1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 2
3 ====== 3 ======
4 ARCnet 4 ARCnet
5 ====== 5 ======
6 6
7 .. note:: 7 .. note::
8 8
9 See also arcnet-hardware.txt in this direct 9 See also arcnet-hardware.txt in this directory for jumper-setting
10 and cabling information if you're like many 10 and cabling information if you're like many of us and didn't happen to get a
11 manual with your ARCnet card. 11 manual with your ARCnet card.
12 12
13 Since no one seems to listen to me otherwise, 13 Since no one seems to listen to me otherwise, perhaps a poem will get your
14 attention:: 14 attention::
15 15
16 This driver's getting fat and 16 This driver's getting fat and beefy,
17 But my cat is still named Fifi 17 But my cat is still named Fifi.
18 18
19 Hmm, I think I'm allowed to call that a poem, 19 Hmm, I think I'm allowed to call that a poem, even though it's only two
20 lines. Hey, I'm in Computer Science, not Engl 20 lines. Hey, I'm in Computer Science, not English. Give me a break.
21 21
22 The point is: I REALLY REALLY REALLY REALLY R 22 The point is: I REALLY REALLY REALLY REALLY REALLY want to hear from you if
23 you test this and get it working. Or if you d 23 you test this and get it working. Or if you don't. Or anything.
24 24
25 ARCnet 0.32 ALPHA first made it into the Linux 25 ARCnet 0.32 ALPHA first made it into the Linux kernel 1.1.80 - this was
26 nice, but after that even FEWER people started 26 nice, but after that even FEWER people started writing to me because they
27 didn't even have to install the patch. <sigh> 27 didn't even have to install the patch. <sigh>
28 28
29 Come on, be a sport! Send me a success report 29 Come on, be a sport! Send me a success report!
30 30
31 (hey, that was even better than my original po 31 (hey, that was even better than my original poem... this is getting bad!)
32 32
33 33
34 .. warning:: 34 .. warning::
35 35
36 If you don't e-mail me about your success/f 36 If you don't e-mail me about your success/failure soon, I may be forced to
37 start SINGING. And we don't want that, do 37 start SINGING. And we don't want that, do we?
38 38
39 (You know, it might be argued that I'm push 39 (You know, it might be argued that I'm pushing this point a little too much.
40 If you think so, why not flame me in a quic 40 If you think so, why not flame me in a quick little e-mail? Please also
41 include the type of card(s) you're using, s 41 include the type of card(s) you're using, software, size of network, and
42 whether it's working or not.) 42 whether it's working or not.)
43 43
44 My e-mail address is: apenwarr@worldvisions 44 My e-mail address is: apenwarr@worldvisions.ca
45 45
46 These are the ARCnet drivers for Linux. 46 These are the ARCnet drivers for Linux.
47 47
48 This new release (2.91) has been put together 48 This new release (2.91) has been put together by David Woodhouse
49 <dwmw2@infradead.org>, in an attempt to tidy up 49 <dwmw2@infradead.org>, in an attempt to tidy up the driver after adding support
50 for yet another chipset. Now the generic suppo 50 for yet another chipset. Now the generic support has been separated from the
51 individual chipset drivers, and the source fil 51 individual chipset drivers, and the source files aren't quite so packed with
52 #ifdefs! I've changed this file a bit, but kep 52 #ifdefs! I've changed this file a bit, but kept it in the first person from
53 Avery, because I didn't want to completely rew 53 Avery, because I didn't want to completely rewrite it.
54 54
55 The previous release resulted from many months 55 The previous release resulted from many months of on-and-off effort from me
56 (Avery Pennarun), many bug reports/fixes and s 56 (Avery Pennarun), many bug reports/fixes and suggestions from others, and in
57 particular a lot of input and coding from Toma 57 particular a lot of input and coding from Tomasz Motylewski. Starting with
58 ARCnet 2.10 ALPHA, Tomasz's all-new-and-improv 58 ARCnet 2.10 ALPHA, Tomasz's all-new-and-improved RFC1051 support has been
59 included and seems to be working fine! 59 included and seems to be working fine!
60 60
61 61
62 Where do I discuss these drivers? 62 Where do I discuss these drivers?
63 --------------------------------- 63 ---------------------------------
64 64
65 Tomasz has been so kind as to set up a new and 65 Tomasz has been so kind as to set up a new and improved mailing list.
66 Subscribe by sending a message with the BODY " 66 Subscribe by sending a message with the BODY "subscribe linux-arcnet YOUR
67 REAL NAME" to listserv@tichy.ch.uj.edu.pl. Th 67 REAL NAME" to listserv@tichy.ch.uj.edu.pl. Then, to submit messages to the
68 list, mail to linux-arcnet@tichy.ch.uj.edu.pl. 68 list, mail to linux-arcnet@tichy.ch.uj.edu.pl.
69 69
70 There are archives of the mailing list at: 70 There are archives of the mailing list at:
71 71
72 http://epistolary.org/mailman/listinfo 72 http://epistolary.org/mailman/listinfo.cgi/arcnet
73 73
74 The people on linux-net@vger.kernel.org (now d 74 The people on linux-net@vger.kernel.org (now defunct, replaced by
75 netdev@vger.kernel.org) have also been known t 75 netdev@vger.kernel.org) have also been known to be very helpful, especially
76 when we're talking about ALPHA Linux kernels t 76 when we're talking about ALPHA Linux kernels that may or may not work right
77 in the first place. 77 in the first place.
78 78
79 79
80 Other Drivers and Info 80 Other Drivers and Info
81 ---------------------- 81 ----------------------
82 82
83 You can try my ARCNET page on the World Wide W 83 You can try my ARCNET page on the World Wide Web at:
84 84
85 http://www.qis.net/~jschmitz/arcnet/ 85 http://www.qis.net/~jschmitz/arcnet/
86 86
87 Also, SMC (one of the companies that makes ARC 87 Also, SMC (one of the companies that makes ARCnet cards) has a WWW site you
88 might be interested in, which includes several 88 might be interested in, which includes several drivers for various cards
89 including ARCnet. Try: 89 including ARCnet. Try:
90 90
91 http://www.smc.com/ 91 http://www.smc.com/
92 92
93 Performance Technologies makes various network 93 Performance Technologies makes various network software that supports
94 ARCnet: 94 ARCnet:
95 95
96 http://www.perftech.com/ or ftp to ftp 96 http://www.perftech.com/ or ftp to ftp.perftech.com.
97 97
98 Novell makes a networking stack for DOS which 98 Novell makes a networking stack for DOS which includes ARCnet drivers. Try
99 FTPing to ftp.novell.com. 99 FTPing to ftp.novell.com.
100 100
101 You can get the Crynwr packet driver collectio 101 You can get the Crynwr packet driver collection (including arcether.com, the
102 one you'll want to use with ARCnet cards) from 102 one you'll want to use with ARCnet cards) from
103 oak.oakland.edu:/simtel/msdos/pktdrvr. It won' 103 oak.oakland.edu:/simtel/msdos/pktdrvr. It won't work perfectly on a 386+
104 without patches, though, and also doesn't like 104 without patches, though, and also doesn't like several cards. Fixed
105 versions are available on my WWW page, or via 105 versions are available on my WWW page, or via e-mail if you don't have WWW
106 access. 106 access.
107 107
108 108
109 Installing the Driver 109 Installing the Driver
110 --------------------- 110 ---------------------
111 111
112 All you will need to do in order to install th 112 All you will need to do in order to install the driver is::
113 113
114 make config 114 make config
115 (be sure to choose ARCnet in t 115 (be sure to choose ARCnet in the network devices
116 and at least one chipset drive 116 and at least one chipset driver.)
117 make clean 117 make clean
118 make zImage 118 make zImage
119 119
120 If you obtained this ARCnet package as an upgr 120 If you obtained this ARCnet package as an upgrade to the ARCnet driver in
121 your current kernel, you will need to first co 121 your current kernel, you will need to first copy arcnet.c over the one in
122 the linux/drivers/net directory. 122 the linux/drivers/net directory.
123 123
124 You will know the driver is installed properly 124 You will know the driver is installed properly if you get some ARCnet
125 messages when you reboot into the new Linux ke 125 messages when you reboot into the new Linux kernel.
126 126
127 There are four chipset options: 127 There are four chipset options:
128 128
129 1. Standard ARCnet COM90xx chipset. 129 1. Standard ARCnet COM90xx chipset.
130 130
131 This is the normal ARCnet card, which you've p 131 This is the normal ARCnet card, which you've probably got. This is the only
132 chipset driver which will autoprobe if not tol 132 chipset driver which will autoprobe if not told where the card is.
133 It following options on the command line:: 133 It following options on the command line::
134 134
135 com90xx=[<io>[,<irq>[,<shmem>]]][,<name>] | < 135 com90xx=[<io>[,<irq>[,<shmem>]]][,<name>] | <name>
136 136
137 If you load the chipset support as a module, t 137 If you load the chipset support as a module, the options are::
138 138
139 io=<io> irq=<irq> shmem=<shmem> device=<name> 139 io=<io> irq=<irq> shmem=<shmem> device=<name>
140 140
141 To disable the autoprobe, just specify "com90x 141 To disable the autoprobe, just specify "com90xx=" on the kernel command line.
142 To specify the name alone, but allow autoprobe 142 To specify the name alone, but allow autoprobe, just put "com90xx=<name>"
143 143
144 2. ARCnet COM20020 chipset. 144 2. ARCnet COM20020 chipset.
145 145
146 This is the new chipset from SMC with support 146 This is the new chipset from SMC with support for promiscuous mode (packet
147 sniffing), extra diagnostic information, etc. 147 sniffing), extra diagnostic information, etc. Unfortunately, there is no
148 sensible method of autoprobing for these cards 148 sensible method of autoprobing for these cards. You must specify the I/O
149 address on the kernel command line. 149 address on the kernel command line.
150 150
151 The command line options are:: 151 The command line options are::
152 152
153 com20020=<io>[,<irq>[,<node_ID>[,backplane[,C 153 com20020=<io>[,<irq>[,<node_ID>[,backplane[,CKP[,timeout]]]]][,name]
154 154
155 If you load the chipset support as a module, t 155 If you load the chipset support as a module, the options are::
156 156
157 io=<io> irq=<irq> node=<node_ID> backplane=<b 157 io=<io> irq=<irq> node=<node_ID> backplane=<backplane> clock=<CKP>
158 timeout=<timeout> device=<name> 158 timeout=<timeout> device=<name>
159 159
160 The COM20020 chipset allows you to set the nod 160 The COM20020 chipset allows you to set the node ID in software, overriding the
161 default which is still set in DIP switches on 161 default which is still set in DIP switches on the card. If you don't have the
162 COM20020 data sheets, and you don't know what 162 COM20020 data sheets, and you don't know what the other three options refer
163 to, then they won't interest you - forget them 163 to, then they won't interest you - forget them.
164 164
165 3. ARCnet COM90xx chipset in IO-mapped mode. 165 3. ARCnet COM90xx chipset in IO-mapped mode.
166 166
167 This will also work with the normal ARCnet car 167 This will also work with the normal ARCnet cards, but doesn't use the shared
168 memory. It performs less well than the above d 168 memory. It performs less well than the above driver, but is provided in case
169 you have a card which doesn't support shared m 169 you have a card which doesn't support shared memory, or (strangely) in case
170 you have so many ARCnet cards in your machine 170 you have so many ARCnet cards in your machine that you run out of shmem slots.
171 If you don't give the IO address on the kernel 171 If you don't give the IO address on the kernel command line, then the driver
172 will not find the card. 172 will not find the card.
173 173
174 The command line options are:: 174 The command line options are::
175 175
176 com90io=<io>[,<irq>][,<name>] 176 com90io=<io>[,<irq>][,<name>]
177 177
178 If you load the chipset support as a module, t 178 If you load the chipset support as a module, the options are:
179 io=<io> irq=<irq> device=<name> 179 io=<io> irq=<irq> device=<name>
180 180
181 4. ARCnet RIM I cards. 181 4. ARCnet RIM I cards.
182 182
183 These are COM90xx chips which are _completely_ 183 These are COM90xx chips which are _completely_ memory mapped. The support for
184 these is not tested. If you have one, please m 184 these is not tested. If you have one, please mail the author with a success
185 report. All options must be specified, except 185 report. All options must be specified, except the device name.
186 Command line options:: 186 Command line options::
187 187
188 arcrimi=<shmem>,<irq>,<node_ID>[,<name>] 188 arcrimi=<shmem>,<irq>,<node_ID>[,<name>]
189 189
190 If you load the chipset support as a module, t 190 If you load the chipset support as a module, the options are::
191 191
192 shmem=<shmem> irq=<irq> node=<node_ID> device 192 shmem=<shmem> irq=<irq> node=<node_ID> device=<name>
193 193
194 194
195 Loadable Module Support 195 Loadable Module Support
196 ----------------------- 196 -----------------------
197 197
198 Configure and rebuild Linux. When asked, answ 198 Configure and rebuild Linux. When asked, answer 'm' to "Generic ARCnet
199 support" and to support for your ARCnet chipse 199 support" and to support for your ARCnet chipset if you want to use the
200 loadable module. You can also say 'y' to "Gene 200 loadable module. You can also say 'y' to "Generic ARCnet support" and 'm'
201 to the chipset support if you wish. 201 to the chipset support if you wish.
202 202
203 :: 203 ::
204 204
205 make config 205 make config
206 make clean 206 make clean
207 make zImage 207 make zImage
208 make modules 208 make modules
209 209
210 If you're using a loadable module, you need to 210 If you're using a loadable module, you need to use insmod to load it, and
211 you can specify various characteristics of you 211 you can specify various characteristics of your card on the command
212 line. (In recent versions of the driver, auto 212 line. (In recent versions of the driver, autoprobing is much more reliable
213 and works as a module, so most of this is now 213 and works as a module, so most of this is now unnecessary.)
214 214
215 For example:: 215 For example::
216 216
217 cd /usr/src/linux/modules 217 cd /usr/src/linux/modules
218 insmod arcnet.o 218 insmod arcnet.o
219 insmod com90xx.o 219 insmod com90xx.o
220 insmod com20020.o io=0x2e0 device=eth1 220 insmod com20020.o io=0x2e0 device=eth1
221 221
222 222
223 Using the Driver 223 Using the Driver
224 ---------------- 224 ----------------
225 225
226 If you build your kernel with ARCnet COM90xx s 226 If you build your kernel with ARCnet COM90xx support included, it should
227 probe for your card automatically when you boo 227 probe for your card automatically when you boot. If you use a different
228 chipset driver complied into the kernel, you m 228 chipset driver complied into the kernel, you must give the necessary options
229 on the kernel command line, as detailed above. 229 on the kernel command line, as detailed above.
230 230
231 Go read the NET-2-HOWTO and ETHERNET-HOWTO for 231 Go read the NET-2-HOWTO and ETHERNET-HOWTO for Linux; they should be
232 available where you picked up this driver. Th 232 available where you picked up this driver. Think of your ARCnet as a
233 souped-up (or down, as the case may be) Ethern 233 souped-up (or down, as the case may be) Ethernet card.
234 234
235 By the way, be sure to change all references f 235 By the way, be sure to change all references from "eth0" to "arc0" in the
236 HOWTOs. Remember that ARCnet isn't a "true" E 236 HOWTOs. Remember that ARCnet isn't a "true" Ethernet, and the device name
237 is DIFFERENT. 237 is DIFFERENT.
238 238
239 239
240 Multiple Cards in One Computer 240 Multiple Cards in One Computer
241 ------------------------------ 241 ------------------------------
242 242
243 Linux has pretty good support for this now, bu 243 Linux has pretty good support for this now, but since I've been busy, the
244 ARCnet driver has somewhat suffered in this re 244 ARCnet driver has somewhat suffered in this respect. COM90xx support, if
245 compiled into the kernel, will (try to) autode 245 compiled into the kernel, will (try to) autodetect all the installed cards.
246 246
247 If you have other cards, with support compiled 247 If you have other cards, with support compiled into the kernel, then you can
248 just repeat the options on the kernel command 248 just repeat the options on the kernel command line, e.g.::
249 249
250 LILO: linux com20020=0x2e0 com20020=0x 250 LILO: linux com20020=0x2e0 com20020=0x380 com90io=0x260
251 251
252 If you have the chipset support built as a loa 252 If you have the chipset support built as a loadable module, then you need to
253 do something like this:: 253 do something like this::
254 254
255 insmod -o arc0 com90xx 255 insmod -o arc0 com90xx
256 insmod -o arc1 com20020 io=0x2e0 256 insmod -o arc1 com20020 io=0x2e0
257 insmod -o arc2 com90xx 257 insmod -o arc2 com90xx
258 258
259 The ARCnet drivers will now sort out their nam 259 The ARCnet drivers will now sort out their names automatically.
260 260
261 261
262 How do I get it to work with...? 262 How do I get it to work with...?
263 -------------------------------- 263 --------------------------------
264 264
265 NFS: 265 NFS:
266 Should be fine linux->linux, just pret 266 Should be fine linux->linux, just pretend you're using Ethernet cards.
267 oak.oakland.edu:/simtel/msdos/nfs has 267 oak.oakland.edu:/simtel/msdos/nfs has some nice DOS clients. There
268 is also a DOS-based NFS server called 268 is also a DOS-based NFS server called SOSS. It doesn't multitask
269 quite the way Linux does (actually, it 269 quite the way Linux does (actually, it doesn't multitask AT ALL) but
270 you never know what you might need. 270 you never know what you might need.
271 271
272 With AmiTCP (and possibly others), you 272 With AmiTCP (and possibly others), you may need to set the following
273 options in your Amiga nfstab: MD 1024 273 options in your Amiga nfstab: MD 1024 MR 1024 MW 1024
274 (Thanks to Christian Gottschling <ferks 274 (Thanks to Christian Gottschling <ferksy@indigo.tng.oche.de>
275 for this.) 275 for this.)
276 276
277 Probably these refer to maximum NFS da 277 Probably these refer to maximum NFS data/read/write block sizes. I
278 don't know why the defaults on the Ami 278 don't know why the defaults on the Amiga didn't work; write to me if
279 you know more. 279 you know more.
280 280
281 DOS: 281 DOS:
282 If you're using the freeware arcether. 282 If you're using the freeware arcether.com, you might want to install
283 the driver patch from my web page. It 283 the driver patch from my web page. It helps with PC/TCP, and also
284 can get arcether to load if it timed o 284 can get arcether to load if it timed out too quickly during
285 initialization. In fact, if you use i 285 initialization. In fact, if you use it on a 386+ you REALLY need
286 the patch, really. 286 the patch, really.
287 287
288 Windows: 288 Windows:
289 See DOS :) Trumpet Winsock works fine 289 See DOS :) Trumpet Winsock works fine with either the Novell or
290 Arcether client, assuming you remember 290 Arcether client, assuming you remember to load winpkt of course.
291 291
292 LAN Manager and Windows for Workgroups: 292 LAN Manager and Windows for Workgroups:
293 These programs use protocols that 293 These programs use protocols that
294 are incompatible with the Internet sta 294 are incompatible with the Internet standard. They try to pretend
295 the cards are Ethernet, and confuse ev 295 the cards are Ethernet, and confuse everyone else on the network.
296 296
297 However, v2.00 and higher of the Linux 297 However, v2.00 and higher of the Linux ARCnet driver supports this
298 protocol via the 'arc0e' device. See 298 protocol via the 'arc0e' device. See the section on "Multiprotocol
299 Support" for more information. 299 Support" for more information.
300 300
301 Using the freeware Samba server and cl 301 Using the freeware Samba server and clients for Linux, you can now
302 interface quite nicely with TCP/IP-bas 302 interface quite nicely with TCP/IP-based WfWg or Lan Manager
303 networks. 303 networks.
304 304
305 Windows 95: 305 Windows 95:
306 Tools are included with Win95 that let 306 Tools are included with Win95 that let you use either the LANMAN
307 style network drivers (NDIS) or Novell 307 style network drivers (NDIS) or Novell drivers (ODI) to handle your
308 ARCnet packets. If you use ODI, you'l 308 ARCnet packets. If you use ODI, you'll need to use the 'arc0'
309 device with Linux. If you use NDIS, t 309 device with Linux. If you use NDIS, then try the 'arc0e' device.
310 See the "Multiprotocol Support" sectio 310 See the "Multiprotocol Support" section below if you need arc0e,
311 you're completely insane, and/or you n 311 you're completely insane, and/or you need to build some kind of
312 hybrid network that uses both encapsul 312 hybrid network that uses both encapsulation types.
313 313
314 OS/2: 314 OS/2:
315 I've been told it works under Warp Con 315 I've been told it works under Warp Connect with an ARCnet driver from
316 SMC. You need to use the 'arc0e' inte 316 SMC. You need to use the 'arc0e' interface for this. If you get
317 the SMC driver to work with the TCP/IP 317 the SMC driver to work with the TCP/IP stuff included in the
318 "normal" Warp Bonus Pack, let me know. 318 "normal" Warp Bonus Pack, let me know.
319 319
320 ftp.microsoft.com also has a freeware 320 ftp.microsoft.com also has a freeware "Lan Manager for OS/2" client
321 which should use the same protocol as 321 which should use the same protocol as WfWg does. I had no luck
322 installing it under Warp, however. Pl 322 installing it under Warp, however. Please mail me with any results.
323 323
324 NetBSD/AmiTCP: 324 NetBSD/AmiTCP:
325 These use an old version of the Intern 325 These use an old version of the Internet standard ARCnet
326 protocol (RFC1051) which is compatible 326 protocol (RFC1051) which is compatible with the Linux driver v2.10
327 ALPHA and above using the arc0s device 327 ALPHA and above using the arc0s device. (See "Multiprotocol ARCnet"
328 below.) ** Newer versions of NetBSD a 328 below.) ** Newer versions of NetBSD apparently support RFC1201.
329 329
330 330
331 Using Multiprotocol ARCnet 331 Using Multiprotocol ARCnet
332 -------------------------- 332 --------------------------
333 333
334 The ARCnet driver v2.10 ALPHA supports three p 334 The ARCnet driver v2.10 ALPHA supports three protocols, each on its own
335 "virtual network device": 335 "virtual network device":
336 336
337 ====== ============================== 337 ====== ===============================================================
338 arc0 RFC1201 protocol, the official 338 arc0 RFC1201 protocol, the official Internet standard which just
339 happens to be 100% compatible 339 happens to be 100% compatible with Novell's TRXNET driver.
340 Version 1.00 of the ARCnet dri 340 Version 1.00 of the ARCnet driver supported _only_ this
341 protocol. arc0 is the fastest 341 protocol. arc0 is the fastest of the three protocols (for
342 whatever reason), and allows l 342 whatever reason), and allows larger packets to be used
343 because it supports RFC1201 "p 343 because it supports RFC1201 "packet splitting" operations.
344 Unless you have a specific nee 344 Unless you have a specific need to use a different protocol,
345 I strongly suggest that you st 345 I strongly suggest that you stick with this one.
346 346
347 arc0e "Ethernet-Encapsulation" which 347 arc0e "Ethernet-Encapsulation" which sends packets over ARCnet
348 that are actually a lot like E 348 that are actually a lot like Ethernet packets, including the
349 6-byte hardware addresses. Th 349 6-byte hardware addresses. This protocol is compatible with
350 Microsoft's NDIS ARCnet driver 350 Microsoft's NDIS ARCnet driver, like the one in WfWg and
351 LANMAN. Because the MTU of 49 351 LANMAN. Because the MTU of 493 is actually smaller than the
352 one "required" by TCP/IP (576) 352 one "required" by TCP/IP (576), there is a chance that some
353 network operations will not fu 353 network operations will not function properly. The Linux
354 TCP/IP layer can compensate in 354 TCP/IP layer can compensate in most cases, however, by
355 automatically fragmenting the 355 automatically fragmenting the TCP/IP packets to make them
356 fit. arc0e also works slightl 356 fit. arc0e also works slightly more slowly than arc0, for
357 reasons yet to be determined. 357 reasons yet to be determined. (Probably it's the smaller
358 MTU that does it.) 358 MTU that does it.)
359 359
360 arc0s The "[s]imple" RFC1051 protoco 360 arc0s The "[s]imple" RFC1051 protocol is the "previous" Internet
361 standard that is completely in 361 standard that is completely incompatible with the new
362 standard. Some software today 362 standard. Some software today, however, continues to
363 support the old standard (and 363 support the old standard (and only the old standard)
364 including NetBSD and AmiTCP. 364 including NetBSD and AmiTCP. RFC1051 also does not support
365 RFC1201's packet splitting, an 365 RFC1201's packet splitting, and the MTU of 507 is still
366 smaller than the Internet "req 366 smaller than the Internet "requirement," so it's quite
367 possible that you may run into 367 possible that you may run into problems. It's also slower
368 than RFC1201 by about 25%, for 368 than RFC1201 by about 25%, for the same reason as arc0e.
369 369
370 The arc0s support was contribu 370 The arc0s support was contributed by Tomasz Motylewski
371 and modified somewhat by me. 371 and modified somewhat by me. Bugs are probably my fault.
372 ====== ============================== 372 ====== ===============================================================
373 373
374 You can choose not to compile arc0e and arc0s 374 You can choose not to compile arc0e and arc0s into the driver if you want -
375 this will save you a bit of memory and avoid c 375 this will save you a bit of memory and avoid confusion when eg. trying to
376 use the "NFS-root" stuff in recent Linux kerne 376 use the "NFS-root" stuff in recent Linux kernels.
377 377
378 The arc0e and arc0s devices are created automa 378 The arc0e and arc0s devices are created automatically when you first
379 ifconfig the arc0 device. To actually use the 379 ifconfig the arc0 device. To actually use them, though, you need to also
380 ifconfig the other virtual devices you need. 380 ifconfig the other virtual devices you need. There are a number of ways you
381 can set up your network then: 381 can set up your network then:
382 382
383 383
384 1. Single Protocol. 384 1. Single Protocol.
385 385
386 This is the simplest way to configure your 386 This is the simplest way to configure your network: use just one of the
387 two available protocols. As mentioned abov 387 two available protocols. As mentioned above, it's a good idea to use
388 only arc0 unless you have a good reason (li 388 only arc0 unless you have a good reason (like some other software, ie.
389 WfWg, that only works with arc0e). 389 WfWg, that only works with arc0e).
390 390
391 If you need only arc0, then the following c 391 If you need only arc0, then the following commands should get you going::
392 392
393 ifconfig arc0 MY.IP.ADD.RESS 393 ifconfig arc0 MY.IP.ADD.RESS
394 route add MY.IP.ADD.RESS arc0 394 route add MY.IP.ADD.RESS arc0
395 route add -net SUB.NET.ADD.RESS arc0 395 route add -net SUB.NET.ADD.RESS arc0
396 [add other local routes here] 396 [add other local routes here]
397 397
398 If you need arc0e (and only arc0e), it's a 398 If you need arc0e (and only arc0e), it's a little different::
399 399
400 ifconfig arc0 MY.IP.ADD.RESS 400 ifconfig arc0 MY.IP.ADD.RESS
401 ifconfig arc0e MY.IP.ADD.RESS 401 ifconfig arc0e MY.IP.ADD.RESS
402 route add MY.IP.ADD.RESS arc0e 402 route add MY.IP.ADD.RESS arc0e
403 route add -net SUB.NET.ADD.RESS arc0e 403 route add -net SUB.NET.ADD.RESS arc0e
404 404
405 arc0s works much the same way as arc0e. 405 arc0s works much the same way as arc0e.
406 406
407 407
408 2. More than one protocol on the same wire. 408 2. More than one protocol on the same wire.
409 409
410 Now things start getting confusing. To eve 410 Now things start getting confusing. To even try it, you may need to be
411 partly crazy. Here's what *I* did. :) Note 411 partly crazy. Here's what *I* did. :) Note that I don't include arc0s in
412 my home network; I don't have any NetBSD or 412 my home network; I don't have any NetBSD or AmiTCP computers, so I only
413 use arc0s during limited testing. 413 use arc0s during limited testing.
414 414
415 I have three computers on my home network; 415 I have three computers on my home network; two Linux boxes (which prefer
416 RFC1201 protocol, for reasons listed above) 416 RFC1201 protocol, for reasons listed above), and one XT that can't run
417 Linux but runs the free Microsoft LANMAN Cl 417 Linux but runs the free Microsoft LANMAN Client instead.
418 418
419 Worse, one of the Linux computers (freedom) 419 Worse, one of the Linux computers (freedom) also has a modem and acts as
420 a router to my Internet provider. The othe 420 a router to my Internet provider. The other Linux box (insight) also has
421 its own IP address and needs to use freedom 421 its own IP address and needs to use freedom as its default gateway. The
422 XT (patience), however, does not have its o 422 XT (patience), however, does not have its own Internet IP address and so
423 I assigned it one on a "private subnet" (as 423 I assigned it one on a "private subnet" (as defined by RFC1597).
424 424
425 To start with, take a simple network with j 425 To start with, take a simple network with just insight and freedom.
426 Insight needs to: 426 Insight needs to:
427 427
428 - talk to freedom via RFC1201 (arc0) p 428 - talk to freedom via RFC1201 (arc0) protocol, because I like it
429 more and it's faster. 429 more and it's faster.
430 - use freedom as its Internet gateway. 430 - use freedom as its Internet gateway.
431 431
432 That's pretty easy to do. Set up insight l 432 That's pretty easy to do. Set up insight like this::
433 433
434 ifconfig arc0 insight 434 ifconfig arc0 insight
435 route add insight arc0 435 route add insight arc0
436 route add freedom arc0 /* I would use 436 route add freedom arc0 /* I would use the subnet here (like I said
437 to in 437 to in "single protocol" above),
438 but th 438 but the rest of the subnet
439 unfort 439 unfortunately lies across the PPP
440 link o 440 link on freedom, which confuses
441 things 441 things. */
442 route add default gw freedom 442 route add default gw freedom
443 443
444 And freedom gets configured like so:: 444 And freedom gets configured like so::
445 445
446 ifconfig arc0 freedom 446 ifconfig arc0 freedom
447 route add freedom arc0 447 route add freedom arc0
448 route add insight arc0 448 route add insight arc0
449 /* and default gateway is configured b 449 /* and default gateway is configured by pppd */
450 450
451 Great, now insight talks to freedom directl 451 Great, now insight talks to freedom directly on arc0, and sends packets
452 to the Internet through freedom. If you di 452 to the Internet through freedom. If you didn't know how to do the above,
453 you should probably stop reading this secti 453 you should probably stop reading this section now because it only gets
454 worse. 454 worse.
455 455
456 Now, how do I add patience into the network 456 Now, how do I add patience into the network? It will be using LANMAN
457 Client, which means I need the arc0e device 457 Client, which means I need the arc0e device. It needs to be able to talk
458 to both insight and freedom, and also use f 458 to both insight and freedom, and also use freedom as a gateway to the
459 Internet. (Recall that patience has a "pri 459 Internet. (Recall that patience has a "private IP address" which won't
460 work on the Internet; that's okay, I config 460 work on the Internet; that's okay, I configured Linux IP masquerading on
461 freedom for this subnet). 461 freedom for this subnet).
462 462
463 So patience (necessarily; I don't have anot 463 So patience (necessarily; I don't have another IP number from my
464 provider) has an IP address on a different 464 provider) has an IP address on a different subnet than freedom and
465 insight, but needs to use freedom as an Int 465 insight, but needs to use freedom as an Internet gateway. Worse, most
466 DOS networking programs, including LANMAN, 466 DOS networking programs, including LANMAN, have braindead networking
467 schemes that rely completely on the netmask 467 schemes that rely completely on the netmask and a 'default gateway' to
468 determine how to route packets. This means 468 determine how to route packets. This means that to get to freedom or
469 insight, patience WILL send through its def 469 insight, patience WILL send through its default gateway, regardless of
470 the fact that both freedom and insight (cou 470 the fact that both freedom and insight (courtesy of the arc0e device)
471 could understand a direct transmission. 471 could understand a direct transmission.
472 472
473 I compensate by giving freedom an extra IP 473 I compensate by giving freedom an extra IP address - aliased 'gatekeeper' -
474 that is on my private subnet, the same subn 474 that is on my private subnet, the same subnet that patience is on. I
475 then define gatekeeper to be the default ga 475 then define gatekeeper to be the default gateway for patience.
476 476
477 To configure freedom (in addition to the co 477 To configure freedom (in addition to the commands above)::
478 478
479 ifconfig arc0e gatekeeper 479 ifconfig arc0e gatekeeper
480 route add gatekeeper arc0e 480 route add gatekeeper arc0e
481 route add patience arc0e 481 route add patience arc0e
482 482
483 This way, freedom will send all packets for 483 This way, freedom will send all packets for patience through arc0e,
484 giving its IP address as gatekeeper (on the 484 giving its IP address as gatekeeper (on the private subnet). When it
485 talks to insight or the Internet, it will u 485 talks to insight or the Internet, it will use its "freedom" Internet IP
486 address. 486 address.
487 487
488 You will notice that we haven't configured 488 You will notice that we haven't configured the arc0e device on insight.
489 This would work, but is not really necessar 489 This would work, but is not really necessary, and would require me to
490 assign insight another special IP number fr 490 assign insight another special IP number from my private subnet. Since
491 both insight and patience are using freedom 491 both insight and patience are using freedom as their default gateway, the
492 two can already talk to each other. 492 two can already talk to each other.
493 493
494 It's quite fortunate that I set things up l 494 It's quite fortunate that I set things up like this the first time (cough
495 cough) because it's really handy when I boo 495 cough) because it's really handy when I boot insight into DOS. There, it
496 runs the Novell ODI protocol stack, which o 496 runs the Novell ODI protocol stack, which only works with RFC1201 ARCnet.
497 In this mode it would be impossible for ins 497 In this mode it would be impossible for insight to communicate directly
498 with patience, since the Novell stack is in 498 with patience, since the Novell stack is incompatible with Microsoft's
499 Ethernet-Encap. Without changing any setti 499 Ethernet-Encap. Without changing any settings on freedom or patience, I
500 simply set freedom as the default gateway f 500 simply set freedom as the default gateway for insight (now in DOS,
501 remember) and all the forwarding happens "a 501 remember) and all the forwarding happens "automagically" between the two
502 hosts that would normally not be able to co 502 hosts that would normally not be able to communicate at all.
503 503
504 For those who like diagrams, I have created 504 For those who like diagrams, I have created two "virtual subnets" on the
505 same physical ARCnet wire. You can picture 505 same physical ARCnet wire. You can picture it like this::
506 506
507 507
508 [RFC1201 NETWORK] 508 [RFC1201 NETWORK] [ETHER-ENCAP NETWORK]
509 (registered Internet subnet) ( 509 (registered Internet subnet) (RFC1597 private subnet)
510 510
511 (IP Masquerade) 511 (IP Masquerade)
512 /---------------\ * 512 /---------------\ * /---------------\
513 | | * 513 | | * | |
514 | +-Freedom-*-Gatekeep 514 | +-Freedom-*-Gatekeeper-+ |
515 | | | * 515 | | | * | |
516 \-------+-------/ | * 516 \-------+-------/ | * \-------+-------/
517 | | 517 | | |
518 Insight | 518 Insight | Patience
519 (Internet) 519 (Internet)
520 520
521 521
522 522
523 It works: what now? 523 It works: what now?
524 ------------------- 524 -------------------
525 525
526 Send mail describing your setup, preferably in 526 Send mail describing your setup, preferably including driver version, kernel
527 version, ARCnet card model, CPU type, number o 527 version, ARCnet card model, CPU type, number of systems on your network, and
528 list of software in use to me at the following 528 list of software in use to me at the following address:
529 529
530 apenwarr@worldvisions.ca 530 apenwarr@worldvisions.ca
531 531
532 I do send (sometimes automated) replies to all 532 I do send (sometimes automated) replies to all messages I receive. My email
533 can be weird (and also usually gets forwarded 533 can be weird (and also usually gets forwarded all over the place along the
534 way to me), so if you don't get a reply within 534 way to me), so if you don't get a reply within a reasonable time, please
535 resend. 535 resend.
536 536
537 537
538 It doesn't work: what now? 538 It doesn't work: what now?
539 -------------------------- 539 --------------------------
540 540
541 Do the same as above, but also include the out 541 Do the same as above, but also include the output of the ifconfig and route
542 commands, as well as any pertinent log entries 542 commands, as well as any pertinent log entries (ie. anything that starts
543 with "arcnet:" and has shown up since the last 543 with "arcnet:" and has shown up since the last reboot) in your mail.
544 544
545 If you want to try fixing it yourself (I stron 545 If you want to try fixing it yourself (I strongly recommend that you mail me
546 about the problem first, since it might alread 546 about the problem first, since it might already have been solved) you may
547 want to try some of the debug levels available 547 want to try some of the debug levels available. For heavy testing on
548 D_DURING or more, it would be a REALLY good id 548 D_DURING or more, it would be a REALLY good idea to kill your klogd daemon
549 first! D_DURING displays 4-5 lines for each p 549 first! D_DURING displays 4-5 lines for each packet sent or received. D_TX,
550 D_RX, and D_SKB actually DISPLAY each packet a 550 D_RX, and D_SKB actually DISPLAY each packet as it is sent or received,
551 which is obviously quite big. 551 which is obviously quite big.
552 552
553 Starting with v2.40 ALPHA, the autoprobe routi 553 Starting with v2.40 ALPHA, the autoprobe routines have changed
554 significantly. In particular, they won't tell 554 significantly. In particular, they won't tell you why the card was not
555 found unless you turn on the D_INIT_REASONS de 555 found unless you turn on the D_INIT_REASONS debugging flag.
556 556
557 Once the driver is running, you can run the ar 557 Once the driver is running, you can run the arcdump shell script (available
558 from me or in the full ARCnet package, if you 558 from me or in the full ARCnet package, if you have it) as root to list the
559 contents of the arcnet buffers at any time. T 559 contents of the arcnet buffers at any time. To make any sense at all out of
560 this, you should grab the pertinent RFCs. (som 560 this, you should grab the pertinent RFCs. (some are listed near the top of
561 arcnet.c). arcdump assumes your card is at 0x 561 arcnet.c). arcdump assumes your card is at 0xD0000. If it isn't, edit the
562 script. 562 script.
563 563
564 Buffers 0 and 1 are used for receiving, and Bu 564 Buffers 0 and 1 are used for receiving, and Buffers 2 and 3 are for sending.
565 Ping-pong buffers are implemented both ways. 565 Ping-pong buffers are implemented both ways.
566 566
567 If your debug level includes D_DURING and you 567 If your debug level includes D_DURING and you did NOT define SLOW_XMIT_COPY,
568 the buffers are cleared to a constant value of 568 the buffers are cleared to a constant value of 0x42 every time the card is
569 reset (which should only happen when you do an 569 reset (which should only happen when you do an ifconfig up, or when Linux
570 decides that the driver is broken). During a 570 decides that the driver is broken). During a transmit, unused parts of the
571 buffer will be cleared to 0x42 as well. This 571 buffer will be cleared to 0x42 as well. This is to make it easier to figure
572 out which bytes are being used by a packet. 572 out which bytes are being used by a packet.
573 573
574 You can change the debug level without recompi 574 You can change the debug level without recompiling the kernel by typing::
575 575
576 ifconfig arc0 down metric 1xxx 576 ifconfig arc0 down metric 1xxx
577 /etc/rc.d/rc.inet1 577 /etc/rc.d/rc.inet1
578 578
579 where "xxx" is the debug level you want. For 579 where "xxx" is the debug level you want. For example, "metric 1015" would put
580 you at debug level 15. Debug level 7 is curre 580 you at debug level 15. Debug level 7 is currently the default.
581 581
582 Note that the debug level is (starting with v1 582 Note that the debug level is (starting with v1.90 ALPHA) a binary
583 combination of different debug flags; so debug 583 combination of different debug flags; so debug level 7 is really 1+2+4 or
584 D_NORMAL+D_EXTRA+D_INIT. To include D_DURING, 584 D_NORMAL+D_EXTRA+D_INIT. To include D_DURING, you would add 16 to this,
585 resulting in debug level 23. 585 resulting in debug level 23.
586 586
587 If you don't understand that, you probably don 587 If you don't understand that, you probably don't want to know anyway.
588 E-mail me about your problem. 588 E-mail me about your problem.
589 589
590 590
591 I want to send money: what now? 591 I want to send money: what now?
592 ------------------------------- 592 -------------------------------
593 593
594 Go take a nap or something. You'll feel bette 594 Go take a nap or something. You'll feel better in the morning.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.