Ghc 8.0 accelerate rewrite

.haskell-ghc-mod.json

@@ -0,0 +1 @@
+{ "suppressErrors": true }

README.md

@@ -54,10 +54,13 @@ To install on Linux or Mac, run the following commands:
     $ cd llvm-3.5.2
     $ mkdir build
     $ cd build
-    $ cmake ..
+    $ ../configure --enable-shared --enable-optimized
     $ make -j5
     $ sudo make install
     ```
+(Configuring with `cmake` won't export the correct shared libraries that llvm-general
+expects.  If you want more debugging capability of the llvm code, removed the `--enable-optimized` flag)
+
 
 1. Any version of BLAS and LAPACK.
 

examples/example0005-accelerate_backend.lhs

@@ -0,0 +1,51 @@
+
+> {-# LANGUAGE NoImplicitPrelude #-}
+> {-# LANGUAGE RebindableSyntax #-}
+> {-# LANGUAGE OverloadedLists #-}
+> {-# LANGUAGE TypeOperators #-}
+> {-# LANGUAGE FlexibleContexts #-}
+> {-# LANGUAGE GADTs #-}
+> {-# LANGUAGE DataKinds #-}
+> import qualified Prelude as P
+> import SubHask
+> import SubHask.Algebra.Accelerate.Vector (ACCVector, ValidBackend(..), mkAccVectorFromList)
+> --import SubHask.Algebra.Matrix (Matrix, unsafeToModuleM)
+> --import SubHask.Algebra.Accelerate.Matrix (ACCMatrix, mmult
+> import SubHask.Algebra.Array
+> import SubHask.Algebra.Vector
+> import qualified Data.Array.Accelerate as A
+> import SubHask.Algebra.Accelerate.AccelerateBackend (Backend(..))
+> import System.IO
+>
+> v  :: ACCVector Interpreter 4 Float
+> v  = mkAccVectorFromList [0.0, 1.0, 2.0, 3.0]
+>
+> v' :: ACCVector Interpreter 4 Float
+> v' = mkAccVectorFromList [0..3]
+> sngtln = (A.constant 2.0)
+> --mmat  :: Matrix (UVector "v" Double) Double "a" "b"
+> --mmat  = unsafeToModuleM 2 [0..5]
+>
+> --m :: ACCMatrix Interpreter (ACCVector Interpreter "v" Double ) "a" "b" Double
+> --m = mkAccMatrixFromMatrix mmat
+>
+> --mm :: ACCMatrix Interpreter (ACCVector Interpreter "v" Double ) "b" "a" Double
+> --mm = mkAccMatrixFromList 5 [0,1,2,3,4,5,6,7,8,9]
+>
+> main :: IO ()
+> main = do
+>   putStrLn $ "v = " + show v
+>   putStrLn $ "v' = " + show v'
+>   putStrLn $ "v + v = " + (show $ runAccVector (v + v))
+>   putStrLn $ "v + v - v = " + (show $ runAccVector(v + v - v'))
+>   putStrLn $ "v * v = " + (show $ runAccVector (v .*. v ))
+>   putStrLn $ "v / v' = " + (show $ runAccVector (v ./. v'))
+>   putStrLn $ "v * v / v' = " + (show $ runAccVector (v .*. v ./. v'))
+>   putStrLn $ "v' .*  2 = " + (show $ runAccVector (v' .* sngtln))
+>   putStrLn $ "v' ./  2 = " + (show $ runAccVector (v' ./ sngtln))
+>   putStrLn $ "v >< v' = " + (show $ runAccVector (v >< v))
+>   putStrLn $ "v**2 = " + (show $ runAccVector (v**v'))
+>   --putStrLn $ "m *  2 = " ++ show (runAccMatrix (m .* 2))
+>   --putStrLn $ "m + 2 = " ++ show (runAccMatrix ((m + 2) - 1 ))
+>   --putStrLn $ "m /  2 = " ++ show (runAccMatrix (m / 2))
+>   --putStrLn $ "m mmult  mm = " ++ show (runAccMatrix (mmult m mm))

src/SubHask/Algebra/Accelerate/AccelerateBackend.hs

@@ -0,0 +1,12 @@
+module SubHask.Algebra.Accelerate.AccelerateBackend
+(
+  Backend(..),
+)
+where
+
+
+data Backend
+    = Interpreter
+    -- | CUDA
+    -- | LLVM
+    -- LLVM has an SoC project slated, so check back in 60 days for non-parial functionality