Re: [lwip-devel] [bug #32417] TCP_OVERSIZE seems to have problems with (fast-)retransmission

[jre1]

I've been trying to figure this out as well.  Could the problem occur
on retransmission if the unsent queue is empty before the unacked 
segment is being put back on it?  The segment may have oversize_left != 0
from before it was first sent, causing the checks to fail later.

This doesn't help me though, I've tried correcting that, but maybe
it causes other problems or isn't the only failure.

It seems to me the unsent and unacked queues will always be in order, right? 

I'm still dealing with 1.4.0-rc1, but the relevant code seems the same as
1.4.0.
I don't see anything in the diffs that jumps out at me as fixing this.  I'll
try
converting to 1.4.0 soon.

   Cheers,
   Joe Eykholt
-- 
View this message in context: 
http://old.nabble.com/-bug--32417--TCP_OVERSIZE-seems-to-have-problems-with-%28fast-%29retransmission-tp30896314p32384799.html
Sent from the lwip-devel mailing list archive at Nabble.com.