On Sun Feb 19 12:27:09 2017, OLIMAUL wrote:
Show quoted text> Am So 19. Feb 2017, 05:19:10, paul@city-fan.org schrieb:
> > On Sat Feb 18 12:59:56 2017, SREZIC wrote:
> > > I see the following failure on my smokers:
> > >
> > > # Failed test 'OO crc32 with addfile 4009136024'
> > > # at t/crc.t line 61.
> > > # Looks like you failed 1 test of 31.
> > > t/crc.t ..
> > > Dubious, test returned 1 (wstat 256, 0x100)
> > > Failed 1/31 subtests
> >
> > This is because the "OO crc32 with addfile" test computes the CRC of
> > the README file, and the test wasn't updated after the "version 0.21"
> > to "version 0.22" change at the top of the file.
>
>
> Thanks. This has been fixed today.
I noticed that the crc64 tests failed on some architectures, with an all-zero CRC value:
$ make test TEST_VERBOSE=1 WITH_CRC64=1
"/usr/bin/perl" -MExtUtils::Command::MM -e 'cp_nonempty' -- CRC.bs blib/arch/auto/Digest/CRC/CRC.bs 644
PERL_DL_NONLAZY=1 "/usr/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(1, 'blib/lib', 'blib/arch')" t/*.t
t/crc.t ..
1..33
# use
ok 1
# crc64 0
not ok 2
# OO crc64 0
not ok 3
# crc32
ok 4
...
I've seen this with i686 and armv7hl, but the correct values were generated for builds on x86_64, ppc64le, aarch64 and ppc64. Is that expected (and why the CRC64 tests are not run by default)?
