Comment # 16
on bug 10503
from Pascal Quantin
(In reply to Jasper Bongertz from comment #12)
> This looks fixed in 1.12.2, at least Wireshark handles the wrap correctly
> (using the trace from DropBox, wrap is seen in frames 571777 to 571778). Can
> anyone confirm?
For me the bug is not related to the sequence number wrap, but to the
reassembly code that gets confused. That's why I updated the bug title
accordingly. But I realize that the way I wrote the sentence could suggest that
the reassembly code is confused when the SN goes to 0 while it's when it sees a
segment number that was already reassembled long time ago (as the SN wrapped).
You are receiving this mail because:
- You are watching all bug changes.