[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #14334 [Tor]: Bugs when registering guard status in connection_or_connect()
#14334: Bugs when registering guard status in connection_or_connect()
-------------------------------------------------+-------------------------
Reporter: asn | Owner: asn
Type: defect | Status:
Priority: Medium | needs_review
Component: Tor | Milestone: Tor:
Severity: Normal | 0.2.9.x-final
Keywords: tor-guard, 027-triaged-1-in, | Version: Tor:
PostFreeze027, TorCoreTeam201604 | 0.2.7
Parent ID: | Resolution:
Reviewer: | Actual Points:
| Points: small
| Sponsor:
| SponsorU-can
-------------------------------------------------+-------------------------
Comment (by asn):
Replying to [comment:12 nickm]:
> Could I have a comment in this case to say why we aren't calling
entry_guard_register_connect_status? Otherwise I worry it will get added
back in.
>
Hm, why do you expect `entry_guard_register_connect_status()` to be called
there? If there is a reason, we ''need to'' add it back in. Otherwise, I
don't know why anyone would add it back in.
My point with comment:10 was that there is no reason to call
`entry_guard_register_connect_status()` right after
`connection_connect()` since we are dealing with asynchronous networking
and by that time we have not even connected to the remote host. The right
place to call that function is at asynchronous connection callbacks like
`circuit_build_failed()` (where we do call it).
If that's not persuasive, then I might be wrong. What do you think?
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/14334#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