Re: [lwip-users] lwip 1.4.1 bug-fix release

[Zayaz Volk]

I would expect it so.
That's one of the purposes of UDP checksum, i guess.

On other hand, if UDP checksum is omitted and under certain conditions (long UDP datagram timeout, high communcation speed and as a result IP overlapping while still within a UDP datagram timeout) - one might expect to get a wrong, but "valid" datagram with IP fragment
belonging to another UDP datagram. - I think, it might happen with IPv4.

> From: address@hidden
> Date: Sat, 18 Feb 2012 14:02:04 +0000
> To: address@hidden
> Subject: Re: [lwip-users] lwip 1.4.1 bug-fix release
>
>
> On 17 Feb 2012, at 13:31, Bill Auerbach wrote:
>
> >> That maximum size of a UDP datagram should only be limited by the
> >> protocol and your resources, so 64K should work, yes.
> >
> > With UDP being unreliable, that implies that one or more fragments could be
> > dropped in a large UDP send, right?
>
> Yes, but if one fragment is lost then the whole datagram should be discarded by the receiving stack.
>
> Kieran
> _______________________________________________
> lwip-users mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/lwip-users