PSOD crash on ESXi Host - PANIC - qlnativefcDmaAlloc Corruption in dlmalloc - QLogic HBA Adapters

Products

• VMware vSphere ESXi

Issue/Introduction

• PSOD backtrace reported as below 
  
  

  Panic_NoSave@vmkernel
  DLM_malloc@vmkernel#nover+0xe86 
  Heap_AlignWithTimeoutAndRA@vmkernel#nover+
  qlnativefcDmaAlloc@(qlnativefc)#
  qlnativefcStartFcNvmeLsReq@(qlnativefc)#
  qlnativefcNvmeResetAdminQueue@(qlnativefc)#
  [e-mail]#
  [e-mail]#
  [e-mail]#
  [e-mail]#
  vmkWorldFunc@vmkernel#
  CpuSched_StartWorld@vmkernel#
  Debug_IsInitialized@vmkernel#
  VMware ESXi 8.0.3 [Releasebuild-25067014 x86_64]ESC[0m
  PANIC bora/vmkernel/main/dlmalloc.c:4774 - Corruption in dlmalloc

   

• VMkernel logs may also report - I/O Aborts and Disconnects reported qlnativefc driver  
  
  

  YYYY-MM-DD:TT:MM:SS.981Z cpu288:2100872)qlnativefc: vmhba8(91:0.0): qlnativefcNvmeAbortAllCmds:140:Enhanced Abort I/O not returned!
  YYYY-MM-DD:TT:MM:SS.981Z cpu288:2100872)qlnativefc: vmhba8(91:0.0): qlnativefcNvmeResetAdminQueue:337:Abort NVMe commands failed
  YYYY-MM-DD:TT:MM:SS.981Z cpu288:2100872)qlnativefc: vmhba8(91:0.0): qlnativefcNvmeResetAdminQueue:348:Disconnect on nqn.2014-08.org.nvmexpress.discovery

   

• Hardware specs 
  
  QLogic Corp QLE2700/QLE2800 32/64G SP/DP Fibre Channel Adapter

Environment

• ESXi 8.x

Cause

• This is a known issue caused by the driver reporting a Memory Corruption and triggering a PSOD on the ESXi Host

Resolution

• Upgrade the driver to the latest version - 5.4.86.0 
  
  
• File name - MRVL-QLogic-FC_5.4.86.0-1OEM.803.0.0.24022510-package.zip
  
  
• Driver Download - Link