[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #23751 [Core Tor/Tor]: [warn] tor_bug_occurred_: Bug: src/common/buffers.c, etc.
#23751: [warn] tor_bug_occurred_: Bug: src/common/buffers.c, etc.
------------------------------------------------+--------------------------
Reporter: Felixix | Owner: (none)
Type: defect | Status: new
Priority: High | Milestone: Tor:
| 0.3.2.x-final
Component: Core Tor/Tor | Version:
Severity: Normal | Resolution:
Keywords: tor-channel, tor-sched, tor-buffer | Actual Points:
Parent ID: | Points:
Reviewer: | Sponsor:
------------------------------------------------+--------------------------
Comment (by dgoulet):
From the stacktrace it seems a channel outbuf was filled up which
triggered the assert when the limit was reached. It then released the
channel but the scheduler couldn't find it in the pending list so it tried
to free the channel and the two last BUG() are basically that. Bottom
line, that channel was never in the scheduler loop.
This is very weird. The fact that the outbuf was completely full is
worrying that is:
1. Was it filled up by then the scheduler?
2. Or we have a bug where in `KISTLite` mode we didn't allow somehow to
let the channel write to kernel?
For (1), currently the only possible way I could find for a channel outbuf
to be used without the scheduler knowing is during the connection
handshake but that shouldn't fill up of `INT_MAX` data...
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/23751#comment:2>
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