Linux* Base Driver for the Intel(R) Ethernet 10 Gigabit PCI Express Family of
Adapters
=============================================================================

April 24, 2012

Contents
========

- Important Note
- In This Release
- Identifying Your Adapter
- Building and Installation
- Command Line Parameters
- Additional Configurations
- Performance Tuning
- Known Issues/Troubleshooting
- Support

IMPORTANT NOTE
==============

WARNING:  The ixgbe driver compiles by default with the LRO (Large Receive
Offload) feature enabled.  This option offers the lowest CPU utilization for
receives, but is completely incompatible with *routing/ip forwarding* and
*bridging*.  If enabling ip forwarding or bridging is a requirement, it is
necessary to disable LRO using compile time options as noted in the LRO
section later in this document.  The result of not disabling LRO when combined
with ip forwarding or bridging can be low throughput or even a kernel panic.

In This Release
===============

This file describes the ixgbe Linux* Base Driver for the 10 Gigabit PCI 
Express Family of Adapters.  This driver supports the 2.6.x and 3.x kernels, 
and includes support for any Linux supported system, including Itanium(R)2, 
x86_64, i686, and PPC.

This driver is only supported as a loadable module at this time.  Intel is
not supplying patches against the kernel source to allow for static linking
of the driver.  A version of the driver may already be included by your 
distribution and/or the kernel.org kernel. For questions related to hardware 
requirements, refer to the documentation supplied with your 10 Gigabit PCI
Express adapter.  All hardware requirements listed apply to use with Linux.

The following features are now available in supported kernels:
 - Native VLANs
 - Channel Bonding (teaming)
 - SNMP
 - Generic Receive Offload
 - Data Center Bridging

Channel Bonding documentation can be found in the Linux kernel source:
/Documentation/networking/bonding.txt

Instructions on updating ethtool can be found in the section "Additional
Configurations" later in this document.

Identifying Your Adapter
========================

The driver in this release is compatible with 82598, 82599 and X540-based Intel 
Ethernet Network Connections.

For more information on how to identify your adapter, go to the Adapter &
Driver ID Guide at:

    http://support.intel.com/support/network/sb/CS-012904.htm

For the latest Intel network drivers for Linux, refer to the following
website. Select the link for your adapter.

    http://support.intel.com/support/go/network/adapter/home.htm

SFP+ Devices with Pluggable Optics
----------------------------------

82599-BASED ADAPTERS  

NOTES: If your 82599-based Intel(R) Network Adapter came with Intel optics, or
is an Intel(R) Ethernet Server Adapter X520-2, then it only supports Intel 
optics and/or the direct attach cables listed below.

When 82599-based SFP+ devices are connected back to back, they should be set to
the same Speed setting via ethtool. Results may vary if you mix speed settings. 
 
Supplier    Type                                             Part Numbers

SR Modules			
Intel 	    DUAL RATE 1G/10G SFP+ SR (bailed)                FTLX8571D3BCV-IT	
Intel	    DUAL RATE 1G/10G SFP+ SR (bailed)                AFBR-703SDZ-IN2
Intel       DUAL RATE 1G/10G SFP+ SR (bailed)                AFBR-703SDDZ-IN1	
LR Modules			
Intel 	    DUAL RATE 1G/10G SFP+ LR (bailed)                FTLX1471D3BCV-IT	
Intel	    DUAL RATE 1G/10G SFP+ LR (bailed)                AFCT-701SDZ-IN2
Intel       DUAL RATE 1G/10G SFP+ LR (bailed)                AFCT-701SDDZ-IN1

The following is a list of 3rd party SFP+ modules that have received some 
testing. Not all modules are applicable to all devices.

Supplier   Type                                              Part Numbers

Finisar    SFP+ SR bailed, 10g single rate                   FTLX8571D3BCL
Avago      SFP+ SR bailed, 10g single rate                   AFBR-700SDZ
Finisar    SFP+ LR bailed, 10g single rate                   FTLX1471D3BCL
		
Finisar    DUAL RATE 1G/10G SFP+ SR (No Bail)	             FTLX8571D3QCV-IT
Avago	   DUAL RATE 1G/10G SFP+ SR (No Bail)	             AFBR-703SDZ-IN1	
Finisar	   DUAL RATE 1G/10G SFP+ LR (No Bail)	             FTLX1471D3QCV-IT
Avago	   DUAL RATE 1G/10G SFP+ LR (No Bail)	             AFCT-701SDZ-IN1
Finisar    1000BASE-T SFP                                    FCLF8522P2BTL
Avago      1000BASE-T SFP                                    ABCU-5710RZ
		
82599-based adapters support all passive and active limiting direct attach 
cables that comply with SFF-8431 v4.1 and SFF-8472 v10.4 specifications.

Laser turns off for SFP+ when ifconfig ethX down
------------------------------------------------
"ifconfig ethX down" turns off the laser for 82599-based SFP+ fiber adapters.
"ifconfig ethX up" turns on the laser.


82598-BASED ADAPTERS

NOTES for 82598-Based Adapters: 
- Intel(R) Ethernet Network Adapters that support removable optical modules 
  only support their original module type (i.e., the Intel(R) 10 Gigabit SR 
  Dual Port Express Module only supports SR optical modules). If you plug in
  a different type of module, the driver will not load.
- Hot Swapping/hot plugging optical modules is not supported.  
- Only single speed, 10 gigabit modules are supported.  
- LAN on Motherboard (LOMs) may support DA, SR, or LR modules. Other module 
  types are not supported. Please see your system documentation for details.  

The following is a list of 3rd party SFP+ modules and direct attach cables that
have received some testing. Not all modules are applicable to all devices.

Supplier   Type                                              Part Numbers

Finisar    SFP+ SR bailed, 10g single rate                   FTLX8571D3BCL
Avago      SFP+ SR bailed, 10g single rate                   AFBR-700SDZ
Finisar    SFP+ LR bailed, 10g single rate                   FTLX1471D3BCL

82598-based adapters support all passive direct attach cables that comply 
with SFF-8431 v4.1 and SFF-8472 v10.4 specifications. Active direct attach 
cables are not supported.	

Third party optic modules and cables referred to above are listed only for the 
purpose of highlighting third party specifications and potential compatibility, 
and are not recommendations or endorsements or sponsorship of any third party's
product by Intel. Intel is not endorsing or promoting products made by any 
third party and the third party reference is provided only to share information
regarding certain optic modules and cables with the above specifications. There
may be other manufacturers or suppliers, producing or supplying optic modules 
and cables with similar or matching descriptions. Customers must use their own 
discretion and diligence to purchase optic modules and cables from any third 
party of their choice. Customers are solely responsible for assessing the 
suitability of the product and/or devices and for the selection of the vendor 
for purchasing any product. INTEL ASSUMES NO LIABILITY WHATSOEVER, AND INTEL
DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF 
SUCH THIRD PARTY PRODUCTS OR SELECTION OF VENDOR BY CUSTOMERS.

Building and Installation
=========================

To build a binary RPM* package of this driver, run 'rpmbuild -tb
<filename.tar.gz>'.  Replace <filename.tar.gz> with the specific filename
of the driver.

NOTE: For the build to work properly, the currently running kernel MUST match
      the version and configuration of the installed kernel sources.  If you
      have just recompiled the kernel, reboot the system before building.

      RPM functionality has only been tested in Red Hat distributions.

To manually build this driver:

1. Move the base driver tar file to the directory of your choice. For example,
   use /home/username/ixgbe or /usr/local/src/ixgbe.

2. Untar/unzip archive:

     tar zxf ixgbe-x.x.x.tar.gz

3. Change to the driver src directory:

     cd ixgbe-x.x.x/src/

4. Compile the driver module:

   make install

   The binary will be installed as:

     /lib/modules/[KERNEL_VERSION]/kernel/drivers/net/ixgbe/ixgbe.[k]o

   The install locations listed above are the default locations.  They might
   not be correct for certain Linux distributions. 

5. Load the module:
   For kernel 2.6.x, use the modprobe command:
     modprobe ixgbe <parameter>=<value>

   Note that for 2.6 kernels the insmod command can be used if the full
   path to the driver module is specified.  For example:

     insmod /lib/modules/<KERNEL VERSION>/kernel/drivers/net/ixgbe/ixgbe.ko

   With 2.6 based kernels also make sure that older ixgbe drivers are
   removed from the kernel, before loading the new module:

     rmmod ixgbe; modprobe ixgbe

6. Assign an IP address to the interface by entering the following, where
   x is the interface number:

     ifconfig ethX <IP_address> netmask <netmask>

7. Verify that the interface works. Enter the following, where <IP_address>
   is the IP address for another machine on the same subnet as the interface
   that is being tested:

     ping  <IP_address>

To build ixgbe driver with DCA:
------------------------------

If your kernel supports DCA, the driver will build by default with DCA enabled. 

Command Line Parameters
=======================

If the driver is built as a module, the  following optional parameters are
used by entering them on the command line with the modprobe command using
this syntax:

     modprobe ixgbe [<option>=<VAL1>,<VAL2>,...]

For example:

     modprobe ixgbe InterruptThrottleRate=16000,16000

The default value for each parameter is generally the recommended setting,
unless otherwise noted.

RSS - Receive Side Scaling (or multiple queues for receives)
------------------------------------------------------------
Valid Range: 0 - 16
0 = Sets the descriptor queue count to whichever is less, number of CPUS or 16
1-16 = Sets the descriptor queue count to 1-16
Default Value: 0
RSS also effects the number of transmit queues allocated on 2.6.23 and      
newer kernels with CONFIG_NETDEVICES_MULTIQUEUE set in the kernel .config file.
CONFIG_NETDEVICES_MULTIQUEUE only exists from 2.6.23 to 2.6.26.  Other options 
enable multiqueue in 2.6.27 and newer kernels.                              

MQ - Multi Queue
----------------
Valid Range: 0, 1
0 = Disables Multiple Queue support
1 = Enables Multiple Queue support (a prerequisite for RSS)
Default Value: 1

DCA - Direct Cache Access
-------------------------
Valid Range: 0, 1
0 = Disables DCA support in the driver
1 = Enables DCA support in the driver
Default Value: 1 (when IXGBE_DCA is enabled)
See the above instructions for enabling DCA.  If the driver is enabled for
DCA this parameter allows load-time control of the feature.

IntMode
-------------
Valid Range: 0 - 2 (0 = Legacy Int, 1 = MSI and 2 = MSI-X)
Default Value: 2
IntMode controls allow load time control over the type of interrupt
registered for by the driver.  MSI-X is required for multiple queue
support, and some kernels and combinations of kernel .config options will
force a lower level of interrupt support.  'cat /proc/interrupts' will show
different values for each type of interrupt.
  
InterruptThrottleRate
---------------------
Valid Range: 956-488281 (0=off, 1=dynamic)
Default Value: 1
     Interrupt Throttle Rate (interrupts/sec)
The ITR parameter controls how many interrupts each interrupt vector can
generate per second.  Increasing ITR lowers latency at the cost of increased 
CPU utilization, though it may help throughput in some circumstances.
1 = Dynamic mode attempts to moderate interrupts per vector while maintaining
very low latency.  This can sometimes cause extra CPU utilization.  If planning
on deploying ixgbe in a latency sensitive environment please consider this
parameter.
0 = Setting InterruptThrottleRate to 0 turns off any interrupt moderation and 
may improve small packet latency, but is generally not suitable for bulk 
throughput traffic due to the increased cpu utilization of the higher interrupt
rate. Please note that on 82599 and X540-based adapters, disabling 
InterruptThrottleRate will also result in the driver disabling HW RSC. On 
82598-based adapters, disabling InterruptThrottleRate will also result in 
disabling LRO (Large Receive Offloads).

LLI (Low Latency Interrupts)
---------------------------- 
LLI allows for immediate generation of an interrupt upon processing receive 
packets that match certain criteria as set by the parameters described below. 
LLI parameters are not enabled when Legacy interrupts are used. You must be 
using MSI or MSI-X (see cat /proc/interrupts) to successfully use LLI.

LLIPort
-------
Valid Range: 0 - 65535
Default Value: 0 (disabled)

  LLI is configured with the LLIPort command-line parameter, which specifies 
  which TCP port should generate Low Latency Interrupts.

  For example, using LLIPort=80 would cause the hardware to generate an 
  immediate interrupt upon receipt of any packet sent to TCP port 80 on the 
  local machine.

WARNING: Enabling LLI can result in an excessive number of interrupts/second 
that may cause problems with the system and in some cases may cause a kernel 
panic.

LLIPush
-------
Valid Range: 0-1
Default Value: 0 (disabled)

  LLIPush can be set to be enabled or disabled (default). It is most 
  effective in an environment with many small transactions.
  NOTE: Enabling LLIPush may allow a denial of service attack.

LLISize
-------
Valid Range: 0-1500
Default Value: 0 (disabled)

  LLISize causes an immediate interrupt if the hardware receives a packet 
  smaller than the specified size.

LLIEType
--------
Valid Range: 0-x8fff
Default Value: 0 (disabled)

  Low Latency Interrupt Ethernet Protocol Type.

LLIVLANP
--------
Valid Range: 0-7
Default Value: 0 (disabled)

  Low Latency Interrupt on VLAN priority threshold.

max_vfs
-------
Valid Range:   1-63
Default Value: 0

  If the value is greater than 0 it will also force the VMDq parameter to be 1
  or more.

  This parameter adds support for SR-IOV.  It causes the driver to spawn up to 
  max_vfs worth of virtual function.  

The parameters for the driver are referenced by position.  So, if you have a 
dual port 82599 or X540-based adapter and you want N virtual functions per 
port, you must specify a number for each port with each parameter separated by
a comma.

For example:
  insmod ixgbe max_vfs=63,63

NOTE: If both 82598 and 82599 or X540-based adapters are installed on the same 
machine, you must be careful in loading the driver with the parameters. 
Depending on system configuration, number of slots, etc. it's impossible to 
predict in all cases where the positions would be on the command line and the 
user will have to specify zero in those positions occupied by an 82598 port.

L2LBen
------
Valid Range: 0 = disable, 1 = enable (default)
Default Value: 1

  This parameter controls the internal switch (L2 loopback between pf and vf).
  By default the switch is enabled.

Flow Control
------------
Ethernet Flow Control (IEEE 802.3x) can be configured with ethtool to enable
receiving and transmitting pause frames for ixgbe. When tx is enabled, PAUSE 
frames are generated when the receive packet buffer crosses a predefined 
threshold.  When rx is enabled, the transmit unit will halt for the time delay
specified when a PAUSE frame is received.

Flow Control is enabled by default. If you want to disable a flow control 
capable link partner, use ethtool:

     ethtool -A ethX autoneg off rx off tx off

NOTE: For 82598 backplane cards entering 1 gig mode, flow control default 
behavior is changed to off.  Flow control in 1 gig mode on these devices can 
lead to Tx hangs. 

Intel(R) Ethernet Flow Director
-------------------------------
NOTE: Flow director parameters are only supported on kernel versions 2.6.30 
or later.

Supports advanced filters that direct receive packets by their flows to 
different queues. Enables tight control on routing a flow in the platform. 
Matches flows and CPU cores for flow affinity. Supports multiple parameters 
for flexible flow classification and load balancing. 

Flow director is enabled only if the kernel is multiple TX queue capable.

An included script (set_irq_affinity.sh) automates setting the IRQ to CPU 
affinity.

You can verify that the driver is using Flow Director by looking at the counter
in ethtool: fdir_miss and fdir_match.

Other ethtool Commands:
To enable Flow Director
	ethtool -K ethX ntuple on
To add a filter
	Use -U switch. e.g., ethtool -U ethX flow-type tcp4 src-ip 0x178000a 
        action 1
To see the list of filters currently present:
	ethtool -u ethX

The following three parameters impact Flow Director.

FdirMode
--------
Valid Range: 0-2 (0=off, 1=ATR, 2=Perfect filter mode)
Default Value: 1

  Flow Director filtering modes.

FdirPballoc
-----------
Valid Range: 0-2 (0=64k, 1=128k, 2=256k)
Default Value: 0

  Flow Director allocated packet buffer size.

AtrSampleRate
--------------
Valid Range: 1-100
Default Value: 20

  Software ATR Tx packet sample rate. For example, when set to 20, every 20th
  packet, looks to see if the packet will create a new flow.

Perfect Filter
-------------- 
Perfect filter is an interface to load the filter table that funnels all flow 
into queue_0 unless an alternative queue is specified using "action". In that 
case, any flow that matches the filter criteria will be directed to the 
appropriate queue.

Support for Virtual Function (VF) is via the user-data field. You must update
to the version of ethtool built for the 2.6.40 kernel.  Perfect Filter is 
supported on all kernels 2.6.30 and later. Rules may be deleted from the table
itself. This is done via "ethtool -U ethX delete N" where N is the rule number
to be deleted.

NOTE: Flow Director Perfect Filters can run in single queue mode, when SR-IOV
is enabled, or when DCB is enabled. 
 
If the queue is defined as -1, filter will drop matching packets. 

To account for filter matches and misses, there are two stats in ethtool: 
fdir_match and fdir_miss. In addition, rx_queue_N_packets shows the number of 
packets processed by the Nth queue.

NOTE: Receive Packet Steering (RPS) and Receive Flow Steering (RFS) are not 
compatible with Flow Director. If Flow Director is enabled, these will be
disabled.

NOTE: For VLAN Masks only 4 masks are supported.

NOTE: Once a rule is defined, you must supply the same fields and masks (if
masks are specified).

Support for UDP RSS
-------------------
This feature adds an ON/OFF switch for hashing over certain flow types. You 
can't turn on anything other than UDP. The default setting is disabled.
We only support enabling/disabling hashing on ports for UDP over IPv4 (udp4)
or IPv6 (udp6).
NOTE: Fragmented packets may arrive out of order when RSS UDP support is 
configured.

Supported Ethtool Commands and Options:

  -n --show-nfc
     Retrieves the receive network flow classification configurations.

   rx-flow-hash tcp4|udp4|ah4|esp4|sctp4|tcp6|udp6|ah6|esp6|sctp6
     Retrieves the hash options for the specified network traffic type.

  -N --config-nfc
     Configures the receive network flow classification.

   rx-flow-hash tcp4|udp4|ah4|esp4|sctp4|tcp6|udp6|ah6|esp6|sctp6 
   m|v|t|s|d|f|n|r...
     Configures the hash options for the specified network traffic type.

     udp4    UDP over IPv4
     udp6    UDP over IPv6

     f   Hash on bytes 0 and 1 of the Layer 4 header of the rx packet.
     n   Hash on bytes 2 and 3 of the Layer 4 header of the rx packet.

The following is an example using udp4 (UDP over IPv4):

  To include UDP port numbers in RSS hashing run:
     ethtool -N ethX rx-flow-hash udp4 sdfn

  To exclude UDP port numbers from RSS hashing run:
     ethtool -N ethX rx-flow-hash udp4 sd

  To display UDP hashing current configuration run:
     ethtool -n ethX rx-flow-hash udp4

  The results of running that call will be the following, if UDP hashing is
  enabled:
  UDP over IPV4 flows use these fields for computing Hash flow key:
    IP SA
    IP DA
    L4 bytes 0 & 1 [TCP/UDP src port]
    L4 bytes 2 & 3 [TCP/UDP dst port]

  The results if UDP hashing is disabled would be:
  UDP over IPV4 flows use these fields for computing Hash flow key:
    IP SA
    IP DA

LRO
---
Enable/disable Large Receive Offload.
Valid Range: 0(off), 1(on)
Default Value: 1

Additional Configurations
=========================

  Configuring the Driver on Different Distributions
  -------------------------------------------------
  Configuring a network driver to load properly when the system is started is
  distribution dependent. Typically, the configuration process involves adding
  an alias line to /etc/modules.conf or /etc/modprobe.conf as well as editing
  other system startup scripts and/or configuration files.  Many popular Linux
  distributions ship with tools to make these changes for you.  To learn the
  proper way to configure a network device for your system, refer to your
  distribution documentation.  If during this process you are asked for the
  driver or module name, the name for the Linux Base Driver for the Intel
  10GbE PCI Express Family of Adapters is ixgbe.

  Viewing Link Messages
  ---------------------
  Link messages will not be displayed to the console if the distribution is
  restricting system messages. In order to see network driver link messages on
  your console, set dmesg to eight by entering the following:

       dmesg -n 8

  NOTE: This setting is not saved across reboots.

  Jumbo Frames
  ------------
  The driver supports Jumbo Frames for all adapters. Jumbo Frames support is
  enabled by changing the MTU to a value larger than the default of 1500.
  The maximum value for the MTU is 16110.  Use the ifconfig command to
  increase the MTU size.  For example:

        ifconfig ethX mtu 9000 up

  The maximum MTU setting for Jumbo Frames is 16110.  This value coincides
  with the maximum Jumbo Frames size of 16128. This driver will attempt to
  use multiple page sized buffers to receive each jumbo packet.  This
  should help to avoid buffer starvation issues when allocating receive
  packets.

  ethtool
  -------
  The driver utilizes the ethtool interface for driver configuration and
  diagnostics, as well as displaying statistical information. The latest 
  ethtool version is required for this functionality.

  The latest release of ethtool can be found from
  http://ftp.kernel.org/pub/software/network/ethtool/.

  NAPI
  ----
  NAPI (Rx polling mode) is supported in the ixgbe driver. 

  See ftp://robur.slu.se/pub/Linux/net-development/NAPI/usenix-paper.tgz for 
  more information on NAPI.

  LRO
  ---
  Large Receive Offload (LRO) is a technique for increasing inbound throughput
  of high-bandwidth network connections by reducing CPU overhead. It works by
  aggregating multiple incoming packets from a single stream into a larger 
  buffer before they are passed higher up the networking stack, thus reducing
  the number of packets that have to be processed. LRO combines multiple 
  Ethernet frames into a single receive in the stack, thereby potentially 
  decreasing CPU utilization for receives. 

  IXGBE_NO_LRO is a compile time flag. The user can enable it at compile
  time to remove support for LRO from the driver. The flag is used by adding 
  CFLAGS_EXTRA="-DIXGBE_NO_LRO" to the make file when it's being compiled. 

     make CFLAGS_EXTRA="-DIXGBE_NO_LRO" install

  You can verify that the driver is using LRO by looking at these counters in 
  ethtool:

  lro_flushed - the total number of receives using LRO.
  lro_aggregated - counts the total number of Ethernet packets that were combined.

  NOTE: IPv6 and UDP are not supported by LRO.

  HW RSC
  ------
  82599 and X540-based adapters support HW based receive side coalescing (RSC) 
  which can merge multiple frames from the same IPv4 TCP/IP flow into a single
  structure that can span one or more descriptors. It works similarly to SW
  Large receive offload technique. By default HW RSC is enabled and SW LRO 
  cannot be used for 82599 or X540-based adapters unless HW RSC is disabled.
 
  IXGBE_NO_HW_RSC is a compile time flag. The user can enable it at compile 
  time to remove support for HW RSC from the driver. The flag is used by adding 
  CFLAGS_EXTRA="-DIXGBE_NO_HW_RSC" to the make file when it's being compiled.
  
     make CFLAGS_EXTRA="-DIXGBE_NO_HW_RSC" install
 
  You can verify that the driver is using HW RSC by looking at the counter in 
  ethtool:
 
     hw_rsc_count - counts the total number of Ethernet packets that were being
     combined.

  rx_dropped_backlog
  ------------------
  When in a non-Napi (or Interrupt) mode, this counter indicates that the stack
  is dropping packets. There is an adjustable parameter in the stack that allows
  you to adjust the amount of backlog. We recommend increasing the 
  netdev_max_backlog if the counter goes up.

  # sysctl -a |grep netdev_max_backlog
  net.core.netdev_max_backlog = 1000

  # sysctl -e net.core.netdev_max_backlog=10000
  net.core.netdev_max_backlog = 10000

  MAC and VLAN anti-spoofing feature
  ----------------------------------
  When a malicious driver attempts to send a spoofed packet, it is dropped by 
  the hardware and not transmitted.  An interrupt is sent to the PF driver 
  notifying it of the spoof attempt.

  When a spoofed packet is detected the PF driver will send the following 
  message to the system log (displayed by  the "dmesg" command):

       ixgbe ethX: ixgbe_spoof_check: n spoofed packets detected

       Where x=the PF interface#  n=the number of spoofed packets

  Setting MAC Address, VLAN and Rate Limit Using IProute2 Tool
  ------------------------------------------------------------
  You can set a MAC address of a Virtual Function (VF), a default VLAN and the
  rate limit using the IProute2 tool. Download the latest version of the 
  iproute2 tool from Sourceforge if your version does not have all the
  features you require.

  (WoL) Wake on LAN Support
  -------------------------
  Some adapters do not support Wake on LAN. To determine if your adapter 
  supports Wake on LAN, run 

        ethtool ethX


Performance Tuning
==================

An excellent article on performance tuning can be found at:

http://www.redhat.com/promo/summit/2008/downloads/pdf/Thursday/Mark_Wagner.pdf

Known Issues/Troubleshooting
============================

Hardware Issues:
----------------
For known hardware and troubleshooting issues, either refer to the "Release 
Notes" in your User Guide, or for more detailed information, go to the 
following website:

    http://www.intel.com

In the search box enter your devices controller ID followed by "spec update" 
(i.e., 82599 spec update). The spec update file has complete information on
known hardware issues.

Software Issues:
----------------

  NOTE: After installing the driver, if your Intel Ethernet Network Connection
  is not working, verify that you have installed the correct driver.

  MSI-X Issues with Kernels between 2.6.19 - 2.6.21 (inclusive)
  ------------------------------------------------------------
  Kernel panics and instability may be observed on any MSI-X hardware if you
  use irqbalance with kernels between 2.6.19 and 2.6.21. 

  If such problems are encountered, you may disable the irqbalance daemon or
  upgrade to a newer kernel. 
                         
  Driver Compilation
  ------------------
  When trying to compile the driver by running make install, the following
  error may occur:

      "Linux kernel source not configured - missing version.h"

  To solve this issue, create the version.h file by going to the Linux source
  tree and entering:

      make include/linux/version.h.

  Do Not Use LRO When Routing or Bridging Packets
  -----------------------------------------------
  Due to a known general compatibility issue with LRO and routing, do not use
  LRO when routing or bridging packets.

  LRO and iSCSI Incompatibility
  ----------------------------------
  LRO is incompatible with iSCSI target or initiator traffic.  A panic may 
  occur when iSCSI traffic is received through the ixgbe driver with LRO 
  enabled.  To workaround this, the driver should be built and installed with:

      # make CFLAGS_EXTRA=-DIXGBE_NO_LRO install

  Performance Degradation with Jumbo Frames
  -----------------------------------------
  Degradation in throughput performance may be observed in some Jumbo frames
  environments.  If this is observed, increasing the application's socket buffer
  size and/or increasing the /proc/sys/net/ipv4/tcp_*mem entry values may help.
  See the specific application manual and /usr/src/linux*/Documentation/
  networking/ip-sysctl.txt for more details.

  Multiple Interfaces on Same Ethernet Broadcast Network
  ------------------------------------------------------
  Due to the default ARP behavior on Linux, it is not possible to have
  one system on two IP networks in the same Ethernet broadcast domain
  (non-partitioned switch) behave as expected.  All Ethernet interfaces
  will respond to IP traffic for any IP address assigned to the system.
  This results in unbalanced receive traffic.

  If you have multiple interfaces in a server, do either of the following:

  - Turn on ARP filtering by entering:
      echo 1 > /proc/sys/net/ipv4/conf/all/arp_filter

  - Install the interfaces in separate broadcast domains - either in
    different switches or in a switch partitioned to VLANs.

  UDP Stress Test Dropped Packet Issue
  --------------------------------------
  Under small packets UDP stress test with 10GbE driver, the Linux system
  may drop UDP packets due to the fullness of socket buffers. 
  You can increase the kernel's default buffer sizes for udp by changing 
  the values in /proc/sys/net/core/rmem_default and rmem_max.

  Unplugging network cable while ethtool -p is running
  ----------------------------------------------------
  In kernel versions 2.5.50 and later (including 2.6 kernel), unplugging 
  the network cable while ethtool -p is running will cause the system to 
  become unresponsive to keyboard commands, except for control-alt-delete.  
  Restarting the system appears to be the only remedy.

  Cisco Catalyst 4948-10GE port resets may cause switch to shut down ports
  ------------------------------------------------------------------------
  82598-based hardware can re-establish link quickly and when connected to some 
  switches, rapid resets within the driver may cause the switch port to become 
  isolated due to "link flap". This is typically indicated by a yellow instead
  of a green link light. Several operations may cause this problem, such as 
  repeatedly running ethtool commands that cause a reset.  

  A potential workaround is to use the Cisco IOS command "no errdisable detect 
  cause all" from the Global Configuration prompt which enables the switch to 
  keep the interfaces up, regardless of errors.

  Installing Red Hat Enterprise Linux 4.7, 5.1, or 5.2 with an Intel(R) 10 
  Gigabit AT Server Adapter may cause kernel panic
  ----------------------------------------------------------------------------
  A known issue may cause a kernel panic or hang after installing an 
  82598AT-based Intel(R) 10 Gigabit AT Server Adapter in a Red Hat Enterprise 
  Linux 4.7, 5.1, or 5.2 system.  The ixgbe driver for both the install kernel 
  and the runtime kernel can create this panic if the 82598AT adapter is 
  installed. Red Hat may release a security update that contains a fix for the 
  panic that you can download using RHN (Red Hat Network) or Intel recommends 
  that you install the ixgbe-1.3.31.5 driver or newer BEFORE installing the 
  hardware.

  Rx Page Allocation Errors
  -------------------------
  Page allocation failure. order:0 errors may occur under stress with kernels 
  2.6.25 and above. This is caused by the way the Linux kernel reports this 
  stressed condition.

  DCB: Generic segmentation offload on causes bandwidth allocation issues
  -----------------------------------------------------------------------
  In order for DCB to work correctly, GSO (Generic Segmentation Offload aka
  software TSO) must be disabled using ethtool.  By default since the hardware
  supports TSO (hardware offload of segmentation) GSO will not be running.  
  The GSO state can be queried with ethtool using ethtool -k ethX.

  When using 82598-based network connections, ixgbe driver only supports 16 
  queues on a platform with more than 16 cores
  -------------------------------------------------------------------------
  Due to known hardware limitations, RSS can only filter in a maximum of 16 
  receive queues.

  82599 and X540-based network connections support up to 64 queues.

  Disable GRO when routing/bridging
  ---------------------------------
  Due to a known kernel issue, GRO must be turned off when routing/bridging. 
  GRO can be turned off via ethtool. 

      ethtool -K ethX gro off

      where ethX is the ethernet interface you're trying to modify.

  Lower than expected performance on dual port and quad port 10GbE devices
  ------------------------------------------------------------------------
  Some PCI-E x8 slots are actually configured as x4 slots. These slots have 
  insufficient bandwidth for full 10Gbe line rate with dual port and quad port
  10GbE devices. In addition, if you put a PCIe Gen 3-capable adapter into a
  PCIe Gen 2  slot, you can not get full bandwidth. The driver can detect this
  situation and will write the following message in the system log: 
  "PCI-Express bandwidth available for this card is not sufficient for optimal
  performance. For optimal performance a x8 PCI-Express slot is required."
  If this error occurs, moving your adapter to a true x8 slot will resolve the 
  issue.

  ethtool may incorrectly display SFP+ fiber module as Direct Attached cable
  --------------------------------------------------------------------------
  Due to kernel limitations, port type can only be correctly displayed on 
  kernel 2.6.33 or greater.

  Under Redhat 5.4 - System May Crash when Closing Guest OS Window after 
  Loading/Unloading Physical Function (PF) Driver
  -------------------------------------------------------------------------
  Do not remove the ixgbe driver from Dom0 while Virtual Functions (VFs) are 
  assigned to guests. VFs must first use the xm "pci-detach" command to 
  hot-plug the VF device out of the VM it is assigned to or else shut down the
  VM.

  Unloading Physical Function (PF) Driver Causes System Reboots When VM is
  Running and VF is Loaded on the VM
  -----------------------------------------------------------------------------
  Do not unload the PF driver (ixgbe) while VFs are assigned to guests.

  Running ethtool -t ethX command causes break between PF and test client
  -----------------------------------------------------------------------
  When there are active VFs, "ethtool -t" will only run the link test.  The
  driver will also log in syslog that VFs should be shut down to run a full
  diags test.

  SLES10 SP3 random system panic when reloading driver
  ---------------------------------------------------
  This is a known SLES-10 SP3 issue. After requesting interrupts for MSI-X 
  vectors, system may panic. 

  Currently the only known workaround is to build the drivers with 
  CFLAGS_EXTRA=-DDISABLE_PCI_MSI if the driver need to be loaded/unloaded.
  Otherwise the driver can be loaded once and will be safe, but unloading it 
  will lead to the issue.

  Enabling SR-IOV in a 32-bit or 64-bit Microsoft* Windows* Server 2008/R2 
  Guest OS using Intel (R) 82576-based GbE, Intel (R) 82599-based or X540-based 
  10GbE controller under KVM
  -----------------------------------------------------------------------------
  KVM Hypervisor/VMM supports direct assignment of a PCIe device to a VM.  This 
  includes traditional PCIe devices, as well as SR-IOV-capable devices using
  Intel 82576-based, 82599-based and X540-based controllers.

  While direct assignment of a PCIe device or an SR-IOV Virtual Function (VF)
  to a Linux-based VM running 2.6.32 or later kernel works fine, there is a
  known issue with Microsoft Windows Server 2008/R2 VM that results in a 
  "yellow bang" error. This problem is within the KVM VMM itself, not the Intel
  driver, or the SR-IOV logic of the VMM, but rather that KVM emulates an older
  CPU model for the guests, and this older CPU model does not support MSI-X
  interrupts, which is a requirement for Intel SR-IOV.  

  If you wish to use the Intel 82576, 82599 or X540-based controllers in SR-IOV
  mode with KVM and a Microsoft Windows Server 2008/R2 guest try the following 
  workaround. The workaround is to tell KVM to emulate a different model of CPU
  when using qemu to create the KVM guest: 

       "-cpu qemu64,model=13"

  Loading ixgbe driver in 3.2.x and above kernels displays kernel tainted 
  message
  -----------------------------------------------------------------------
  Due to recent kernel changes, loading an out of tree driver will cause the
  kernel to be tainted.

  Unable to obtain DHCP lease on boot with RedHat
  -----------------------------------------------
  For configurations where the auto-negotiation process takes more than 5 
  seconds, the boot script may fail with the following message:

      "ethX: failed. No link present. Check cable?"

  If this error appears even though the presence of a link can be confirmed
  using ethtool ethX, try setting

      "LINKDELAY=5" in /etc/sysconfig/network-scripts/ifcfg-ethX.
      NOTE: Link time can take up to 30 seconds. Adjust LINKDELAY value 
      accordingly.

  Host may reboot after removing PF when VF is active in guest
  ------------------------------------------------------------
  Using kernel versions earlier than 3.2, do not unload the PF driver with 
  active VFs. Doing this will cause your VFs to stop working until you reload
  the PF driver and may cause a spontaneous reboot of your system.

  Software bridging does not work with SR-IOV Virtual Functions
  -------------------------------------------------------------
  SR-IOV Virtual Functions are unable to send or receive traffic between VMs
  using emulated connections on a Linux Software bridge and connections that 
  use SR-IOV VFs.


Support
=======

For general information, go to the Intel support website at:

    www.intel.com/support/

or the Intel Wired Networking project hosted by Sourceforge at:

    http://sourceforge.net/projects/e1000

If an issue is identified with the released source code on the supported
kernel with a supported adapter, email the specific information related
to the issue to e1000-devel@lists.sf.net

License
=======

Intel 10 Gigabit PCI Express Linux driver.
Copyright(c) 1999 - 2012 Intel Corporation.

This program is free software; you can redistribute it and/or modify it
under the terms and conditions of the GNU General Public License,
version 2, as published by the Free Software Foundation.

This program is distributed in the hope it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
more details.

You should have received a copy of the GNU General Public License along with
this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.

The full GNU General Public License is included in this distribution in
the file called "COPYING".

Trademarks
==========

Intel, Itanium, and Pentium are trademarks or registered trademarks of
Intel Corporation or its subsidiaries in the United States and other
countries.

* Other names and brands may be claimed as the property of others.