I'm having absolutely no luck building a working multicast routing capable kernel with the eepro100.c driver. The driver works beautifully in a 'normal' kernel (in both cases, I'm actually dynamically loading the eepro100 module at boot). The machine boots fine; no error messages when the eepro100 module loads, but-- once up-- the machine passes NO traffic across the network interface [with or without the 'standard' multicast route active within the routing table]. Looking through the various system logs and through /proc, I have found nothing to indicate why the traffic is not passing through the interface. The loopback address works fine. Looking at the back of the machine, none of the lights on the pro/100 card are lit-- normally, it appears that at least one should be lit to indicate a viable connection to the hub? I am working with version 2.0.23 of the linux kernel under a debian-based system. It is a Dell Dimension XPS 200. No SCSI controller. Eventually the machine produces the following error on console: eth0: Transmit timed out: status 0050 command 0020. eth0: Tx timout fill index 22 scavenge index 0 Tx queue 000c0000 000c0000 000c0000 000c0000 000c0000 400c0000 0000a000 Rx ring 00000001 ... repeats ...
This message shows up about once every minute after the first one; which often takes about 10-15 minutes to show. I would be happy to pursue debugging further, but don't know where to start. Any pointers would be appreciated. b.bum -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]