testsuite/tests/typecheck/should_compile/T12466a.hs shouldn't compile
Test T12466a was introduced as a test case in response to #12466 (comment 123011) which used code excerpted from
constraints. However, in the discussion following on from #11066 (comment 134986) that code was determined to be buggy or at least dodgy, and now
constraints uses a different definition of
Gcd as seen at https://hackage.haskell.org/package/constraints-0.10.1/docs/src/Data.Constraint.Nat.html#Gcd ...
It might not be necessary to reject the current T12466a, but it probably shouldn't be necessary to accept it either.
Steps to reproduce
Create a patch to reenable warnings or errors for unreachable code at the top level. Watch this test fail by producing a totally legitimate warning or error.
Should probably be testing something more similar to the current
- GHC version used: f9a513e0 on master