
> >>>> Type
> >>>> "show copying"
> >>>>
> >>>> 12:45:59.349 §and "show warranty" for details.
> >>>>
> >>>> 12:45:59.350 §This GDB was configured as "x86_64-redhat-linux-gnu".
> >>>>
> >>>> 12:45:59.351 §For bug reporting instructions, please see:
> >>>>
> >>>> 12:45:59.351 §http://www.gnu.org/software/gdb/bugs/...
> >>>>
> >>>> 12:45:59.415 §Reading symbols from
> >>>> /home/itwsadmin/dpdk-20.11/build/app/dpdk-testpmd...(no
> debugging
> >>>> symbols found)...done.
> >>>>
> >>>> 12:46:11.492 §(gdb) r -l 0-5 -n4 -- -i
> >>>>
> >>>> 12:46:11.493 §Starting program:
> >>>> /home/itwsadmin/dpdk-20.11/build/app/./dpdk-testpmd -l 0-5 -n4 --
> >>>> -i
> >>>>
> >>>> 12:46:11.717 §[Thread debugging using libthread_db enabled]
> >>>>
> >>>> 12:46:11.719 §Using host libthread_db library
> >>>> "/lib64/libthread_db.so.1".
> >>>>
> >>>> 12:46:11.768 §EAL: Detected 6 lcore(s)
> >>>>
> >>>> 12:46:11.769 §EAL: Detected 1 NUMA nodes
> >>>>
> >>>> 12:46:11.769 §[New Thread 0x7ffff6a71700 (LWP 23184)]
> >>>>
> >>>> 12:46:11.769 §EAL: Multi-process socket /var/run/dpdk/rte/mp_socket
> >>>>
> >>>> 12:46:11.770 §[New Thread 0x7ffff6270700 (LWP 23185)]
> >>>>
> >>>> 12:46:11.792 §EAL: Selected IOVA mode 'PA'
> >>>>
> >>>> 12:46:11.792 §EAL: No free hugepages reported in hugepages-2048kB
> >>>>
> >>>> 12:46:11.792 §EAL: No available hugepages reported in
> >>>> hugepages-2048kB
> >>>>
> >>>> 12:46:11.854 §EAL: Probing VFIO support...
> >>>>
> >>>> 12:46:11.855 §[New Thread 0x7ffff5a6f700 (LWP 23186)]
> >>>>
> >>>> 12:46:11.855 §[New Thread 0x7ffff526e700 (LWP 23187)]
> >>>>
> >>>> 12:46:11.855 §[New Thread 0x7ffff4a6d700 (LWP 23188)]
> >>>>
> >>>> 12:46:11.855 §[New Thread 0x7ffff426c700 (LWP 23189)]
> >>>>
> >>>> 12:46:11.856 §[New Thread 0x7ffff3a6b700 (LWP 23190)]
> >>>>
> >>>> 12:46:11.896 §EAL: Probe PCI driver: net_qede (1077:8070) device:
> >>>> 0000:65:00.0 (socket 0)
> >>>>
> >>>> 12:46:11.915 §
> >>>>
> >>>> 12:46:11.915 §Program received signal SIGSEGV, Segmentation fault.
> >>>>
> >>>> 12:46:11.916 §0x0000000000ec5b20 in ecore_fw_overlay_mem_alloc ()
> >>>>
> >>>> 12:46:11.916 §Missing separate debuginfos, use: debuginfo-install
> >>>> glibc-2.17-324.el7_9.x86_64 libgcc-4.8.5-44.el7.x86_64
> >>>> libpcap-1.5.3-12.el7.x86_64 numactl-libs-2.0.12-5.el7.x86_64
> >>>> zlib-1.2.7-19.el7_9.x86_64
> >>>>
> >>>> 12:46:14.542 §(gdb) bt
> >>>>
> >>>> 12:46:14.548 §#0  0x0000000000ec5b20 in
> ecore_fw_overlay_mem_alloc
> >>>> ()
> >>>>
> >>>> 12:46:14.548 §#1  0x0000000000eb811e in ecore_hw_init ()
> >>>>
> >>>> 12:46:14.549 §#2  0x0000000000ef7397 in qed_slowpath_start ()
> >>>>
> >>>> 12:46:14.549 §#3  0x0000000000ef0483 in qede_common_dev_init ()
> >>>>
> >>>> 12:46:14.549 §#4  0x0000000000ef1113 in qede_eth_dev_pci_probe ()
> >>>>
> >>>> 12:46:14.549 §#5  0x0000000000a4450b in pci_probe ()
> >>>>
> >>>> 12:46:14.550 §#6  0x00000000009dbc6c in rte_bus_probe ()
> >>>>
> >>>> 12:46:14.550 §#7  0x00000000009fa91c in rte_eal_init ()
> >>>>
> >>>> 12:46:14.617 §#8  0x0000000000425db1 in main ()
> >>>>
> >>>> 12:46:17.052 §(gdb) quit
> >>>>
> >>>> 12:46:17.056 §A debugging session is active.
> >>>>
> >>>> 12:46:17.056 §
> >>>>
> >>>> 12:46:17.057 §        Inferior 1 [process 23180] will be killed.
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> 2. *DPDK 21.05:*
> >>>>
> >>>>    ./dpdk-testpmd -l 0-5 -n4 -- -i
> >>>>
> >>>> 12:46:46.257 §EAL: Detected 6 lcore(s)
> >>>>
> >>>> 12:46:46.257 §EAL: Detected 1 NUMA nodes
> >>>>
> >>>> 12:46:46.258 §EAL: Detected static linkage of DPDK
> >>>>
> >>>> 12:46:46.260 §EAL: Multi-process socket /var/run/dpdk/rte/mp_socket
> >>>>
> >>>> 12:46:46.285 §EAL: Selected IOVA mode 'PA'
> >>>>
> >>>> 12:46:46.285 §EAL: No free 2048 kB hugepages reported on node 0
> >>>>
> >>>> 12:46:46.288 §EAL: No available 2048 kB hugepages reported
> >>>>
> >>>> 12:46:46.382 §EAL: Probe PCI driver: net_qede (1077:8070) device:
> >>>> 0000:65:00.0 (socket 0)
> >>>>
> >>>> 12:46:46.403 §[QEDE PMD: ()]ecore_load_mcp_offsets:The address of
> >>>> the MCP scratch-pad is not configured
> >>>>
> >>>> 12:46:46.403 §[QEDE PMD: ()]ecore_mcp_cmd_init:MCP is not
> >>>> initialized
> >>>>
> >>>> 12:46:46.404 §[QEDE PMD: ()]ecore_mcp_cmd_and_union:MFW is not
> >>>> initialized!
> >>>>
> >>>> 12:46:46.404 §[QEDE PMD: ()]ecore_hw_get_nvm_info:Shared
> memory not
> >>>> initialized
> >>>>
> >>>> 12:46:46.404 §[QEDE PMD: ()]ecore_hw_prepare_single:Failed to get
> >>>> HW information
> >>>>
> >>>> 12:46:46.404 §[qed_probe:74(65:00.0:dpdk-port-0)]hw prepare failed
> >>>>
> >>>> 12:46:46.404 §[qede_common_dev_init:2566(65:00.0:dpdk-port-
> 0)]qede
> >>>> probe failed rc -3
> >>>>
> >>>> 12:46:46.406 §EAL: Requested device 0000:65:00.0 cannot be used
> >>>>
> >>>> 12:46:46.408 §TELEMETRY: No legacy callbacks, legacy socket not
> >>>> created
> >>>>
> >>>> 12:46:46.423 §testpmd: No probed ethernet devices
> >>>>
> >>>> 12:46:46.461 §Interactive-mode selected
> >>>>
> >>>> 12:46:46.519 §testpmd: create a new mbuf pool <mb_pool_0>:
> >>>> n=187456, size=2176, socket=0
> >>>>
> >>>> 12:46:46.522 §testpmd: preferred mempool ops selected: ring_mp_mc
> >>>>
> >>>> 12:46:46.540 §Done
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> System Information:
> >>>>
> >>>> *ethtool -i p2p1*
> >>>> driver: qede
> >>>> version: 8.33.0.20
> >>>> *firmware-version: mfw 8.24.44.0 storm 8.33.11.0*
> >>>> expansion-rom-version:
> >>>> bus-info: 0000:65:00.0
> >>>> supports-statistics: yes
> >>>> supports-test: yes
> >>>> supports-eeprom-access: no
> >>>> supports-register-dump: yes
> >>>> supports-priv-flags: yes
> >>>>
> >>>>
> >>>>
> >>>> *ethtool -i p2p2 *driver: qede
> >>>> version: 8.33.0.20
> >>>> firmware-version: mfw 8.24.44.0 storm 8.33.11.0
> >>>> expansion-rom-version:
> >>>> bus-info: 0000:65:00.1
> >>>> supports-statistics: yes
> >>>> supports-test: yes
> >>>> supports-eeprom-access: no
> >>>> supports-register-dump: yes
> >>>> supports-priv-flags: yes
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>> *lspci -nn | grep Ether*
> >>>>
> >>>> 12:42:50.579 §04:00.0 Ethernet controller [0200]: Broadcom Inc. and
> >>>> subsidiaries NetXtreme BCM5720 Gigabit Ethernet PCIe [14e4:165f]
> >>>>
> >>>> 12:42:50.581 §04:00.1 Ethernet controller [0200]: Broadcom Inc. and
> >>>> subsidiaries NetXtreme BCM5720 Gigabit Ethernet PCIe [14e4:165f]
> >>>>
> >>>> 12:42:50.582 §65:00.0 Ethernet controller [0200]: QLogic Corp.
> >>>> FastLinQ
> >>>> QL41000 Series 10/25/40/50GbE Controller [1077:8070] (rev 02)
> >>>>
> >>>> 12:42:50.583 §65:00.1 Ethernet controller [0200]: QLogic Corp.
> >>>> FastLinQ
> >>>> QL41000 Series 10/25/40/50GbE Controller [1077:8070] (rev 02)
> >>>>
> >>>>
> >>>>
> >>>> cd /lib/firmware/qed/
> >>>>
> >>>> ls
> >>>>
> >>>> 12:48:27.269 §qed_init_values-8.10.9.0.bin
> >>>> qed_init_values-8.30.12.0.bin  qed_init_values_zipped-8.10.5.0.bin
> >>>> qed_init_values_zipped-8.33.11.0.bin
> >>>> qed_init_values_zipped-8.7.3.0.bin
> >>>>
> >>>> 12:48:27.272 §qed_init_values-8.14.6.0.bin
> >>>> qed_init_values-8.33.12.0.bin  qed_init_values_zipped-8.15.3.0.bin
> >>>> qed_init_values_zipped-8.37.2.0.bin
> >>>>
> >>>> 12:48:27.273 §qed_init_values-8.18.9.0.bin
> >>>> qed_init_values-8.40.33.0.bin  qed_init_values_zipped-8.20.0.0.bin
> >>>> qed_init_values_zipped-8.37.7.0.bin
> >>>>
> >>>> 12:48:27.275 §qed_init_values-8.20.0.0.bin
> >>>> qed_init_values_zipped-8.10.10.0.bin
> >>>> qed_init_values_zipped-8.33.1.0.bin
> >>>> qed_init_values_zipped-8.4.2.0.bin
> >>>>
> >>>>
> >>>> Best Regards
> >>>>
> >>>> Puneet
> >>>>
> >>>
Previous message (by thread): [dpdk-dev] [dpdk-users] DPDK issue with Marvell NIC QLogic Corp. FastLinQ QL41000
Next message (by thread): [dpdk-dev] [PATCH] net/vmxnet3: add interrupt support in MSI-X
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the dev mailing list