[tor-bugs] #26398 [Core Tor/Tor]: feature gate testing C code from Rust for now

Tor Bug Tracker & Wiki blackhole at torproject.org
Fri Jul 13 19:48:40 UTC 2018


#26398: feature gate testing C code from Rust for now
----------------------------------------------+----------------------------
 Reporter:  isis                              |          Owner:  isis
     Type:  defect                            |         Status:
                                              |  needs_review
 Priority:  Very High                         |      Milestone:  Tor:
                                              |  0.3.4.x-final
Component:  Core Tor/Tor                      |        Version:  Tor:
                                              |  0.3.3.1-alpha
 Severity:  Normal                            |     Resolution:
 Keywords:  rust, tor-test, fast-fix, tor-ci  |  Actual Points:
Parent ID:                                    |         Points:  1
 Reviewer:                                    |        Sponsor:
                                              |  Sponsor8-can
----------------------------------------------+----------------------------
Changes (by catalyst):

 * status:  merge_ready => needs_review


Comment:

 Replying to [comment:6 isis]:
 > Looking at the current status of #25386 it seems there's still going to
 be linker issues, particularly from doctests, for the time being until
 everything has rustc 1.27 stable (when the `-C` option [https://github.com
 /rust-
 lang/rust/pull/49956/commits/1a0d7a8207907e2459db8df72a8f604cdcc8285f was
 added] to rustdoc). But also, feel free to disagree.
 >
 > It might be good to have a configurable way to disable this stuff until
 all the issues are resolved, that way teor and I can continue writing
 tests as we would normally for now.  In the future, it also gives us an
 easier way to enable running the tests all the time, when we remove C code
 (e.g. if I were to write a sha2 or keccak implementation and we used that
 instead of the C code).
 After thinking about the current situation some more, I think we want to
 adopt something  like this approach.  Right now there is a restricted set
 of Rust testing configurations that we run in Travis, instead of the more
 comprehensive set that we used to do.  Basically, on master, none of the
 older Travis subjobs with Rust enabled pass tests because of linker
 problems with `--enable-fragile-hardening`.

 I put this back in "needs_review" because I want to make sure we do
 something reasonable with the doctests in the crypto crate, and a few
 potentially conflicting changes to Rust tests have happened on master
 since the original pull request.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/26398#comment:7>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online


More information about the tor-bugs mailing list