Summary: | sysutils/e2fsprogs: 1.44.0, test failure on 12-CURRENT | ||
---|---|---|---|
Product: | Ports & Packages | Reporter: | Charlie Li <vishwin> |
Component: | Individual Port(s) | Assignee: | Matthias Andree <mandree> |
Status: | Closed Not A Bug | ||
Severity: | Affects Some People | CC: | cem, w.schwarzenfeld |
Priority: | --- | Flags: | mandree:
maintainer-feedback+
|
Version: | Latest | ||
Hardware: | amd64 | ||
OS: | Any |
Description
Charlie Li
2018-03-18 15:04:48 UTC
Can't help here, am not running 12-CURRENT. Since things work on 11-STABLE and 10-STABLE releases for me, I am not going to forward this upstream. Might be libc bugs in 12-CURRENT. Please do not CC huge mailing lists on bugs. The clocks are consistently 25 seconds earlier on CURRENT, relative to the test. I'm guessing this test uses a local time and CURRENT got a timezone update in whatever that local time is to shift by 25 seconds. Just a guess, anyway. In that case I do have WITH_ZONEINFO_LEAPSECONDS_SUPPORT set in src.conf. This setting also breaks databases/postgresql-server but it's a known quantity there. Also could someone remove current@ again? I accidentally used a cached page to comment and didn't pay attention to the CC list. (In reply to Charlie Li from comment #5) I think that explains it. There were 25 leapseconds between 1970 and 2014: https://en.wikipedia.org/wiki/Leap_second#Insertion_of_leap_seconds . (Since 2014, there have been two more.) |