path: root/src/Data/Array/Nested/Internal/Arith.hs

{-# LANGUAGE DataKinds #-}
{-# LANGUAGE LambdaCase #-}
{-# LANGUAGE ScopedTypeVariables #-}
{-# LANGUAGE TemplateHaskell #-}
{-# LANGUAGE TypeApplications #-}
{-# LANGUAGE TypeOperators #-}
{-# LANGUAGE ViewPatterns #-}
{-# OPTIONS_GHC -fplugin GHC.TypeLits.KnownNat.Solver #-}
module Data.Array.Nested.Internal.Arith where

import Control.Monad (forM, guard)
import qualified Data.Array.Internal as OI
import qualified Data.Array.Internal.RankedG as RG
import qualified Data.Array.Internal.RankedS as RS
import Data.Bits
import Data.Int
import Data.List (sort)
import qualified Data.Vector.Storable as VS
import qualified Data.Vector.Storable.Mutable as VSM
import Foreign.C.Types
import Foreign.Ptr
import Foreign.Storable (Storable)
import GHC.TypeLits
import Language.Haskell.TH
import System.IO.Unsafe

import Data.Array.Nested.Internal.Arith.Foreign
import Data.Array.Nested.Internal.Arith.Lists


liftVEltwise1 :: Storable a
              => SNat n
              -> (VS.Vector a -> VS.Vector a)
              -> RS.Array n a -> RS.Array n a
liftVEltwise1 SNat f arr@(RS.A (RG.A sh (OI.T strides offset vec)))
  | Just prefixSz <- stridesDense sh strides =
      let vec' = f (VS.slice offset prefixSz vec)
      in RS.A (RG.A sh (OI.T strides 0 vec'))
  | otherwise = RS.fromVector sh (f (RS.toVector arr))

liftVEltwise2 :: Storable a
              => SNat n
              -> (Either a (VS.Vector a) -> Either a (VS.Vector a) -> VS.Vector a)
              -> RS.Array n a -> RS.Array n a -> RS.Array n a
liftVEltwise2 SNat f
    arr1@(RS.A (RG.A sh1 (OI.T strides1 offset1 vec1)))
    arr2@(RS.A (RG.A sh2 (OI.T strides2 offset2 vec2)))
  | sh1 /= sh2 = error $ "liftVEltwise2: shapes unequal: " ++ show sh1 ++ " vs " ++ show sh2
  | product sh1 == 0 = arr1  -- if the arrays are empty, just return one of the empty inputs
  | otherwise = case (stridesDense sh1 strides1, stridesDense sh2 strides2) of
      (Just 1, Just 1) ->  -- both are a (potentially replicated) scalar; just apply f to the scalars
        let vec' = f (Left (vec1 VS.! offset1)) (Left (vec2 VS.! offset2))
        in RS.A (RG.A sh1 (OI.T strides1 0 vec'))
      (Just 1, Just n) ->  -- scalar * dense
        RS.fromVector sh1 (f (Left (vec1 VS.! offset1)) (Right (VS.slice offset2 n vec2)))
      (Just n, Just 1) ->  -- dense * scalar
        RS.fromVector sh1 (f (Right (VS.slice offset1 n vec1)) (Left (vec2 VS.! offset2)))
      (_, _) ->  -- fallback case
        RS.fromVector sh1 (f (Right (RS.toVector arr1)) (Right (RS.toVector arr2)))

-- | Given the shape vector and the stride vector, return whether this vector
-- of strides uses a dense prefix of its backing array. If so, the number of
-- elements in this prefix is returned.
-- This excludes any offset.
stridesDense :: [Int] -> [Int] -> Maybe Int
stridesDense sh _ | any (<= 0) sh = Just 0
stridesDense sh str =
  -- sort dimensions on their stride, ascending, dropping any zero strides
  case dropWhile ((== 0) . fst) (sort (zip str sh)) of
    [] -> Just 1
    (1, n) : (unzip -> (str', sh')) -> checkCover n sh' str'
    _ -> Nothing  -- if the smallest stride is not 1, it will never be dense
  where
    -- Given size of currently densely covered region at beginning of the
    -- array, the remaining shape vector and the corresponding remaining stride
    -- vector, return whether this all together covers a dense prefix of the
    -- array. If it does, return the number of elements in this prefix.
    checkCover :: Int -> [Int] -> [Int] -> Maybe Int
    checkCover block [] [] = Just block
    checkCover block (n : sh') (s : str') = guard (s <= block) >> checkCover (max block (n * s)) sh' str'
    checkCover _ _ _ = error "Orthotope array's shape vector and stride vector have different lengths"

{-# NOINLINE vectorOp1 #-}
vectorOp1 :: forall a b. Storable a
          => (Ptr a -> Ptr b)
          -> (Int64 -> Ptr b -> Ptr b -> IO ())
          -> VS.Vector a -> VS.Vector a
vectorOp1 ptrconv f v = unsafePerformIO $ do
  outv <- VSM.unsafeNew (VS.length v)
  VSM.unsafeWith outv $ \poutv ->
    VS.unsafeWith v $ \pv ->
      f (fromIntegral (VS.length v)) (ptrconv poutv) (ptrconv pv)
  VS.unsafeFreeze outv

-- | If two vectors are given, assumes that they have the same length.
{-# NOINLINE vectorOp2 #-}
vectorOp2 :: forall a b. Storable a
          => (a -> b)
          -> (Ptr a -> Ptr b)
          -> (a -> a -> a)
          -> (Int64 -> Ptr b -> b -> Ptr b -> IO ())  -- sv
          -> (Int64 -> Ptr b -> Ptr b -> b -> IO ())  -- vs
          -> (Int64 -> Ptr b -> Ptr b -> Ptr b -> IO ())  -- vv
          -> Either a (VS.Vector a) -> Either a (VS.Vector a) -> VS.Vector a
vectorOp2 valconv ptrconv fss fsv fvs fvv = \cases
  (Left x) (Left y) -> VS.singleton (fss x y)

  (Left x) (Right vy) ->
    unsafePerformIO $ do
      outv <- VSM.unsafeNew (VS.length vy)
      VSM.unsafeWith outv $ \poutv ->
        VS.unsafeWith vy $ \pvy ->
          fsv (fromIntegral (VS.length vy)) (ptrconv poutv) (valconv x) (ptrconv pvy)
      VS.unsafeFreeze outv

  (Right vx) (Left y) ->
    unsafePerformIO $ do
      outv <- VSM.unsafeNew (VS.length vx)
      VSM.unsafeWith outv $ \poutv ->
        VS.unsafeWith vx $ \pvx ->
          fvs (fromIntegral (VS.length vx)) (ptrconv poutv) (ptrconv pvx) (valconv y)
      VS.unsafeFreeze outv

  (Right vx) (Right vy)
    | VS.length vx == VS.length vy ->
        unsafePerformIO $ do
          outv <- VSM.unsafeNew (VS.length vx)
          VSM.unsafeWith outv $ \poutv ->
            VS.unsafeWith vx $ \pvx ->
              VS.unsafeWith vy $ \pvy ->
                fvv (fromIntegral (VS.length vx)) (ptrconv poutv) (ptrconv pvx) (ptrconv pvy)
          VS.unsafeFreeze outv
    | otherwise -> error $ "vectorOp: unequal lengths: " ++ show (VS.length vx) ++ " /= " ++ show (VS.length vy)

-- TODO: test all the weird cases of this function
-- | Reduce along the inner dimension
{-# NOINLINE vectorRedInnerOp #-}
vectorRedInnerOp :: forall a b n. (Num a, Storable a)
                 => SNat n
                 -> (a -> b)
                 -> (Ptr a -> Ptr b)
                 -> (Int64 -> Ptr b -> b -> Ptr b -> IO ())  -- ^ scale by constant
                 -> (Int64 -> Ptr Int64 -> Ptr Int64 -> Ptr b -> Ptr b -> IO ())  -- ^ reduction kernel
                 -> RS.Array (n + 1) a -> RS.Array n a
vectorRedInnerOp sn@SNat valconv ptrconv fscale fred (RS.A (RG.A sh (OI.T strides offset vec)))
  | null sh = error "unreachable"
  | last sh <= 0 = RS.stretch (init sh) (RS.fromList (map (const 1) (init sh)) [0])
  | any (<= 0) (init sh) = RS.A (RG.A (init sh) (OI.T (map (const 0) (init strides)) 0 VS.empty))
  -- now the input array is nonempty
  | last sh == 1 = RS.A (RG.A (init sh) (OI.T (init strides) offset vec))
  | last strides == 0 =
      liftVEltwise1 sn
        (vectorOp1 id (\n pout px -> fscale n (ptrconv pout) (valconv (fromIntegral (last sh))) (ptrconv px)))
        (RS.A (RG.A (init sh) (OI.T (init strides) offset vec)))
  -- now there is useful work along the inner dimension
  | otherwise =
      let -- filter out zero-stride dimensions; the reduction kernel need not concern itself with those
          (shF, stridesF) = unzip $ filter ((/= 0) . snd) (zip sh strides)
          ndimsF = length shF
      in unsafePerformIO $ do
           outv <- VSM.unsafeNew (product (init shF))
           VSM.unsafeWith outv $ \poutv ->
             VS.unsafeWith (VS.fromListN ndimsF (map fromIntegral shF)) $ \pshF ->
               VS.unsafeWith (VS.fromListN ndimsF (map fromIntegral stridesF)) $ \pstridesF ->
                 VS.unsafeWith (VS.slice offset (VS.length vec - offset) vec) $ \pvec ->
                   fred (fromIntegral ndimsF) pshF pstridesF (ptrconv poutv) (ptrconv pvec)
           RS.fromVector (init sh) <$> VS.unsafeFreeze outv

flipOp :: (Int64 -> Ptr a -> a -> Ptr a -> IO ())
       ->  Int64 -> Ptr a -> Ptr a -> a -> IO ()
flipOp f n out v s = f n out s v

class NumElt a where
  numEltAdd :: SNat n -> RS.Array n a -> RS.Array n a -> RS.Array n a
  numEltSub :: SNat n -> RS.Array n a -> RS.Array n a -> RS.Array n a
  numEltMul :: SNat n -> RS.Array n a -> RS.Array n a -> RS.Array n a
  numEltNeg :: SNat n -> RS.Array n a -> RS.Array n a
  numEltAbs :: SNat n -> RS.Array n a -> RS.Array n a
  numEltSignum :: SNat n -> RS.Array n a -> RS.Array n a
  numEltSum1Inner :: SNat n -> RS.Array (n + 1) a -> RS.Array n a
  numEltProduct1Inner :: SNat n -> RS.Array (n + 1) a -> RS.Array n a

$(fmap concat . forM typesList $ \arithtype -> do
    let ttyp = conT (atType arithtype)
    fmap concat . forM binopsList $ \arithop -> do
      let name = mkName (aboName arithop ++ "Vector" ++ nameBase (atType arithtype))
          cnamebase = "c_" ++ aboName arithop ++ "_" ++ atCName arithtype
          c_ss = varE (aboScalFun arithop arithtype)
          c_sv = varE $ mkName (cnamebase ++ "_sv")
          c_vs | aboComm arithop == NonComm = varE $ mkName (cnamebase ++ "_vs")
               | otherwise = [| flipOp $c_sv |]
          c_vv = varE $ mkName (cnamebase ++ "_vv")
      sequence [SigD name <$>
                     [t| forall n. SNat n -> RS.Array n $ttyp -> RS.Array n $ttyp -> RS.Array n $ttyp |]
               ,do body <- [| \sn -> liftVEltwise2 sn (vectorOp2 id id $c_ss $c_sv $c_vs $c_vv) |]
                   return $ FunD name [Clause [] (NormalB body) []]])

$(fmap concat . forM typesList $ \arithtype -> do
    let ttyp = conT (atType arithtype)
    fmap concat . forM unopsList $ \arithop -> do
      let name = mkName (auoName arithop ++ "Vector" ++ nameBase (atType arithtype))
          c_op = varE $ mkName ("c_" ++ auoName arithop ++ "_" ++ atCName arithtype)
      sequence [SigD name <$>
                     [t| forall n. SNat n -> RS.Array n $ttyp -> RS.Array n $ttyp |]
               ,do body <- [| \sn -> liftVEltwise1 sn (vectorOp1 id $c_op) |]
                   return $ FunD name [Clause [] (NormalB body) []]])

$(fmap concat . forM typesList $ \arithtype -> do
    let ttyp = conT (atType arithtype)
    fmap concat . forM redopsList $ \redop -> do
      let name = mkName (aroName redop ++ "Vector" ++ nameBase (atType arithtype))
          c_op = varE $ mkName ("c_" ++ aroName redop ++ "_" ++ atCName arithtype)
          c_scale_op = varE $ mkName ("c_mul_" ++ atCName arithtype ++ "_sv")
      sequence [SigD name <$>
                     [t| forall n. SNat n -> RS.Array (n + 1) $ttyp -> RS.Array n $ttyp |]
               ,do body <- [| \sn -> vectorRedInnerOp sn id id $c_scale_op $c_op |]
                   return $ FunD name [Clause [] (NormalB body) []]])

-- This branch is ostensibly a runtime branch, but will (hopefully) be
-- constant-folded away by GHC.
intWidBranch1 :: forall i n. (FiniteBits i, Storable i)
              => (Int64 -> Ptr Int32 -> Ptr Int32 -> IO ())
              -> (Int64 -> Ptr Int64 -> Ptr Int64 -> IO ())
              -> (SNat n -> RS.Array n i -> RS.Array n i)
intWidBranch1 f32 f64 sn
  | finiteBitSize (undefined :: i) == 32 = liftVEltwise1 sn (vectorOp1 @i @Int32 castPtr f32)
  | finiteBitSize (undefined :: i) == 64 = liftVEltwise1 sn (vectorOp1 @i @Int64 castPtr f64)
  | otherwise = error "Unsupported Int width"

intWidBranch2 :: forall i n. (FiniteBits i, Storable i, Integral i)
              => (i -> i -> i)  -- ss
                 -- int32
              -> (Int64 -> Ptr Int32 -> Int32 -> Ptr Int32 -> IO ())  -- sv
              -> (Int64 -> Ptr Int32 -> Ptr Int32 -> Int32 -> IO ())  -- vs
              -> (Int64 -> Ptr Int32 -> Ptr Int32 -> Ptr Int32 -> IO ())  -- vv
                 -- int64
              -> (Int64 -> Ptr Int64 -> Int64 -> Ptr Int64 -> IO ())  -- sv
              -> (Int64 -> Ptr Int64 -> Ptr Int64 -> Int64 -> IO ())  -- vs
              -> (Int64 -> Ptr Int64 -> Ptr Int64 -> Ptr Int64 -> IO ())  -- vv
              -> (SNat n -> RS.Array n i -> RS.Array n i -> RS.Array n i)
intWidBranch2 ss sv32 vs32 vv32 sv64 vs64 vv64 sn
  | finiteBitSize (undefined :: i) == 32 = liftVEltwise2 sn (vectorOp2 @i @Int32 fromIntegral castPtr ss sv32 vs32 vv32)
  | finiteBitSize (undefined :: i) == 64 = liftVEltwise2 sn (vectorOp2 @i @Int64 fromIntegral castPtr ss sv64 vs64 vv64)
  | otherwise = error "Unsupported Int width"

intWidBranchRed :: forall i n. (FiniteBits i, Storable i, Integral i)
                => -- int32
                   (Int64 -> Ptr Int32 -> Int32 -> Ptr Int32 -> IO ())  -- ^ scale by constant
                -> (Int64 -> Ptr Int64 -> Ptr Int64 -> Ptr Int32 -> Ptr Int32 -> IO ())  -- ^ reduction kernel
                   -- int64
                -> (Int64 -> Ptr Int64 -> Int64 -> Ptr Int64 -> IO ())  -- ^ scale by constant
                -> (Int64 -> Ptr Int64 -> Ptr Int64 -> Ptr Int64 -> Ptr Int64 -> IO ())  -- ^ reduction kernel
                -> (SNat n -> RS.Array (n + 1) i -> RS.Array n i)
intWidBranchRed fsc32 fred32 fsc64 fred64 sn
  | finiteBitSize (undefined :: i) == 32 = vectorRedInnerOp @i @Int32 sn fromIntegral castPtr fsc32 fred32
  | finiteBitSize (undefined :: i) == 64 = vectorRedInnerOp @i @Int64 sn fromIntegral castPtr fsc64 fred64
  | otherwise = error "Unsupported Int width"

instance NumElt Int32 where
  numEltAdd = addVectorInt32
  numEltSub = subVectorInt32
  numEltMul = mulVectorInt32
  numEltNeg = negVectorInt32
  numEltAbs = absVectorInt32
  numEltSignum = signumVectorInt32
  numEltSum1Inner = sum1VectorInt32
  numEltProduct1Inner = product1VectorInt32

instance NumElt Int64 where
  numEltAdd = addVectorInt64
  numEltSub = subVectorInt64
  numEltMul = mulVectorInt64
  numEltNeg = negVectorInt64
  numEltAbs = absVectorInt64
  numEltSignum = signumVectorInt64
  numEltSum1Inner = sum1VectorInt64
  numEltProduct1Inner = product1VectorInt64

instance NumElt Float where
  numEltAdd = addVectorFloat
  numEltSub = subVectorFloat
  numEltMul = mulVectorFloat
  numEltNeg = negVectorFloat
  numEltAbs = absVectorFloat
  numEltSignum = signumVectorFloat
  numEltSum1Inner = sum1VectorFloat
  numEltProduct1Inner = product1VectorFloat

instance NumElt Double where
  numEltAdd = addVectorDouble
  numEltSub = subVectorDouble
  numEltMul = mulVectorDouble
  numEltNeg = negVectorDouble
  numEltAbs = absVectorDouble
  numEltSignum = signumVectorDouble
  numEltSum1Inner = sum1VectorDouble
  numEltProduct1Inner = product1VectorDouble

instance NumElt Int where
  numEltAdd = intWidBranch2 @Int (+) c_add_i32_sv (flipOp c_add_i32_sv) c_add_i32_vv c_add_i64_sv (flipOp c_add_i64_sv) c_add_i64_vv
  numEltSub = intWidBranch2 @Int (-) c_sub_i32_sv (flipOp c_sub_i32_sv) c_sub_i32_vv c_sub_i64_sv (flipOp c_sub_i64_sv) c_sub_i64_vv
  numEltMul = intWidBranch2 @Int (*) c_mul_i32_sv (flipOp c_mul_i32_sv) c_mul_i32_vv c_mul_i64_sv (flipOp c_mul_i64_sv) c_mul_i64_vv
  numEltNeg = intWidBranch1 @Int c_neg_i32 c_neg_i64
  numEltAbs = intWidBranch1 @Int c_abs_i32 c_abs_i64
  numEltSignum = intWidBranch1 @Int c_signum_i32 c_signum_i64
  numEltSum1Inner = intWidBranchRed @Int c_mul_i32_sv c_sum1_i32 c_mul_i64_sv c_sum1_i64
  numEltProduct1Inner = intWidBranchRed @Int c_mul_i32_sv c_product1_i32 c_mul_i64_sv c_product1_i64

instance NumElt CInt where
  numEltAdd = intWidBranch2 @CInt (+) c_add_i32_sv (flipOp c_add_i32_sv) c_add_i32_vv c_add_i64_sv (flipOp c_add_i64_sv) c_add_i64_vv
  numEltSub = intWidBranch2 @CInt (-) c_sub_i32_sv (flipOp c_sub_i32_sv) c_sub_i32_vv c_sub_i64_sv (flipOp c_sub_i64_sv) c_sub_i64_vv
  numEltMul = intWidBranch2 @CInt (*) c_mul_i32_sv (flipOp c_mul_i32_sv) c_mul_i32_vv c_mul_i64_sv (flipOp c_mul_i64_sv) c_mul_i64_vv
  numEltNeg = intWidBranch1 @CInt c_neg_i32 c_neg_i64
  numEltAbs = intWidBranch1 @CInt c_abs_i32 c_abs_i64
  numEltSignum = intWidBranch1 @CInt c_signum_i32 c_signum_i64
  numEltSum1Inner = intWidBranchRed @CInt c_mul_i32_sv c_sum1_i32 c_mul_i64_sv c_sum1_i64
  numEltProduct1Inner = intWidBranchRed @CInt c_mul_i32_sv c_product1_i32 c_mul_i64_sv c_product1_i64