Bug #41487 for Test-Refcount: Test::Refcount-0.02 all tests fail

Sat Dec 06 12:31:12 2008 david [...] davidfavor.com - Ticket created

Subject:	Test::Refcount-0.02 all tests fail
Date:	Sat, 06 Dec 2008 11:30:53 -0600
To:	bug-Test-Refcount [...] rt.cpan.org
From:	David Favor <david [...] davidfavor.com>

PERL_DL_NONLAZY=1 /runtime/perl-5.10.0-35017/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/00use.........ok t/01count....... # Failed test 'two refs to object fails to be 1' # at t/01count.t line 39. # STDERR is: # # Failed test 'one ref' # # at t/01count.t line 38. # # expected 1 references, found 2 # # Some::Class=HASH(0x86621c0) [refcount 2] is # # +- referenced by REF(0x866ff38) [refcount 1], which is # # | in the lexical '$newref' in CODE(0x856aa08) [refcount 1], which is # # | the main body of the program. # # +- referenced by REF(0x866fe98) [refcount 1], which is # # in the lexical '$object' in CODE(0x856aa08) [refcount 1], which was seen before. # # not: # /#\s+Failed\ test.*?\n?.*?at\ t\/01count\.t line 38.*\n?/ # # # expected 1 references, found 2 # # (?-xism:^# Some::Class=HASH$0x[0-9a-f]+$ is\n) # (?m-xis:(?:^#.*\n){1,}) # as expected Failed 1/8 subtests t/02one......... # Failed test 'two refs to object fails to be 1' # at t/02one.t line 29. # STDERR is: # # Failed test 'one ref' # # at t/02one.t line 28. # # expected 1 references, found 2 # # Some::Class=HASH(0x9b64150) [refcount 2] is # # +- referenced by REF(0x9b94ad0) [refcount 1], which is # # | in the lexical '$newref' in CODE(0x9aa6a08) [refcount 1], which is # # | the main body of the program. # # +- referenced by REF(0x9ac0578) [refcount 1], which is # # in the lexical '$object' in CODE(0x9aa6a08) [refcount 1], which was seen before. # # not: # /#\s+Failed\ test.*?\n?.*?at\ t\/02one\.t line 28.*\n?/ # # # expected 1 references, found 2 # # (?-xism:^# Some::Class=HASH$0x[0-9a-f]+$ is\n) # (?m-xis:(?:^#.*\n){1,}) # as expected Failed 1/3 subtests t/03weak........ # Failed test 'two refs to object fails to be 1' # at t/03weak.t line 21. # STDERR is: # # Failed test 'one ref' # # at t/03weak.t line 20. # # expected 1 references, found 2 # # Some::Class=HASH(0x8ed0888) [refcount 2] is # # +- referenced by REF(0x8eea598) [refcount 1], which is # # | in the lexical '$newref' in CODE(0x8ed0a08) [refcount 1], which is # # | the main body of the program. # # +- referenced by REF(0x8eea528) [refcount 1], which is # # in the lexical '$object' in CODE(0x8ed0a08) [refcount 1], which was seen before. # # not: # /#\s+Failed\ test.*?\n?.*?at\ t\/03weak\.t line 20.*\n?/ # # # expected 1 references, found 2 # # (?-xism:^# Some::Class=HASH$0x[0-9a-f]+$ is\n) # (?m-xis:(?:^#.*\n){1,}) # as expected Failed 1/2 subtests t/04reftypes....ok Test Summary Report ------------------- t/01count (Wstat: 0 Tests: 8 Failed: 1) Failed test: 5 t/02one (Wstat: 0 Tests: 3 Failed: 1) Failed test: 3 t/03weak (Wstat: 0 Tests: 2 Failed: 1) Failed test: 1 Files=5, Tests=20, 1 wallclock secs ( 0.05 usr 0.01 sys + 0.23 cusr 0.02 csys = 0.31 CPU) Result: FAIL Failed 3/5 test programs. 3/20 subtests failed. make: *** [test_dynamic] Error 255 -- Love feeling your best ever, all day, every day? Click http://RadicalHealth.com/join for the easy way.

Sun Dec 07 08:00:39 2008 leonerd-cpan [...] leonerd.org.uk - Correspondence added

Yes, I've been looking over these recently, and I can't offhand see what's up with it. It's a long text output from Test::Builder::Tester that seems to fail a regexp match built in the test script. Not an easy one to eyeball. These all used to work; I suspect it may be a change in behaviour of one of the test modules. Can you send me a list of version numbers of the following that you used to test this?: Test::More Test::Builder::Tester It works for me when I use: $ ./Build test t/00use.........ok t/01count.......ok t/02one.........ok t/03weak........ok t/04reftypes....ok All tests successful. Files=5, Tests=20, 1 wallclock secs ( 0.05 usr 0.02 sys + 0.39 cusr 0.06 csys = 0.52 CPU) Result: PASS $ perlmodversion Test::More 0.8 $ perlmodversion Test::Builder::Tester 1.13 Perhaps one of these is newer in a way that upset its? Thanks, -- Paul Evans

Sun Dec 07 08:01:02 2008 The RT System itself - Status changed from 'new' to 'open'

Sun Dec 07 08:05:17 2008 leonerd-cpan [...] leonerd.org.uk - Correspondence added

On Sun Dec 07 08:00:39 2008, PEVANS wrote: Show quoted text

> Perhaps one of these is newer in a way that upset its?

Actually, I've now updated libtest-simple-perl to latest in debian/unstable, and still it passes: $ ./Build test t/00use.........ok t/01count.......ok t/02one.........ok t/03weak........ok t/04reftypes....ok All tests successful. Files=5, Tests=20, 0 wallclock secs ( 0.04 usr 0.01 sys + 0.37 cusr 0.08 csys = 0.50 CPU) Result: PASS $ perlmodversion Test::More 0.86 $ perlmodversion Test::Builder::Tester 1.18 So perhaps it's not that... -- Paul Evans

Sun Dec 07 12:57:48 2008 david [...] davidfavor.com - Correspondence added

Subject:	Re: [rt.cpan.org #41487] Test::Refcount-0.02 all tests fail
Date:	Sun, 07 Dec 2008 11:57:35 -0600
To:	bug-Test-Refcount [...] rt.cpan.org
From:	David Favor <david [...] davidfavor.com>

Paul Evans via RT wrote: Show quoted text

> <URL: http://rt.cpan.org/Ticket/Display.html?id=41487 > > > Yes, I've been looking over these recently, and I can't offhand see > what's up with it. It's a long text output from Test::Builder::Tester > that seems to fail a regexp match built in the test script. Not an easy > one to eyeball. > > These all used to work; I suspect it may be a change in behaviour of one > of the test modules. Can you send me a list of version numbers of the > following that you used to test this?: > > Test::More > Test::Builder::Tester > > It works for me when I use: > > $ ./Build test > t/00use.........ok > t/01count.......ok > t/02one.........ok > t/03weak........ok > t/04reftypes....ok > All tests successful. > Files=5, Tests=20, 1 wallclock secs ( 0.05 usr 0.02 sys + 0.39 cusr > 0.06 csys = 0.52 CPU) > Result: PASS > > $ perlmodversion Test::More > 0.8 > > $ perlmodversion Test::Builder::Tester > 1.13 > > > Perhaps one of these is newer in a way that upset its? > > Thanks, >

Here you go... net1#perldetail System : Linux 2.6.27.5-41.fc9.i686 perl : /runtime/perl-5.10.0-35051/bin/perl (5.10.0-35051) Linux::Inotify2 : 1.2 DBI : 1.607 DBD::SQLite : 1.14001 SQLite::DBMS : 3.6.1 Danga::Socket : 1.61 w/epoll EV : 3.49 Coro : 5.12 Sys::Syscall : 0.22 IO::Compress::Base : 2.015 IO::Lambda : 0.45 version : 0.76 Class::MOP : 0.71 Template : 2.20 Test::Simple : 0.86 Test::More : 0.86 Test::Builder::Tester : 1.18 Pod::Simple : 3.07 Pod::Man : 2.21 Sys::Syslog : 0.27 LWP::UserAgent : 5.822 WWW::Mechanize : 1.52 -- Love feeling your best ever, all day, every day? Click http://RadicalHealth.com/join for the easy way.

Mon Dec 08 12:46:04 2008 leonerd-cpan [...] leonerd.org.uk - Correspondence added

Just spotted it: On Sat Dec 06 12:31:12 2008, david@davidfavor.com wrote: Show quoted text

> # # Some::Class=HASH(0x86621c0) [refcount 2] is

... Show quoted text

> # (?-xism:^# Some::Class=HASH$0x[0-9a-f]+$ is\n)

The new version of Devel::FindRef, 1.4, now puts this [refcount 2] in its output. I've now adjusted the test regexps to optionally match that too, so it should cope with either 1.31 or 1.4. Find Test-Refcount-0.03 on CPAN (when the mirrors update it; I uploaded it around 2 minutes ago :) ). -- Paul Evans

Sat Dec 20 09:55:30 2008 leonerd-cpan [...] leonerd.org.uk - Correspondence added

Any luck with 0.03? Now on CPAN: http://search.cpan.org/~pevans/Test-Refcount-0.03/ -- Paul Evans

Mon Jan 26 09:29:46 2009 leonerd-cpan [...] leonerd.org.uk - Correspondence added

0.04 is now on CPAN, and shows 38 pass, no fails. Closing ticket. -- Paul Evans

Mon Jan 26 09:29:47 2009 leonerd-cpan [...] leonerd.org.uk - Status changed from 'open' to 'resolved'