[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [tor-bugs] #11792 [Tor]: Consider directory connections zlib buffers when handling OOM
#11792: Consider directory connections zlib buffers when handling OOM
-------------------------+-------------------------------------------------
Reporter: nickm | Owner:
Type: defect | Status: needs_review
Priority: normal | Milestone: Tor: 0.2.6.x-final
Component: Tor | Version:
Resolution: | Keywords: tor-relay, dir, zlib, oom,
Actual Points: | 026-triaged-1
Points: | Parent ID:
-------------------------+-------------------------------------------------
Comment (by nickm):
Replying to [comment:4 andrea]:
> Begin code review:
>
> * ec59167cae1f5b3057ed722857d78ec78239e991:
> - This looks fine to me, I think. I presume you've checked that
we're not at risk of multiple-counting by also freeing the linked conn
later and adjusting for its memory usage again?
The linked conn can't be another stream ... and we don't consider linked
directory connections later on, only the linked ones. So I think it can't
be double-counted.
> * c5260f5b78c434a223dc82fd3153cd9c8a4f63bf:
> - I think this is okay.
>
> * 55556568283d44219e8439d0dea7d01bb182623d:
> - I think this looks reasonable, but conn_get_buffer_age() defined
the age of a circuit with nothing buffered in either direction as zero
isn't totally intuitive to me. It leads to reasonable behavior, but maybe
a sentence explicitly stating that it does that in its comment is in
order?
Will do.
> * 58020ed079af3794a0869f4a6bfa34ccaecc4b61:
> - This looks fine to me.
>
> End code review.
Thanks!
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/11792#comment:5>
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