[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #6572 [Tor]: âBug: Circuit somehow completed a hop while the network was not live.â
#6572: âBug: Circuit somehow completed a hop while the network was not live.â
---------------------------------------+------------------------------------
Reporter: bpmcontrol | Owner: mikeperry
Type: defect | Status: needs_information
Priority: normal | Milestone: Tor: 0.2.4.x-final
Component: Tor | Version:
Keywords: tor-client 024-deferrable | Parent:
Points: | Actualpoints:
---------------------------------------+------------------------------------
Comment(by mikeperry):
There are only 4 places where timestamp_began gets updated in the source:
1. circuit_deliver_create_cell() - When we first decide to send the
initial CREATE cell
2. pathbias_send_usable_probe() - At circuit close when we decide to probe
it
3. circuit_extend_to_new_exit() - Some cannibalization cases
4. circuit_launch_by_extend_info() - Other cannibalization cases
1 should not cause this issue, because in this case the circuit still has
not completed any hops.
2 should not cause this issue, because the circuit is already opened, and
this log message is in a code block that checks has_opened.
3 and 4 should not cause this issue, again because they should only be
caused for has_opened circuits. Additionally, they also increment the
desired_path_len, which this code block also checks is exactly 3...
I suppose we could beef up this log line to make extra sure none of the
above have happened somehow, but otherwise I am at a loss...
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/6572#comment:13>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@xxxxxxxxxxxxxxxxxxxx
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs