pkg/44416: perl5 fails it's self-test on amd64 and i386 (at least)

[ignatios%random87.localdomain@localhost]

>Number:         44416
>Category:       pkg
>Synopsis:       perl5 fails it's self-test on amd64 and i386 (at least)
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    pkg-manager
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Wed Jan 19 11:50:00 +0000 2011
>Originator:     ignatios%random87.localdomain@localhost
>Release:        NetBSD 5.0.2
>Organization:
computer science department, university of Bonn, Germany
>Environment:
        
        
System: NetBSD random87 5.0.2 NetBSD 5.0.2 (GENERIC) #0: Mon May 10 17:27:16 
CEST 2010 
ignatios@random85:/var/itch/sources/5.0/oamd64/sys/arch/amd64/compile/GENERIC 
amd64
Architecture: x86_64
Machine: amd64
12 CPUs.
>Description:
        "make test" shows 7 failed tests.
        Running the slightly more detailed test outlined in the
        error message results in this summary:

Test Summary Report
-------------------
../cpan/Math-BigInt/t/bare_mbf.t                                (Wstat: 0 
Tests: 2308 Failed: 1)
  Failed test:  202
../cpan/Math-BigInt/t/bare_mbi.t                                (Wstat: 0 
Tests: 3273 Failed: 1)
  Failed test:  2733
../cpan/Math-BigInt/t/bigfltpm.t                                (Wstat: 0 
Tests: 2313 Failed: 1)
  Failed test:  207
../cpan/Math-BigInt/t/bigintpm.t                                (Wstat: 0 
Tests: 3279 Failed: 1)
  Failed test:  2739
../cpan/Math-BigInt/t/sub_mbf.t                                 (Wstat: 0 
Tests: 2314 Failed: 1)
  Failed test:  202
../cpan/Math-BigInt/t/sub_mbi.t                                 (Wstat: 0 
Tests: 3278 Failed: 1)
  Failed test:  2733
../cpan/Math-BigInt/t/with_sub.t                                (Wstat: 0 
Tests: 2309 Failed: 1)
  Failed test:  203
Files=1804, Tests=357760, 891 wallclock secs (46.10 usr  8.50 sys + 398.59 cusr 
55.09 csys = 508.28 CPU)
Result: FAIL

I've also run make test on a 32bit one-CPU machine, and saw a similar 
count of failed tests. from Math-BigInt

>How-To-Repeat:
        make test
>Fix:

        

>Unformatted: