[En-Nut-Discussion] Can anyone check httpd example on AT91SAM7X-EK?

[Ulrich Prinz]

Hey, this is an emergency call :)

I had an old archive of an early 4.9.9.0 trunk where I did some other 
work at on my private at91sam7x-ek for improving DHCP. I flashed the 
httpd binary that was already available form my experiments and it works.

Then I merged the bugfixes arpcache.c and tcpsm.c from Harald and 
hwmaier. Rebuild, Recompile, Flash:

Nut/OS 4.9.9.0 HTTP Daemon...Registering device failed
Configure eth0...failed
192.168.10.130 ready

Looking at the LEDs the PHY isn't initialized correctly or something 
like that. It will be the same behavior I had at work, where the first 
packet sent to the board throws the EMAC or the PHY. But in my example 
the DHCP is active and therefore the things explode right at the beginning.

Ok, I reverted the two fixes and recompiled again...

Nut/OS 4.9.9.0 HTTP Daemon...Registering device failed
Configure eth0...failed
192.168.10.130 ready

So, now I cannot recreate the first original working version.
All compile runs where with arm-elf-gcc version 4.4.2
I switched to arm-none-eabi-gcc version 4.5.1 and tried again: Not 
working too.

Actually I have no idea what is going wrong there but I need to get the 
at91sam7x-ek working...

Any idea or help?

Am 02.11.2010 15:18, schrieb Ulrich Prinz:
> Hi!
>
> I have a problem to get the httpd example running on AT91SAM7X-EK
> board.
> It only works a bit if I set the PHY code in dev->Ethernet PHYceiver to
> the right code of the DM9161A what should be done by the
> at91sam7x-ek.conf.
> Then DHCP fails while the LEDs of the PHY flash only every 4 seconds.
>
> I disabled DHCP but it always takes the defaults for IP/Mask/Gateway
> that where written in ConfNet Flash area. But the PHY LEDs show
> something that might be correct. But with the defaults I cannot test.
>
> If I disable DHCP and force to take my presets of IP/Mask/gateway I
> have the 4 second flashing of the LEDs back and the board doesn't react
> to the IP I set.
>
> According the SVN Log the last changes where made by declaring
> at91sam7x_emac.c obsolete and using at91_emac.c for all at91 series
> chips in May of this year. Didn't anyone use the AT91SAM7X since then?
>
> Best Regards
> Ulrich
> _______________________________________________
> http://lists.egnite.de/mailman/listinfo/en-nut-discussion