]> bbs.cooldavid.org Git - net-next-2.6.git/blame - Documentation/networking/e1000e.txt
ipv6: AF_INET6 link address family
[net-next-2.6.git] / Documentation / networking / e1000e.txt
CommitLineData
da8c01c4
JK
1Linux* Driver for Intel(R) Network Connection
2===============================================================
3
4Intel Gigabit Linux driver.
5Copyright(c) 1999 - 2010 Intel Corporation.
6
7Contents
8========
9
10- Identifying Your Adapter
11- Command Line Parameters
12- Additional Configurations
13- Support
14
15Identifying Your Adapter
16========================
17
18The e1000e driver supports all PCI Express Intel(R) Gigabit Network
19Connections, except those that are 82575, 82576 and 82580-based*.
20
21* NOTE: The Intel(R) PRO/1000 P Dual Port Server Adapter is supported by
22 the e1000 driver, not the e1000e driver due to the 82546 part being used
23 behind a PCI Express bridge.
24
25For more information on how to identify your adapter, go to the Adapter &
26Driver ID Guide at:
27
28 http://support.intel.com/support/go/network/adapter/idguide.htm
29
30For the latest Intel network drivers for Linux, refer to the following
31website. In the search field, enter your adapter name or type, or use the
32networking link on the left to search for your adapter:
33
34 http://support.intel.com/support/go/network/adapter/home.htm
35
36Command Line Parameters
37=======================
38
39The default value for each parameter is generally the recommended setting,
40unless otherwise noted.
41
42NOTES: For more information about the InterruptThrottleRate,
43 RxIntDelay, TxIntDelay, RxAbsIntDelay, and TxAbsIntDelay
44 parameters, see the application note at:
45 http://www.intel.com/design/network/applnots/ap450.htm
46
47InterruptThrottleRate
48---------------------
49Valid Range: 0,1,3,4,100-100000 (0=off, 1=dynamic, 3=dynamic conservative,
50 4=simplified balancing)
51Default Value: 3
52
53The driver can limit the amount of interrupts per second that the adapter
54will generate for incoming packets. It does this by writing a value to the
55adapter that is based on the maximum amount of interrupts that the adapter
56will generate per second.
57
58Setting InterruptThrottleRate to a value greater or equal to 100
59will program the adapter to send out a maximum of that many interrupts
60per second, even if more packets have come in. This reduces interrupt
61load on the system and can lower CPU utilization under heavy load,
62but will increase latency as packets are not processed as quickly.
63
64The driver has two adaptive modes (setting 1 or 3) in which
65it dynamically adjusts the InterruptThrottleRate value based on the traffic
66that it receives. After determining the type of incoming traffic in the last
67timeframe, it will adjust the InterruptThrottleRate to an appropriate value
68for that traffic.
69
70The algorithm classifies the incoming traffic every interval into
71classes. Once the class is determined, the InterruptThrottleRate value is
72adjusted to suit that traffic type the best. There are three classes defined:
73"Bulk traffic", for large amounts of packets of normal size; "Low latency",
74for small amounts of traffic and/or a significant percentage of small
75packets; and "Lowest latency", for almost completely small packets or
76minimal traffic.
77
78In dynamic conservative mode, the InterruptThrottleRate value is set to 4000
79for traffic that falls in class "Bulk traffic". If traffic falls in the "Low
80latency" or "Lowest latency" class, the InterruptThrottleRate is increased
81stepwise to 20000. This default mode is suitable for most applications.
82
83For situations where low latency is vital such as cluster or
84grid computing, the algorithm can reduce latency even more when
85InterruptThrottleRate is set to mode 1. In this mode, which operates
86the same as mode 3, the InterruptThrottleRate will be increased stepwise to
8770000 for traffic in class "Lowest latency".
88
89In simplified mode the interrupt rate is based on the ratio of Tx and
90Rx traffic. If the bytes per second rate is approximately equal the
91interrupt rate will drop as low as 2000 interrupts per second. If the
92traffic is mostly transmit or mostly receive, the interrupt rate could
93be as high as 8000.
94
95Setting InterruptThrottleRate to 0 turns off any interrupt moderation
96and may improve small packet latency, but is generally not suitable
97for bulk throughput traffic.
98
99NOTE: InterruptThrottleRate takes precedence over the TxAbsIntDelay and
100 RxAbsIntDelay parameters. In other words, minimizing the receive
101 and/or transmit absolute delays does not force the controller to
102 generate more interrupts than what the Interrupt Throttle Rate
103 allows.
104
105NOTE: When e1000e is loaded with default settings and multiple adapters
106 are in use simultaneously, the CPU utilization may increase non-
107 linearly. In order to limit the CPU utilization without impacting
108 the overall throughput, we recommend that you load the driver as
109 follows:
110
111 modprobe e1000e InterruptThrottleRate=3000,3000,3000
112
113 This sets the InterruptThrottleRate to 3000 interrupts/sec for
114 the first, second, and third instances of the driver. The range
115 of 2000 to 3000 interrupts per second works on a majority of
116 systems and is a good starting point, but the optimal value will
117 be platform-specific. If CPU utilization is not a concern, use
118 RX_POLLING (NAPI) and default driver settings.
119
120RxIntDelay
121----------
122Valid Range: 0-65535 (0=off)
123Default Value: 0
124
125This value delays the generation of receive interrupts in units of 1.024
126microseconds. Receive interrupt reduction can improve CPU efficiency if
127properly tuned for specific network traffic. Increasing this value adds
128extra latency to frame reception and can end up decreasing the throughput
129of TCP traffic. If the system is reporting dropped receives, this value
130may be set too high, causing the driver to run out of available receive
131descriptors.
132
133CAUTION: When setting RxIntDelay to a value other than 0, adapters may
134 hang (stop transmitting) under certain network conditions. If
135 this occurs a NETDEV WATCHDOG message is logged in the system
136 event log. In addition, the controller is automatically reset,
137 restoring the network connection. To eliminate the potential
138 for the hang ensure that RxIntDelay is set to 0.
139
140RxAbsIntDelay
141-------------
142Valid Range: 0-65535 (0=off)
143Default Value: 8
144
145This value, in units of 1.024 microseconds, limits the delay in which a
146receive interrupt is generated. Useful only if RxIntDelay is non-zero,
147this value ensures that an interrupt is generated after the initial
148packet is received within the set amount of time. Proper tuning,
149along with RxIntDelay, may improve traffic throughput in specific network
150conditions.
151
152TxIntDelay
153----------
154Valid Range: 0-65535 (0=off)
155Default Value: 8
156
157This value delays the generation of transmit interrupts in units of
1581.024 microseconds. Transmit interrupt reduction can improve CPU
159efficiency if properly tuned for specific network traffic. If the
160system is reporting dropped transmits, this value may be set too high
161causing the driver to run out of available transmit descriptors.
162
163TxAbsIntDelay
164-------------
165Valid Range: 0-65535 (0=off)
166Default Value: 32
167
168This value, in units of 1.024 microseconds, limits the delay in which a
169transmit interrupt is generated. Useful only if TxIntDelay is non-zero,
170this value ensures that an interrupt is generated after the initial
171packet is sent on the wire within the set amount of time. Proper tuning,
172along with TxIntDelay, may improve traffic throughput in specific
173network conditions.
174
175Copybreak
176---------
177Valid Range: 0-xxxxxxx (0=off)
178Default Value: 256
179
180Driver copies all packets below or equaling this size to a fresh Rx
181buffer before handing it up the stack.
182
183This parameter is different than other parameters, in that it is a
184single (not 1,1,1 etc.) parameter applied to all driver instances and
185it is also available during runtime at
186/sys/module/e1000e/parameters/copybreak
187
188SmartPowerDownEnable
189--------------------
190Valid Range: 0-1
191Default Value: 0 (disabled)
192
193Allows PHY to turn off in lower power states. The user can set this parameter
194in supported chipsets.
195
196KumeranLockLoss
197---------------
198Valid Range: 0-1
199Default Value: 1 (enabled)
200
201This workaround skips resetting the PHY at shutdown for the initial
202silicon releases of ICH8 systems.
203
204IntMode
205-------
206Valid Range: 0-2 (0=legacy, 1=MSI, 2=MSI-X)
207Default Value: 2
208
209Allows changing the interrupt mode at module load time, without requiring a
210recompile. If the driver load fails to enable a specific interrupt mode, the
211driver will try other interrupt modes, from least to most compatible. The
212interrupt order is MSI-X, MSI, Legacy. If specifying MSI (IntMode=1)
213interrupts, only MSI and Legacy will be attempted.
214
215CrcStripping
216------------
217Valid Range: 0-1
218Default Value: 1 (enabled)
219
220Strip the CRC from received packets before sending up the network stack. If
221you have a machine with a BMC enabled but cannot receive IPMI traffic after
222loading or enabling the driver, try disabling this feature.
223
224WriteProtectNVM
225---------------
226Valid Range: 0-1
227Default Value: 1 (enabled)
228
229Set the hardware to ignore all write/erase cycles to the GbE region in the
230ICHx NVM (non-volatile memory). This feature can be disabled by the
231WriteProtectNVM module parameter (enabled by default) only after a hardware
232reset, but the machine must be power cycled before trying to enable writes.
233
234Note: the kernel boot option iomem=relaxed may need to be set if the kernel
235config option CONFIG_STRICT_DEVMEM=y, if the root user wants to write the
236NVM from user space via ethtool.
237
238Additional Configurations
239=========================
240
241 Jumbo Frames
242 ------------
243 Jumbo Frames support is enabled by changing the MTU to a value larger than
244 the default of 1500. Use the ifconfig command to increase the MTU size.
245 For example:
246
247 ifconfig eth<x> mtu 9000 up
248
249 This setting is not saved across reboots.
250
251 Notes:
252
253 - The maximum MTU setting for Jumbo Frames is 9216. This value coincides
254 with the maximum Jumbo Frames size of 9234 bytes.
255
256 - Using Jumbo Frames at 10 or 100 Mbps is not supported and may result in
257 poor performance or loss of link.
258
259 - Some adapters limit Jumbo Frames sized packets to a maximum of
260 4096 bytes and some adapters do not support Jumbo Frames.
261
262
263 Ethtool
264 -------
265 The driver utilizes the ethtool interface for driver configuration and
266 diagnostics, as well as displaying statistical information. We
267 strongly recommend downloading the latest version of Ethtool at:
268
269 http://sourceforge.net/projects/gkernel.
270
271 Speed and Duplex
272 ----------------
273 Speed and Duplex are configured through the Ethtool* utility. For
274 instructions, refer to the Ethtool man page.
275
276 Enabling Wake on LAN* (WoL)
277 ---------------------------
278 WoL is configured through the Ethtool* utility. For instructions on
279 enabling WoL with Ethtool, refer to the Ethtool man page.
280
281 WoL will be enabled on the system during the next shut down or reboot.
282 For this driver version, in order to enable WoL, the e1000e driver must be
283 loaded when shutting down or rebooting the system.
284
285 In most cases Wake On LAN is only supported on port A for multiple port
286 adapters. To verify if a port supports Wake on LAN run ethtool eth<X>.
287
288
289Support
290=======
291
292For general information, go to the Intel support website at:
293
294 www.intel.com/support/
295
296or the Intel Wired Networking project hosted by Sourceforge at:
297
298 http://sourceforge.net/projects/e1000
299
300If an issue is identified with the released source code on the supported
301kernel with a supported adapter, email the specific information related
302to the issue to e1000-devel@lists.sf.net