Windows build broken (again)

[Herbert Valerio Riedel]

On 2014-10-03 at 17:29:31 +0200, Simon Peyton Jones wrote:
> Perhaps, yes, it is Python 3. I don't know.  Could someone revert to
> make it work again, please?

Fyi, I can't reproduce this specific problem on Cygwin at least (I don't
have any working pure Msys2 environment yet (still working on it), as
this may exactly be the kind of failure I'd expect Msys2 to be prone to
while Cygwin to be unaffected by). 

What I tried in order to reproduce:

  $ git rev-parse HEAD
  084d241b316bfa12e41fc34cae993ca276bf0730  # <-- this is the Py3/testsuite commit 

  $ make TEST=tc012 WAY=normal
  ...
  =====> tc012(normal) 3039 of 4088 [0, 0, 0] 
  cd ./typecheck/should_compile && 'C:/cygwin64/home/ghc/ghc-hvr/inplace/bin/ghc-stage2.exe' -fforce-recomp -dcore-lint -dcmm-lint -dno-debug-output -no-user-package-db -rtsopts -fno-ghci-history -c tc012.hs   -fno-warn-incomplete-patterns >tc012.comp.stderr 2>&1
   
  OVERALL SUMMARY for test run started at Fri Oct  3 15:42:04 2014 GMT
   0:00:03 spent to go through
      4088 total tests, which gave rise to
     12360 test cases, of which
     12359 were skipped
   
         0 had missing libraries
         1 expected passes
         0 expected failures
  ...


And btw, with the latest GHC HEAD commit (and I suspect the recent
HEAP_ALLOCED-related commits to be responsible for that), I get a ton of
testsuite failures due to such errors:

  T8639_api.exe: Unknown PEi386 section name `staticclosures' (while processing: C:\cygwin64\home\ghc\ghc-hvr\libraries\ghc-prim\dist-install\build\HSghcpr_BE58KUgBe9ELCsPXiJ1Q2r.o)