[8u232-b08] New test failures due to tzdata-updates

[Andrew John Hughes]

On 15/10/2019 13:27, Severin Gehwolf wrote:
> Hi,
> 
> The latest EA tags of the jdk8u/jdk8u tree (jdk8u232-b08) have new test
> failures in. jdk8u232-b07 was OK:
> 
> sun/util/calendar/zi/TestZoneInfo310.java
> 
> The failure looks like this:
> 
> Compiling tz files!
> testing!
> OLD.getZoneInfoOld()[1]=648
> OLD.getZoneInfoOld()[2]=194
> OLD.getZoneInfoOld()[3]=169
> OLD.getAvailableIDs()=746, total=628
> OLD.getAliasTable()=493, total=228
> OLD.TotalTZ()=80 (ms)
> NEW.getTimeZone()[1]=70
> NEW.getTimeZone()[2]=16
> NEW.getTimeZone()[3]=9
> NEW.getAvailableIDs()=18, total=628
> NEW.getAliasTable()=74, total=228
> NEW.TotalTZ()=1 (ms)
> ******************************
> America/Campo_Grande : America/Campo_Grande
>     offset=-14400000,dstSavings=0,useDaylight=false,transitions=93,offsets=4,checksum=-718118701,gmtChanged=false
> [NG]offset=-14400000,dstSavings=3600000,useDaylight=false,transitions=129,offsets=4,checksum=374743623,gmtChanged=false
>     -------------
>     (NG) Different trans size :93, 129
>                   1900-01-01T00:00  [utc=-2208988800   raw=fffffdfdae01dc00, offset=-13108/-03:38:28, saving=0]
>     (OK)          1900-01-01T00:00  [utc=-2208988800   raw=fffffdfdae01dc00, offset=-13108/-03:38:28, saving=0]
>             -----
> 
> Failures are also observed at AdoptOpenJDK:
> https://ci.adoptopenjdk.net/view/Test_upstream/job/Test_openjdk8_hs_sanity.openjdk_x86-64_windows_upstream/19/
> 
> These failures are due to these two changesets:
> http://hg.openjdk.java.net/jdk8u/jdk8u/jdk/rev/d355fd566378
> http://hg.openjdk.java.net/jdk8u/jdk8u/jdk/rev/8560bc534080
> 
> Both of them would need to update tzdata copies in the test subtree for
> the tests to pass again.
> 
> The addendum for JDK-8228469 (rev d355fd566378) would be this:
> http://cr.openjdk.java.net/~sgehwolf/webrevs/JDK-8228469-8u-fixup/01/webrev/
> 
> The addendum for JDK-8231098 (rev 8560bc534080) would be this:
> http://cr.openjdk.java.net/~sgehwolf/webrevs/JDK-8231098-8u-fixup/02/webrev/
> 
> The current code in jdk8u232-b08 compares tzdata2019a (test code) with
> tzdata2019c (jdk code). Thus the test failure(s).
> 
> Shall I create 8u-only bugs for them or is there another, preferred,
> solution to this?
> 
> Thanks,
> Severin
> 

Yes, this appears to have been missed because the test changes were not
part of the 11u version, having been removed by the switch to vanguard
format.

I think it would be preferable to just move straight to backporting the
vanguard support once the CPU is done and dusted. Fixing the tests
before that would just make backporting the vanguard fix (based on
2019a) more difficult with no clear benefit.

Thanks for catching this though,
-- 
Andrew :)

Senior Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: ed25519/0xCFDA0F9B35964222 (hkp://keys.gnupg.net)
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222
https://keybase.io/gnu_andrew