Closed me-and closed 4 years ago
Going back to the oldest commit that I can build on Cygwin (v1.8.3-rc0-1-g380395d09), and cherry-picking 612c49e94, there are fewer failures:
test | 380395d09 | 612c49e94 |
---|---|---|
t0301.12 | fail | fail |
t0301.24 | fail | fail |
t0301.25 | pass | fail |
t0301.26 | fail | fail |
t0301.27 | pass | fail |
t0301.28 | pass | fail |
The 25, 27 and 28 subtests were all broken by v2.1.0-28-gf5e3c0b9d.
This has been fixed as of v2.12.0-268-g1f180e5eb, which is currently in the master, next and pu branches.
Fixed here in 8db9305bc8ec2e7aac7db7a855441cd089f1c72b
As of v2.12.0-267-g612c49e94 in the main Git repository, t0301 is expected to pass but fails on Cygwin regardless of whether
configure
is run.This commit adds new tests rather than new function, which implies this doesn't indicate a regression.