The following lists many cards in alphabetical order by vendor name and then product identifier. Beside each product ID, you will see either `Supported', `Semi-Supported' or `Not Supported'.
Supported means that a driver for that card exists, and many people are happily using it and it seems quite reliable.
Semi-Supported means that a driver exists, but at least one of the following descriptions is true: (1) The driver and/or hardware are buggy, which may cause poor performance, failing connections or even crashes. (2) The driver is new or the card is fairly uncommon, and hence the driver has seen very little use/testing and the driver author has had very little feedback. Obviously (2) is preferable to (1), and the individual description of the card/driver should make it clear which one holds true. In either case, you will probably have to answer `Y' when asked ``Prompt for development and/or incomplete code/drivers?'' when running make config.
Not Supported means there is not a driver currently available for that card. This could be due to a lack of interest in hardware that is rare/uncommon, or because the vendors won't release the hardware documentation required to write a driver.
Note that the difference between `Supported' and `Semi-Supported' is rather subjective, and is based on user feedback observed in newsgroup postings and mailing list messages. (After all, it is impossible for one person to test all drivers with all cards for each kernel version!!!) So be warned that you may find a card listed as semi-supported works perfectly for you (which is great), or that a card listed as supported gives you no end of troubles and problems (which is not so great).
After the status, the name of the driver given in the linux kernel is listed. This will also be the name of the driver module that would be used in the alias eth0 driver_name line that is found in the /etc/conf.modules module configuration file.
If you are not sure what your card is, but you think it is a 3Com card, you can probably figure it out from the assembly number. 3Com has a document `Identifying 3Com Adapters By Assembly Number' (ref 24500002) that would most likely clear things up. See Technical Information from 3Com for info on how to get documents from 3Com.
Also note that 3Com has a WWW/FTP site with various goodies: www.3Com.com that you may want to check out. They even have linux drivers for some of their cards there you may wish to test out. It has been reported that their drivers are not stable and/or unuseable on SMP and non ix86 based machines, so you may want to keep that in mind.
Status: Semi-Supported, Driver Name: 3c501
This obsolete stone-age 8 bit card is really too brain-damaged to use. Avoid it like the plague. Do not purchase this card, even as a joke. It's performance is horrible, and it breaks in many ways.
For those not yet convinced, the 3c501 can only do one thing at a time -- while you are removing one packet from the single-packet buffer it cannot receive another packet, nor can it receive a packet while loading a transmit packet. This was fine for a network between two 8088-based computers where processing each packet and replying took 10's of msecs, but modern networks send back-to-back packets for almost every transaction.
AutoIRQ works, DMA isn't used, the autoprobe only looks at 0x280 and 0x300, and the debug level is set with the third boot-time argument.
Once again, the use of a 3c501 is strongly discouraged! Even more so with a IP multicast kernel, as you will grind to a halt while listening to all multicast packets. See the comments at the top of the source code for more details.
Status: Supported, Driver Name: 3c503 (+8390)
The 3c503 does not have ``EEPROM setup'', so a diagnostic/setup program isn't needed before running the card with Linux. The shared memory address of the 3c503 is set using jumpers that are shared with the boot PROM address. This is confusing to people familiar with other ISA cards, where you always leave the jumper set to ``disable'' unless you have a boot PROM.
These cards should be about the same speed as the same bus width WD80x3, but turn out to be actually a bit slower. These shared-memory ethercards also have a programmed I/O mode that doesn't use the 8390 facilities (their engineers found too many bugs!) The Linux 3c503 driver can also work with the 3c503 in programmed-I/O mode, but this is slower and less reliable than shared memory mode. Also, programmed-I/O mode is not as well tested when updating the drivers. You shouldn't use the programmed-I/O mode unless you need it for MS-DOS compatibility.
The 3c503's IRQ line is set in software, with no hints from an EEPROM. Unlike the MS-DOS drivers, the Linux driver has capability to autoIRQ: it uses the first available IRQ line in {5,2/9,3,4}, selected each time the card is ifconfig'ed. (Older driver versions selected the IRQ at boot time.) The ioctl() call in `ifconfig' will return EAGAIN if no IRQ line is available at that time.
Some common problems that people have with the 503 are discussed in Problems with....
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Note that some old diskless 386 workstations have an on board 3c503 (made by 3Com and sold under different names, like `Bull') but the vendor ID is not a 3Com ID and so it won't be detected. More details can be found in the Etherboot package, which you will need anyways to boot these diskless boxes.
Status: Semi-Supported, Driver Name: 3c505
These cards use the i82586 chip but are not that many of them about. It is included in the standard kernel, but it is classed as an alpha driver. See Alpha Drivers for important information on using alpha-test ethernet drivers with Linux.
There is also the file /usr/src/linux/drivers/net/README.3c505 that you should read if you are going to use one of these cards. It contains various options that you can enable/disable.
Status: Semi-Supported, Driver Name: 3c507
This card uses one of the Intel chips, and the development of the driver is closely related to the development of the Intel Ether Express driver. The driver is included in the standard kernel release, but as an alpha driver. See Alpha Drivers for important information on using alpha-test ethernet drivers with Linux.
Status: Supported, Driver Name: 3c509
This card is fairly inexpensive and has good performance for an ISA non-bus-master design. The drawbacks are that the original 3c509 requires very low interrupt latency. The 3c509B shouldn't suffer from the same problem, due to having a larger buffer. (See below.) These cards use PIO transfers, similar to a ne2000 card, and so a shared memory card such as a wd8013 will be more efficient in comparison.
The original 3c509 has a small packet buffer (4kB total, 2kB Rx, 2kB Tx), causing the driver to occasionally drop a packet if interrupts are masked for too long. To minimize this problem, you can try unmasking interrupts during IDE disk transfers (see man hdparm) and/or increasing your ISA bus speed so IDE transfers finish sooner.
The newer model 3c509B has 8kB on board, and the buffer can be split 4/4, 5/3 or 6/2 for Rx/Tx. This setting is changed with the DOS configuration utility, and is stored on the EEPROM. This should alleviate the above problem with the original 3c509.
3c509B users should use either the supplied DOS utility to disable the plug and play support, and to set the output media to what they require. The linux driver currently does not support the Autodetect media setting, so you have to select 10Base-T or 10Base-2 or AUI. Note that to turn off PnP entirely, you should do a 3C5X9CFG /PNP:DISABLE and then follow that with a hard reset to ensure that it takes effect.
Some people ask about the ``Server or Workstation'' and ``Highest Modem Speed'' settings presented in the DOS configuration utility. Donald writes ``These are only hints to the drivers, and the Linux driver does not use these parameters: it always optimizes for high throughput rather than low latency (`Server'). Low latency was critically important for old, non-windowed, IPX throughput. To reduce the latency the MS-DOS driver for the 3c509 disables interrupts for some operations, blocking serial port interrupts. Thus the need for the `modem speed' setting. The Linux driver avoids the need to disable interrupts for long periods by operating only on whole packets e.g. by not starting to transmit a packet until it is completely transferred to the card.''
Note that the ISA card detection uses a different method than most cards. Basically, you ask the cards to respond by sending data to an ID_PORT (port 0x100 to 0x1ff on intervals of 0x10). This detection method means that a particular card will always get detected first in a multiple ISA 3c509 configuration. The card with the lowest hardware ethernet address will always end up being eth0. This shouldn't matter to anyone, except for those people who want to assign a 6 byte hardware address to a particular interface. If you have multiple 3c509 cards, it is best to append ether=0,0,ethN commands without the I/O port specified (i.e. use I/O=zero) and allow the probe to sort out which card is first. Using a non-zero I/O value will ensure that it does not detect all your cards, so don't do it.
If this really bothers you, have a look at Donald's latest driver, as you may be able to use a 0x3c509 value in the unused mem address fields to order the detection to suit your needs.
Status: Supported, Driver Name: 3c515
This is 3Com's ISA 100Mbps offering, codenamed ``CorkScrew''. A relatively new driver from Donald for these cards is included in the v2.2 kernels. For the most up to date information, you should probably look on the Vortex page:
Vortex
Status: Semi-Supported, Driver Name: 3c523
This MCA bus card uses the i82586, and Chris Beauregard has modified the ni52 driver to work with these cards. The driver for it can be found in the v2.2 kernel source tree.
More details can be found on the MCA-Linux page at http://glycerine.cetmm.uni.edu/mca/
Status: Not Supported.
Yes, another MCA card. No, not too much interest in it. Better chances with the 3c529 if you are stuck with MCA.
Status: Supported, Driver Name: 3c509
This card actually uses the same chipset as the 3c509. Donald actually put hooks into the 3c509 driver to check for MCA cards after probing for EISA cards, and before probing for ISA cards, long before MCA support was added to the kernel. The required MCA probe code is included in the driver shipped with v2.2 kernels. More details can be found on the MCA-Linux page at:
http://glycerine.cetmm.uni.edu/mca/
Status: Supported, Driver Name: 3c589 (distributed separately)
This PCMCIA card is the combination of a 3c589B ethernet card with a modem. The modem appears as a standard modem to the end user. The only difficulty is getting the two separate linux drivers to share one interrupt. There are a couple of new registers and some hardware interrupt sharing support. You need to use a v2.0 or newer kernel that has the support for interrupt sharing.
Thanks again to Cameron for getting a sample unit and documentation sent off to David Hinds. Look for support in David's PCMCIA package release.
See PCMCIA Support for more info on PCMCIA chipsets, socket enablers, etc.
Status: Unknown.
A driver for this PCMCIA card is under development and hopefully will be included in David's PCMCIA package in the future. Best to check the PCMCIA package to get the current status.
Status: Supported, Driver Name: 3c509
The EISA version of the 509. The current EISA version uses the same 16 bit wide chip rather than a 32 bit interface, so the performance increase isn't stunning. Make sure the card is configured for EISA addressing mode. Read the above 3c509 section for info on the driver.
Status: Semi-Supported, Driver Name: 3c589
Many people have been using this PCMCIA card for quite some time now. Note that support for it is not (at present) included in the default kernel source tree. The "B" in the name means the same here as it does for the 3c509 case.
There are drivers available on Donald's ftp site and in David Hinds PCMCIA package. You will also need a supported PCMCIA controller chipset. See PCMCIA Support for more info on PCMCIA drivers, chipsets, socket enablers, etc.
Status: Supported, Driver Name: 3c59x
These ``Vortex'' cards are for PCI bus machines, with the '590 being 10Mbps and the '595 being 3Com's 100Mbs offering. Also note that you can run the '595 as a '590 (i.e. in a 10Mbps mode). The driver is included in the v2.0 kernel source, but is also continually being updated. If you have problems with the driver in the v2.0 kernel, you can get an updated driver from the following URL:
Vortex
Note that there are two different 3c590 cards out there, early models that had 32kB of on-board memory, and later models that only have 8kB of memory. Chances are you won't be able to buy a new 3c59x for much longer, as it is being replaced with the 3c90x card. If you are buying a used one off somebody, try and get the 32kB version. The 3c595 cards have 64kB, as you can't get away with only 8kB RAM at 100Mbps!
A thanks to Cameron Spitzer and Terry Murphy of 3Com for sending cards and documentation to Donald so he could write the driver.
Status: Supported, Driver Name: 3c59x
These are the EISA versions of the 3c59x series of cards. The 3c592/3c597 (aka Demon) should work with the vortex driver discussed above.
Status: Supported, Driver Name: 3c59x
These cards (aka `Boomerang', aka EtherLink III XL) have been released to take over the place of the 3c590/3c595 cards.
The support for the Cyclone `B' revision was only recently added. To use this card with older v2.0 kernels, you must obtain the updated 3c59x.c driver from Donald's site at:
Vortex-Page
Status: Supported, Driver Name: acenic
This driver, by Jes Sorensen, is available in v2.2 kernels It supports several other Gigabit cards in addition to the 3Com model.
Status: Supported, Driver Name: ne (+8390)
Don't let the name fool you. This is still supposed to be a NE2000 compatible card, and should work with the ne2000 driver.
Status: Supported, Driver Name: de4x5, tulip, OR rtl8139
Apparently there have been several revisions of the EN1207 (A through D) with A, B, and C being tulip based and the D revision being RealTek 8139 based (different driver). So as with all purchases, you should try and make sure you can return it if it doesn't work for you.
Status: Semi-Supported, Driver Name: ?
A driver for these parallel port adapters is available but not yet part of the 2.0 or 2.1 kernel source. You have to get the driver from:
http://www.unix-ag.uni-siegen.de/~nils/accton_linux.html
Status: Semi-Supported, Driver Name: ?
David Hinds has been working on a driver for this card, and you are best to check the latest release of his PCMCIA package to see what the present status is.
Status: Supported, Driver Name: lance
These are a series of low-cost ethercards using the 79C960 version of the AMD LANCE. These are bus-master cards, and hence one of the faster ISA bus ethercards available.
DMA selection and chip numbering information can be found in AMD LANCE.
More technical information on AMD LANCE based Ethernet cards can be found in Notes on AMD....
Status: Supported, Driver Name: at1700
Note that to access this driver during make config you still have to answer `Y' when asked ``Prompt for development and/or incomplete code/drivers?'' at the first. This is simply due to lack of feedback on the driver stability due to it being a relatively rare card. If you have problems with the driver that ships with the kernel then you may be interested in the alternative driver available at: http://www.cc.hit-u.ac.jp/nagoya/at1700/
The Allied Telesis AT1700 series ethercards are based on the Fujitsu MB86965. This chip uses a programmed I/O interface, and a pair of fixed-size transmit buffers. This allows small groups of packets to be sent back-to-back, with a short pause while switching buffers.
A unique feature is the ability to drive 150ohm STP (Shielded Twisted Pair) cable commonly installed for Token Ring, in addition to 10baseT 100ohm UTP (unshielded twisted pair). A fibre optic version of the card (AT1700FT) exists as well.
The Fujitsu chip used on the AT1700 has a design flaw: it can only be fully reset by doing a power cycle of the machine. Pressing the reset button doesn't reset the bus interface. This wouldn't be so bad, except that it can only be reliably detected when it has been freshly reset. The solution/work-around is to power-cycle the machine if the kernel has a problem detecting the AT1700.
Status: Supported, Driver Name: ne, ne2k-pci (+8390)
Yet another PCI NE2000 clone card. This one is based on the RealTek 8029 chip.
Status: Supported, Driver Name: pcnet32
This is the PCI version of the AT1500, and it doesn't suffer from the problems that the Boca 79c970 PCI card does. DMA selection and chip numbering information can be found in AMD LANCE.
More technical information on AMD LANCE based Ethernet cards can be found in Notes on AMD....
Status: Semi-Supported, Driver Name: rtl8139
This card uses the RealTek 8139 chip - see the section RealTek 8139.
Status: Semi-Supported, Driver Name: eepro100
This card uses the i82557 chip, and hence may/should work with the eepro100 driver. If you try this please send in a report so this information can be updated.
Carl Ching of AMD was kind enough to provide a very detailed description of all the relevant AMD ethernet products which helped clear up this section.
Status: Supported, Driver Name: lance
There really is no AMD ethernet card. You are probably reading this because the only markings you could find on your card said AMD and the above number. The 7990 is the original `LANCE' chip, but most stuff (including this document) refer to all these similar chips as `LANCE' chips. (...incorrectly, I might add.)
These above numbers refer to chips from AMD that are the heart of many ethernet cards. For example, the Allied Telesis AT1500 (see AT1500) and the NE1500/2100 (see NE1500) use these chips.
The 7990/79c90 have long been replaced by newer versions. The 79C960 (a.k.a. PCnet-ISA) essentially contains the 79c90 core, along with all the other hardware support required, which allows a single-chip ethernet solution. The 79c961 (PCnet-ISA+) is a jumperless Plug and Play version of the '960. The final chip in the ISA series is the 79c961A (PCnet-ISA II), which adds full duplex capabilities. All cards with one of these chips should work with the lance.c driver, with the exception of very old cards that used the original 7990 in a shared memory configuration. These old cards can be spotted by the lack of jumpers for a DMA channel.
One common problem people have is the `busmaster arbitration failure' message. This is printed out when the LANCE driver can't get access to the bus after a reasonable amount of time has elapsed (50us). This usually indicates that the motherboard implementation of bus-mastering DMA is broken, or some other device is hogging the bus, or there is a DMA channel conflict. If your BIOS setup has the `GAT option' (for Guaranteed Access Time) then try toggling/altering that setting to see if it helps.
Also note that the driver only looks at the addresses: 0x300, 0x320, 0x340, 0x360 for a valid card, and any address supplied by an ether= boot argument is silently ignored (this will be fixed) so make sure your card is configured for one of the above I/O addresses for now.
The driver will still work fine, even if more than 16MB of memory is installed, since low-memory `bounce-buffers' are used when needed (i.e. any data from above 16MB is copied into a buffer below 16MB before being given to the card to transmit.)
The DMA channel can be set with the low bits of the otherwise-unused dev->mem_start value (a.k.a. PARAM_1). (see PARAM_1) If unset it is probed for by enabling each free DMA channel in turn and checking if initialization succeeds.
The HP-J2405A board is an exception: with this board it's easy to read the EEPROM-set values for the IRQ, and DMA.
See Notes on AMD... for more info on these chips.
Status: Supported, Driver Name: pcnet32
This is the PCnet-32 -- a 32 bit bus-master version of the original LANCE chip for VL-bus and local bus systems. chip. While these chips can be operated with the standard lance.c driver, a 32 bit version (pcnet32.c) is also available that does not have to concern itself with any 16MB limitations associated with the ISA bus.
Status: Supported, Driver Name: pcnet32
This is the PCnet-PCI -- similar to the PCnet-32, but designed for PCI bus based systems. Please see the above PCnet-32 information. This means that you need to build a kernel with PCI BIOS support enabled. The '970A adds full duplex support along with some other features to the original '970 design.
Note that the Boca implementation of the 79C970 fails on fast Pentium machines. This is a hardware problem, as it affects DOS users as well. See the Boca section for more details.
Status: Supported, Driver Name: pcnet32
This is AMD's 100Mbit chip for PCI systems, which also supports full duplex operation. It was introduced in June 1996.
Status: Supported, Driver Name: pcnet32
This has been confirmed to work just like the '971.
Status: Supported, Driver Name: pcnet32
This is the PCnet-SCSI -- which is basically treated like a '970 from an Ethernet point of view. Also see the above information. Don't ask if the SCSI half of the chip is supported -- this is the Ethernet-HowTo, not the SCSI-HowTo.
Status: Semi-Supported, Driver Name: ac3200
Note that to access this driver during make config you still have to answer `Y' when asked ``Prompt for development and/or incomplete code/drivers?'' at the first. This is simply due to lack of feedback on the driver stability due to it being a relatively rare card.
This driver is included in the present kernel as an alpha test driver. It is based on the common NS8390 chip used in the ne2000 and wd80x3 cards. Please see Alpha Drivers in this document for important information regarding alpha drivers.
If you use it, let one of us know how things work out, as feedback has been low, even though the driver has been in the kernel since v1.1.25.
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Semi-Supported, Driver Name: apricot
This on board ethernet uses an i82596 bus-master chip. It can only be at I/O address 0x300. By looking at the driver source, it appears that the IRQ is also hardwired to 10.
Earlier versions of the driver had a tendency to think that anything living at 0x300 was an apricot NIC. Since then the hardware address is checked to avoid these false detections.
Status: Supported, Driver Name: arcnet (arc-rimi, com90xx, com20020)
With the very low cost and better performance of ethernet, chances are that most places will be giving away their Arcnet hardware for free, resulting in a lot of home systems with Arcnet.
An advantage of Arcnet is that all of the cards have identical interfaces, so one driver will work for everyone. It also has built in error handling so that it supposedly never loses a packet. (Great for UDP traffic!) Note that the arcnet driver uses `arc0' as its name instead of the usual `eth0' for ethernet devices.
There are information files contained in the standard kernel for setting jumpers, general hints and where to mail bug reports.
Supposedly the driver also works with the 100Mbs ARCnet cards as well!
Note that AT&T's StarLAN is an orphaned technology, like SynOptics LattisNet, and can't be used in a standard 10Base-T environment, without a hub that `speaks' both.
Status: Not Supported.
These StarLAN cards use an interface similar to the i82586 chip. At one point, Matthijs Melchior ([email protected]) was playing with the 3c507 driver, and almost had something useable working. Haven't heard much since that.
Yes, they make more than just multi-port serial cards.
Status: Supported, Driver Name: ne (+8390)
Apparently this is a NE2000 clone, using a VIA VT86C916 chip.
Status: Supported, Driver Name: lance, pcnet32
These cards are based on AMD's PCnet chips. Perspective buyers should be warned that many users have had endless problems with these VLB/PCI cards. Owners of fast Pentium systems have been especially hit. Note that this is not a driver problem, as it hits DOS/Win/NT users as well. Boca's technical support number is (407) 241-8088, and you can also reach them at [email protected]. The older ISA cards don't appear to suffer the same problems.
Boca was offering a `warranty repair' for affected owners, which involved adding one of the missing capacitors, but it appears that this fix didn't work 100 percent for most people, although it helps some.
More general information on the AMD chips can be found in AMD LANCE.
More technical information on AMD LANCE based Ethernet cards can be found in Notes on AMD....
Donald writes: `Yes, another one of these companies that won't release its programming information. They waited for months before actually confirming that all their information was proprietary, deliberately wasting my time. Avoid their cards like the plague if you can. Also note that some people have phoned Cabletron, and have been told things like `a D. Becker is working on a driver for linux' -- making it sound like I work for them. This is NOT the case.'
Apparently Cabletron has changed their policy with respect to programming information (like Xircom) since Donald made the above comment several years ago -- send e-mail to [email protected] if you want to verify this or ask for programming information. However, at this point in time, there is little demand for modified/updated drivers for the older E20xx and E21xx cards.
Status: Semi-Supported, Driver Name: ne (+8390)
These are NEx000 almost-clones that are reported to work with the standard NEx000 drivers, thanks to a ctron-specific check during the probe. If there are any problems, they are unlikely to be fixed, as the programming information is unavailable.
Status: Semi-Supported, Driver Name: e2100 (+8390)
Again, there is not much one can do when the programming information is proprietary. The E2100 is a poor design. Whenever it maps its shared memory in during a packet transfer, it maps it into the whole 128K region! That means you can't safely use another interrupt-driven shared memory device in that region, including another E2100. It will work most of the time, but every once in a while it will bite you. (Yes, this problem can be avoided by turning off interrupts while transferring packets, but that will almost certainly lose clock ticks.) Also, if you mis-program the board, or halt the machine at just the wrong moment, even the reset button won't bring it back. You will have to turn it off and leave it off for about 30 seconds.
Media selection is automatic, but you can override this with the low bits of the dev->mem_end parameter. See PARAM_2. Module users can specify an xcvr=N value as an option in the /etc/conf.modules file.
Also, don't confuse the E2100 for a NE2100 clone. The E2100 is a shared memory NatSemi DP8390 design, roughly similar to a brain-damaged WD8013, whereas the NE2100 (and NE1500) use a bus-mastering AMD LANCE design.
There is an E2100 driver included in the standard kernel. However, seeing as programming info isn't available, don't expect bug-fixes. Don't use one unless you are already stuck with the card.
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Semi-Supported, Driver Name: lance
According to information in a Cabletron Tech Bulletin, these cards use the standard AMD PC-Net chipset (see AMD PC-Net) and should work with the generic lance driver.
Here is where and how to reach them:
Cogent Data Technologies, Inc.
175 West Street, P.O. Box 926
Friday Harbour, WA 98250, USA.
Cogent Sales
15375 S.E. 30th Place, Suite 310
Bellevue, WA 98007, USA.
Technical Support:
Phone (360) 378-2929 between 8am and 5pm PST
Fax (360) 378-2882
Compuserve GO COGENT
Bulletin Board Service (360) 378-5405
Internet: [email protected]
Status: Semi-Supported, Driver Name: smc9194
These cards use the SMC 91c100 chip and may work with the SMC 91c92 driver, but this has yet to be verified.
Status: Supported, Driver Name: de4x5, tulip
These are yet another DEC 21040 implementation that should hopefully work fine with the standard 21040 driver.
The EM400 and the EM964 are four port cards using a DEC 21050 bridge and 4 21040 chips.
See DEC 21040 for more information on these cards, and the present driver situation.
Compaq aren't really in the business of making ethernet cards, but a lot of their systems have embedded ethernet controllers on the motherboard.
Status: Supported, Driver Name: pcnet32
Machines such as the XL series have an AMD 79c97x PCI chip on the mainboard that can be used with the standard LANCE driver. But before you can use it, you have to do some trickery to get the PCI BIOS to a place where Linux can see it. Frank Maas was kind enough to provide the details:
`` The problem with this Compaq machine however is that the PCI directory is loaded in high memory, at a spot where the Linux kernel can't (won't) reach. Result: the card is never detected nor is it usable (sideline: the mouse won't work either) The workaround (as described thoroughly in http://www-c724.uibk.ac.at/XL/) is to load MS-DOS, launch a little driver Compaq wrote and then load the Linux kernel using LOADLIN. Ok, I'll give you time to say `yuck, yuck', but for now this is the only working solution I know of. The little driver simply moves the PCI directory to a place where it is normally stored (and where Linux can find it).''
More general information on the AMD chips can be found in AMD LANCE.
Status: Supported, Driver Name: tlan
These systems use a Texas Instruments ThunderLAN chip Information on the ThunderLAN driver can be found in ThunderLAN.
Status: Supported, Driver Name: eepro100
Check your card - if it has part number 323551-821 and/or an intel 82558 chip on it then it is another Intel EEPro100 based card.
Status: Supported, Driver Name: de4x5, tulip
Yet another card based on the DEC 21040 chip, reported to work fine, and at a relatively cheap price.
See DEC 21040 for more information on these cards, and the present driver situation.
Status: Supported, Driver Name: ne (+8390)
Some of the early D-Link cards didn't have the 0x57 PROM signature, but the ne2000 driver knows about them. For the software configurable cards, you can get the config program from www.dlink.com. The DE2** cards were the most widely reported as having the spurious transfer address mismatch errors with early versions of linux. Note that there are also cards from Digital (DEC) that are also named DE100 and DE200, but the similarity stops there.
Status: Supported, Driver Name: pcnet32
This is a PCI card using the PCI version of AMD's LANCE chip. DMA selection and chip numbering information can be found in AMD LANCE.
More technical information on AMD LANCE based Ethernet cards can be found in Notes on AMD....
Status: Supported, Driver Name: ne, ne2k-pci (+8390)
Apparently D-Link have also started making PCI NE2000 clones.
Status: Supported, Driver Name: de4x5, tulip
This is a generic DEC 21040 PCI chip implementation, and is reported to work with the generic 21040 tulip driver. Note that this is NOT the DFE-530.
See DEC 21040 for more information on these cards, and the present driver situation.
Status: Supported, Driver Name: de600
Laptop users and other folk who might want a quick way to put their computer onto the ethernet may want to use this. The driver is included with the default kernel source tree. Expect about 180kb/s transfer speed from this via the parallel port. You should read the README.DLINK file in the kernel source tree.
Note that the device name that you pass to ifconfig is now eth0 and not the previously used dl0.
If your parallel port is not at the standard 0x378 then you will have to recompile, as the address is compiled directly into the driver. Also note that some laptops implement the on-board parallel port at 0x3bc which is where the parallel ports on monochrome cards were/are.
Status: Supported, Driver Name: de620
Same as the DE-600, only with two output formats. Bjorn has written a driver for this model, for kernel versions 1.1 and above. See the above information on the DE-600.
Status: Semi-Supported, Driver Name: de650 (?)
Some people have been using this PCMCIA card for some time now with their notebooks. It is a basic 8390 design, much like a NE2000. The LinkSys PCMCIA card and the IC-Card Ethernet are supposedly DE-650 clones as well. Note that at present, this driver is not part of the standard kernel, and so you will have to do some patching. See PCMCIA Support in this document.
Status Supported, Driver Name: via-rhine
Another card using the VIA Rhine chipset. (see VIA Rhine) Don't confuse this with the DE-530 which is a tulip based card.
Status Supported, Driver Name: rtl8139, 8139too
This card uses the RealTek 8139 chip - see the section RealTek 8139.
Status: Supported, Driver Name: ne (+8390)
Yet another poor NE clone card - these use `DFI' in the first 3 bytes of the prom, instead of using 0x57 in bytes 14 and 15, which is what all the NE1000 and NE2000 cards should use. (The 300 is an 8 bit pseudo NE1000 clone, and the 400 is a pseudo NE2000 clone.)
Status: Supported, Driver Name: depca
There is documentation included in the source file `depca.c', which includes info on how to use more than one of these cards in a machine. Note that the DE422 is an EISA card. These cards are all based on the AMD LANCE chip. See AMD LANCE for more info. A maximum of two of the ISA cards can be used, because they can only be set for 0x300 and 0x200 base I/O address. If you are intending to do this, please read the notes in the driver source file depca.c in the standard kernel source tree.
This driver will also work on Alpha CPU based machines, and there are various ioctl()s that the user can play with.
Status: Supported, Driver Name: ewrk3
These cards use a proprietary chip from DEC, as opposed to the LANCE chip used in the earlier cards like the DE200. These cards support both shared memory or programmed I/O, although you take about a 50%performance hit if you use PIO mode. The shared memory size can be set to 2kB, 32kB or 64kB, but only 2 and 32 have been tested with this driver. David says that the performance is virtually identical between the 2kB and 32kB mode. There is more information (including using the driver as a loadable module) at the top of the driver file ewrk3.c and also in README.ewrk3. Both of these files come with the standard kernel distribution. This driver has Alpha CPU support like depca.c does.
The standard driver has a number of interesting ioctl() calls that can be used to get or clear packet statistics, read/write the EEPROM, change the hardware address, and the like. Hackers can see the source code for more info on that one.
David has also written a configuration utility for this card (along the lines of the DOS program NICSETUP.EXE) along with other tools. These can be found on most Linux FTP sites in the directory /pub/Linux/system/Network/management -- look for the file ewrk3tools-X.XX.tar.gz.
Status: Supported, Driver Name: de4x5, tulip
These cards are based on the 21040 chip mentioned below. The DE500 uses the 21140 chip to provide 10/100Mbs ethernet connections. Have a read of the 21040 section below for extra info. There are also some compile-time options available for non-DEC cards using this driver. Have a look at README.de4x5 for details.
All the Digital cards will autoprobe for their media (except, temporarily, the DE500 due to a patent issue).
This driver is also Alpha CPU ready and supports being loaded as a module. Users can access the driver internals through ioctl() calls - see the 'ewrk3' tools and the de4x5.c sources for information about how to do this.
Status: Supported, Driver Name: de4x5, tulip
The DEC 21040 is a bus-mastering single chip ethernet solution from Digital, similar to AMD's PCnet chip. The 21040 is specifically designed for the PCI bus architecture. Apparently these chips are no longer being produced, as Intel has bought the semiconductor portion of DEC and is favouring their own ethernet chip(s).
You have a choice of two drivers for cards based on this chip. There is the DE425 driver discussed above, and the generic 21040 `tulip' driver.
Warning: Even though your card may be based upon this chip, the drivers may not work for you. David C. Davies writes:
``There are no guarantees that either `tulip.c' OR `de4x5.c' will run any DC2114x based card other than those they've been written to support. WHY?? You ask. Because there is a register, the General Purpose Register (CSR12) that (1) in the DC21140A is programmable by each vendor and they all do it differently (2) in the DC21142/3 this is now an SIA control register (a la DC21041). The only small ray of hope is that we can decode the SROM to help set up the driver. However, this is not a guaranteed solution since some vendors (e.g. SMC 9332 card) don't follow the Digital Semiconductor recommended SROM programming format."
In non-technical terms, this means that if you aren't sure that an unknown card with a DC2114x chip will work with the linux driver(s), then make sure you can return the card to the place of purchase before you pay for it.
The 21041 chip is also found in place of the 21040 on most of the later SMC EtherPower cards. The 21140 is for supporting 100Base-T and works with the Linux drivers for the 21040 chip. To use David's de4x5 driver with non-DEC cards, have a look at README.de4x5 for details.
If you are having trouble with the tulip driver, you can try the newest version from Donald's ftp/WWW site.
Tulip Driver
There is also a (non-exhaustive) list of various cards/vendors that use the 21040 chip.
Farallon sells EtherWave adaptors and transceivers. This device allows multiple 10baseT devices to be daisy-chained.
Status: Supported, Driver Name: 3c509
This is reported to be a 3c509 clone that includes the EtherWave transceiver. People have used these successfully with Linux and the present 3c509 driver. They are too expensive for general use, but are a great option for special cases. Hublet prices start at $125, and Etherwave adds $75-$100 to the price of the board -- worth it if you have pulled one wire too few, but not if you are two network drops short.
Status: Supported, Driver Name: de4x5, tulip
It has been reported that this card was detected with the de4x5 driver.
Unlike many network chip manufacturers, Fujitsu have also made and sold some network cards based upon their chip.
Status: Supported, Driver Name: fmv18x
According to the driver, these cards are a straight forward Fujitsu MB86965 implementation, which would make them very similar to the Allied Telesis AT1700 cards.
The 272** cards use programmed I/O, similar to the NE*000 boards, but the data transfer port can be `turned off' when you aren't accessing it, avoiding problems with autoprobing drivers.
Thanks to Glenn Talbott for helping clean up the confusion in this section regarding the version numbers of the HP hardware.
Status: Supported, Driver Name: pcnet32
Apparently these cards use the AMD 79C972 chip.
Status: Supported, Driver Name: hp (+8390)
8 bit 8390 based 10BaseT, not recommended for all the 8 bit reasons. It was re-designed a couple years ago to be highly integrated which caused some changes in initialization timing which only affected testing programs, not LAN drivers. (The new card is not `ready' as soon after switching into and out of loopback mode.)
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Supported, Driver Name: hp+ (+8390)
The HP PC Lan+ is different to the standard HP PC Lan card. This driver was added to the list of drivers in the standard kernel during the v1.1.x development cycle. It can be operated in either a PIO mode like a ne2000, or a shared memory mode like a wd8013.
The 47B is a 16 Bit 8390 based 10BaseT w/AUI, and the 52A is a 16 Bit 8390 based ThinLAN w/AUI. These cards have 32K onboard RAM for Tx/Rx packet buffering instead of the usual 16KB, and they both offer LAN connector autosense.
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Supported, Driver Name: lance
These are lower priced, and slightly faster than the 27247/27252A, but are missing some features, such as AUI, ThinLAN connectivity, and boot PROM socket. This is a fairly generic LANCE design, but a minor design decision makes it incompatible with a generic `NE2100' driver. Special support for it (including reading the DMA channel from the board) is included thanks to information provided by HP's Glenn Talbott.
More technical information on LANCE based cards can be found in Notes on AMD...
Status: Supported, Driver Name: lance
The HP-Vectra has an AMD PCnet chip on the motherboard. DMA selection and chip numbering information can be found in AMD LANCE.
More technical information on LANCE based cards can be found in Notes on AMD...
Status: Supported, Driver Name: hp100
This driver also supports some of the Compex VG products. Since the driver supports ISA, EISA and PCI cards, it is found under ISA cards when running make config on a kernel source.
Status: Supported, Driver Name: eepro100
Apparently these are just a rebadged Intel EtherExpress Pro 10/100B card. See the Intel section for more information.
Status: Supported, Driver Name: znet
This is compatible with the Intel based Zenith Z-note. See Z-note for more info.
Supposedly this site has a comprehensive database of useful stuff for newer versions of the Thinkpad. I haven't checked it out myself yet.
Thinkpad-info
For those without a WWW browser handy, try peipa.essex.ac.uk:/pub/tp750/
Status: Semi-Supported, Driver Name: ? (distributed separately)
People have been using this PCMCIA card with Linux as well. Similar points apply, those being that you need a supported PCMCIA chipset on your notebook, and that you will have to patch the PCMCIA support into the standard kernel. See PCMCIA Support in this document.
Status: Supported, Driver Name: eepro100
This card is reported to be compatible with the Intel EtherExpress Pro 100 driver.
Status: Semi-Supported, Driver Name: ibmtr
To support token ring requires more than only writing a device driver, it also requires writing the source routing routines for token ring. It is the source routing that would be the most time comsuming to write.
Initial driver development was done with IBM ISA and MCA token ring cards, and tested on an MCA 16/4 Megabit Token Ring board, but it should work with other Tropic based boards.
Status: Supported, Driver Name: eth16i
This driver supports both the ISA (16i) and EISA (32) versions of the card. It uses the Fujitsu MB86965 chip that is also used on the at1700 cards.
Note that the naming of the various Intel cards is ambiguous and confusing at best. If in doubt, then check the i8xxxx number on the main chip on the card or for PCI cards, use the PCI information in the /proc directory and then compare that to the numbers listed here.
Status: Supported, Driver Name: eexpress
This card uses the intel i82586. Earlier versions of this driver (in v1.2 kernels) were classed as alpha-test, as it didn't work well for most people. The driver in the v2.0 kernel seems to work much better for those who have tried it, although the driver source still lists it as experimental and more problematic on faster machines.
The comments at the top of the driver source list some of the problems (and fixes!) associated with these cards. The slowdown hack of replacing all the outb with outb_p in the driver has been reported to avoid lockups for at least one user. Also check that the size of the RAM buffer reported by the driver matches what the Intel configuration utility reports.
Status: Supported, Driver Name: eepro
Bao Chau Ha has written a driver for these cards that has been included into early 1.3.x kernels. It may also work with some of the Compaq built-in ethernet systems that are based on the i82595 chip. You may have to use the configuration utility that came with the card to disable PnP support where applicable.
Status: Semi-Supported, Driver Name: ? (distributed separately)
There is a driver for the PCI version that is distributed separately from the default kernel. These cards use the PLX9036 PCI interface chip with the Intel i82596 LAN controller chip. If your card has the i82557 chip, then you don't have this card, but rather the version discussed next, and hence want the EEPro100 driver instead.
You can get the alpha driver for the PRO/10 PCI card, along with instructions on how to use it at:
EEPro10 Driver
If you have the EISA card, you will probably have to hack the driver a bit to account for the different (PCI vs. EISA) detection mechanisms that are used in each case.
Status: Supported, Driver Name: eepro100
Note that this driver will not work with the older 100A cards. The chip numbers listed in the driver are i82557/i82558. For driver updates and/or driver support, have a look at:
EEPro-100B Page
Kingston make various cards, including NE2000+, AMD PCnet based cards, and DEC tulip based cards. Most of these cards should work fine with their respective driver. See Kingston Web Page
LinkSys make a handful of different NE2000 clones, some straight ISA cards, some ISA plug and play and some even ne2000-PCI clones based on one of the supported ne2000-PCI chipsets. There are just too many models to list here.
LinkSys are linux-friendly, with a linux specific WWW support page, and even have Linux printed on the boxes of some of their products. Have a look at:
http://www.linksys.com/support/solution/nos/linux.htm
Status: Supported, Driver Name: tulip
Note that with these cards there have been several `revisions' (i.e. different chipset used) all with the same card name. The 1st used the DEC chipset. The 2nd revision used the Lite-On PNIC 82c168 PCI Network Interface Controller, and the 3rd revision of the card uses a LinkSys 82c169 NIC chip. Support for the latter two has been merged into the standard tulip driver -- you may need a driver upgrade to get support for them depending on how old your current driver version is.
More PNIC information is available at:
http://www.scyld.com/linux/drivers/pnic.html
More information on the various versions of these cards can be found at the LinkSys WWW site mentioned above.
Status: Supported, Driver Name: de620
This is supposedly a DE-620 clone, and is reported to work well with that driver. See DE-620 for more information.
Status: Supported, Driver Name: de650 (?)
This is supposed to be a re-badged DE-650. See DE-650 for more information.
Eagle Technology (aka Novell cards) was sold to Microdyne. If you can't find your card listed here, check the Novell section of this document. While Microdyne are not actively selling network cards anymore, there is still some stuff relating to their products on their site at ftp.mcdy.com
Status: Semi-Supported, Driver Name: ?
Another i82586 based card. Dirk Niggemann [email protected] has written a driver that he classes as ``pre-alpha'' that he would like people to test. Mail him for more details.
Mylex can be reached at the following numbers, in case anyone wants to ask them anything.
MYLEX CORPORATION, Fremont
Sales: 800-77-MYLEX, (510) 796-6100
FAX: (510) 745-8016.
They also have a web site: Mylex WWW Site
Status: Supported, Driver Name: lne390 (+8390)
These are fairly old EISA cards that make use of a shared memory implementation similar to the wd80x3. A driver for these cards is available in the current 2.1.x series of kernels. Ensure you set the shared memory address below 1MB or above the highest address of the physical RAM installed in the machine.
Status: Supported, Driver Name: de4x5, tulip
This is a PCI card that is based on DEC's 21040 chip. It is selectable between 10BaseT, 10Base2 and 10Base5 output. The LNP101 card has been verified to work with the generic 21040 driver.
See the section on the 21040 chip ( DEC 21040) for more information.
Status: Semi-Supported, Driver Name: de4x5, tulip
The LNP104 uses the DEC 21050 chip to deliver four independent 10BaseT ports. It should work with recent 21040 drivers that know how to share IRQs, but nobody has reported trying it yet (that I am aware of).
The prefix `NE' came from Novell Ethernet. Novell followed the cheapest NatSemi databook design and sold the manufacturing rights (spun off?) Eagle, just to get reasonably-priced ethercards into the market. (The now ubiquitous NE2000 card.)
Status: Supported, Driver Name: ne (+8390)
The ne2000 is now a generic name for a bare-bones design around the NatSemi 8390 chip. They use programmed I/O rather than shared memory, leading to easier installation but slightly lower performance and a few problems. Some of the more common problems that arise with NE2000 cards are listed in Problems with...
Some NE2000 clones use the National Semiconductor `AT/LANTic' 83905 chip, which offers a shared memory mode similar to the wd8013 and EEPROM software configuration. The shared memory mode will offer less CPU usage (i.e. more efficient) than the programmed I/O mode.
In general it is not a good idea to put a NE2000 clone at I/O address 0x300 because nearly every device driver probes there at boot. Some poor NE2000 clones don't take kindly to being prodded in the wrong areas, and will respond by locking your machine. Also 0x320 is bad because SCSI drivers probe into 0x330.
Donald has written a NE2000 diagnostic program (ne2k.c) for all ne2000 cards. See Diagnostic Programs for more information.
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Supported, Driver Name: ne, ne2k-pci (+8390)
Yes, believe it or not, people are making PCI cards based on the more than ten year old interface design of the ne2000. At the moment nearly all of these cards are based on the RealTek 8029 chip, or the Winbond 89c940 chip. The Compex, KTI, VIA and Netvin cards apparently also use these chips, but have a different PCI ID.
The latest v2.0 kernel has support to automatically detect all these cards and use them. (If you are using a kernel v2.0.34 or older, you should upgrade to ensure your card will be detected.) There are now two drivers to choose from; the original ISA/PCI ne.c driver, and a relatively new PCI-only ne2k-pci.c driver.
To use the original ISA/PCI driver you have to say `Y' to the `Other ISA cards' option when running make config as you are actually using the same NE2000 driver as the ISA cards use. (That should also give you a hint that these cards aren't anywhere as intelligent as say a PCNet-PCI or DEC 21040 card...)
The newer PCI-only driver differs from the ISA/PCI driver in that all the support for old NE1000 8 bit cards has been removed and that data is moved to/from the card in bigger blocks, without any intervening pauses that the older ISA-NE2000's required for reliable operation. The result is a driver that is slightly smaller and slightly more efficient, but don't get too excited as the difference will not be obvious under normal use. (If you really wanted maximum efficiency/low CPU use, then a PCI-NE2000 is simply a very poor choice.) Driver updates and more information can be found at:
http://www.scyld.com/linux/drivers/ne2k-pci.html
If you have a NE2000 PCI card that is not detected by the most current version of the driver, please contact the maintainer of the NE2000 driver as listed in /usr/src/linux/MAINTAINERS along with the output from a cat /proc/pci and dmesg so that support for your card can also be added to the driver.
Also note that various card makers have been known to put `NE2000 Compatible' stickers on their product boxes even when it is completely different (e.g. PCNet-PCI or RealTek 8139). If in doubt check the main chip number against this document.
Status: Not Supported.
These are ISA 100Mbps cards based on the National Semiconductor DP83800 and DP83840 chips. There is currently no driver support, nor has anyone reported that they are working on a driver. Apparently documentation on the chip is unavailable with the exception of a single PDF file that doesn't give enough details for a driver.
Status: Supported, Driver Name: lance
These cards use the original 7990 LANCE chip from AMD and are supported using the Linux lance driver. Newer NE2100 clones use the updated PCnet/ISA chip from AMD.
Some earlier versions of the lance driver had problems with getting the IRQ line via autoIRQ from the original Novell/Eagle 7990 cards. Hopefully this is now fixed. If not, then specify the IRQ via LILO, and let us know that it still has problems.
DMA selection and chip numbering information can be found in AMD LANCE.
More technical information on LANCE based cards can be found in Notes on AMD...
Status: Semi-Supported, Driver Name: ne2
There were a few NE2000 microchannel cards made by various companies. This driver, available in v2.2 kernels, will detect the following MCA cards: Novell Ethernet Adapter NE/2, Compex ENET-16 MC/P, and the Arco Ethernet Adapter AE/2.
Status: Not Supported.
This old EISA card uses a 8MHz 80186 in conjunction with an i82586. Nobody is working on a driver for it, as there is no information available on the card, and no real demand for a driver either.
Status: Supported, Driver Name: ne3210 (+8390)
This EISA card is completely different from the NE3200, as it uses a Nat Semi 8390 chip. The driver can be found in the v2.2 kernel source tree. Ensure you set the shared memory address below 1MB or above the highest address of the physical RAM installed in the machine.
Status: Supported, Driver Name: pcnet32
These are just AMD PCnet-PCI cards ('970A) chips. More information on LANCE/PCnet based cards can be found in AMD LANCE.
Status: Supported, Driver Name: ne (+8390)
Apparently this is a NE2000 clone, and works fine with Linux.
Status: Supported, Driver Name: de4x5, tulip
This is yet another PCI card that is based on DEC's Tulip chip. It has been reported to work fine with Linux.
See the section on the 21040 chip ( DEC 21040) for more driver information.
Status: Supported, Driver Name: wd (+8390)
The PureData PDUC8028 and PDI8023 series of cards are `almost clones' of the wd80x3 cards - there is special code in the wd.c driver to probe for these cards.
Racal Interlan can be reached via WWW at www.interlan.com. I believe they were also known as MiCom-Interlan at one point in the past.
Status: Semi-Supported, Driver Name: es3210
This is an EISA 8390 based shared memory card. An experimetal driver is shipped with v2.2 kernels and it is reported to work fine, but the EISA IRQ and shared memory address detection appears not to work with (at least) the early revision cards. (This problem is not unique to the Linux world either...) In that case, you have to supply them to the driver. For example, card at IRQ 5 and shared memory 0xd0000, with a modular driver, add options es3210 irq=5 mem=0xd0000 to /etc/conf.modules. Or with the driver compiled into the kernel, supply at boot ether=5,0,0xd0000,eth0 The I/O base is automatically detected and hence a value of zero should be used.
Status: Semi-Supported, Driver Name: ni5010
You used to have to go get the driver for these old 8 bit MiCom-Interlan cards separately, but now it is shipped with the v2.2 kernels as an experimental driver.
Status: Semi-Supported, Driver Name: ni52
This card also uses one of the Intel chips. Michael Hipp has written a driver for this card. It is included in the standard kernel as an `alpha' driver. Michael would like to hear feedback from users that have this card. See Alpha Drivers for important information on using alpha-test ethernet drivers with Linux.
Status: Semi-Supported, Driver Name: ni65
There is also a driver for the LANCE based NI6510, and it is also written by Michael Hipp. Again, it is also an `alpha' driver. For some reason, this card is not compatible with the generic LANCE driver. See Alpha Drivers for important information on using alpha-test ethernet drivers with Linux.
Status: Supported, Driver Name: lance
As of kernel 1.3.23, the generic LANCE driver had a check added to it for the 0x52, 0x44 NI6510EB specific signature. Others have reported that this signature is not the same for all NI6510EB cards however, which will cause the lance driver to not detect your card. If this happens to you, you can change the probe (at about line 322 in lance.c) to printk() out what the values are for your card and then use them instead of the 0x52, 0x44 defaults.
The cards should probably be run in `high-performance' mode and not in the NI6510 compatible mode when using the lance driver.
Status: Supported, Driver Name: atp
This is a generic, low-cost OEM pocket adaptor being sold by AT-Lan-Tec, and (likely) a number of other suppliers. A driver for it is included in the standard kernel. Note that there is substantial information contained in the driver source file `atp.c'.
Note that the device name that you pass to ifconfig was not eth0 but atp0 for earlier versions of this driver.
Status: Supported, Driver Name: ne (+8390)
This is an ISA NE2000 clone, and is reported to work fine with the linux NE2000 driver. The rset8009.exe program can be obtained from RealTek's WWW site at http://www.realtek.com.tw - or via ftp from the same site.
Status: Supported, Driver Name: ne (+8390)
This is a Plug and Pray version of the above. Use the DOS software to disable PnP and enable jumperless configuration; set the card to a sensible I/O address and IRQ and you should be ready to go. (If using the driver as a module, don't forget to add an io=0xNNN option to /etc/conf.modules). The rset8019.exe program can be obtained from RealTek's WWW site at http://www.realtek.com.tw - or via ftp from the same site.
Status: Supported, Driver Name: ne, ne2k-pci (+8390)
This is a PCI single chip implementation of a NE2000 clone. Various vendors are now selling cards with this chip. See NE2000-PCI for information on using any of these cards. Note that this is still a 10+ year old design just glued onto a PCI bus. Performance won't be staggeringly better than the equivalent ISA model.
Status: Semi-Supported, Driver Name: rtl8139, 8139too
Another PCI single chip ethernet solution from RealTek. A driver for cards based upon this chip was included in the v2.0.34 release of linux. You currently have to answer `Y' when asked if you want experimental drivers for v2.2 kernels to get access to this driver.
Donald says that cards based on this chip are around the same price (or even cheaper in places - 13 bucks!) as a PCI NE2000 clone card, and while the 8139 design is not the best 10/100 board, it is better than a PCI NE2000 clone card.
The 2.4.x kernels have another driver called 8139too which was based on rtl8139 but tries to adress some of the more common problems people were reporting, so you may wish to try that if using a 2.4 kernel.
Status: Semi-Supported, Driver Name: 3c501
This is just a 3c501 clone, with a different S.A. PROM prefix. I assume it is equally as brain dead as the original 3c501 as well. The driver checks for the NP943 I.D. and then just treats it as a 3c501 after that. See 3Com 3c501 for all the reasons as to why you really don't want to use one of these cards.
Status: Supported, Driver Name: sk_g16
This driver was included into the v1.1 kernels, and it was written by PJD Weichmann and SWS Bern. It appears that the SK G16 is similar to the NI6510, in that it is based on the first edition LANCE chip (the 7990). Once again, it appears as though this card won't work with the generic LANCE driver.
Status: Supported, Driver Name: seeq8005
There is little information about the card included in the driver, and hence little information to be put here. If you have a question, you are probably best trying to e-mail the driver author as listed in the source.
The ethernet part of Western Digital was bought out by SMC many years ago when the wd8003 and wd8013 were the main product. Since then SMC has continued making 8390 based ISA cards (Elite16, Ultra, EtherEZ) and also added several PCI products to their range.
Contact information for SMC:
SMC / Standard Microsystems Corp., 80 Arkay Drive, Hauppage, New York, 11788, USA. Technical Support via phone: 800-992-4762 (USA) or 800-433-5345 (Canada) or 516-435-6250 (Other Countries). Literature requests: 800-SMC-4-YOU (USA) or 800-833-4-SMC (Canada) or 516-435-6255 (Other Countries). Technical Support via E-mail: [email protected]. FTP Site: ftp.smc.com. WWW Site: SMC.
Status: Supported, Driver Name: wd (+8390)
These are the 8-bit versions of the card. The 8 bit 8003 is slightly less expensive, but only worth the savings for light use. Note that some of the non-EEPROM cards (clones with jumpers, or old old old wd8003 cards) have no way of reporting the IRQ line used. In this case, auto-irq is used, and if that fails, the driver silently assings IRQ 5. You can get the SMC setup/driver disks from SMC's ftp site. Note that some of the newer SMC `SuperDisk' programs will fail to detect the real old EEPROM-less cards. The file SMCDSK46.EXE seems to be a good all-round choice. Also the jumper settings for all their cards are in an ASCII text file in the aforementioned archive. The latest (greatest?) version can be obtained from ftp.smc.com.
As these are basically the same as their 16 bit counterparts (WD8013 / SMC Elite16), you should see the next section for more information.
Status: Supported, Driver Name: wd (+8390)
Over the years the design has added more registers and an EEPROM. (The first wd8003 cards appeared about ten years ago!) Clones usually go by the `8013' name, and usually use a non-EEPROM (jumpered) design. Late model SMC cards will have the SMC 83c690 chip instead of the original Nat Semi DP8390 found on earlier cards. The shared memory design makes the cards a bit faster than PIO cards, especially with larger packets. More importantly, from the driver's point of view, it avoids a few bugs in the programmed-I/O mode of the 8390, allows safe multi-threaded access to the packet buffer, and it doesn't have a programmed-I/O data register that hangs your machine during warm-boot probes.
Non-EEPROM cards that can't just read the selected IRQ will attempt auto-irq, and if that fails, they will silently assign IRQ 10. (8 bit versions will assign IRQ 5)
Cards with a non standard amount of memory on board can have the memory size specified at boot (or as an option in /etc/conf.modules if using modules). The standard memory size is 8kB for an 8bit card and 16kB for a 16bit card. For example, the older WD8003EBT cards could be jumpered for 32kB memory. To make full use of that RAM, you would use something like (for I/O=0x280 and IRQ 9):
LILO: linux ether=9,0x280,0xd0000,0xd8000,eth0
Also see 8013 problems for some of the more common problems and frequently asked questions that pop up often.
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Supported, Driver Name: smc-ultra (+8390)
This ethercard is based on the 83c790 chip from SMC, which has a few new features over the 83c690. While it has a mode that is similar to the older SMC ethercards, it's not entirely compatible with the old WD80*3 drivers. However, in this mode it shares most of its code with the other 8390 drivers, while operating slightly faster than a WD8013 clone.
Since part of the Ultra looks like an 8013, the Ultra probe is supposed to find an Ultra before the wd8013 probe has a chance to mistakenly identify it.
Donald mentioned that it is possible to write a separate driver for the Ultra's `Altego' mode which allows chaining transmits at the cost of inefficient use of receive buffers, but that will probably not happen.
Bus-Master SCSI host adaptor users take note: In the manual that ships with Interactive UNIX, it mentions that a bug in the SMC Ultra will cause data corruption with SCSI disks being run from an aha-154X host adaptor. This will probably bite aha-154X compatible cards, such as the BusLogic boards, and the AMI-FastDisk SCSI host adaptors as well.
SMC has acknowledged the problem occurs with Interactive, and older Windows NT drivers. It is a hardware conflict with early revisions of the card that can be worked around in the driver design. The current Ultra driver protects against this by only enabling the shared memory during data transfers with the card. Make sure your kernel version is at least 1.1.84, or that the driver version reported at boot is at least smc-ultra.c:v1.12 otherwise you are vulnerable.
If you intend on using this driver as a loadable module you should probably see Using the Ethernet Drivers as Modules for module specific information.
Status: Supported, Driver Name: smc-ultra32 (+8390)
This EISA card shares a lot in common with its ISA counterpart. A working (and stable) driver is included in both v2.0 and v2.2 kernels. Thanks go to Leonard Zubkoff for purchasing some of these cards so that linux support could be added for them.
Status: Supported, Driver Name: smc-ultra (+8390)
This card uses SMC's 83c795 chip and supports the Plug 'n Play specification. It also has an SMC Ultra compatible mode, which allows it to be used with the Linux Ultra driver. For best results, use the SMC supplied program (avail. from their www/ftp site) to disable PnP and configure it for shared memory mode. See the above information for notes on the Ultra driver.
For v1.2 kernels, the card had to be configured for shared memory operation. However v2.0 kernels can use the card in shared memory or programmed I/O mode. Shared memory mode will be slightly faster, and use less CPU resources as well.
Status: Supported, Driver Name: de4x5, tulip
NB: The EtherPower II is an entirely different card. See below! These cards are a basic DEC 21040 implementation, i.e. one big chip and a couple of transceivers. Donald has used one of these cards for his development of the generic 21040 driver (aka tulip.c). Thanks to Duke Kamstra, once again, for supplying a card to do development on.
Some of the later revisons of this card use the newer DEC 21041 chip, which may cause problems with older versions of the tulip driver. If you have problems, make sure you are using the latest driver release, which may not yet be included in the current kernel source tree.
See DEC 21040 for more details on using one of these cards, and the current status of the driver.
Apparently, the latest revision of the card, the EtherPower-II uses the 9432 chip. It is unclear at the moment if this one will work with the present driver. As always, if unsure, check that you can return the card if it doesn't work with the linux driver before paying for the card.
Status: Semi-Supported, Driver Name: epic100
These cards, based upon the SMC 83c170 chip, are entirely different than the Tulip based cards. A new driver has been included in kernels v2.0 and v2.2 to support these cards. For more details, see:
http://www.scyld.com/linux/drivers/epic100.html
Status: Semi-Supported, Driver Name: rtl8139
Apparently SMC is no longer the same company that brought you cards like the Ultra and the EPIC. The chip design part is now called SMSC and you will see the SMC name stuck on low end OEM boards like this one - a RealTek 8139 with a modified EEPROM.
Status: Not Supported.
These 8 bit cards are based on the Fujitsu MB86950, which is an ancient version of the MB86965 used in the Linux at1700 driver. Russ says that you could probably hack up a driver by looking at the at1700.c code and his DOS packet driver for the Tiara card (tiara.asm). They are not very common.
Status: Not Supported.
These are 16bit I/O mapped 8390 cards, much similar to a generic NE2000 card. If you can get the specifications from SMC, then porting the NE2000 driver would probably be quite easy. They are not very common.
Status: Supported, Driver Name: smc9194
The SMC9000 is a VLB card based on the 91c92 chip. The 91c92 appears on a few other brand cards as well, but is fairly uncommon.
Status: Semi-Supported, Driver Name: smc9194
The SMC 91c92 driver is supposed to work for cards based on this 100Base-T chip, but at the moment this is unverified.
Status: Supported, Driver Name: tlan
This driver covers many Compaq built-in ethernet devices, including the NetFlex and Netelligent groups. It also supports the Olicom 2183, 2185, 2325 and 2326 products.
This is yet another PCI card that is based on DEC's 21040 chip.
See the section on the 21040 chip ( DEC 21040) for more information.
You probably won't see a VIA networking card, as VIA make several networking chips that are then used by others in the construction of an ethernet card. They have a WWW site at:
http://www.via.com.tw/
Status: Supported, Driver Name: ne, ne2k-pci (+8390)
This controller chip is VIA's PCI-NE2000 offering. You can choose between the ISA/PCI ne.c driver or the PCI-only ne2k-pci.c driver. See the PCI-NE2000 section for more details.
Status Supported, Driver Name: via-rhine
This relatively new driver can be found in current 2.0 and 2.1 kernels. It is an improvement over the 86C926 NE2000 chip in that it supports bus master transfers, but strict 32 bit buffer alignment requirements limit the benefit gained from this. For more details and driver updates, see:
http://www.scyld.com/linux/drivers/via-rhine.html
Please see SMC for information on SMC cards. (SMC bought out Western Digital's network card section many years ago.)
Winbond don't really make and sell complete cards to the general public -- instead they make single chip ethernet solutions that other companies buy, stick onto a PCI board with their own name and then sell through retail stores. Some setup programs and tech support is available at:
http://www.winbond.com.tw
Status: Semi-Supported, Driver Name: winbond-840
This chip has been described as `the mutant spawn of a NE2000 and a Tulip clone' -- see the driver notes for more details. The driver isn't currently shipped with the kernel, as it is in the testing phase (as of Sept 1998). It is available at:
http://www.scyld.com/linux/drivers/test/winbond-840.c
Status: Supported, Driver Name: ne (+8390)
These are Winbond's ISA 10Mbps ne2000 compatible ethernet chips. Setup programs are available at the Winbond site.
Status: Supported, Driver Name: ne, ne2k-pci (+8390)
This chip is one of the two commonly found on the low price PCI ne2000 cards sold by lots of manufacturers. Note that this is still a 10+ year old design just glued onto a PCI bus. Performance won't be staggeringly better than the equivalent ISA model.
For the longest time, Xircom wouldn't release the programming information required to write a driver, unless you signed your life away. Apparently enough linux users have pestered them for driver support (they claim to support all popular networking operating systems...) so that they have changed their policy to allow documentation to be released without having to sign a non-disclosure agreement. Some people have said they they will release the source code to the SCO driver, while others have been told that they are no longer providing information on `obsolete' products like the earlier PE models. If you are interested and want to check into this yourself, you can reach Xircom at 1-800-874-7875, 1-800-438-4526 or +1-818-878-7600.
Status: Not Supported.
Not to get your hopes up, but if you have one of these parallel port adaptors, you may be able to use it in the DOS emulator with the Xircom-supplied DOS drivers. You will have to allow DOSEMU access to your parallel port, and will probably have to play with SIG (DOSEMU's Silly Interrupt Generator).
Status: Semi-Supported, Driver Name: ????
Some of the Xircom PCMCIA card(s) have drivers that are available with David Hinds PCMCIA package. Check there for the most up to date indformation
Status: Supported, Driver Name: znet
The built-in Z-Note network adaptor is based on the Intel i82593 using two DMA channels. There is an (alpha?) driver available in the present kernel version. As with all notebook and pocket adaptors, it is under the `Pocket and portable adaptors' section when running make config. Also note that the IBM ThinkPad 300 is compatible with the Z-Note.
Status: Supported, Driver Name: de4x5, tulip
You have a choice of two drivers for cards based on this chip. There is the DE425 driver written by David, and the generic 21040 driver that Donald has written.
Note that as of 1.1.91, David has added a compile time option that may allow non-DEC cards (such as the Znyx cards) to work with this driver. Have a look at README.de4x5 for details.
See DEC 21040 for more information on these cards, and the present driver situation.
Okay, so your uncle's cousin's neighbour's friend had a brother who found an old ISA ethernet card in the AT case he was using as a cage for his son's pet hampster. Somehow you ended up with the card and want to try and use it with linux, but nobody has a clue what the card is and there isn't any documentation.
First of all, look for any obvious model numbers that might give a clue. Any model number that contains 2000 will most likely be a NE2000 clone. Any cards with 8003 or 8013 on them somewhere will be Western/Digital WD80x3 cards or SMC Elite cards or clones of them.
Look for the biggest chip on the card. This will be the network controller (NIC) itself, and most can be identified by the part number. If you know which NIC is on the card, the following might be able to help you figure out what card it is.
Probably still the most common NIC is the National Semiconductor DP8390 aka NS32490 aka DP83901 aka DP83902 aka DP83905 aka DP83907. And those are just the ones made by National! Other companies such as Winbond and UMC make DP8390 and DP83905 clone parts, such as the Winbond 89c904 (DP83905 clone) and the UMC 9090. If the card has some form of 8390 on it, then chances are it is a ne1000 or ne2000 clone card. The second most common 8390 based card are wd80x3 cards and clones. Cards with a DP83905 can be configured to be an ne2000 or a wd8013. Never versions of the genuine wd80x3 and SMC Elite cards have an 83c690 in place of the original DP8390. The SMC Ultra cards have an 83c790, and use a slightly different driver than the wd80x3 cards. The SMC EtherEZ cards have an 83c795, and use the same driver as the SMC Ultra. All BNC cards based on some sort of 8390 or 8390 clone will usually have an 8392 (or 83c692, or ???392) 16 pin DIP chip very close to the BNC connector.
Another common NIC found on older cards is the Intel i82586. Cards having this NIC include the 3c505, 3c507, 3c523, Intel EtherExpress-ISA, Microdyne Exos-205T, and the Racal-Interlan NI5210.
The original AMD LANCE NIC was numbered AM7990, and newer revisions include the 79c960, 79c961, 79c965, 79c970, and 79c974. Most cards with one of the above will work with the Linux LANCE driver, with the exception of the old Racal-Interlan NI6510 cards that have their own driver.
Newer PCI cards having a DEC 21040, 21041, 21140, or similar number on the NIC should be able to use the linux tulip or de4x5 driver.
Other PCI cards having a big chip marked RTL8029 or 89C940 or 86C926 are ne2000 clone cards, and the ne driver in linux version v2.0 and up should automatically detect these cards at boot.
Each ethernet card has its own six byte address that is unique to that card. The first three bytes of that address are the same for each card made by that particular manufacturer. For example all SMC cards start with 00:00:c0. The last three are assigned by the manufacturer uniquely to each individual card as they are produced.
If your card has a sticker on it giving all six bits of its address, you can look up the vendor from the first three. However it is more common to see only the last three bytes printed onto a sticker attached to a socketed PROM, which tells you nothing.
You can determine which vendors have which assigned addresses from RFC-1340. Apparently there is a more up to date listing available in various places as well. Try a WWW or FTP search for EtherNet-codes or Ethernet-codes and you will find something.
As part of the certification process a card typically has to pass before being sold to the user, it gets tested by the FCC, and from this gets a FCC ID which is supposed to be printed on the card somewhere. For example, a card has on it FCC ID: J158013EWC - and this card happens to be a SMC/WD8013-EWC. Some web sites like www.driverguide.com and drdriver.com make use of listings of FCC IDs that may help with less obvious ID numbers.
If you are still not sure what the card is, but have at least narrowed it down some, then you can build a kernel with a whole bunch of drivers included, and see if any of them autodetect the card at boot.
If the kernel doesn't detect the card, it may be that the card is not configured to one of the addresses that the driver probes when looking for a card. In this case, you might want to try getting scanport.tar.gz from your local linux ftp site, and see if that can locate where your card is jumpered for. It scans ISA I/O space from 0x100 to 0x3ff looking for devices that aren't registered in /proc/ioports. If it finds an unknown device starting at some particular address, you can then explicity point the ethernet probes at that address with an ether= boot argument.
If you manage to get the card detected, you can then usually figure out the unknown jumpers by changing them one at a time and seeing at what I/O base and IRQ that the card is detected at. The IRQ settings can also usually be determined by following the traces on the back of the card to where the jumpers are soldered through. Counting the `gold fingers' on the backside, from the end of the card with the metal bracket, you have IRQ 9, 7, 6, 5, 4, 3, 10, 11, 12, 15, 14 at fingers 4, 21, 22, 23, 24, 25, 34, 35, 36, 37, 38 respectively. Eight bit cards only have up to finger 31.
Jumpers that appear to do nothing usually are for selecting the memory address of an optional boot ROM. Other jumpers that are located near the BNC or RJ-45 or AUI connectors are usually to select the output media. These are also typically near the `black box' voltage converters marked YCL, Valor, or Fil-Mag.
A nice collection of jumper settings for various cards can be found at the following URL:
Ethercard Settings
There are a few other drivers that are in the linux source that present an ethernet-like device to network programs, while not really being ethernet. These are briefly listed here for completeness.
dummy.c - The purpose of this driver is to provide a device to point a route through, but not to actually transmit packets.
eql.c - Load Equalizer, enslaves multiple devices (usually modems) and balances the Tx load across them while presenting a single device to the network programs.
ibmtr.c - IBM Token Ring, which is not really ethernet. Broken-Ring requires source routing and other uglies.
loopback.c - Loopback device, for which all packets from your machine and destined for your own machine go. It essentially just moves the packet off the Tx queue and onto the Rx queue.
pi2.c - Ottawa Amateur Radio Club PI and PI2 interface.
plip.c - Parallel Line Internet Protocol, allows two computers to send packets to each other over two joined parallel ports in a point-to-point fashion.
ppp.c - Point-to-Point Protocol (RFC1331), for the Transmission of Multi-protocol Datagrams over a Point-to-Point Link (again usually modems).
slip.c - Serial Line Internet Protocol, allows two computers to send packets to each other over two joined serial ports (usually via modems) in a point-to-point fashion.
tunnel.c - Provides an IP tunnel through which you can tunnel network traffic transparently across subnets
wavelan.c - An Ethernet-like radio transceiver controlled by the Intel 82586 coprocessor which is used on other ethercards such as the Intel EtherExpress.