BlueField-2 DPU can't find representor

I tried to test DOCA file compression application, by DPU ARM as server and host as client. However, it cannot find proper representor like below

  1. DPU
$ lspci
00:00.0 PCI bridge: Mellanox Technologies MT42822 BlueField-2 SoC Crypto enabled (rev 01)
01:00.0 PCI bridge: Mellanox Technologies MT42822 Family [BlueField-2 SoC PCIe Bridge] (rev 01)
02:00.0 PCI bridge: Mellanox Technologies MT42822 Family [BlueField-2 SoC PCIe Bridge] (rev 01)
03:00.0 Ethernet controller: Mellanox Technologies MT42822 BlueField-2 integrated ConnectX-6 Dx network controller (rev 01)
03:00.1 Ethernet controller: Mellanox Technologies MT42822 BlueField-2 integrated ConnectX-6 Dx network controller (rev 01)
$ sudo lshw -c network -businfo
Bus info          Device          Class      Description
========================================================
pci@0000:03:00.0  p0              network    MT42822 BlueField-2 integrated ConnectX-6 Dx network controller
pci@0000:03:00.1  p1              network    MT42822 BlueField-2 integrated ConnectX-6 Dx network controller
virtio@1          tmfifo_net0     network    Ethernet interface
                  oob_net0        network    Ethernet interface
pci@0000:03:00.0  pf0hpf          network    Ethernet interface
pci@0000:03:00.1  pf1hpf          network    Ethernet interface
pci@0000:03:00.0  en3f0pf0sf0     network    Ethernet interface
pci@0000:03:00.0  enp3s0f0s0      network    Ethernet interface
pci@0000:03:00.1  en3f1pf1sf0     network    Ethernet interface
pci@0000:03:00.1  enp3s0f1s0      network    Ethernet interface

$ /tmp/build/file_compression/doca_file_compression -p 03:00.0 -r 40:00.0 -f result.gz
[02:45:17:995504][1506792][DOCA][WRN][common.c:297][open_doca_device_rep_with_pci] Matching device not found
[02:45:17:995541][1506792][DOCA][ERR][comch_utils.c:390][comch_utils_fast_path_init] Failed to open Comm Channel DOCA device representor based on PCI address: Requested Resource Not Found
[02:45:17:995586][1506792][DOCA][ERR][file_compression.c:111][main] Failed to initialize a comch: Requested Resource Not Found

$ /tmp/build/file_compression/doca_file_compression -p 03:00.0 -r 03:00.0 -f result.gz
[01:37:28:926366][1480122][DOCA][INF][comch_utils.c:464][comch_utils_fast_path_init] Server waiting on a client to connect
  1. Host
$ lspci | grep Mell
3f:00.0 Ethernet controller: Mellanox Technologies MT28908 Family [ConnectX-6]
3f:00.1 Ethernet controller: Mellanox Technologies MT28908 Family [ConnectX-6]
40:00.0 Ethernet controller: Mellanox Technologies MT42822 BlueField-2 integrated ConnectX-6 Dx network controller (rev 01)
40:00.1 Ethernet controller: Mellanox Technologies MT42822 BlueField-2 integrated ConnectX-6 Dx network controller (rev 01)
40:00.2 DMA controller: Mellanox Technologies MT42822 BlueField-2 SoC Management Interface (rev 01)
$ sudo lshw -c network -businfo
Bus info          Device        Class          Description
==========================================================
pci@0000:03:00.0  enp3s0f0      network        Ethernet Controller X710 for 10GBASE-T
pci@0000:03:00.1  enp3s0f1      network        Ethernet Controller X710 for 10GBASE-T
pci@0000:3f:00.0  enp63s0f0np0  network        MT28908 Family [ConnectX-6]
pci@0000:3f:00.1  enp63s0f1np1  network        MT28908 Family [ConnectX-6]
pci@0000:40:00.0  enp64s0f0np0  network        MT42822 BlueField-2 integrated ConnectX-6 Dx network controller
pci@0000:40:00.1  enp64s0f1np1  network        MT42822 BlueField-2 integrated ConnectX-6 Dx network controller

$ /tmp/build/file_compression/doca_file_compression -p 40:00.0 -f benchmark/datasets/fpsingle/msg_sp.sp
[10:38:20:676733][139058][DOCA][WRN][doca_compress.cpp:191][doca_compress_cap_task_compress_deflate_is_supported] devinfo 0x556557a67530: Failed to check if compress_deflate_task is supported: compress_deflate is not supported by the device
[10:38:20:680540][139058][DOCA][WRN][doca_compress.cpp:191][doca_compress_cap_task_compress_deflate_is_supported] devinfo 0x556557a6c820: Failed to check if compress_deflate_task is supported: compress_deflate is not supported by the device
[10:38:20:994800][139058][DOCA][ERR][linux_devx_obj.cpp:73][priv_doca_devx_object] Failed to create devx object with syndrome=0xe5300
[10:38:20:994863][139058][DOCA][ERR][doca_dev.cpp:2184][priv_doca_dev_devx_object_create] Failed to create devx object: failed to allocate devx object wrapper with exception:
[10:38:20:994877][139058][DOCA][ERR][doca_dev.cpp:2184][priv_doca_dev_devx_object_create] DOCA exception [DOCA_ERROR_DRIVER] with message DEVX object failed to create
[10:38:20:994886][139058][DOCA][ERR][cc_devx_2.cpp:265][cc_devx_2_register_channel_connection] Failed to create channel connection object with error DOCA_ERROR_DRIVER
[10:38:20:994890][139058][DOCA][ERR][qp_channel_2.cpp:993][cc_2_qp_register_client] client registration failed for send side
[10:38:20:994903][139058][DOCA][ERR][doca_comm_channel_2.cpp:847][doca_comm_channel_2_ep_client_connect] client registration failed for doca_comm_channel_2_ep_client_connect()
[10:38:20:994907][139058][DOCA][ERR][doca_comch_pe.cpp:393][priv_doca_comch_pe_start_cb] failed to connect on client with error = DOCA_ERROR_CONNECTION_ABORTED
[10:38:21:000534][139058][DOCA][ERR][doca_pe.cpp:1092][start_context] Progress engine 0x556557a6e740: Failed to start context=0x556557a70910. err=DOCA_ERROR_CONNECTION_ABORTED
[10:38:21:000542][139058][DOCA][ERR][comch_utils.c:535][comch_utils_fast_path_init] Failed to start comch client context: Connection aborted
[10:38:21:000555][139058][DOCA][ERR][file_compression.c:111][main] Failed to initialize a comch: Connection aborted

It is DPU mode and I tested on both DOCA 2.9.2 LTS and 2.10.0.

Device #1:
----------

  Device Type:      BlueField2
  Part Number:      MBF2H516A-EEEO_Ax_Bx
  Description:      BlueField-2 DPU 100GbE/EDR/HDR100 VPI Dual-Port QSFP56; PCIe Gen4 x16; Crypto Enabled; 16GB on-board DDR; 1GbE OOB management; FHHL
  PSID:             MT_0000000704
  PCI Device Name:  0000:40:00.0
  Base MAC:         b8cef6d13e9c
  Versions:         Current        Available
     FW             24.43.2566     N/A
     FW (Running)   24.44.1036     N/A
     PXE            3.7.0500       N/A
     UEFI           14.37.0014     N/A
     UEFI Virtio blk   22.4.0014      N/A
     UEFI Virtio net   21.4.0013      N/A

  Status:           No matching image found
  1. Should I make a new SF for comch?
  2. Should I change status from ETH to Infiniband?

Hi, I meet the same error and solve it due to the wrong representor PCI address. Can you use the command " /opt/mellanox/doca/tools/doca_caps --list-rep-devs" instead to find the representor PCI address in the DPU side? You can find the documents here NVIDIA DOCA Capabilities Print Tool - NVIDIA Docs. For other versions of DOCA, it should be similar.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.