Re: [lwip-users] Testing the internet app based on lwIP-1.3.0 and Mini-OS

[Kieran Mansley]

On Mon, 2010-10-18 at 20:29 +0800, Bei Guan wrote:
> 
> 
> 2010/10/18 Kieran Mansley <address@hidden>
>         On Mon, 2010-10-18 at 19:00 +0800, Bei Guan wrote:
>         > But if I set 192.168.122.20 as Mini-OS IP address, data
>         caught by
>         > tcpdump is
>         > like this:
>         >
>         > 16:30:33.161162 arp who-has 192.168.122.20 tell
>         192.168.122.20
>         > 16:30:35.911512 arp who-has 192.168.122.20 tell
>         192.168.122.20
>         > 16:30:38.911843 arp who-has 192.168.122.20 tell
>         192.168.122.20
>         > 16:30:41.912160 arp who-has 192.168.122.20 tell
>         192.168.122.20
>         > 16:30:44.912541 arp who-has 192.168.122.20 tell
>         192.168.122.20
>         >
>         > And I cannot find the variable  LWIP_NETIF_LOOPBACK in opt.h
>         in lwIP
>         > 1.3.0.
>         
>         
>         That doesn't make 0.0.0.0 a valid IP address.  It means you
>         have an old
>         version of lwIP that probably doesn't support the feature you
>         want.  I
>         think most of your problems are to do with how lwIP integrates
>         with
>         miniOS.  I know this has been done with lwIP - XenSource (used
>         to?) use
>         it for that purpose - so perhaps you will have most success if
>         you can
>         find someone who is expert on miniOS to recommend what to do?
>         
>         
>         
> 
> 
> Yes, all the problems are to do with how lwIP integrates with mini-os.
> But some question with lwIP may not be given proper answer from the
> miniOS experts. That's why I came to lwIP mail list :)

I think your question does not require detailed knowledge about lwIP and
that someone who had used lwIP with miniOS would be in a much better
position to answer and help you.

> In my experiments, the client use the API  netconn_connect to connect
> to the server. But I always get the error return value -4, which is
>  definition as  "Connection aborted.",    /* ERR_ABRT       -4  */.
>  What's the most reason to cause this error? 

Are you sure about that definition?  I don't have a copy of 1.3.0 to
hand but in 1.3.1 an error of -4 was 

#define ERR_RTE        -4    /* Routing problem.         */

This makes a lot more sense for a connect error than ERR_ABRT which I
would normally only expect for listening and receiving errors.

Kieran