AW: [lwip-users] About the use of ethernetif_input

[Steffen Damelang]

Hi ...

I didn't know if anybody has answered this question.

So I will do ... My application runs also in a single-threaded environment. 
There is a good document by Adam Dunkels which describes how to do a single 
threaded lwip application.

http://www.sics.se/~adam/lwip/os.html

You have to poll on all network devices in your main loop. If there was a 
packet received you will call the ip_input to process the packet. That is
the way I use the ethernetif_input. I don't have a function get_packet. I call 
ethernetif_input if a packet was received.

Best regards
Steffen Damelang

-----Ursprüngliche Nachricht-----
Von: address@hidden [mailto:address@hidden Im Auftrag von Jeffery Du
Gesendet: Freitag, 17. März 2006 13:31
An: lwip-users
Betreff: [lwip-users] About the use of ethernetif_input

hi all,

         As reminded by Kieran, I decided to change my receiving mechanism to 
make a single thread application. 
         Then I reviewed the original code of lwip1.1.0 again. Honestly, I 
didn't quite undestand the use of "ethernetif_input" in ethernetif.c when I 
researched this code for the first time. It was called nowhere. So I revised 
the rx mechanism by my method:
    deivce driver rx thread -> ethernetif_input -> ip_input/arp_input ...
    Then can anyone tell me how to use ethernetif_input properly?

Best regards
Jeffery Du
2006-03-17



_______________________________________________
lwip-users mailing list
address@hidden
http://lists.nongnu.org/mailman/listinfo/lwip-users