NIC FW Error - Timeout Issues

Preface

In NVIDIA® BlueField®-3 NIC mode, the Arm cores are put to sleep during UEFI boot. The Arm does not boot an OS except during BFB installation. In this mode, most of the DRAM (normally allocated to the OS in DPU mode) is reassigned to NIC firmware. This memory region is known as static ICM.

Note

The information in this section applies only to BlueField-3 NIC mode.

2. Command Cheat Sheet

Command	Description
`obmc-console-client`	Access the BlueField console via the DPU BMC
`echo "DISPLAY_LEVEL 2" > /dev/rshim0/misc`	Set RShim debug log level to 2
`cat /dev/rhim0/misc`	Dump the RShim log
`for i in {1..3}; do mstdump -full /dev/mst/mt41692_pciconf0 > mstdump_"$i".log; done`	Get three full NIC firmware mstdumps
`echo "SW_RESET 1" > /dev/rshim0/misc`	Trigger a software reset of the Arm cores

3. Logging and Counters

There are no hardware counters available for NIC mode debugging on the Arm side.

For troubleshooting, always collect the RShim log using the command cat /dev/rhim0/misc. It provides the primary source of diagnostic information in NIC mode.

4. Debug Info Package

N/A

5. Scenarios

TBD

On This Page