Rob Landley d6a0b3031f Have "make tests" fail if any of the tests failed.
This is sort of a hack because I want to move the individual tests to
subshells, which means they don't have a shared shell context. Functions
and variables not persisting into later tests is a good thing (test environments
polluting each other), but it makes accumulating FAILCOUNT more awkward.
2020-06-16 21:14:35 -05:00
..
2020-06-08 11:49:44 -05:00
2020-05-01 10:14:12 -05:00
2019-06-29 12:31:22 -05:00
2020-03-13 15:59:26 -05:00