Bug 1296507 - perl-Crypt-GPG-1.64-4.fc24 FTBFS: tests fail
Summary: perl-Crypt-GPG-1.64-4.fc24 FTBFS: tests fail
Keywords:
Status: CLOSED RAWHIDE
Alias: None
Product: Fedora
Classification: Fedora
Component: perl-Crypt-GPG
Version: rawhide
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
Assignee: Robert Scheck
QA Contact: Fedora Extras Quality Assurance
URL: http://koji.fedoraproject.org/koji/ta...
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2016-01-07 12:01 UTC by Petr Pisar
Modified: 2016-01-09 05:54 UTC (History)
2 users (show)

Fixed In Version: perl-Crypt-GPG-1.64-5.fc24
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2016-01-09 05:54:19 UTC
Type: Bug
Embargoed:


Attachments (Terms of Use)

Description Petr Pisar 2016-01-07 12:01:19 UTC
perl-Crypt-GPG-1.64-4.fc24 fails to build in F24 because tests fail:

+ make test
PERL_DL_NONLAZY=1 "/usr/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
NOTE: If the tests are skipped, you may need to install gpg,
      and/or set the environment variable GPGBIN to the 
      location of the gpg binary.
      If the keygen test takes a long time you may need to 
      generate more randomness on your computer (by running a 
      recursive directory listing in the background, for 
      example).
t/01-keygen.t .. ok
Use of uninitialized value in string eq at t/02-import.t line 40.
# Failed test 1 in t/02-import.t at line 44
t/02-import.t .. 
Failed 1/1 subtests 
t/03-export.t .. ok
t/04-encdec.t .. ok
t/05-sigver.t .. ok
Use of uninitialized value in string eq at t/06-keyops.t line 38.
Can't return outside a subroutine at t/06-keyops.t line 38.
t/06-keyops.t .. 
Dubious, test returned 10 (wstat 2560, 0xa00)
Failed 30/30 subtests 
Test Summary Report
-------------------
t/02-import.t (Wstat: 0 Tests: 1 Failed: 1)
  Failed test:  1
t/06-keyops.t (Wstat: 2560 Tests: 0 Failed: 0)
  Non-zero exit status: 10
  Parse errors: Bad plan.  You planned 30 tests but ran 0.
Files=6, Tests=55, 8816 wallclock secs ( 0.05 usr  0.01 sys +  0.82 cusr  0.12 csys =  1.00 CPU)
Result: FAIL

Difference between working and failing build root is:

        gnupg 	1.4.19-3.fc23 	> 	1.4.20-1.fc24
	libarchive 	3.1.2-13.fc23 	> 	3.1.2-14.fc24
	sqlite 	3.9.0-2.fc24 	> 	3.9.2-1.fc24

Comment 1 Petr Pisar 2016-01-08 09:59:25 UTC
This is caused by upgrading gnupg from 1.4.19-3.fc23 to 1.4.20-1.fc24.

Comment 2 Petr Pisar 2016-01-08 11:55:36 UTC
gnupg 1.4.20 refuses MD5. One has to add --allow-weak-digest-algos gnupg option to enable it, or regenerate the test keys with stronger hash algorithm.

Comment 3 Robert Scheck 2016-01-08 21:58:06 UTC
Thank you for the analysis. I tried --allow-weak-digest-algos, but that
breaks with older branches, because older gnupg versions don't support the
option.

Comment 4 Robert Scheck 2016-01-09 05:54:19 UTC
I now replaced the test keys with the package signing keys of EPEL 7, which
works with GnuPG 1.4.19 and 1.4.20. Thanks for the report and the analysis!


Note You need to log in before you can comment on or make changes to this bug.