[tor-bugs] #18943 [Core Tor/Tor]: sha3 testsuite fails on big endian hosts

[Tor Bug Tracker & Wiki]

#18943: sha3 testsuite fails on big endian hosts
-------------------------------------------------+-------------------------
 Reporter:  weasel                               |          Owner:  yawning
     Type:  defect                               |         Status:
 Priority:  High                                 |  needs_review
Component:  Core Tor/Tor                         |      Milestone:  Tor:
 Severity:  Major                                |  0.2.8.x-final
 Keywords:  tor-core, crypto, TorCoreTeam201605  |        Version:  Tor:
Parent ID:                                       |  0.2.8.2-alpha
 Reviewer:                                       |     Resolution:
                                                 |  Actual Points:
                                                 |         Points:  small-
                                                 |  remaining
                                                 |        Sponsor:
-------------------------------------------------+-------------------------

Comment (by yawning):

 Replying to [comment:12 nickm]:
 > Hi, Yawning!  This looks fine to me; there are optimizations we might
 make, but they can wait till we discover whether this is critical-path.

 Yeah, I thought about optimizing for the cases where the host is little
 endian, and where the host allows unaligned qword access but decided to
 get something that will basically always work, and this makes it obvious
 where byte swapping is needed.

 > Also given that this is causing failing tests in 0.2.8, I'm considering
 cherry-picking this patch to 0.2.8. Please let me know if you think that's
 a poor idea.  (It should be harmless.)

 Go for it.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/18943#comment:13>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online