| Subject: | t/01-encoding.t fails with -Duselongdouble |
If perl was compiled with -Duselongdouble for increased floating point accuracy, then the test suite fails. Seen on linux and freebsd:
...
# Failed test 'decode(7FG49Q00+): 20.375,2.775 == 20.375,2.77499999999999999'
# at t/01-encoding.t line 47.
# Failed test 'decode(7FG49QCJ+2V): 20.3700625,2.7821875 == 20.3700625,2.78218749999999999'
# at t/01-encoding.t line 47.
# Failed test 'decode(8FVC2222+22): 47.0000625,8.0000625 == 47.0000625,8.00006249999999999'
# at t/01-encoding.t line 47.
# Looks like you failed 3 tests of 31.
t/01-encoding.t ...
Dubious, test returned 3 (wstat 768, 0x300)
Failed 3/31 subtests
...