[tor-bugs] #31652 [Core Tor/Tor]: hs-v3: Service circuit retry limit should not close a valid circuit
Tor Bug Tracker & Wiki
blackhole at torproject.org
Thu Sep 26 09:19:38 UTC 2019
#31652: hs-v3: Service circuit retry limit should not close a valid circuit
---------------------------------------------+-----------------------------
Reporter: dgoulet | Owner: neel
Type: defect | Status:
| needs_revision
Priority: Medium | Milestone: Tor:
| 0.4.2.x-final
Component: Core Tor/Tor | Version:
Severity: Normal | Resolution:
Keywords: tor-hs, tor-circuit, 042-should | Actual Points:
Parent ID: #30200 | Points: 0.1
Reviewer: asn | Sponsor:
| Sponsor27-must
---------------------------------------------+-----------------------------
Changes (by asn):
* status: needs_review => needs_revision
Comment:
Neel, I'm still confused here...
Why does a function called `should_not_retry_intro_point()` allow the
possibility of a retry when we have gone over the number of maximum
retries? In my view, it should always return `false` in that case.
In particular the following block:
{{{
/* If we have gone over the number of retried circuits, make sure we
don't
* already have an established circuit. */
if (ip->circuit_retries > MAX_INTRO_POINT_CIRCUIT_RETRIES) {
return !ip->circuit_established || hs_circ_service_get_intro_circ(ip);
}
}}}
is confusing me even tho I've read it a few times. When I'm reading that
function my logic would be "If we have gone over the number of retried
circuits, we only allow retries if ...", but then why would we allow
retries if we are past the max? Also the comment mentions the established
circuit clause, but not the `hs_circ_service_get_intro_circ(ip);` one. Can
you please clarify that logic further? :/
We seem to have left some comments in `cleanup_intro_points()` that are
only relevant to `should_not_retry_intro_point()`.
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/31652#comment:16>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
More information about the tor-bugs
mailing list