]> git.proxmox.com Git - mirror_ubuntu-bionic-kernel.git/blame - Documentation/networking/e1000e.txt
Merge git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending
[mirror_ubuntu-bionic-kernel.git] / Documentation / networking / e1000e.txt
CommitLineData
d7064f4c
JK
1Linux* Driver for Intel(R) Ethernet Network Connection
2======================================================
da8c01c4
JK
3
4Intel Gigabit Linux driver.
d7064f4c 5Copyright(c) 1999 - 2013 Intel Corporation.
da8c01c4
JK
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
072ed34f
JK
64The default behaviour of the driver previously assumed a static
65InterruptThrottleRate value of 8000, providing a good fallback value for
66all traffic types, but lacking in small packet performance and latency.
67The hardware can handle many more small packets per second however, and
68for this reason an adaptive interrupt moderation algorithm was implemented.
69
da8c01c4
JK
70The driver has two adaptive modes (setting 1 or 3) in which
71it dynamically adjusts the InterruptThrottleRate value based on the traffic
72that it receives. After determining the type of incoming traffic in the last
73timeframe, it will adjust the InterruptThrottleRate to an appropriate value
74for that traffic.
75
76The algorithm classifies the incoming traffic every interval into
77classes. Once the class is determined, the InterruptThrottleRate value is
78adjusted to suit that traffic type the best. There are three classes defined:
79"Bulk traffic", for large amounts of packets of normal size; "Low latency",
80for small amounts of traffic and/or a significant percentage of small
81packets; and "Lowest latency", for almost completely small packets or
82minimal traffic.
83
84In dynamic conservative mode, the InterruptThrottleRate value is set to 4000
85for traffic that falls in class "Bulk traffic". If traffic falls in the "Low
86latency" or "Lowest latency" class, the InterruptThrottleRate is increased
87stepwise to 20000. This default mode is suitable for most applications.
88
89For situations where low latency is vital such as cluster or
90grid computing, the algorithm can reduce latency even more when
91InterruptThrottleRate is set to mode 1. In this mode, which operates
92the same as mode 3, the InterruptThrottleRate will be increased stepwise to
9370000 for traffic in class "Lowest latency".
94
072ed34f
JK
95In simplified mode the interrupt rate is based on the ratio of TX and
96RX traffic. If the bytes per second rate is approximately equal, the
da8c01c4
JK
97interrupt rate will drop as low as 2000 interrupts per second. If the
98traffic is mostly transmit or mostly receive, the interrupt rate could
99be as high as 8000.
100
101Setting InterruptThrottleRate to 0 turns off any interrupt moderation
102and may improve small packet latency, but is generally not suitable
103for bulk throughput traffic.
104
105NOTE: InterruptThrottleRate takes precedence over the TxAbsIntDelay and
106 RxAbsIntDelay parameters. In other words, minimizing the receive
107 and/or transmit absolute delays does not force the controller to
108 generate more interrupts than what the Interrupt Throttle Rate
109 allows.
110
111NOTE: When e1000e is loaded with default settings and multiple adapters
112 are in use simultaneously, the CPU utilization may increase non-
113 linearly. In order to limit the CPU utilization without impacting
114 the overall throughput, we recommend that you load the driver as
115 follows:
116
117 modprobe e1000e InterruptThrottleRate=3000,3000,3000
118
119 This sets the InterruptThrottleRate to 3000 interrupts/sec for
120 the first, second, and third instances of the driver. The range
121 of 2000 to 3000 interrupts per second works on a majority of
122 systems and is a good starting point, but the optimal value will
123 be platform-specific. If CPU utilization is not a concern, use
124 RX_POLLING (NAPI) and default driver settings.
125
126RxIntDelay
127----------
128Valid Range: 0-65535 (0=off)
129Default Value: 0
130
131This value delays the generation of receive interrupts in units of 1.024
132microseconds. Receive interrupt reduction can improve CPU efficiency if
133properly tuned for specific network traffic. Increasing this value adds
134extra latency to frame reception and can end up decreasing the throughput
135of TCP traffic. If the system is reporting dropped receives, this value
136may be set too high, causing the driver to run out of available receive
137descriptors.
138
139CAUTION: When setting RxIntDelay to a value other than 0, adapters may
140 hang (stop transmitting) under certain network conditions. If
141 this occurs a NETDEV WATCHDOG message is logged in the system
142 event log. In addition, the controller is automatically reset,
143 restoring the network connection. To eliminate the potential
144 for the hang ensure that RxIntDelay is set to 0.
145
146RxAbsIntDelay
147-------------
148Valid Range: 0-65535 (0=off)
149Default Value: 8
150
151This value, in units of 1.024 microseconds, limits the delay in which a
152receive interrupt is generated. Useful only if RxIntDelay is non-zero,
153this value ensures that an interrupt is generated after the initial
154packet is received within the set amount of time. Proper tuning,
155along with RxIntDelay, may improve traffic throughput in specific network
156conditions.
157
158TxIntDelay
159----------
160Valid Range: 0-65535 (0=off)
161Default Value: 8
162
163This value delays the generation of transmit interrupts in units of
1641.024 microseconds. Transmit interrupt reduction can improve CPU
165efficiency if properly tuned for specific network traffic. If the
166system is reporting dropped transmits, this value may be set too high
167causing the driver to run out of available transmit descriptors.
168
169TxAbsIntDelay
170-------------
171Valid Range: 0-65535 (0=off)
172Default Value: 32
173
174This value, in units of 1.024 microseconds, limits the delay in which a
175transmit interrupt is generated. Useful only if TxIntDelay is non-zero,
176this value ensures that an interrupt is generated after the initial
177packet is sent on the wire within the set amount of time. Proper tuning,
178along with TxIntDelay, may improve traffic throughput in specific
179network conditions.
180
181Copybreak
182---------
183Valid Range: 0-xxxxxxx (0=off)
184Default Value: 256
185
072ed34f 186Driver copies all packets below or equaling this size to a fresh RX
da8c01c4
JK
187buffer before handing it up the stack.
188
189This parameter is different than other parameters, in that it is a
190single (not 1,1,1 etc.) parameter applied to all driver instances and
191it is also available during runtime at
192/sys/module/e1000e/parameters/copybreak
193
194SmartPowerDownEnable
195--------------------
196Valid Range: 0-1
197Default Value: 0 (disabled)
198
199Allows PHY to turn off in lower power states. The user can set this parameter
200in supported chipsets.
201
202KumeranLockLoss
203---------------
204Valid Range: 0-1
205Default Value: 1 (enabled)
206
207This workaround skips resetting the PHY at shutdown for the initial
208silicon releases of ICH8 systems.
209
210IntMode
211-------
212Valid Range: 0-2 (0=legacy, 1=MSI, 2=MSI-X)
213Default Value: 2
214
215Allows changing the interrupt mode at module load time, without requiring a
216recompile. If the driver load fails to enable a specific interrupt mode, the
217driver will try other interrupt modes, from least to most compatible. The
218interrupt order is MSI-X, MSI, Legacy. If specifying MSI (IntMode=1)
219interrupts, only MSI and Legacy will be attempted.
220
221CrcStripping
222------------
223Valid Range: 0-1
224Default Value: 1 (enabled)
225
226Strip the CRC from received packets before sending up the network stack. If
227you have a machine with a BMC enabled but cannot receive IPMI traffic after
228loading or enabling the driver, try disabling this feature.
229
230WriteProtectNVM
231---------------
072ed34f
JK
232Valid Range: 0,1
233Default Value: 1
234
235If set to 1, configure the hardware to ignore all write/erase cycles to the
236GbE region in the ICHx NVM (in order to prevent accidental corruption of the
237NVM). This feature can be disabled by setting the parameter to 0 during initial
238driver load.
239NOTE: The machine must be power cycled (full off/on) when enabling NVM writes
240via setting the parameter to zero. Once the NVM has been locked (via the
241parameter at 1 when the driver loads) it cannot be unlocked except via power
242cycle.
da8c01c4
JK
243
244Additional Configurations
245=========================
246
247 Jumbo Frames
248 ------------
249 Jumbo Frames support is enabled by changing the MTU to a value larger than
250 the default of 1500. Use the ifconfig command to increase the MTU size.
251 For example:
252
253 ifconfig eth<x> mtu 9000 up
254
255 This setting is not saved across reboots.
256
257 Notes:
258
259 - The maximum MTU setting for Jumbo Frames is 9216. This value coincides
260 with the maximum Jumbo Frames size of 9234 bytes.
261
d7064f4c 262 - Using Jumbo frames at 10 or 100 Mbps is not supported and may result in
da8c01c4
JK
263 poor performance or loss of link.
264
265 - Some adapters limit Jumbo Frames sized packets to a maximum of
266 4096 bytes and some adapters do not support Jumbo Frames.
267
d7064f4c
JK
268 - Jumbo Frames cannot be configured on an 82579-based Network device, if
269 MACSec is enabled on the system.
270
271 ethtool
da8c01c4
JK
272 -------
273 The driver utilizes the ethtool interface for driver configuration and
274 diagnostics, as well as displaying statistical information. We
68f20d94 275 strongly recommend downloading the latest version of ethtool at:
da8c01c4 276
68f20d94 277 http://ftp.kernel.org/pub/software/network/ethtool/
da8c01c4 278
d7064f4c
JK
279 NOTE: When validating enable/disable tests on some parts (82578, for example)
280 you need to add a few seconds between tests when working with ethtool.
281
da8c01c4
JK
282 Speed and Duplex
283 ----------------
68f20d94
JK
284 Speed and Duplex are configured through the ethtool* utility. For
285 instructions, refer to the ethtool man page.
da8c01c4
JK
286
287 Enabling Wake on LAN* (WoL)
288 ---------------------------
68f20d94
JK
289 WoL is configured through the ethtool* utility. For instructions on
290 enabling WoL with ethtool, refer to the ethtool man page.
da8c01c4
JK
291
292 WoL will be enabled on the system during the next shut down or reboot.
293 For this driver version, in order to enable WoL, the e1000e driver must be
294 loaded when shutting down or rebooting the system.
295
296 In most cases Wake On LAN is only supported on port A for multiple port
68f20d94 297 adapters. To verify if a port supports Wake on Lan run ethtool eth<X>.
da8c01c4
JK
298
299Support
300=======
301
302For general information, go to the Intel support website at:
303
304 www.intel.com/support/
305
306or the Intel Wired Networking project hosted by Sourceforge at:
307
308 http://sourceforge.net/projects/e1000
309
310If an issue is identified with the released source code on the supported
311kernel with a supported adapter, email the specific information related
312to the issue to e1000-devel@lists.sf.net