PSOD in 6.5/6.7 with igbn driver

Products

• VMware vSphere ESXi

Issue/Introduction

Panic Message

@BlueScreen: #PF Exception 14 in world 2580078:vmnic5-0-tx IP ########### addr #####
Backtrace:
  0x451a5959bce0:[0x418011039133]igbn_XmitPacket@(igbn)#<None>+0xef stack: 0x0, 0x0, 0x0, 0x1e66341bba23ae, 0x45a2bcebc898
  0x451a5959be10:[0x41801104b6e2]indrv_UplinkTx@(igbn)#<None>+0x4b stack: 0x43048f2bc780, 0x43042fd15d40, 0x1, 0x418010656415, 0x0
  0x451a5959be40:[0x418010656414]UplinkDevTransmit@vmkernel#nover+0x4cd stack: 0x3f, 0x89, 0xb84f, 0x16de94, 0x451a5959bf20
  0x451a5959bf10:[0x41801117d384]NetSchedFIFORunLocked@(netsched_fifo)#<None>+0xcd stack: 0x0, 0x45a2bceaa2c0, 0x459ab36e3940, 0x4180000000c7, 0x0
  0x451a5959bf80:[0x41801117d534]NetSchedFIFORun@(netsched_fifo)#<None>+0x31 stack: 0x80000000, 0x4180106011b1, 0x0, 0x43042fd15d40, 0x451a49aa3100
  0x451a5959bfa0:[0x4180106011b0]NetSchedDevQueueWorld@vmkernel#nover+0x89 stack: 0x451a49aa3100, 0x451a595a3000, 0x451a49aa3100, 0x451a595a3100, 0x0
  0x451a5959bfe0:[0x418010709112]CpuSched_StartWorld@vmkernel#nover+0x77 stack: 0x0, 0x0, 0x0, 0x0, 0x0

Environment

VMware vSphere ESXi 6.5
VMware vSphere ESXi 6.7
VMware vSphere ESXi 6.0

Cause

The PSOD is because of a race between reset and data path. There are two issues here:

• Tx queue xmit and completion path ran into a race
• The above race caused queue stuck and uplink reset, which ran into the PSOD.

Resolution

Intel fixed the problem in driver igbn version 1.4.10 and later.

• Broadcom Compatibility Guide 

For instructions on how to install the driver, see Installing async drivers in ESXi using VMware Update Manager