Inconsistency in CoreSubst invariant

Richard Eisenberg rae at
Fri May 25 01:47:42 UTC 2018

> On May 24, 2018, at 8:21 AM, Joachim Breitner <mail at> wrote:
> Which one is it?

See Note [The substitution invariant] in TyCoRep. That applies to types, not terms, but I'd be shocked if terms had a different situation. That would suggest that the answer is (A) (and that the WARNing is correct).


More information about the ghc-devs mailing list