mirror of
https://git.FreeBSD.org/src.git
synced 2024-12-25 11:37:56 +00:00
063efed28c
tree used it incorrectly, which lead to inaccurate overrated
if_obytes accounting. The drbr(9) used to update ifnet stats on
drbr_enqueue(), which is not accurate since enqueuing doesn't
imply successful processing by driver. Dequeuing neither mean
that. Most drivers also called drbr_stats_update() which did
accounting again, leading to doubled if_obytes statistics. And
in case of severe transmitting, when a packet could be several
times enqueued and dequeued it could have been accounted several
times.
o Thus, make drbr(9) API thinner. Now drbr(9) merely chooses between
ALTQ queueing or buf_ring(9) queueing.
- It doesn't touch the buf_ring stats any more.
- It doesn't touch ifnet stats anymore.
- drbr_stats_update() no longer exists.
o buf_ring(9) handles its stats itself:
- It handles br_drops itself.
- br_prod_bytes stats are dropped. Rationale: no one ever
reads them but update of a common counter on every packet
negatively affects performance due to excessive cache
invalidation.
- buf_ring_enqueue_bytes() reduced to buf_ring_enqueue(), since
we no longer account bytes.
o Drivers handle their stats theirselves: if_obytes, if_omcasts.
o mlx4(4), igb(4), em(4), vxge(4), oce(4) and ixv(4) no longer
use drbr_stats_update(), and update ifnet stats theirselves.
o bxe(4) was the most correct driver, it didn't call
drbr_stats_update(), thus it was the only driver accurate under
moderate load. Now it also maintains stats itself.
o ixgbe(4) had already taken stats from hardware, so just
- drop software stats updating.
- take multicast packet count from hardware as well.
o mxge(4) just no longer needs NO_SLOW_STATS define.
o cxgb(4), cxgbe(4) need no change, since they obtain stats
from hardware.
Reviewed by: jfv, gnn
|
||
|---|---|---|
| .. | ||
| ixgbe_82598.c | ||
| ixgbe_82598.h | ||
| ixgbe_82599.c | ||
| ixgbe_82599.h | ||
| ixgbe_api.c | ||
| ixgbe_api.h | ||
| ixgbe_common.c | ||
| ixgbe_common.h | ||
| ixgbe_mbx.c | ||
| ixgbe_mbx.h | ||
| ixgbe_osdep.h | ||
| ixgbe_phy.c | ||
| ixgbe_phy.h | ||
| ixgbe_type.h | ||
| ixgbe_vf.c | ||
| ixgbe_vf.h | ||
| ixgbe_x540.c | ||
| ixgbe_x540.h | ||
| ixgbe.c | ||
| ixgbe.h | ||
| ixv.c | ||
| ixv.h | ||
| LICENSE | ||
| README | ||
FreeBSD Driver for Intel(R) Ethernet 10 Gigabit PCI Express Server Adapters
============================================================================
/*$FreeBSD$*/
November 12, 2010
Contents
========
- Overview
- Supported Adapters
- Building and Installation
- Additional Configurations and Tuning
- Known Limitations
Overview
========
This file describes the FreeBSD* driver for the Intel(R) Ethernet 10 Gigabit
Family of Adapters. Driver has been developed for use with FreeBSD 7.2 or later.
For questions related to hardware requirements, refer to the documentation
supplied with your Intel 10GbE adapter. All hardware requirements listed
apply to use with FreeBSD.
Supported Adapters
==================
The driver in this release is compatible with 82598 and 82599-based Intel
Network Connections.
SFP+ Devices with Pluggable Optics
----------------------------------
82599-BASED ADAPTERS
NOTE: If your 82599-based Intel(R) Ethernet 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 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 FTLX8571D3BCV-IT
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
Finistar 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 down
--------------------------------------------------------
"ifconfig down" turns off the laser for 82599-based SFP+ fiber adapters.
"ifconfig up" turns on the later.
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. Customer 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.
Configuration and Tuning
========================
The driver supports Transmit/Receive Checksum Offload and Jumbo Frames on
all 10 Gigabit adapters.
Jumbo Frames
------------
To enable Jumbo Frames, use the ifconfig utility to increase the MTU
beyond 1500 bytes.
NOTES:
- The Jumbo Frames setting on the switch must be set to at least
22 bytes larger than that of the adapter.
- There are known performance issues with this driver when running
UDP traffic with Jumbo Frames.
The Jumbo Frames MTU range for Intel Adapters is 1500 to 16114. The default
MTU range is 1500. To modify the setting, enter the following:
ifconfig ix<interface_num> <hostname or IP address> mtu 9000
To confirm an interface's MTU value, use the ifconfig command. To confirm
the MTU used between two specific devices, use:
route get <destination_IP_address>
VLANs
-----
To create a new VLAN pseudo-interface:
ifconfig <vlan_name> create
To associate the VLAN pseudo-interface with a physical interface and
assign a VLAN ID, IP address, and netmask:
ifconfig <vlan_name> <ip_address> netmask <subnet_mask> vlan
<vlan_id> vlandev <physical_interface>
Example:
ifconfig vlan10 10.0.0.1 netmask 255.255.255.0 vlan 10 vlandev ixgbe0
In this example, all packets will be marked on egress with 802.1Q VLAN
tags, specifying a VLAN ID of 10.
To remove a VLAN pseudo-interface:
ifconfig <vlan_name> destroy
Checksum Offload
----------------
Checksum offloading supports both TCP and UDP packets and is
supported for both transmit and receive.
Checksum offloading can be enabled or disabled using ifconfig.
Both transmit and receive offloading will be either enabled or
disabled together. You cannot enable/disable one without the other.
To enable checksum offloading:
ifconfig <interface_num> rxcsum
To disable checksum offloading:
ifconfig <interface_num> -rxcsum
To confirm the current setting:
ifconfig <interface_num>
TSO
---
TSO is enabled by default.
To disable:
ifconfig <interface_num> -tso
To re-enable:
ifconfig <interface_num> tso
LRO
---
Large Receive Offload is available in the driver; it is on by default.
It can be disabled by using:
ifconfig <interface_num> -lro
To enable:
ifconfig <interface_num> lro
Important system configuration changes:
---------------------------------------
When there is a choice run on a 64bit OS rather than 32, it makes a
significant difference in improvement.
The default scheduler SCHED_4BSD is not smart about SMP locality issues.
Significant improvement can be achieved by switching to the ULE scheduler.
This is done by changing the entry in the config file from SCHED_4BSD to
SCHED_ULE. Note that this is only advisable on FreeBSD 7, on 6.X there have
been stability problems with ULE.
The interface can generate high number of interrupts. To avoid running
into the limit set by the kernel, adjust hw.intr_storm_threshold
setting using sysctl:
sysctl hw.intr_storm_threshold=9000 (the default is 1000)
For this change to take effect on boot, edit /etc/sysctl.conf and add the
line:
hw.intr_storm_threshold=9000
If you still see Interrupt Storm detected messages, increase the limit to a
higher number.
Best throughput results are seen with a large MTU; use 9000 if possible.
The default number of descriptors is 1024, increasing this to 2K or even
4K may improve performance in some workloads, but change carefully.
Known Limitations
=================
For known hardware and troubleshooting issues, refer to the following website.
http://support.intel.com/support/go/network/adapter/home.htm
Either select the link for your adapter or perform a search for the adapter
number. The adapter's page lists many issues. For a complete list of hardware
issues download your adapter's user guide and read the Release Notes.
UDP stress test with 10GbE driver
---------------------------------
Under small packets UDP stress test with 10GbE driver, the FreeBSD system
will drop UDP packets due to the fullness of socket buffers. You may want
to change the driver's Flow Control variables to the minimum value for
controlling packet reception.
Attempting to configure larger MTUs with a large numbers of processors may
generate the error message "ix0:could not setup receive structures"
--------------------------------------------------------------------------
When using the ixgbe driver with RSS autoconfigured based on the number of
cores (the default setting) and that number is larger than 4, increase the
memory resources allocated for the mbuf pool as follows:
Add to the sysctl.conf file for the system:
kern.ipc.nmbclusters=262144
kern.ipc.nmbjumbop=262144
Lower than expected performance on dual 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 10GbE devices.
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.
Support
=======
For general information and support, go to the Intel support website at:
www.intel.com/support/
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 freebsd@intel.com
License
=======
This software program is released under the terms of a license agreement
between you ('Licensee') and Intel. Do not use or load this software or any
associated materials (collectively, the 'Software') until you have carefully
read the full terms and conditions of the LICENSE located in this software
package. By loading or using the Software, you agree to the terms of this
Agreement. If you do not agree with the terms of this Agreement, do not
install or use the Software.
* Other names and brands may be claimed as the property of others.