Mailing List ArchiveSupport open source code!
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]Re: ethner card problem
- To: tlug@example.com
- Subject: Re: ethner card problem
- From: Frank BENNETT <bennett@example.com>
- Date: Mon, 18 Sep 2000 20:28:11 +0900
- Content-Transfer-Encoding: 7bit
- Content-Type: text/plain; charset=iso-2022-jp
- In-Reply-To: <E13avlw-00044T-00@example.com>; from ulrike@example.com on Mon, Sep 18, 2000 at 04:55:28PM +0900
- References: <E13avlw-00044T-00@example.com>
- Reply-To: tlug@example.com
- Resent-From: tlug@example.com
- Resent-Message-ID: <tKLSJD.A.ijG.h6fx5@example.com>
- Resent-Sender: tlug-request@example.com
On Mon, Sep 18, 2000 at 04:55:28PM +0900, ulrike@example.com wrote: > I bought an ethernetcard, a Corega Ether PCC-T, and although it had a little > penguin on the box I am having difficultes with getting it recognized. In fact, > yesterday cardinfo claimed it was some kind of other ethernetcard, forgot the > name, today it says it is "Anonymous Memory". Does that mean it is broken now? I was bitten by this same card just today, as it happens. Here's what to do. What I found was that the card manager chirped happily when I slotted the card into the bus, but (a) I got a link light but no connection, and (b) in X, the display went blank immediately if I stopped moving the cursor. Neat, I guess. Not useful, but maybe neat. Definitely weird. Doing "lsmod" showed that the "cb_enabler" and "tulip_cb" modules were loaded. Tulip? The PCC-T is supposed to be a 10baseT-only card. Tulip is a 100baseT-capable chip. This had to be wrong. cardctl gets its information directly from the card, not from the /etc/pcmcia/config card list. As a consequence, cardctl ident "lies" about what it thinks the card is. To "cardctl ident", it shows up like this: Socket 0: product info: "corega K.K.", "corega EtherII PCC-T" manfid: 0xc00f, 0x0000 function: 6 (network) None of the "corega" listings in the config file are bound to the tulip_cb driver, so the text description of the card in the config file was being ignored. Setting "tail -f /var/log/syslog" while inserting the card told a different story: Sep 18 11:27:19 vaio cardmgr[136]: initializing socket 0 Sep 18 11:27:19 vaio cardmgr[136]: socket 0: Allied Telesyn AT-2800 10/100 Fast Ethernet Sep 18 11:27:19 vaio cardmgr[136]: executing: 'insmod /lib/modules/2.2.17/pcmcia/cb_enabler.o' Sep 18 11:27:19 vaio cardmgr[136]: executing: 'insmod /lib/modules/2.2.17/pcmcia/tulip_cb.o' Sep 18 11:27:19 vaio kernel: tulip_reap() Sep 18 11:27:19 vaio kernel: tulip_attach(bus 0, function 0) Sep 18 11:27:19 vaio kernel: tulip.c:v0.91g-ppc 7/16/99 becker@example.com (modified by danilo@example.com for XIRCOM CBE, fixed by Doug Ledford) Sep 18 11:27:19 vaio kernel: eth0: Digital DS21143 Tulip rev 1 at 0x0, EEPROM not present, 00:4C:69:6E:75:79, IRQ 0. Sep 18 11:27:19 vaio kernel: eth0: Missing EEPROM, this interface may not work correctly! Sep 18 11:27:19 vaio cardmgr[136]: executing: './network start eth0' Sep 18 11:27:19 vaio cardmgr[136]: + ioctl: Operation not supported Sep 18 11:27:19 vaio cardmgr[136]: + SIOCSIFFLAGS: Resource temporarily unavailable Sep 18 11:27:19 vaio cardmgr[136]: + SIOCSIFFLAGS: Resource temporarily unavailable Sep 18 11:27:20 vaio cardmgr[136]: + SIOCADDRT: Network is down Sep 18 11:27:20 vaio cardmgr[136]: + SIOCADDRT: Network is unreachable Darn. That's not worth 2,800 yen. Checking /etc/pcmcia/config turned up the following entry for the Allied Telesyn card that uses the tulip driver: card "Allied Telesyn AT-2800 10/100 Fast Ethernet" manfid 0xc00f, 0x0000 bind "tulip_cb" The identification was turning on the "manfid" line --- note that the numbers there for the PCC-T and for the AT-2800 are the same. It may be that "manfid" is given priority, on the assumption that numbers are less ambiguous than text strings. Oops. In any case, things worked when I commented out the AT-2800 entry (completely), and added the following entry for the PCC-T: card "Corega PCC-T Ethernet" # version "corega K.K.", "corega EtherII PCC-T" manfid 0xc00f, 0x0000 bind "pcnet_cs" I don't think you need to have the card version string commented out like that -- that is left over from my tour of possibilities. The key seemed to be to eliminate the conflicting entry for the AT-2800. Once you've made these changes to /etc/pcmcia/config, restart card services with (in Debian): /etc/init.d/pcmcia restart Now when the card manager chirps, you should get a net connection. Let us know if this doesn't work. But I'm 99.9% sure that the above steps will put you in business with that card (this message is being sent through one of them :-) Cheers, ---- -x80 Frank G Bennett, Jr @@ Faculty of Law, Nagoya Univ () email: bennett@example.com Tel: +81[(0)52]789-2239 ()
- Follow-Ups:
- Re: ethner card problem
- From: Frank BENNETT <bennett@example.com>
- References:
- ethner card problem
- From: ulrike@example.com
Home | Main Index | Thread Index
- Prev by Date: APT::Force-LoopBreak option
- Next by Date: Re: ethner card problem
- Prev by thread: ethner card problem
- Next by thread: Re: ethner card problem
- Index(es):
Home Page Mailing List Linux and Japan TLUG Members Links