Re: [lwip-users] Asynchronous events.

[address@hidden]

Nick Thomas wrote:
> On the second attempt I now get:
>
> Assertion "tcp_input: pcb->next != pcb (before cache)" failed at line 181 in
> \blah\blah\core\tcp_in.c
>   
This is an indication pcb-list tcp_active_pcbs is corrupted (since there 
is a loop in the list: a next pointer of one pcb points to itself). This 
is most often caused by multiple threads working with the lwIP core code 
at the same time (other threads than tcpip_thread may only use the 
netconn or sockets api). You really should get to the bottom of this, it 
will cause other problems, too!
> [...]
>         /* ASSERT caused by the next line (I think) on second entry into this
> function. */
>         err = netconn_connect(test_conn, &ipaddr, 80);
>   
This will not be the actual problem here, but where do you call 
netcon_new()? I think the code does not allow reusing a netconn if one 
connection failed...

Simon