Re: D-Link DFE-680TXD: "The transmitter stopped"

Donald Becker <[email protected]>
Newsgroups gmane.linux.drivers.tulip.general
Message-ID <[email protected]>
On 7 Nov 2002, Leif Nixon wrote:

> Donald Becker <[email protected]> writes:
> > What does 'tulip-diag -ee' report?
> I've attached the output below. Please note that I use -t 4, since
> tulip-diag didn't find anything recognizable in /proc/pci.
> 
> > I'm guessing that this is an ADMtek Comet chip, not a 21143.
> Now, that's interesting.

Did you use an unmodified tulip driver?
Or a version where someone else added an entry.
I don't see how my driver release would have misdetected this as
a 21143 instead of the Comet.

> But tulip.c supports both chips, right? So shouldn't things just work
> even if cardmgr is confused, as long as it modprobes the correct
> driver?

...only if the detection table is correct.

> > What is the CardBus ID and the PCI ID?
> >   CardBus ID: cardctl or read the lines in 'dmesg'.
>   Socket 1:
>     product info: "D-Link", "DFE-680TXD", "V1.0", ""
>     manfid: 0x13d1, 0xab02

This is in the driver.  It maps to an Abocom design with a ADMtek
Centaur-C/Comet chip.

>     function: 6 (network)
>     PCI id: 0x1186, 0x1541

This ID was not in the driver -- I just added it.

> >   PCI ID: Use 'lspci ....'  and/or 'cat /proc/pci'
> lspci says:
>   pcilib: Cannot open /proc/bus/pci/05/00.0
>   lspci: Unable to read 64 bytes of configuration space.

Bogus.

> and in /proc/pci the card turns up as just:
>   Bus  5, device   0, function  0:
>     Ethernet controller:  (rev 17).

Even more bogus.

> There seems to be conflicting opinions as to whether the card is a
> D-Link card, a Linksys card or simply an unknown card...

It's a D-Link PCI ID, with an Abocom (www.Abocom.com.tw) CardBus ID.

> EEPROM contents (256 words):
> 0x00:  0985 0002 0000 0000 5000 82ba 63e6 0000
> 0x08:  0000 0400 0000 0000 0000 0000 0000 0100
> 0x10:  1541 1186 1541 1186 ffff 0202 0000 80cc

The card loads its PCI ID from the configuration EEPROM.
There is nothing hard-wired that identifies the chip.

> 0xa0:  0313 4943 2053 d104 0213 21ab 0602 2201
> 0xa8:  0102 2202 0205 9680 0098 0522 0002 f5e1
> 0xb0:  2205 0302 1501 051c 4400 4c2d 6e69 006b
> 0xb8:  4644 2d45 3836 5430 4458 5600 2e31 0030
> 0xc0:  ff00 0400 0306 2a01 0000 0700 1106 0000
> 0xc8:  0001 0700 0206 0000 0001 0700 0706 0000
> 0xd0:  0002 0500 410a 0139 1eb5 b002 fefc ff84
> 0xd8:  ff84 ff84 ff84 ffff ffff ffff ffff ffff
> 0xe0:  ffff ffff ffff ffff ffff ffff ffff ffff

Pass "-eee" to the see the text in the CIS table.

-- 
Donald Becker				[email protected]
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Scyld Beowulf cluster system
Annapolis MD 21403			410-990-9993

_______________________________________________
tulip mailing list, [email protected]
To change to digest mode or unsubscribe visit
http://www.scyld.com/mailman/listinfo/tulip
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.