1Linux and the 3Com EtherLink III Series Ethercards (driver v1.18c and higher)
   4This file contains the instructions and caveats for v1.18c and higher versions
   5of the 3c509 driver. You should not use the driver without reading this file.
   7release 1.0
   828 February 2002
   9Current maintainer (corrections to):
  10  David Ruggiero <>
  14(0) Introduction
  16The following are notes and information on using the 3Com EtherLink III series
  17ethercards in Linux. These cards are commonly known by the most widely-used
  18card's 3Com model number, 3c509. They are all 10mb/s ISA-bus cards and shouldn't
  19be (but sometimes are) confused with the similarly-numbered PCI-bus "3c905"
  20(aka "Vortex" or "Boomerang") series.  Kernel support for the 3c509 family is
  21provided by the module 3c509.c, which has code to support all of the following
  24  3c509 (original ISA card)
  25  3c509B (later revision of the ISA card; supports full-duplex)
  26  3c589 (PCMCIA)
  27  3c589B (later revision of the 3c589; supports full-duplex)
  28  3c579 (EISA)
  30Large portions of this documentation were heavily borrowed from the guide
  31written the original author of the 3c509 driver, Donald Becker. The master
  32copy of that document, which contains notes on older versions of the driver,
  33currently resides on Scyld web server:
  36(1) Special Driver Features
  38Overriding card settings
  40The driver allows boot- or load-time overriding of the card's detected IOADDR,
  41IRQ, and transceiver settings, although this capability shouldn't generally be
  42needed except to enable full-duplex mode (see below). An example of the syntax
  43for LILO parameters for doing this:
  45    ether=10,0x310,3,0x3c509,eth0 
  47This configures the first found 3c509 card for IRQ 10, base I/O 0x310, and
  48transceiver type 3 (10base2). The flag "0x3c509" must be set to avoid conflicts
  49with other card types when overriding the I/O address. When the driver is
  50loaded as a module, only the IRQ may be overridden. For example,
  51setting two cards to IRQ10 and IRQ11 is done by using the irq module
  54   options 3c509 irq=10,11
  57(2) Full-duplex mode
  59The v1.18c driver added support for the 3c509B's full-duplex capabilities.
  60In order to enable and successfully use full-duplex mode, three conditions
  61must be met: 
  63(a) You must have a Etherlink III card model whose hardware supports full-
  64duplex operations. Currently, the only members of the 3c509 family that are
  65positively known to support full-duplex are the 3c509B (ISA bus) and 3c589B
  66(PCMCIA) cards. Cards without the "B" model designation do *not* support
  67full-duplex mode; these include the original 3c509 (no "B"), the original
  683c589, the 3c529 (MCA bus), and the 3c579 (EISA bus).
  70(b) You must be using your card's 10baseT transceiver (i.e., the RJ-45
  71connector), not its AUI (thick-net) or 10base2 (thin-net/coax) interfaces.
  72AUI and 10base2 network cabling is physically incapable of full-duplex
  75(c) Most importantly, your 3c509B must be connected to a link partner that is
  76itself full-duplex capable. This is almost certainly one of two things: a full-
  77duplex-capable  Ethernet switch (*not* a hub), or a full-duplex-capable NIC on
  78another system that's connected directly to the 3c509B via a crossover cable.
  80Full-duplex mode can be enabled using 'ethtool'.
  82/////Extremely important caution concerning full-duplex mode/////
  83Understand that the 3c509B's hardware's full-duplex support is much more
  84limited than that provide by more modern network interface cards. Although
  85at the physical layer of the network it fully supports full-duplex operation,
  86the card was designed before the current Ethernet auto-negotiation (N-way)
  87spec was written. This means that the 3c509B family ***cannot and will not
  88auto-negotiate a full-duplex connection with its link partner under any
  89circumstances, no matter how it is initialized***. If the full-duplex mode
  90of the 3c509B is enabled, its link partner will very likely need to be
  91independently _forced_ into full-duplex mode as well; otherwise various nasty
  92failures will occur - at the very least, you'll see massive numbers of packet
  93collisions. This is one of very rare circumstances where disabling auto-
  94negotiation and forcing the duplex mode of a network interface card or switch
  95would ever be necessary or desirable.
  98(3) Available Transceiver Types
 100For versions of the driver v1.18c and above, the available transceiver types are:
 1020  transceiver type from EEPROM config (normally 10baseT); force half-duplex
 1031  AUI (thick-net / DB15 connector)
 1042  (undefined)
 1053  10base2 (thin-net == coax / BNC connector)
 1064  10baseT (RJ-45 connector); force half-duplex mode
 1078  transceiver type and duplex mode taken from card's EEPROM config settings
 10812 10baseT (RJ-45 connector); force full-duplex mode
 110Prior to driver version 1.18c, only transceiver codes 0-4 were supported. Note
 111that the new transceiver codes 8 and 12 are the *only* ones that will enable
 112full-duplex mode, no matter what the card's detected EEPROM settings might be.
 113This insured that merely upgrading the driver from an earlier version would
 114never automatically enable full-duplex mode in an existing installation;
 115it must always be explicitly enabled via one of these code in order to be
 118The transceiver type can be changed using 'ethtool'.
 121(4a) Interpretation of error messages and common problems
 123Error Messages
 125eth0: Infinite loop in interrupt, status 2011. 
 126These are "mostly harmless" message indicating that the driver had too much
 127work during that interrupt cycle. With a status of 0x2011 you are receiving
 128packets faster than they can be removed from the card. This should be rare
 129or impossible in normal operation. Possible causes of this error report are:
 131   - a "green" mode enabled that slows the processor down when there is no
 132     keyboard activity. 
 134   - some other device or device driver hogging the bus or disabling interrupts.
 135     Check /proc/interrupts for excessive interrupt counts. The timer tick
 136     interrupt should always be incrementing faster than the others. 
 138No received packets 
 139If a 3c509, 3c562 or 3c589 can successfully transmit packets, but never
 140receives packets (as reported by /proc/net/dev or 'ifconfig') you likely
 141have an interrupt line problem. Check /proc/interrupts to verify that the
 142card is actually generating interrupts. If the interrupt count is not
 143increasing you likely have a physical conflict with two devices trying to
 144use the same ISA IRQ line. The common conflict is with a sound card on IRQ10
 145or IRQ5, and the easiest solution is to move the 3c509 to a different
 146interrupt line. If the device is receiving packets but 'ping' doesn't work,
 147you have a routing problem.
 149Tx Carrier Errors Reported in /proc/net/dev 
 150If an EtherLink III appears to transmit packets, but the "Tx carrier errors"
 151field in /proc/net/dev increments as quickly as the Tx packet count, you
 152likely have an unterminated network or the incorrect media transceiver selected. 
 1543c509B card is not detected on machines with an ISA PnP BIOS. 
 155While the updated driver works with most PnP BIOS programs, it does not work
 156with all. This can be fixed by disabling PnP support using the 3Com-supplied
 157setup program. 
 1593c509 card is not detected on overclocked machines 
 160Increase the delay time in id_read_eeprom() from the current value, 500,
 161to an absurdly high value, such as 5000. 
 164(4b) Decoding Status and Error Messages
 166The bits in the main status register are: 
 168value   description
 1690x01    Interrupt latch
 1700x02    Tx overrun, or Rx underrun
 1710x04    Tx complete
 1720x08    Tx FIFO room available
 1730x10    A complete Rx packet has arrived
 1740x20    A Rx packet has started to arrive
 1750x40    The driver has requested an interrupt
 1760x80    Statistics counter nearly full
 178The bits in the transmit (Tx) status word are: 
 180value   description
 1810x02    Out-of-window collision.
 1820x04    Status stack overflow (normally impossible).
 1830x08    16 collisions.
 1840x10    Tx underrun (not enough PCI bus bandwidth).
 1850x20    Tx jabber.
 1860x40    Tx interrupt requested.
 1870x80    Status is valid (this should always be set).
 190When a transmit error occurs the driver produces a status message such as 
 192   eth0: Transmit error, Tx status register 82
 194The two values typically seen here are:
 197Out of window collision. This typically occurs when some other Ethernet
 198host is incorrectly set to full duplex on a half duplex network. 
 20116 collisions. This typically occurs when the network is exceptionally busy
 202or when another host doesn't correctly back off after a collision. If this
 203error is mixed with 0x82 errors it is the result of a host incorrectly set
 204to full duplex (see above).
 206Both of these errors are the result of network problems that should be
 207corrected. They do not represent driver malfunction.
 210(5) Revision history (this file)
 21228Feb02 v1.0  DR   New; major portions based on Becker original 3c509 docs
 214 kindly hosted by Redpill Linpro AS, provider of Linux consulting and operations services since 1995.