Install DOCA on bluefield 2 failed

Hi, I follow the DOCA SDK DOCUMENTATION(v1.4.0)to install software for my BlueFiled 2 DPU (MBF2M516A-EEEO).
But I have some problems at step 3.3.1 Installing Full DOCA Image On DPU, following is the logs
1. the bfd-install failed to execute with “cat: write error: Connection timed out”

changzihao@dl28:~$ sudo bfb-install --bfb ./DOCA_1.4.0_BSP_3.9.2_Ubuntu_20.04-4.signed.bfb --config bf.cfg --rshim rshim0
 [sudo] password for changzihao: 
Pushing bfb + cfg 
cat: write error: Connection timed out        ]  
832KiB 0:01:50 [7.51KiB/s] [         <=>          ] 
Failed to push BFB

2. rshim console shows “Memory Device: 0 BIST Failed” and “DDR BIST POST failed!”

changzihao@dl28:~$ sudo cat /dev/rshim0/console 115200
Mellanox BlueField-2 A1 BL1 V1.1
NOTICE:  No CDI passed to Riot core!
NOTICE:  BL2R: v2.2(release):3.9.2-3-gacd025e
NOTICE:  BL2R: Built : 00:31:25, Jul 25 2022
NOTICE:  BL2R built for hw (ver 1)
NOTICE:  BL2R: Booting BL2
NOTICE:  BL2: v2.2(release):3.9.2-3-gacd025e
NOTICE:  BL2: Built : 00:31:25, Jul 25 2022
NOTICE:  BL2 built for hw (ver 1)
NOTICE:  Running as MBF2M516A-EEEO system
NOTICE:  No SPD detected on MSS0 DIMM0
NOTICE:  No SPD detected on MSS0 DIMM1
NOTICE:  Finished initializing DDR
  Memory Device: 0 BIST Failed
 -----------------------------
 Error Address = 0x799ca1
 Error Row = 0xf339
 Error Column = 0x21
 Error Bank = 0x1 :
 Error Physical Rank = 0x0 :
 Error Logical Rank = 0x0 :
 Error Chunk = 0x0 :
 Number of errors = 0x138 :
 SRAM entry 0 = 0xad :

     Expected Data:
   0xff   0x20   0x20   0xdf   0xff   0x40   0x40   0xbf   0xff   0x80   0x80   0x7f   0xfe   0x00  [0x00]  0xfe   0xfd   0x02
   0x02   0xfd   0xfb   0x04   0x04   0xfb   0xf7   0x08   0x08   0xf7   0xef   0x10   0x10   0xef   0xdf   0x20   0x20   0xdf

     Error Data:
   0xff   0x20   0x20   0xdf   0xff   0x40   0x40   0xbf   0xff   0x80   0x80   0x7f   0xfe   0x00  [0x01]  0xfe   0xfd   0x02
   0x02   0xfd   0xfb   0x04   0x04   0xfb   0xf7   0x08   0x08   0xf7   0xef   0x10   0x10   0xef   0xdf   0x20   0x20   0xdf
Byte 14: Expected 0x0(0) Received 0x1(1)
  DQ 24 (bit 112)  Second Rising  (0 --> 1)
ERROR:   DDR BIST POST failed!
ERROR:   Verify DDR

3. my host os is ubuntu 20.04 with Linux 5.4.0-26-generic kernel

changzihao@dl28:~$ lsb_release -a
No LSB modules are available.
Distributor ID:	Ubuntu
Description:	Ubuntu 20.04.1 LTS
Release:	20.04
Codename:	focal
changzihao@dl28:~$ uname -a
Linux dl28 5.4.0-26-generic #30-Ubuntu SMP Mon Apr 20 16:58:30 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

Thank you for your reply.

Could you try POWER CYCLE BF2 AND SERVER, then re-burn by bfb-install.

Thanks a lot.
I have rebooted the host and even put the BF2 on other hosts but still getting the same error, I think maybe there is something wrong with my BF2.

Does BF2 can boot normally? Or just can’t burn new DOCA image?

I think the BF2 do not work normally, after setting ip to tmfifo_net0, I can not ssh to BF2

OK, please check below,

cat /dev/rshim<>/misc

If not “DPU is ready”, you can open support ticket from ESP portal for RMA,

https://enterprise-support.nvidia.com/s/

Ok, thank you

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