[Prev][Next][Index][Thread]

DEC 21140 Fast Ethernet card still don't work



-----Messaggio originale-----
Da: Kevin Van Maren <vanmaren@fast.cs.utah.edu>
A: lettieri@iet.unipi.it <lettieri@iet.unipi.it>; oskit@fast.cs.utah.edu
<oskit@fast.cs.utah.edu>
Data: lunedì 14 giugno 1999 20.54
Oggetto: Re: problem with a DEC 21140 Fast Ethernet card


>> Hi,
>>
>> We have a problem with our ethernet card. We are using a DEC PCI Fast =
>> Ethernet DECchip 21140 and Flux OSkit version 0.97.=20
>> =20
>> We have set TULIP_PORT in ...linux/src/drivers/net/tulip.c equal to =
>> TULIP_100TP_PORT.
>> =20
>> This is what "pingreply" prints during boot:
>> =20
>> Probing devices...
>> tulip.c:v0.10 8/11/95 becker@cesdis.gsfc.nasa.gov
>>  +0.72 4/17/96 http://www.dsl.tutics.tut.ac.jp/~linux/tulip
>>  +0.72 12/15/96 mjacob@feral.com (2.0.72)
>> eth0: 21140 (DEC 21140 Tulip) at 0xd000, 00:40:05:a4:86:1f, IRQ 9
>> Current hardware tree:
>>   isa.isa   Industry Standard Architecture (ISA) Bus
>>     0xd000.tulip  DEC 21040 Ethernet Adaptor
>> 1 Ethernet adaptor found:
>>     Tulip DEC 21040 Ethernet adaptor  00:40:05:a4:86:1f
>> eth0: enabling 100TX port.
>> bootp: Trying bootp with the tulip card (0:40:5:a4:86:1f)...
>> =20
>> ...then it makes no progress. The bootp server (10.1.1.21) does not =
>> receive any packet.
>>     We have tried all other values for TULIP_PORT. The enabled port =
>> changes accordingly, but the outcame is always the same. =20
>
>
>Well, I notice that the driver does not identify which card you
>have.  The problem could be that the driver is not handling
>the particular card you have correctly.  This fits in with the
>bootp server not receiving a request.
>
>For example, with the old SMC EtherPower 10/100 card, it does:
>    tulip.c:v0.10 8/11/95 becker@cesdis.gsfc.nasa.gov
>            +0.72 4/17/96 http://www.dsl.tutics.tut.ac.jp/~linux/tulip
>            +0.02 12/15/96 mjacob@feral.com (2.0.27)
>    eth0: smc9332 (DEC 21140 Tulip) at 0x6100, 00:00:c0:06:d2:e4, IRQ 11
>    eth0: enabling 10TP port.
>
>What brand/model networking card are you attempting to use?
>Can you verify that the card works under Linux 2.0.29?
>
>Another option would be to take the current if_de.c driver
>from 2.0.3x and see if a "drop in replacement" works any better.
>There may be some changes required to get a newer driver working,
>but they should be minimal for a 2.0.X driver.  A more
>recent version of the driver may support your networking card
>better.
>
>Kevin Van Maren
>University of Utah
>

Thanks for the answer.
The card is a TRENDnet PCI COMBO DIGITAL CHIP
This is what we have tried to do after your suggestions:

- we have installed a red-hat linux distribution featuring 2.0.32 kernel:
the card works well

- we didn't had a linux distribution installing kernel 2.0.29. We rather
downloaded kernel 2.0.29 and use it in the same red-hat distribution as
before (!). The card doesn't work (we sometimes receive an error from eth0
saying "too much work during interrupt", and anyway ping hangs up)

- we tried to replace "tulip.c" file from oskit with the corresponding file
in 2.0.32 kernel: this required two little changes to compile and link:
    - we have deleted 'skb->free = 1' at line 2537. Oskit doesn't seem    to
use this field
    - char x86 seems to be undefined in oskit. We have temporarily defined
it in "tulip.c" to be 5 (i.e. pentium)

... now pingreply says:

Probing devices....
tulip.c: v0.79 9/3/97 becker@cesdis.gsfc.nasa.gov
eth0: DEC DS21140 Tulip at 0xd000, 00 40 05 a4 86 1f, IRQ 9.
    the following verbose information is emitted for
<6> bug reports on media selection.
eth0: EEPROM default media type Autosense.
eth0: Index #0 - Media MII (#11) described by a 21140 MII PHY (1) block.
eth0: MII transciever found at MDI0 address 0.
Current hardware tree:
  isa.isa   Industry Standard Architecture (ISA) Bus
    0xd000.tulip  DEC 21040 Ethernet Adaptor
1 Ethernet adaptor found:
    Tulip DEC 21040 Ethernet adaptor  00:40:05:a4:86:1f
bootp: Trying bootp with the tulip card (0:40:5:a4:86:1f)...

...and nothing more, as before (sigh)

Before giving up, we would like to know if something else could be done