man/man4/igb.4

.\" Copyright (c) 2001-2003, Intel Corporation
.\" All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions are met:
.\"
.\" 1. Redistributions of source code must retain the above copyright notice,
.\"    this list of conditions and the following disclaimer.
.\"
.\" 2. Redistributions in binary form must reproduce the above copyright
.\"    notice, this list of conditions and the following disclaimer in the
.\"    documentation and/or other materials provided with the distribution.
.\"
.\" 3. Neither the name of the Intel Corporation nor the names of its
.\"    contributors may be used to endorse or promote products derived from
.\"    this software without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
.\" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
.\" LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
.\" CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
.\" SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
.\" INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
.\" CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
.\" ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
.\" POSSIBILITY OF SUCH DAMAGE.
.\"
.\" * Other names and brands may be claimed as the property of others.
.\"
.\" $FreeBSD: src/share/man/man4/igb.4,v 1.2 2010/05/14 20:11:30 marius Exp $
.\"
.Dd November 28, 2014
.Dt IGB 4
.Os
.Sh NAME
.Nm igb
.Nd "Intel(R) PRO/1000 PCI Express Gigabit Ethernet adapter driver"
.Sh SYNOPSIS
To compile this driver into the kernel,
place the following line in your
kernel configuration file:
.Bd -ragged -offset indent
.Cd "device ig_hal"
.Cd "device igb"
.Ed
.Pp
Alternatively, to load the driver as a
module at boot time, place the following line in
.Xr loader.conf 5 :
.Bd -literal -offset indent
ig_hal_load="YES"
if_igb_load="YES"
.Ed
.Sh DESCRIPTION
The
.Nm
driver provides support for PCI Express Gigabit Ethernet adapters based on
the Intel
82575,
82576,
82580,
I210,
I211,
I350 and I354
Ethernet controller chips.
The
.Nm
driver supports:
.Pp
.Bl -item -offset indent -compact
.It
Transmit/Receive checksum offload for IP/UDP/TCP.
.\"Jumbo Frames.
.It
Interrupt moderation
.It
TCP segmentation offload (TSO)
.It
Receive side scaling (RSS)
.It
Multiple tranmission queues
.It
Multiple vector MSI-X
.It
VLAN tag stripping and inserting
.El
.Pp
If
.Xr polling 4
or MSI-X is used,
by default,
the
.Nm
driver will try enabling as many reception queues and transmission queues
as are allowed by the number of CPUs in the system.
.Pp
If multiple transmission queues are used,
the round-robin arbitration is performed among the transmission queues.
It should be noted that
if both TSO and multiple transmission queues are used,
for 82575, the round-robin arbitration between transmission queues is done
at the TSO packet boundary;
for the reset of the hardware, the round-robin arbitration
between transmission queues is done at the TCP segment boundary after
the hardware segmentation is performed.
.Pp
82575 supports 4 reception queues and 4 transmission queues.
MSI-X is not enabled due to hardware errata.
Under MSI or legacy interrupt mode,
2 reception queues are enabled for hardware RSS hash
and only 1 transmission queue is enable.
.Pp
82576 supports 16 reception queues and 16 transmission queues.
MSI-X is enable by default.
However,
due to the number of MSI-X vectors is 10,
at most 8 reception queues and 8 transmission queues will be enabled
under MSI-X mode.
When
.Xr polling 4
is enabled on the devices,
at most 16 reception queues and 16 transmission queues will be enabled.
.Pp
82580,
I350 and I354 support 8 reception queues and 8 transmission queues.
MSI-X is enabled by default.
.Pp
I210 supports 4 reception queues and 4 transmission queues.
MSI-X is enabled by default.
.Pp
I211 supports 2 reception queues and 2 transmission queues.
MSI-X is enabled by default.
.Pp
.\"For questions related to hardware requirements, refer to the
.\"documentation supplied with your Intel PRO/1000 adapter.
.\"All hardware requirements listed apply to use with
.\".Dx .
.\".Pp
.\"Support for Jumbo Frames is provided via the interface MTU setting.
.\"Selecting an MTU larger than 1500 bytes with the
.\".Xr ifconfig 8
.\"utility configures the adapter to receive and transmit Jumbo Frames.
.\"The maximum MTU size for Jumbo Frames is 9216.
.\".Pp
The
.Nm
driver supports the following media types:
.Bl -tag -width ".Cm 10baseT/UTP"
.It Cm autoselect
Enables auto-negotiation for speed and duplex.
.It Cm 10baseT/UTP
Sets 10Mbps operation.
Use the
.Cm mediaopt
option to select
.Cm full-duplex
mode.
.It Cm 100baseTX
Sets 100Mbps operation.
Use the
.Cm mediaopt
option to select
.Cm full-duplex
mode.
.It Cm 1000baseSX
Sets 1000Mbps operation.
Only
.Cm full-duplex
mode is supported at this speed.
.It Cm 1000baseT
Sets 1000Mbps operation.
Only
.Cm full-duplex
mode is supported at this speed.
.El
.Pp
The
.Nm
driver supports the following media options:
.Bl -tag -width ".Cm full-duplex"
.It Cm full-duplex
Forces full-duplex operation
.It Cm half-duplex
Forces half-duplex operation.
.El
.Pp
Only use
.Cm mediaopt
to set the driver to
.Cm full-duplex .
If
.Cm mediaopt
is not specified, the driver defaults to
.Cm half-duplex .
.Pp
For more information on configuring this device, see
.Xr ifconfig 8 .
The
.Nm
driver supports
.Xr polling 4 .
.Sh HARDWARE
The
.Nm
driver supports Gigabit Ethernet adapters based on the Intel
82575,
82576,
82580,
I210,
I211,
I350 and I354
controller chips:
.Pp
.Bl -bullet -compact
.It
Intel 82575EB Gigabit Ethernet Controller
.It
Intel 82576 Gigabit Ethernet Controller
.It
Intel 82580EB Gigabit Ethernet Controller
.It
Intel Ethernet Controller I210 Series
.It
Intel Ethernet Controller I211 Series
.It
Intel Ethernet Controller I350
.It
Intel Ethernet Controller I354
.It
Intel Ethernet Server Adapter I210-T1
.It
Intel Ethernet Server Adapter I340-F4
.It
Intel Ethernet Server Adapter I340-T4
.It
Intel Ethernet Server Adapter I350-F2
.It
Intel Ethernet Server Adapter I350-F4
.It
Intel Ethernet Server Adapter I350-T2
.It
Intel Ethernet Server Adapter I350-T4
.It
Intel Gigabit EF Dual Port Server Adapter
.It
Intel Gigabit ET Dual Port Server Adapter
.It
Intel Gigabit ET Quad Port Server Adapter
.It
Intel Gigabit ET2 Quad Port Server Adapter
.It
Intel Gigabit VT Quad Port Server Adapter
.El
.Sh TUNABLES
Tunables can be set at the
.Xr loader 8
prompt before booting the kernel or stored in
.Xr loader.conf 5 .
.Em X
is the device unit number.
.Bl -tag -width ".Va hw.igbX.irq.unshared"
.It Va hw.igb.rxd Va hw.igbX.rxd
Number of receive descriptors allocated by the driver.
The default value is 512.
The minimum is 256,
and the maximum is 4096.
.It Va hw.igb.txd Va hw.igbX.txd
Number of transmit descriptors allocated by the driver.
The default value is 1024.
The minimum is 256,
and the maximum is 4096.
.It Va hw.igb.rxr Va hw.igbX.rxr
This tunable specifies the number of reception queues could be enabled.
Maximum allowed value for these tunables is device specific
and it must be power of 2 aligned.
Setting these tunables to 0 allows the driver to make
as many reception queues ready-for-use as allowed by the number of CPUs.
.It Va hw.igb.txr Va hw.igbX.txr
This tunable specifies the number of transmission queues could be enabled.
Maximum allowed value for these tunables is device specific
and it must be power of 2 aligned.
Setting these tunables to 0 allows the driver to make
as many transmission queues ready-for-use as allowed by the number of CPUs.
.It Va hw.igb.msix.enable Va hw.igbX.msix.enable
By default,
the driver will use MSI-X if it is supported.
This behaviour can be turned off by setting this tunable to 0.
.It Va hw.igbX.msix.off
If MSI-X is used
and the number of MSI-X vectors is not enough to
put transmission queue processing and reception queue processing
onto independent MSI-X vector,
this tunable specifies the leading target CPU for
transmission and reception queues processing.
The value specificed must be aligned to the maximum of
the number of reception queues
and the number of transmission queues enabled,
and must be less than the power of 2 number of CPUs.
.It Va hw.igbX.msix.rxoff
If MSI-X is used
and the number of MSI-X vectors is enough to
put transmission queue processing and reception queue processing
onto independent MSI-X vector,
this tunable specifies the leading target CPU for reception queues processing.
The value specificed must be aligned to the number of reception queues enabled
and must be less than the power of 2 number of CPUs.
.It Va hw.igbX.msix.txoff
If MSI-X is used
and the number of MSI-X vectors is enough to
put transmission queue processing and reception queue processing
onto independent MSI-X vector,
this tunable specifies the leading target CPU
for transmission queues processing.
The value specificed must be aligned to
the number of transmission queues enabled
and must be less than the power of 2 number of CPUs.
.It Va hw.igb.msi.enable Va hw.igbX.msi.enable
If MSI-X is disabled and MSI is supported,
the driver will use MSI.
This behavior can be turned off by setting this tunable to 0.
.It Va hw.igbX.msi.cpu
If MSI is used,
it specifies the MSI's target CPU.
.It Va hw.igbX.irq.unshared
If legacy interrupt is used,
by default,
the driver assumes the interrupt could be shared.
Setting this tunable to 1 allows the driver to perform certain
optimization based on the knowledge that the interrupt is not shared.
.It Va hw.igbX.npoll.txoff
This tunable specifies the leading target CPU for
transmission queue
.Xr polling 4
processing.
The value specificed must be aligned to the number of transmission queues
enabled and must be less than the power of 2 number of CPUs.
.It Va hw.igbX.npoll.rxoff
This tunable specifies the leading target CPU for
reception queue
.Xr polling 4
processing.
The value specificed must be aligned to the number of reception queues
enabled and must be less than the power of 2 number of CPUs.
.El
.Sh MIB Variables
A number of per-interface variables are implemented in the
.Va dev.igb. Ns Em X
branch of the
.Xr sysctl 3
MIB.
.Bl -tag -width "tx_intr_nsegs"
.It Va rxr
Number of reception queues could be enabled (read-only).
Use the tunable
.Va hw.igb.rxr
or
.Va hw.igbX.rxr
to configure it.
.It Va rxr_inuse
Number of reception queues being used (read-only).
.It Va txr
Number of transmission queues could be enabled (read-only).
Use the tunable
.Va hw.igb.txr
or
.Va hw.igbX.txr
to configure it.
.It Va txr_inuse
Number of transmission queues being used (read-only).
.It Va rxd
Number of descriptors per reception queue (read-only).
Use the tunable
.Va hw.igb.rxd
or
.Va hw.igbX.rxd
to configure it.
.It Va txd
Number of descriptors per transmission queue (read-only).
Use the tunable
.Va hw.igb.txd
or
.Va hw.igbX.txd
to configure it.
.It Va intr_rate
If MSI or legacy interrupt is used,
this sysctl controls the highest possible frequency
that interrupt could be generated by the device.
It is 6000 by default (~150us).
.It Va msixY_rate
If MSI-X is used,
this sysctl controls the highest possible frequency
that interrupt could be generated by the
.Em Y
MSI-X vector.
For reception processing only MSI-X vector,
the default value is 6000 (~150us).
For transmission processing only MSI-X vector,
the default value is 4000 (250us).
For MSI-X vector which handles both reception and transmission,
the default value is 6000 (~150us).
.It Va tx_intr_nsegs
Transmission interrupt is asked to be generated upon every
.Va tx_intr_nsegs
transmission descritors having been setup.
The default value is 1/16 of the number of transmission descriptors per queue.
.It Va tx_wreg_nsegs
The number of transmission descriptors should be setup
before the hardware register is written.
Setting this value too high will have negative effect
on transmission timeliness.
Setting this value too low will hurt overall transmission performance
due to the frequent hardware register writing.
The default value is 8.
.It Va rx_wreg_nsegs
The number of reception descriptors should be setup
before the hardware register is written.
Setting this value too high will make device drop incoming packets.
Setting this value too low will hurt overall reception performance
due to the frequent hardware register writing.
The default value is 32.
.It Va npoll_rxoff
See the tunable
.Va hw.igbX.npoll.rxoff .
The set value will take effect the next time
.Xr polling 4
is enabled on the device.
.It Va npoll_txoff
See the tunable
.Va hw.igbX.npoll.txoff .
The set value will take effect the next time
.Xr polling 4
is enabled on the device.
.El
.\".Sh SUPPORT
.\"For general information and support,
.\"go to the Intel support website at:
.\".Pa http://support.intel.com .
.\".Pp
.\"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
.\".Aq Mt freebsdnic@mailbox.intel.com .
.Sh SEE ALSO
.Xr altq 4 ,
.Xr arp 4 ,
.Xr ifmedia 4 ,
.Xr netintro 4 ,
.Xr ng_ether 4 ,
.Xr polling 4 ,
.Xr vlan 4 ,
.Xr ifconfig 8
.Sh HISTORY
The
.Nm
device driver first appeared in
.Fx 7.1 .
.Sh AUTHORS
The
.Nm
driver was written by
.An Intel Corporation Aq Mt freebsdnic@mailbox.intel.com .