[E1000-devel] Hardware Unit Hang, followed by Reset adapter on e1000e - 82583V

[Padam J. S. <pad...@in...>]

Hello,

I am using this appliance as a router, and am facing this in the kernel logs:


e1000e 0000:01:00.0: eth0: Detected Hardware Unit Hang:
  TDH                  <458>
  TDT                  <173>
  next_to_use          <173>
  next_to_clean        <457>
buffer_info[next_to_clean]:
  time_stamp           <10030f499>
  next_to_watch        <45a>
  jiffies              <10030f538>
  next_to_watch.status <0>
MAC Status             <80783>
PHY Status             <796d>
PHY 1000BASE-T Status  <3800>
PHY Extended Status    <3000>
PCI Status             <10>
e1000e 0000:01:00.0: eth0: Detected Hardware Unit Hang:
  TDH                  <458>
  TDT                  <173>
  next_to_use          <173>
  next_to_clean        <457>
buffer_info[next_to_clean]:
  time_stamp           <10030f499>
  next_to_watch        <45a>
  jiffies              <10030f600>
  next_to_watch.status <0>
MAC Status             <80783>
PHY Status             <796d>
PHY 1000BASE-T Status  <3800>
PHY Extended Status    <3000>
PCI Status             <10>
e1000e 0000:01:00.0: eth0: Detected Hardware Unit Hang:
  TDH                  <458>
  TDT                  <173>
  next_to_use          <173>
  next_to_clean        <457>
buffer_info[next_to_clean]:
  time_stamp           <10030f499>
  next_to_watch        <45a>
  jiffies              <10030f6c8>
  next_to_watch.status <0>
MAC Status             <80783>
PHY Status             <796d>
PHY 1000BASE-T Status  <3800>
PHY Extended Status    <3000>
PCI Status             <10>
e1000e 0000:01:00.0: eth0: Detected Hardware Unit Hang:
  TDH                  <458>
  TDT                  <173>
  next_to_use          <173>
  next_to_clean        <457>
buffer_info[next_to_clean]:
  time_stamp           <10030f499>
  next_to_watch        <45a>
  jiffies              <10030f790>
  next_to_watch.status <0>
MAC Status             <80783>
PHY Status             <796d>
PHY 1000BASE-T Status  <3800>
PHY Extended Status    <3000>
PCI Status             <10>
e1000e 0000:01:00.0: eth0: Reset adapter
e1000e 0000:01:00.0: eth0: Detected Hardware Unit Hang:
  TDH                  <0>
  TDT                  <0>
  next_to_use          <0>
  next_to_clean        <0>
buffer_info[next_to_clean]:
  time_stamp           <0>
  next_to_watch        <0>
  jiffies              <10030f864>
  next_to_watch.status <0>
MAC Status             <80780>
PHY Status             <7949>
PHY 1000BASE-T Status  <0>
PHY Extended Status    <3000>
PCI Status             <10>
e1000e: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None
e1000e 0000:01:00.0: eth0: Reset adapter
e1000e: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: None



Kernel version is 2.6.34 and driver version is 2.0.0-NAPI

The kernel parameters are:

ro root=/dev/sda2 console=ttyS0,115200n8 console=tty1 ide=nodma pcie_aspm=off loglevel=0 acpi=off

The adapter resets itself, and the frequency with which this happens increases proportionally to the device's uptime.

I tried the Kernel 3.8 with latest e1000e driver from sourceforge, but as soon as the network traffic starts on the device it reboots.

Could this be a faulty Ethernet chip or is a driver issue?

Thanks,
Padam