util/time test failure on Jenkins
Our 'util/time' test fails on our tor-ci-mingwcross-master-test builder with:
util/time: Mar 27 13:29:43.773 [warn] tor_gmtime_r(): Bug: gmtime(-1) failed with error Invalid argument: Rounding up to 1970 (on Tor 0.4.1.0-alpha-dev )
[time FAILED]
- Show closed items
Activity
-
Newest first Oldest first
-
Show all activity Show comments only Show history only
- Alexander Hansen Færøy changed milestone to %Tor: 0.3.5.x-final
changed milestone to %Tor: 0.3.5.x-final
- Alexander Hansen Færøy added 035-backport 040-backport 040-must actualpoints::0.5 component::core tor/tor milestone::Tor: 0.3.5.x-final owner::nickm points::1 priority::medium resolution::fixed reviewer::teor severity::normal sponsor::31-can status::closed tor-ci-fail type::defect version::tor 0.3.5.8 labels
I think this failure might be caused by the "fail tests on bugs and errors" fix. So we need to fix these bug warnings somehow.
We must have fixed them on master on other platforms, and in Appveyor. So I wonder why the fix isn't working in jenkins?
I'll try to look into this tomorrow (~18 hours).
Trac:
Owner: N/A to teor
Keywords: N/A deleted, 040-must added
Status: new to assigned
Points: N/A to 1It's possible that this is a portability issue -- I have noticed that different Windows implementations tend to give different libc timeval manipulation functions.
Hi Nick, I tried to fix this issue, but it's hard without a backtrace on Windows. I couldn't really work out where to start.
Do you think we should:
- downgrade the bugs to warnings in src/lib/encoding/time_fmt.c
- fix the util/time unit test?
- handle the bugs in the util/time unit test? Or something else?
Trac:
Actualpoints: N/A to 0.2
Version: Tor: unspecified to Tor: 0.3.5.8Likely fix in my branch
bug29922_035
with PR at https://github.com/torproject/tor/pull/915 .I won't be absolutely sure till we merge it, but it seems okay to me so far. Setting teor as reviewer in hopes we can merge this week.
Trac:
Status: assigned to needs_review
Actualpoints: 0.2 to 0.3
Reviewer: N/A to teorIf we tried really hard, we might be able to squeeze this into Sponsor 31's modularity support, because it came from our unit test error-handling impovements. Better tests are important for modularity.
Looks fine to me.
Trac:
Status: needs_review to merge_ready
Keywords: N/A deleted, asn-merge added
Actualpoints: 0.3 to 0.4
Sponsor: N/A to Sponsor31-canmerged to 040 and forward. leaving it open for backports.
Trac:
Keywords: asn-merge deleted, N/A addedTrac:
Milestone: Tor: 0.4.0.x-final to Tor: 0.3.5.x-finalAppears not to have worked on jenkins: it was looking for the wrong error. I added another fix here and merged it to 0.4.0 and later.
Merged to 0.3.5 and merged forward.
Merged #29922 (moved) with #30041 (moved).
Trac:
Actualpoints: 0.4 to 0.5
Resolution: N/A to fixed
Status: merge_ready to closed- Trac closed
closed
- Trac changed time estimate to 8h
changed time estimate to 8h
- Trac added 4h of time spent
added 4h of time spent
- teor mentioned in issue #30041 (moved)
mentioned in issue #30041 (moved)
- Trac moved to tpo/core/tor#29922 (closed)
moved to tpo/core/tor#29922 (closed)