Neptune354

Marco,

Firstly, many congratulations on some excellent work. I am using Neptune 354 as I want to connect an aging Apple Mac SE/30 to my V1 WAG354. Unfortunately the auto-negotiation on the wired connection does not correctly sync with the Mac.

I can successfully connect the Mac to a Thinkpad using a crossover cable, providing I turn off auto-negotiation and force the Thinkpad's ethernet port to half duplex 10 Baset using ethtool. So it looks as though the hardware in the Mac is working, it is just a problem with the negotiation.

Having installed Neptune 354 v0.1 I can telnet onto the router. However, I am a little confused as to the network configuration of the router.

Am I right in thinking that eth0 is the device that all wired ports are connected to (this suggests that I may only be able to force all wired ports to a given setting)? Is it correct that both wlan0 and eth0 are bridged by br0?

Do you have any suggestions as to where to find information about the wired network chipset used by the WAG354 (/var/log/messages is not terrible helpful) or the current speed and duplex state of the 4 wired connections (I cannot see anything helpful in /proc)?

Thanks for any help you can provide and keep up the good work.


Ian

Yes, there is a single ethernet interface, eth0, that is connected to a 5 port switch, so you have 4 ports available.




Yes, this is how an access point is supposed to act, that is work in the same way with ethernet or wireless.
It is possible to un-bridge the interface, but you have to specify 2 different subnets, and eventually routing and natting.
With bridge instead you have to configure only the bridge.



the chip used in the wag354g is the ADM6996L, that is a 6 port switch. All its functions are well documented on the datasheet. The problem is that at now there is nothing much you can do because the firmware lacks of the drivers to control it.
There are sources for the adm6996l, because it is the same chip used in the wrt54g, which uses on its port as wan port using vlan. However sources must be ported, because the gpio lines and the functions to control them are different from the original version.
So the first thing to do is to find out if the control pins on the ADM are linked to some gpio pins of the processor, and take note of the gpio number. Then, go through the code and substitute the function calls to control the gpio with the functions used by our firmware.
That is, the work that this guy has done: http://www.bettina-attack.de/jonny/view ... 96_driver/