ne2000 PCI clone: HW problem?

ne2000 PCI clone: HW problem?

Post by and » Sun, 28 Feb 1999 04:00:00

Hi all,

when adding a new box to a ethernetwork I encountered some problem
that looks to me like a hardware problem:

The network:  used to consist of two linux boxes with ISA ne2000
        cards and 10base 2 cable. It used to work fine (and still
        does for those two boxes).

The nex box: has a rtl8029 (ne2000 compatible) PCI card. It is running
        a SUSE 6.0 Linux with the included 2.0.36 kernel included.

The symptoms: the card is correctly recognized by the driver, ifconfig
        works as I expect and shows me

        eth0      Link encap:10Mbps Ethernet  HWaddr 00:E0:7D:50:29:FA
                inet addr:  Bcast:  Mask:
                UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
                RX packets:0 errors:0 dropped:0 overruns:0
                TX packets:18 errors:0 dropped:0 overruns:0
                Interrupt:11 Base address:0x6300

        what looks fine to me.

        Now what looks wrong?

        - there are two LEDs on the card, one to show I/O traffic and
          one to show  `network connected' (LED on) or `network disconnected'
          (LED off). And this second LED is allways off.

        - when I try to ping another box in the network, the number of
          transmitted packages slowly goes up, but ping gets nothing back
          Ping claims to have transmitted a lot more packages than
          the statistics from ifconfig or netstat -i

        - arp -a  prints instead the ethernet address of the pinged box

        - the LED on the card signalling traffic indicates no traffic

The included diagnistics software for the card finds no errors when checking
the cards components. Unfortunately I cannot test the network conection
with this software because this needs two boxes running the software; but
this software detects unplugging the cable.

I used two different cables and exchanged the T-connectors between the
ethernet cards of the three boxes without changing the symptoms.

Well this looks to me like a hardware problem but I would like to be sure
before going back to the dealer... Is there any other possibility?

Thanks a lot for any help!



1. NE2000 clone: Hw. address read/write mismap 0

I have a Genius GE2000III Series ISA ethernet card.
It refuses to work under linux.

It is PnP compatible, but is in soft-jumpered mode. The NE*000 driver
(modular) correctly detects it on boot, with the correct address and IRQ
(0x300, IRQ10 - They work in Win98) but immediately afterwards, I start
getting messages like this:

Hw. address read/write mismap 0
eth0: trigger_send() called with the transmitter busy.

These messages pop up on the console every few seconds.
The following line also appears afterwards in the dmesg log, but not on the
screen (t changes):

eth0: Tx timed out, cable problem? TSR=0x3, ISR=0x0, t=715.

I know that there is nothing wrong with the cable (or the card) as they work
in Windows.
Ifconfig reports that it is recieving packets, but that all transmission
failed. Again, this indicates that it is not a cable fault as the cable is
10Base2 - transmit and recieve use the same wire.
/proc/ioports and /proc/intrerrupts report that the NE2000 is using the
correct resources. My PnP sound card works and is using different resouces.
I don't think I have any other hardware that might conflict.
My kernel is version 2.2.9.

Can anyone help?
Please reply by email

2. HP Omnibook Laptop

3. PCI NE2000 clone problem

4. There's clearly a bug SOMEWHERE. What to do?

5. Problems with NE2000 Clone PCI card.

6. n_tty.c buglet breaks error returns

7. ne2000-pci clone is clown

8. sockets - newbie question

9. patch for ne2000 pci clones 2.1.93/94

10. Caldera base 1.1- support for PCI NE2000 clone card

11. Trying to install PCI NE2000 clone

12. Configure PCI in the BIOS for a NE2000 Clone

13. Sharing IRQ on ne2000 PCI clones ?