Travis can't build on Windows, I'm afraid. I'm sort of mitigating another crisis right now, but here's a shot in the dark - based on my Very Amazing Deductive Skills[1], completed in mere seconds, I'm guessing that perhaps e68195a96529cf1cc2d9cc6a9bc05183fce5ecea is relevant, one of Simon's latest commits? e.g., just try, as a shot in the dark $ git revert e68195a96529cf1cc2d9cc6a9bc05183fce5ecea And build again. [1] Note: my Very Amazing Deductive Skills in this case, are as follows. Look at git log for the last few days. See that some C code was changed. Immediately become suspicious that is the culprit, because it's C code. So, yes, a complete shot in the dark, but it's one that's served me well, too. :) On Wed, Apr 27, 2016 at 3:36 PM, Simon Peyton Jones wrote:

Does anyone feel able to look at

https://ghc.haskell.org/trac/ghc/ticket/11988

The symptom is that every single GHC-compiled binary now seg-faults on Windows (with HEAD).

That’s pretty serious.

Can Travis pinpoint the offending commit?

Simon

-- Regards, Austin Seipp, Haskell Consultant Well-Typed LLP, http://www.well-typed.com/