1 Documentation/networking/vortex.txt
2 Andrew Morton <andrewm@uow.edu.au>
6 This document describes the usage and errata of the 3Com "Vortex" device
7 driver for Linux, 3c59x.c.
9 The driver was written by Donald Becker <becker@scyld.com>
11 Don is no longer the prime maintainer of this version of the driver.
12 Please report problems to one or more of:
14 Andrew Morton <andrewm@uow.edu.au>
15 Netdev mailing list <netdev@oss.sgi.com>
16 Linux kernel mailing list <linux-kernel@vger.kernel.org>
18 Please note the 'Reporting and Diagnosing Problems' section at the end
22 Since kernel 2.3.99-pre6, this driver incorporates the support for the
23 3c575-series Cardbus cards which used to be handled by 3c575_cb.c.
25 This driver supports the following hardware:
28 3c592 EISA 10mbps Demon/Vortex
29 3c597 EISA Fast Demon/Vortex
30 3c595 Vortex 100baseTx
31 3c595 Vortex 100baseT4
32 3c595 Vortex 100base-MII
34 3c900 Boomerang 10baseT
35 3c900 Boomerang 10Mbps Combo
36 3c900 Cyclone 10Mbps TPO
37 3c900B Cyclone 10Mbps T
38 3c900 Cyclone 10Mbps Combo
39 3c900 Cyclone 10Mbps TPC
40 3c900B-FL Cyclone 10base-FL
41 3c905 Boomerang 100baseTx
42 3c905 Boomerang 100baseT4
43 3c905B Cyclone 100baseTx
44 3c905B Cyclone 10/100/BNC
45 3c905B-FX Cyclone 100baseFx
48 3cSOHO100-TX Hurricane
49 3c555 Laptop Hurricane
50 3c575 Boomerang CardBus
51 3CCFE575 Cyclone CardBus
52 3CCFE575CT Cyclone CardBus
53 3CCFE656 Cyclone CardBus
54 3CCFEM656 Cyclone CardBus
61 There are several parameters which may be provided to the driver when
62 its module is loaded. These are usually placed in /etc/modprobe.conf
63 (/etc/modules.conf in 2.4). Example:
65 options 3c59x debug=3 rx_copybreak=300
67 If you are using the PCMCIA tools (cardmgr) then the options may be
68 placed in /etc/pcmcia/config.opts:
70 module "3c59x" opts "debug=3 rx_copybreak=300"
73 The supported parameters are:
77 Where N is a number from 0 to 7. Anything above 3 produces a lot
78 of output in your system logs. debug=1 is default.
82 Each number in the list provides an option to the corresponding
83 network card. So if you have two 3c905's and you wish to provide
84 them with option 0x204 you would use:
88 The individual options are composed of a number of bitfields which
89 have the following meanings:
91 Possible media type settings
98 6 MII (Media Independent Interface)
99 7 Use default setting from EEPROM
102 10 Use default setting from EEPROM
104 When generating a value for the 'options' setting, the above media
105 selection values may be OR'ed (or added to) the following:
107 0x8000 Set driver debugging level to 7
108 0x4000 Set driver debugging level to 2
109 0x0400 Enable Wake-on-LAN
110 0x0200 Force full duplex mode.
111 0x0010 Bus-master enable bit (Old Vortex cards only)
115 insmod 3c59x options=0x204
117 will force full-duplex 100base-TX, rather than allowing the usual
122 Sets the `options' parameter for all 3c59x NICs in the machine.
123 Entries in the `options' array above will override any setting of
126 full_duplex=N1,N2,N3...
128 Similar to bit 9 of 'options'. Forces the corresponding card into
129 full-duplex mode. Please use this in preference to the `options'
132 In fact, please don't use this at all! You're better off getting
133 autonegotiation working properly.
135 global_full_duplex=N1
137 Sets full duplex mode for all 3c59x NICs in the machine. Entries
138 in the `full_duplex' array above will override any setting of this.
140 flow_ctrl=N1,N2,N3...
142 Use 802.3x MAC-layer flow control. The 3com cards only support the
143 PAUSE command, which means that they will stop sending packets for a
144 short period if they receive a PAUSE frame from the link partner.
146 The driver only allows flow control on a link which is operating in
149 This feature does not appear to work on the 3c905 - only 3c905B and
150 3c905C have been tested.
152 The 3com cards appear to only respond to PAUSE frames which are
153 sent to the reserved destination address of 01:80:c2:00:00:01. They
154 do not honour PAUSE frames which are sent to the station MAC address.
158 The driver preallocates 32 full-sized (1536 byte) network buffers
159 for receiving. When a packet arrives, the driver has to decide
160 whether to leave the packet in its full-sized buffer, or to allocate
161 a smaller buffer and copy the packet across into it.
163 This is a speed/space tradeoff.
165 The value of rx_copybreak is used to decide when to make the copy.
166 If the packet size is less than rx_copybreak, the packet is copied.
167 The default value for rx_copybreak is 200 bytes.
171 The driver's interrupt service routine can handle many receive and
172 transmit packets in a single invocation. It does this in a loop.
173 The value of max_interrupt_work governs how mnay times the interrupt
174 service routine will loop. The default value is 32 loops. If this
175 is exceeded the interrupt service routine gives up and generates a
176 warning message "eth0: Too much work in interrupt".
178 hw_checksums=N1,N2,N3,...
180 Recent 3com NICs are able to generate IPv4, TCP and UDP checksums
181 in hardware. Linux has used the Rx checksumming for a long time.
182 The "zero copy" patch which is planned for the 2.4 kernel series
183 allows you to make use of the NIC's DMA scatter/gather and transmit
184 checksumming as well.
186 The driver is set up so that, when the zerocopy patch is applied,
187 all Tornado and Cyclone devices will use S/G and Tx checksums.
189 This module parameter has been provided so you can override this
190 decision. If you think that Tx checksums are causing a problem, you
191 may disable the feature with `hw_checksums=0'.
193 If you think your NIC should be performing Tx checksumming and the
194 driver isn't enabling it, you can force the use of hardware Tx
195 checksumming with `hw_checksums=1'.
197 The driver drops a message in the logfiles to indicate whether or
198 not it is using hardware scatter/gather and hardware Tx checksums.
200 Scatter/gather and hardware checksums provide considerable
201 performance improvement for the sendfile() system call, but a small
202 decrease in throughput for send(). There is no effect upon receive
209 "Variables to work-around the Compaq PCI BIOS32 problem"....
213 Sets the time duration (in milliseconds) after which the kernel
214 decides that the transmitter has become stuck and needs to be reset.
215 This is mainly for debugging purposes, although it may be advantageous
216 to increase this value on LANs which have very high collision rates.
217 The default value is 5000 (5.0 seconds).
219 enable_wol=N1,N2,N3,...
221 Enable Wake-on-LAN support for the relevant interface. Donald
222 Becker's `ether-wake' application may be used to wake suspended
225 Also enables the NIC's power management support.
229 Sets enable_wol mode for all 3c59x NICs in the machine. Entries in
230 the `enable_wol' array above will override any setting of this.
235 A number of the older NICs such as the 3c590 and 3c900 series have
236 10base2 and AUI interfaces.
238 Prior to January, 2001 this driver would autoeselect the 10base2 or AUI
239 port if it didn't detect activity on the 10baseT port. It would then
240 get stuck on the 10base2 port and a driver reload was necessary to
241 switch back to 10baseT. This behaviour could not be prevented with a
242 module option override.
244 Later (current) versions of the driver _do_ support locking of the
245 media type. So if you load the driver module with
247 modprobe 3c59x options=0
249 it will permanently select the 10baseT port. Automatic selection of
250 other media types does not occur.
253 Transmit error, Tx status register 82
254 -------------------------------------
256 This is a common error which is almost always caused by another host on
257 the same network being in full-duplex mode, while this host is in
258 half-duplex mode. You need to find that other host and make it run in
259 half-duplex mode or fix this host to run in full-duplex mode.
261 As a last resort, you can force the 3c59x driver into full-duplex mode
264 options 3c59x full_duplex=1
266 but this has to be viewed as a workaround for broken network gear and
267 should only really be used for equipment which cannot autonegotiate.
273 Details of the device driver implementation are at the top of the source file.
275 Additional documentation is available at Don Becker's Linux Drivers site:
277 http://www.scyld.com/network/vortex.html
279 Donald Becker's driver development site:
281 http://www.scyld.com/network
283 Donald's vortex-diag program is useful for inspecting the NIC's state:
285 http://www.scyld.com/diag/#pci-diags
287 Donald's mii-diag program may be used for inspecting and manipulating
288 the NIC's Media Independent Interface subsystem:
290 http://www.scyld.com/diag/#mii-diag
292 Donald's wake-on-LAN page:
294 http://www.scyld.com/expert/wake-on-lan.html
296 3Com's documentation for many NICs, including the ones supported by
297 this driver is available at
299 http://support.3com.com/partners/developer/developer_form.html
301 3Com's DOS-based application for setting up the NICs EEPROMs:
303 ftp://ftp.3com.com/pub/nic/3c90x/3c90xx2.exe
305 Driver updates and a detailed changelog for the modifications which
306 were made for the 2.3/2,4 series kernel is available at
308 http://www.uow.edu.au/~andrewm/linux/#3c59x-2.3
311 Autonegotiation notes
312 ---------------------
314 The driver uses a one-minute heartbeat for adapting to changes in
315 the external LAN environment. This means that when, for example, a
316 machine is unplugged from a hubbed 10baseT LAN plugged into a
317 switched 100baseT LAN, the throughput will be quite dreadful for up
318 to sixty seconds. Be patient.
320 Cisco interoperability note from Walter Wong <wcw+@CMU.EDU>:
322 On a side note, adding HAS_NWAY seems to share a problem with the
323 Cisco 6509 switch. Specifically, you need to change the spanning
324 tree parameter for the port the machine is plugged into to 'portfast'
325 mode. Otherwise, the negotiation fails. This has been an issue
326 we've noticed for a while but haven't had the time to track down.
328 Cisco switches (Jeff Busch <jbusch@deja.com>)
330 My "standard config" for ports to which PC's/servers connect directly:
332 interface FastEthernet0/N
333 description machinename
335 spanning-tree portfast
337 If autonegotiation is a problem, you may need to specify "speed
338 100" and "duplex full" as well (or "speed 10" and "duplex half").
340 WARNING: DO NOT hook up hubs/switches/bridges to these
341 specially-configured ports! The switch will become very confused.
344 Reporting and diagnosing problems
345 ---------------------------------
347 Maintainers find that accurate and complete problem reports are
348 invaluable in resolving driver problems. We are frequently not able to
349 reproduce problems and must rely on your patience and efforts to get to
350 the bottom of the problem.
352 If you believe you have a driver problem here are some of the
353 steps you should take:
355 - Is it really a driver problem?
357 Eliminate some variables: try different cards, different
358 computers, different cables, different ports on the switch/hub,
359 different versions of the kernel or ofthe driver, etc.
361 - OK, it's a driver problem.
363 You need to generate a report. Typically this is an email to the
364 maintainer and/or linux-net@vger.kernel.org. The maintainer's
365 email address will be inthe driver source or in the MAINTAINERS file.
367 - The contents of your report will vary a lot depending upon the
368 problem. If it's a kernel crash then you should refer to the
371 But for most problems it is useful to provide the following:
373 o Kernel version, driver version
375 o A copy of the banner message which the driver generates when
376 it is initialised. For example:
378 eth0: 3Com PCI 3c905C Tornado at 0xa400, 00:50:da:6a:88:f0, IRQ 19
379 8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate interface.
380 MII transceiver found at address 24, status 782d.
381 Enabling bus-master transmits and whole-frame receives.
383 NOTE: You must provide the `debug=2' modprobe option to generate
384 a full detection message. Please do this:
386 modprobe 3c59x debug=2
388 o If it is a PCI device, the relevant output from 'lspci -vx', eg:
390 00:09.0 Ethernet controller: 3Com Corporation 3c905C-TX [Fast Etherlink] (rev 74)
391 Subsystem: 3Com Corporation: Unknown device 9200
392 Flags: bus master, medium devsel, latency 32, IRQ 19
393 I/O ports at a400 [size=128]
394 Memory at db000000 (32-bit, non-prefetchable) [size=128]
395 Expansion ROM at <unassigned> [disabled] [size=128K]
396 Capabilities: [dc] Power Management version 2
397 00: b7 10 00 92 07 00 10 02 74 00 00 02 08 20 00 00
398 10: 01 a4 00 00 00 00 00 db 00 00 00 00 00 00 00 00
399 20: 00 00 00 00 00 00 00 00 00 00 00 00 b7 10 00 10
400 30: 00 00 00 00 dc 00 00 00 00 00 00 00 05 01 0a 0a
402 o A description of the environment: 10baseT? 100baseT?
403 full/half duplex? switched or hubbed?
405 o Any additional module parameters which you may be providing to the driver.
407 o Any kernel logs which are produced. The more the merrier.
408 If this is a large file and you are sending your report to a
409 mailing list, mention that you have the logfile, but don't send
410 it. If you're reporting direct to the maintainer then just send
413 To ensure that all kernel logs are available, add the
414 following line to /etc/syslog.conf:
416 kern.* /var/log/messages
418 Then restart syslogd with:
420 /etc/rc.d/init.d/syslog restart
422 (The above may vary, depending upon which Linux distribution you use).
424 o If your problem is reproducible then that's great. Try the
427 1) Increase the debug level. Usually this is done via:
429 a) modprobe driver debug=7
430 b) In /etc/modprobe.conf (or /etc/modules.conf for 2.4):
431 options driver debug=7
433 2) Recreate the problem with the higher debug level,
434 send all logs to the maintainer.
436 3) Download you card's diagnostic tool from Donald
437 Backer's website http://www.scyld.com/diag. Download
438 mii-diag.c as well. Build these.
440 a) Run 'vortex-diag -aaee' and 'mii-diag -v' when the card is
441 working correctly. Save the output.
443 b) Run the above commands when the card is malfunctioning. Send
446 Finally, please be patient and be prepared to do some work. You may end up working on
447 this problem for a week or more as the maintainer asks more questions, asks for more
448 tests, asks for patches to be applied, etc. At the end of it all, the problem may even