[tor-bugs] #12464 [Tor]: When Tor 0.2.6.x is closer to done, profile relays running Tor 0.2.6.x and optimize accordingly

[Tor Bug Tracker & Wiki]

#12464: When Tor 0.2.6.x is closer to done, profile relays running Tor 0.2.6.x and
optimize accordingly
------------------------+-------------------------------------------------
     Reporter:  nickm   |      Owner:  dgoulet
         Type:  defect  |     Status:  assigned
     Priority:  normal  |  Milestone:  Tor: 0.2.6.x-final
    Component:  Tor     |    Version:
   Resolution:          |   Keywords:  tor-relay performance 026-triaged-1
Actual Points:          |  Parent ID:
       Points:          |
------------------------+-------------------------------------------------

Comment (by nickm):

 I tend to agree there, though if we ''do'' someday decide to optimize
 memcpy, I'd like to see call graph info to see which memcpy really
 matters; I bet we might be surprised.

 wrt crypto speed improvements, that's an 0.2.7 issue at earliest.

 The other functions that appear above in the profile don't matter a lot
 individually, and they don't look like bottlenecks per se, but I bet
 there's some opportunity to make them faster.  Not for 0.2.6, though.

 Two questions:

 1. Let's think about the elephant in ring 0: that 37% of our time that
 we're spending in the kernel. I don't think this is a regression, though,
 but we should check.  And some time, we should investigate where the time
 is going.

 2. Can anybody profile... an exit?  a busy hidden service? Some other
 relevant configuration?

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/12464#comment:10>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online