#429 82579V Detected Hardware Unit Hang

open
None
in-kernel_driver
1
2015-08-20
2014-08-05
Volker Braun
No

I'm running a NAT with ports

  • WAN (em1): 82579V, driver e1000e version 2.3.2-k
  • LAN (p6p1,p6p2): 82546GB, driver e1000 version 7.3.21-k8-NAPI

The WAN port (em1) resets with "Detected Hardware Unit Hang" when I transfer large (about 1GB) files from a LAN connected machine through the NAT server. I can easily reproduce the failure, happens in less than a minute. Disabling EEE (ethtool -set-eee em1 off) does not help.

Workarounds:

  • Transferring the file directly from the NAT server to the WAN works
  • Disabling TSO works (ethtool -K em1 tso off)
  • Reducing speed works (ethtool -s em1 speed 100 duplex full)

Discussion

  • Volker Braun
    Volker Braun
    2014-08-05

    full syslog output:

    Aug 06 00:17:32 desktop kernel: e1000e 0000:00:19.0 em1: Detected Hardware Unit Hang:
                                  TDH                  <f2>
                                  TDT                  <2>
                                  next_to_use          <2>
                                  next_to_clean        <f0>
                                buffer_info[next_to_clean]:
                                  time_stamp           <107afb31b>
                                  next_to_watch        <f2>
                                  jiffies              <107afb918>
                                  next_to_watch.status <0>
                                MAC Status             <40080083>
                                PHY Status             <796d>
                                PHY 1000BASE-T Status  <3c00>
                                PHY Extended Status    <3000>
                                PCI Status             <10>
    Aug 06 00:17:34 desktop kernel: e1000e 0000:00:19.0 em1: Detected Hardware Unit Hang:
                                  TDH                  <f2>
                                  TDT                  <2>
                                  next_to_use          <2>
                                  next_to_clean        <f0>
                                buffer_info[next_to_clean]:
                                  time_stamp           <107afb31b>
                                  next_to_watch        <f2>
                                  jiffies              <107afc0e8>
                                  next_to_watch.status <0>
                                MAC Status             <40080083>
                                PHY Status             <796d>
                                PHY 1000BASE-T Status  <3c00>
                                PHY Extended Status    <3000>
                                PCI Status             <10>
    Aug 06 00:17:36 desktop kernel: e1000e 0000:00:19.0 em1: Detected Hardware Unit Hang:
                                  TDH                  <f2>
                                  TDT                  <2>
                                  next_to_use          <2>
                                  next_to_clean        <f0>
                                buffer_info[next_to_clean]:
                                  time_stamp           <107afb31b>
                                  next_to_watch        <f2>
                                  jiffies              <107afc8b8>
                                  next_to_watch.status <0>
                                MAC Status             <40080083>
                                PHY Status             <796d>
                                PHY 1000BASE-T Status  <3c00>
                                PHY Extended Status    <3000>
                                PCI Status             <10>
    Aug 06 00:17:38 desktop kernel: e1000e 0000:00:19.0 em1: Detected Hardware Unit Hang:
                                  TDH                  <f2>
                                  TDT                  <2>
                                  next_to_use          <2>
                                  next_to_clean        <f0>
                                buffer_info[next_to_clean]:
                                  time_stamp           <107afb31b>
                                  next_to_watch        <f2>
                                  jiffies              <107afd088>
                                  next_to_watch.status <0>
                                MAC Status             <40080083>
                                PHY Status             <796d>
                                PHY 1000BASE-T Status  <3c00>
                                PHY Extended Status    <3000>
                                PCI Status             <10>
    Aug 06 00:17:39 desktop kernel: e1000e 0000:00:19.0 em1: Reset adapter unexpectedly
    
     
  • Todd Fujinaka
    Todd Fujinaka
    2015-05-12

    • assigned_to: Yanir Lubetkin
     
  • Todd Fujinaka
    Todd Fujinaka
    2015-08-20

    • assigned_to: Yanir Lubetkin --> Raanan Avargil