On Tue, 2009-08-11 at 14:37 +0200, Kesten, Joerg wrote: > Does anyone with further insight have some suggestion how this should > be handled and/or if this really is a bug or just some > misinterpretation by me?
Although I've just spotted another issue there that won't yet be handled by my fix: by shrinking the new segment we should also be removing the FIN flag from it (as it no longer contains the FIN sequence number). This gets unpleasant as tcp_process() may already have done things based on it having the FIN in it. I think the best course of action would be to say if an in-order received segment contains a FIN, we should bin the out-of-order sequence queue at that time, because there can be no data after the FIN and so logically anything that was out of order must have now been correctly received. That makes things a bit simpler. The new code that trims frames outside of the receive window will also need to handle FINs better. I'll re-work the patch for BUG27209 Kieran _______________________________________________ lwip-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/lwip-users
