[tor-bugs] #23753 [Core Tor/Tor]: sched: Implement a SCHED_BUG() that prints more information

[Tor Bug Tracker & Wiki]

#23753: sched: Implement a SCHED_BUG() that prints more information
--------------------------+------------------------------------
 Reporter:  dgoulet       |          Owner:  dgoulet
     Type:  defect        |         Status:  needs_review
 Priority:  Medium        |      Milestone:  Tor: 0.3.2.x-final
Component:  Core Tor/Tor  |        Version:
 Severity:  Normal        |     Resolution:
 Keywords:  tor-sched     |  Actual Points:
Parent ID:                |         Points:
 Reviewer:                |        Sponsor:  SponsorV
--------------------------+------------------------------------
Changes (by dgoulet):

 * status:  needs_revision => needs_review


Comment:

 Replying to [comment:6 nickm]:
 > Also, are we going to regret the transition away from IF_BUG_ONCE? Will
 this flood people's logs now?

 (First, fixup commit added for comment:5)

 I don't think so. All of them are highly unlikely to happen but if they
 do, the channel shouldn't get stuck in the KIST scheduler thus not being
 called over and over again.

 Actually, that made me think about it and I think the `IF_BUG_ONCE()` is
 not that helpful that is if we hit it once, it is a problem but if we hit
 a bug non stop, we learn something more, which that something is stuck and
 that gives us a hint.

 So that being said, I've added a ratelimit to the log_warn in
 `scheduler_bug_occurred()`. Let me know what you think.

 Branch: `ticket23753_032_02`

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/23753#comment:11>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online