Bug #39342 for Cache-FastMmap: Cache::FastMmap-1.28 fails leak test same as 1.27

Tue Sep 16 18:58:34 2008 david [...] davidfavor.com - Ticket created

Subject:	Cache::FastMmap-1.28 fails leak test same as 1.27
Date:	Tue, 16 Sep 2008 17:58:11 -0500
To:	bug-Cache-FastMmap [...] rt.cpan.org
From:	David Favor <david [...] davidfavor.com>

net1#PERL_DL_NONLAZY=1 /common/pkgs/perl-5.10.0-34372/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(1, 'blib/lib', 'blib/arch')" t/6.t t/6.... 1..10 ok 1 - use Cache::FastMmap; not ok 2 - leak test 128 > 30k # Failed test 'leak test 128 > 30k' # at t/6.t line 93. ok 3 - leak test 0 > 30k ok 4 - leak test 0 > 30k ok 5 - leak test 0 > 30k ok 6 ok 7 - leak test 0 > 30k ok 8 - leak test 0 > 30k ok 9 - leak test 0 > 30k ok 10 - leak test 0 > 30k # Looks like you failed 1 test of 10. Dubious, test returned 1 (wstat 256, 0x100) Failed 1/10 subtests Test Summary Report ------------------- t/6 (Wstat: 256 Tests: 10 Failed: 1) Failed test: 2 Non-zero exit status: 1 Files=1, Tests=10, 4 wallclock secs ( 0.04 usr 0.01 sys + 3.58 cusr 1.18 csys = 4.81 CPU) Result: FAIL Failed 1/1 test programs. 1/10 subtests failed. -- Love feeling your best ever, all day, every day? Click http://RadicalHealth.com/join for the easy way.

Tue Sep 28 23:43:43 2010 cpan [...] robm.fastmail.fm - Correspondence added

I never was able to reproduce this, and I haven't seen any other reports. If you can still reliably recreate it, please re-open this bug.

Tue Sep 28 23:43:44 2010 The RT System itself - Status changed from 'new' to 'open'

Tue Sep 28 23:43:44 2010 cpan [...] robm.fastmail.fm - Status changed from 'open' to 'resolved'

Thu May 19 10:28:45 2016 jplesnik [...] redhat.com - Status changed from 'resolved' to 'open'

Thu May 19 11:04:43 2016 jplesnik [...] redhat.com - Correspondence added

From:

jplesnik [...] redhat.com

On Tue Sep 28 23:43:43 2010, ROBM wrote: Show quoted text

> I never was able to reproduce this, and I haven't seen any other > reports. If you can still reliably recreate it, please re-open this bug.

I found the similar issue against Perl 5.24 on x86_64 arch It fails randomly. Cache::FastMmap 1.43 GTop 0.18 Fedora Rawhide (f25) # PERL_DL_NONLAZY=1 "/usr/bin/perl" -Iblib/{lib,arch} t/6.t 1..10 ok 1 - use Cache::FastMmap; ok 2 - leak test 0 > 30k - new - 1 ok 3 - leak test 0 > 30k - new - 2 ok 4 - leak test 0 > 30k - new2 - 1 ok 5 - leak test 0 > 30k - new2 - 2 ok 6 ok 7 - leak test 0 > 30k - set ok 8 - leak test 0 > 30k - get ok 9 - leak test 0 > 30k - set2 not ok 10 - leak test 136 > 30k - list # Failed test 'leak test 136 > 30k - list' # at t/6.t line 96. # Looks like you failed 1 test of 10. CPAN Testers could not find the issue, because GTop is optional dependency.

Fri May 20 03:12:08 2016 SREZIC [...] cpan.org - Correspondence added

RT-Send-CC:

jplesnik [...] redhat.com

On 2016-05-19 11:04:43, jplesnik wrote: Show quoted text

> On Tue Sep 28 23:43:43 2010, ROBM wrote:

> > I never was able to reproduce this, and I haven't seen any other > > reports. If you can still reliably recreate it, please re-open this bug.

> > I found the similar issue against Perl 5.24 on x86_64 arch > It fails randomly. > > Cache::FastMmap 1.43 > GTop 0.18 > Fedora Rawhide (f25) > > # PERL_DL_NONLAZY=1 "/usr/bin/perl" -Iblib/{lib,arch} t/6.t > 1..10 > ok 1 - use Cache::FastMmap; > ok 2 - leak test 0 > 30k - new - 1 > ok 3 - leak test 0 > 30k - new - 2 > ok 4 - leak test 0 > 30k - new2 - 1 > ok 5 - leak test 0 > 30k - new2 - 2 > ok 6 > ok 7 - leak test 0 > 30k - set > ok 8 - leak test 0 > 30k - get > ok 9 - leak test 0 > 30k - set2 > not ok 10 - leak test 136 > 30k - list > # Failed test 'leak test 136 > 30k - list' > # at t/6.t line 96. > # Looks like you failed 1 test of 10. > > > CPAN Testers could not find the issue, because GTop is optional dependency.

I see this problem also on a Debian/jessie system with perl 5.24.0: # Failed test 'leak test 156 > 30k - list' # at t/6.t line 96. # Looks like you failed 1 test of 10. t/6.t ... Dubious, test returned 1 (wstat 256, 0x100) Failed 1/10 subtests FAIL report is on the way to CPAN Testers...

Tue Jul 11 22:23:36 2017 cpan [...] robm.fastmail.fm - Correspondence added

I've pushed a new version to github that adds more slop to the leak tests since they're heuristic anyway.

Tue Jul 11 22:23:37 2017 cpan [...] robm.fastmail.fm - Status changed from 'open' to 'resolved'