PSOD due to parity errors in the qfle3 driver version: 1.4.35.0

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms:

PSOD with the following backtrace,

#PF Exception 14 in world 2097436:HELPER_UPLIN IP 0x4200200888c4 addr 0xa
Backtrace:
  0x453948e1b8a0:[0x4200200888c4]qfle3_sp_helper_func@(qfle3)#<None>+0x124 stack: 0x8000000041000080, 0x383522d2d1fb8, 0x0, 0x0, 0x42001f3aed50
  0x453948e1b940:[0x4200200ed8b3]ecore_state_wait_enhanced@(qfle3)#<None>+0x168 stack: 0x0, 0x4313b5378110, 0x0, 0x0, 0x0
  0x453948e1b9c0:[0x4200200ea9eb]ecore_config_vlan_mac@(qfle3)#<None>+0x190 stack: 0x2726252827262528, 0x4313b5378110, 0x453948e1bc40, 0x1, 0x453948e1bb18
  0x453948e1ba10:[0x42002006f368]qfle3_set_mac_one@(qfle3)#<None>+0x149 stack: 0x25, 0x42001f3a8e3e, 0x4313b4d7caf8, 0x3834ae85fcc7a, 0x420041405610
  0x453948e1bac0:[0x4200200949a1]qfle3_apply_queue_mac_filter@(qfle3)#<None>+0x3e2 stack: 0x453948e1bc40, 0x453948e1bc40, 0x4313b5382c90, 0x9640, 0x453948e1be48
  0x453948e1bb60:[0x420020095807]qfle3_queue_apply_filter@(qfle3)#<None>+0x24c stack: 0x0, 0x4307ae03e940, 0x453948e1bc50, 0x4307ae04c8a0, 0x453948e1be48
  0x453948e1bbb0:[0x42001f36df79]UplinkNetq_ApplyHwRxFilter@vmkernel#nover+0x8a stack: 0x453948e10035, 0x4200200955bc, 0x42001f36df5e, 0x0, 0x58c
  0x453948e1bc00:[0x42001f249eb5]NetqueueBalActivatePendingFilters@vmkernel#nover+0x17e stack: 0x4307b8a46df0, 0x453948e1bce8, 0x4307b8a46c10, 0x4307040000e8, 0x453948e1bce8
  0x453948e1bd40:[0x42001f24c7ed]NetqueueBalRxQueueCommitChanges@vmkernel#nover+0x56 stack: 0xff, 0x4307ae03e940, 0x8, 0x0, 0x0
  0x453948e1bd80:[0x42001f25104f]UplinkNetqueueBal_BalanceCB@vmkernel#nover+0x1d50 stack: 0x4307b8a46c58, 0x0, 0x43063aa41e00, 0x4306e9606d40, 0x4307ae03ed90
  0x453948e1bf10:[0x42001f35a6c4]UplinkAsyncProcessCallsHelperCB@vmkernel#nover+0x119 stack: 0x3834d505210f4, 0x430228401220, 0x430a2ca01750, 0x42001f0da29a, 0x6
  0x453948e1bf40:[0x42001f0da299]HelperQueueFunc@vmkernel#nover+0x1d2 stack: 0x453948e20b48, 0x430a2ca01238, 0x453948e1f000, 0x453948e1f000, 0x430a2ca012c0
  0x453948e1bfe0:[0x42001f3b3ad9]CpuSched_StartWorld@vmkernel#nover+0x86 stack: 0x0, 0x42001f0c4d80, 0x0, 0x0, 0x0
  0x453948e1c000:[0x42001f0c4d7f]Debug_IsInitialized@vmkernel#nover+0xc stack: 0x0, 0x0, 0x0, 0x0, 0x0

 

Logs Snippets in vmkernel.log

2023-06-15T04:27:10.354Z cpu1:2098258)WARNING: qfle3: qfle3_parity_attn:16971: [vmnic4] Parity errors detected in blocks
2023-06-15T04:27:10.354Z cpu1:2098258)WARNING: qfle3: _print_next_block:16602: XSEMI
2023-06-15T04:27:10.354Z cpu1:2098258)WARNING: qfle3: _print_parity:16596:  [0x00000000]
2023-06-15T04:27:10.354Z cpu1:2098258)WARNING: qfle3: _print_parity:16596:  [0x00000020]
 

Environment

VMware vSphere ESXi 7.0.x

Cause

Parity error is for VFC block.

Resolution

Marvell has a recovery mechanism implemented in the qfle3 driver to correct parity errors (when possible). If these parity errors are still being observed after driver/firmware upgrades it may be a result of hardware degradation. If these are isolated incidents, contact the hardware vendor to have the adapter replaced.