Very helpful thanks.  TcCoercibleFail is known to time out (for decent reasons) with –DDEBUG (see comments in the dsource file)

I’l look into the PatSyn thing

Simon

From: Thomas Miedema [mailto:thomasmiedema@gmail.com]
Sent: 18 February 2016 18:41
To: Simon Peyton Jones <simonpj@microsoft.com>; ghc-devs@haskell.org
Subject: Build failures with -DDEBUG

Simon,

the commits you pushed today don't validate with -DDEBUG. 

Unexpected failures:

   patsyn/should_compile  MoreEx [exit code non-0] (normal)
   patsyn/should_compile  T11224b [exit code non-0] (normal)
   polykinds              MonoidsTF [exit code non-0] (normal)
   polykinds              T11480b [exit code non-0] (normal)
   polykinds              T11523 [exit code non-0] (normal)
   typecheck/should_fail  TcCoercibleFail [stderr mismatch] (normal)
 
TcCoercibleFail timed out, and the others all hit the following debug assert:

=====> T11224b(normal) 2510 of 5029 [0, 0, 0] 

Compile failed (status 256) errors were:

ghc-stage2: panic! (the 'impossible' happened)

  (GHC version 8.1.20160218 for x86_64-unknown-linux):

          ASSERT failed!

  CallStack (from HasCallStack):

  assertPprPanic, called at compiler/types/TyCoRep.hs:1932:56 in ghc:TyCoRep

  checkValidSubst, called at compiler/types/TyCoRep.hs:1991:17 in ghc:TyCoRep

  substTys, called at compiler/types/TyCoRep.hs:2012:14 in ghc:TyCoRep

  substTheta, called at compiler/typecheck/TcPatSyn.hs:255:20 in ghc:TcPatSyn

  in_scope InScope {d_ap0 c_apv}

  tenv [ap1 :-> c_apv[tau:5]]

  tenvFVs [aps :-> t_aps[tau:1], apv :-> c_apv[tau:5]]

  cenv []

  cenvFVs []

  tys []

  cos []

Complete log from Travis: https://s3.amazonaws.com/archive.travis-ci.org/jobs/110115377/log.txt

-- Thomas