[lwip-users] Re: [lwip] TCP ack segment versus OOSEQ handling

[Andy Valencia]

[Kieran Mansley <address@hidden> writes:]

>Sounds like you've found a bug.  Out of order segments are pretty rare
>things, which is why this won't have been spotted before.

I had some test code, and the test code itself had a bug, which is why I was
causing a lot of this.

>Replacing the
>
>pbuf_chain(pcb->recv_data, cseg->p);
>
>with
>
>if(pcb->recv_data)
>  pbuf_chain(pcb->recv_data, cseg->p);
>else
>  pcb->recv_data = cseg->p;
>
>would seem to be a suitable fix.

I actually just avoided doing the OOSEQ scan if there wasn't data received,
since you can't be changing the situation with respect to needing to trim
things and/or move them to the received queue, if you don't have any new
data in the first place?

>Hope that helps,

Thanks for the quick answer.

Regards,
Andy Valencia
[This message was sent through the lwip discussion list.]