Properly cast values when writing/reading unboxed sums.wip/fix-ubx-cast

Unboxed sums might store a Int8# value as Int64#. This patch
makes sure we keep track of the actual value type.

See Note [Casting slot arguments] for the details.

17 files changed, 791 insertions, 82 deletions

diff --git a/compiler/GHC/Builtin/PrimOps/Casts.hs b/compiler/GHC/Builtin/PrimOps/Casts.hs
new file mode 100644
index 0000000000..ced3ace986
--- /dev/null
+++ b/compiler/GHC/Builtin/PrimOps/Casts.hs
@@ -0,0 +1,212 @@
+{-
+This module contains helpers to cast variables
+between different Int/WordReps in StgLand.
+
+-}
+
+module GHC.Builtin.PrimOps.Casts
+    ( getCasts )
+where
+
+import GHC.Prelude
+
+import GHC.Core.TyCon
+import GHC.Utils.Outputable
+import GHC.Utils.Panic
+import GHC.Utils.Panic.Plain
+import GHC.Types.RepType
+import GHC.Core.Type
+import GHC.Builtin.Types.Prim
+
+import GHC.Builtin.PrimOps
+import GHC.Plugins (HasDebugCallStack)
+
+{- Note [PrimRep based casting]
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+This module contains a number of utility functions useful when
+converting between variables of differing PrimReps.
+
+The general pattern is:
+* We have two primReps `from_rep` and `to_rep`.
+* We want a list of PrimOps we can apply to a variable of rep `from_rep`.
+Applying the list of primOps in order takes us to `to_rep` from `from_rep` giving
+us a variable of the returned type at each step.
+
+E.g. we call `getCasts from_rep to_rep` and get back [(op1#,ty1),(op2#,ty2)].
+We can use this result to construct a function of type
+`StgExpr -> StgExpr` by construction an expression
+
+    case op1# <from> of (x' :: ty1) -> case op2# x' of x' -> <rhs_hole>
+
+Ideally backends will compile the sequence of PrimOps to a no-op. E.g. by reusing
+the same register but just relabeling it as another width.
+However this is might not always be possible or the required optimizations
+simply not implemented in the backend. This means currently many of these casts
+will be cheap but not all of them will be completely zero-cost.
+
+-}
+
+-- | `getCasts from_rep to_rep` gives us a list of primops which when applied in order convert from_rep to to_rep.
+-- See Note [PrimRep based casting]
+getCasts :: PrimRep -> PrimRep -> [(PrimOp,Type)]
+getCasts from_rep to_rep
+  -- No-op
+  | -- pprTrace "getCasts" (ppr (from_rep,to_rep)) $
+    to_rep == from_rep
+  = []
+
+  -- Float <-> Double
+  | to_rep == FloatRep =
+    assertPpr (from_rep == DoubleRep) (ppr from_rep <+> ppr to_rep) $
+    [(DoubleToFloatOp,floatPrimTy)]
+  | to_rep == DoubleRep =
+    assertPpr (from_rep == FloatRep) (ppr from_rep <+> ppr to_rep) $
+    [(FloatToDoubleOp,doublePrimTy)]
+
+  -- Addr <-> Word/Int
+  | to_rep == AddrRep = wordOrIntToAddrRep from_rep
+  | from_rep == AddrRep = addrToWordOrIntRep to_rep
+
+  -- Int* -> Int*
+  | primRepIsInt from_rep
+  , primRepIsInt to_rep
+  = sizedIntToSizedInt from_rep to_rep
+
+  -- Word* -> Word*
+  | primRepIsWord from_rep
+  , primRepIsWord to_rep
+  = sizedWordToSizedWord from_rep to_rep
+
+  -- Word* -> Int*
+  | primRepIsWord from_rep
+  , primRepIsInt to_rep
+  = let (op1,r1) = wordToIntRep from_rep
+    in (op1,primRepToType r1):sizedIntToSizedInt r1 to_rep
+
+  -- Int* -> Word*
+  | primRepIsInt from_rep
+  , primRepIsWord to_rep
+  = let (op1,r1) = intToWordRep from_rep
+    in (op1,primRepToType r1):sizedWordToSizedWord r1 to_rep
+
+  | otherwise = pprPanic "getCasts:Unexpect rep combination"
+                          (ppr (from_rep,to_rep))
+
+wordOrIntToAddrRep :: HasDebugCallStack => PrimRep -> [(PrimOp,Type)]
+wordOrIntToAddrRep AddrRep = [] -- No-op argument is already AddrRep
+wordOrIntToAddrRep IntRep = [(IntToAddrOp, addrPrimTy)]
+wordOrIntToAddrRep WordRep = [(WordToIntOp,intPrimTy), (IntToAddrOp,addrPrimTy)]
+wordOrIntToAddrRep r
+    | primRepIsInt r = (intToMachineInt r,intPrimTy):[(IntToAddrOp,addrPrimTy)]
+    | primRepIsWord r =
+        let (op1,r1) = wordToIntRep r
+        in (op1, primRepToType r1):[(intToMachineInt r1,intPrimTy), (IntToAddrOp,addrPrimTy)]
+    | otherwise = pprPanic "Rep not word or int rep" (ppr r)
+
+addrToWordOrIntRep :: HasDebugCallStack => PrimRep -> [(PrimOp,Type)]
+-- Machine sizes
+addrToWordOrIntRep IntRep = [(AddrToIntOp, intPrimTy)]
+addrToWordOrIntRep WordRep = [(AddrToIntOp,intPrimTy), (IntToWordOp,wordPrimTy)]
+-- Explicitly sized reps
+addrToWordOrIntRep r
+    | primRepIsWord r = (AddrToIntOp,intPrimTy) : (IntToWordOp,wordPrimTy) : sizedWordToSizedWord WordRep r
+    | primRepIsInt r = (AddrToIntOp,intPrimTy) : sizedIntToSizedInt IntRep r
+    | otherwise = pprPanic "Target rep not word or int rep" (ppr r)
+
+
+-- WordX# -> IntX# (same size), argument is source rep
+wordToIntRep :: HasDebugCallStack => PrimRep -> (PrimOp,PrimRep)
+wordToIntRep rep
+    = case rep of
+        (WordRep) -> (WordToIntOp, IntRep)
+        (Word8Rep) -> (Word8ToInt8Op, Int8Rep)
+        (Word16Rep) -> (Word16ToInt16Op, Int16Rep)
+        (Word32Rep) -> (Word32ToInt32Op, Int32Rep)
+        (Word64Rep) -> (Word64ToInt64Op, Int64Rep)
+        _ -> pprPanic "Rep not a wordRep" (ppr rep)
+
+-- IntX# -> WordX#, argument is source rep
+intToWordRep :: HasDebugCallStack => PrimRep -> (PrimOp,PrimRep)
+intToWordRep rep
+    = case rep of
+        (IntRep) -> (IntToWordOp, WordRep)
+        (Int8Rep) -> (Int8ToWord8Op, Word8Rep)
+        (Int16Rep) -> (Int16ToWord16Op, Word16Rep)
+        (Int32Rep) -> (Int32ToWord32Op, Word32Rep)
+        (Int64Rep) -> (Int64ToWord64Op, Word64Rep)
+        _ -> pprPanic "Rep not a wordRep" (ppr rep)
+
+-- Casts between any size int to any other size of int
+sizedIntToSizedInt :: HasDebugCallStack => PrimRep -> PrimRep -> [(PrimOp,Type)]
+sizedIntToSizedInt r1 r2
+    | r1 == r2 = []
+-- Cast to Int#
+sizedIntToSizedInt r IntRep = [(intToMachineInt r,intPrimTy)]
+-- Cast from Int#
+sizedIntToSizedInt IntRep r = [(intFromMachineInt r,primRepToType r)]
+-- Sized to differently sized must go over machine word.
+sizedIntToSizedInt r1 r2 = (intToMachineInt r1,intPrimTy) : [(intFromMachineInt r2,primRepToType r2)]
+
+-- Casts between any size Word to any other size of Word
+sizedWordToSizedWord :: HasDebugCallStack => PrimRep -> PrimRep -> [(PrimOp,Type)]
+sizedWordToSizedWord r1 r2
+    | r1 == r2 = []
+-- Cast to Word#
+sizedWordToSizedWord r WordRep = [(wordToMachineWord r,wordPrimTy)]
+-- Cast from Word#
+sizedWordToSizedWord WordRep r = [(wordFromMachineWord r, primRepToType r)]
+-- Conversion between different non-machine sizes must go via machine word.
+sizedWordToSizedWord r1 r2 = (wordToMachineWord r1,wordPrimTy) : [(wordFromMachineWord r2, primRepToType r2)]
+
+
+-- Prefer the definitions above this line if possible
+----------------------
+
+
+-- Int*# to Int#
+{-# INLINE intToMachineInt #-}
+intToMachineInt :: HasDebugCallStack => PrimRep -> PrimOp
+intToMachineInt r =
+    assertPpr (primRepIsInt r) (ppr r) $
+    case r of
+        (Int8Rep) -> Int8ToIntOp
+        (Int16Rep) -> Int16ToIntOp
+        (Int32Rep) -> Int32ToIntOp
+        (Int64Rep) -> Int64ToIntOp
+        _ -> pprPanic "Source rep not int" $ ppr r
+
+-- Int# to Int*#
+{-# INLINE intFromMachineInt #-}
+intFromMachineInt :: HasDebugCallStack => PrimRep -> PrimOp
+intFromMachineInt r =
+    assertPpr (primRepIsInt r) (ppr r) $
+    case r of
+        Int8Rep -> IntToInt8Op
+        Int16Rep -> IntToInt16Op
+        Int32Rep -> IntToInt32Op
+        Int64Rep -> IntToInt64Op
+        _ -> pprPanic "Dest rep not sized int" $ ppr r
+
+-- Word# to Word*#
+{-# INLINE wordFromMachineWord #-}
+wordFromMachineWord :: HasDebugCallStack => PrimRep -> PrimOp
+wordFromMachineWord r =
+    assert (primRepIsWord r) $
+    case r of
+        Word8Rep -> WordToWord8Op
+        Word16Rep -> WordToWord16Op
+        Word32Rep -> WordToWord32Op
+        Word64Rep -> WordToWord64Op
+        _ -> pprPanic "Dest rep not sized word" $ ppr r
+
+-- Word*# to Word#
+{-# INLINE wordToMachineWord #-}
+wordToMachineWord :: HasDebugCallStack => PrimRep -> PrimOp
+wordToMachineWord r =
+    assertPpr (primRepIsWord r) (text "Not a word rep:" <> ppr r) $
+    case r of
+        Word8Rep -> Word8ToWordOp
+        Word16Rep -> Word16ToWordOp
+        Word32Rep -> Word32ToWordOp
+        Word64Rep -> Word64ToWordOp
+        _ -> pprPanic "Dest rep not sized word" $ ppr r
\ No newline at end of file
diff --git a/compiler/GHC/Cmm/CLabel.hs b/compiler/GHC/Cmm/CLabel.hs
index 96f78b6789..c492e8e7c1 100644
--- a/compiler/GHC/Cmm/CLabel.hs
+++ b/compiler/GHC/Cmm/CLabel.hs
@@ -1448,7 +1448,8 @@ instance OutputableP Platform CLabel where
   pdoc !platform lbl = getPprStyle $ \pp_sty ->
                         case pp_sty of
                           PprDump{} -> pprCLabel platform lbl
-                          _         -> pprPanic "Labels in code should be printed with pprCLabel or pprAsmLabel" (pprCLabel platform lbl)
+                          _         -> let lbl_doc = (pprCLabel platform lbl)
+                                       in pprTraceUserWarning (text "Labels in code should be printed with pprCLabel or pprAsmLabel" <> lbl_doc) lbl_doc
 
 pprCLabelStyle :: forall doc. IsLine doc => Platform -> LabelStyle -> CLabel -> doc
 pprCLabelStyle !platform !sty lbl = -- see Note [Bangs in CLabel]
diff --git a/compiler/GHC/Core/TyCon.hs b/compiler/GHC/Core/TyCon.hs
index 2a26363c7b..4c1957dd1f 100644
--- a/compiler/GHC/Core/TyCon.hs
+++ b/compiler/GHC/Core/TyCon.hs
@@ -130,6 +130,8 @@ module GHC.Core.TyCon(
         primRepIsFloat,
         primRepsCompatible,
         primRepCompatible,
+        primRepIsWord,
+        primRepIsInt,
 
 ) where
 
@@ -1787,6 +1789,24 @@ primRepIsFloat  DoubleRep    = Just True
 primRepIsFloat  (VecRep _ _) = Nothing
 primRepIsFloat  _            = Just False
 
+-- Rep is one of the word reps.
+primRepIsWord :: PrimRep -> Bool
+primRepIsWord WordRep = True
+primRepIsWord (Word8Rep) = True
+primRepIsWord (Word16Rep) = True
+primRepIsWord (Word32Rep) = True
+primRepIsWord (Word64Rep) = True
+primRepIsWord _ = False
+
+-- Rep is one of the int reps.
+primRepIsInt :: PrimRep -> Bool
+primRepIsInt (IntRep) = True
+primRepIsInt (Int8Rep) = True
+primRepIsInt (Int16Rep) = True
+primRepIsInt (Int32Rep) = True
+primRepIsInt (Int64Rep) = True
+primRepIsInt _ = False
+
 {-
 ************************************************************************
 *                                                                      *
diff --git a/compiler/GHC/Driver/Flags.hs b/compiler/GHC/Driver/Flags.hs
index b1154b6398..7970131212 100644
--- a/compiler/GHC/Driver/Flags.hs
+++ b/compiler/GHC/Driver/Flags.hs
@@ -424,6 +424,7 @@ data GeneralFlag
    -- variables that have otherwise identical names.
    | Opt_SuppressUniques
    | Opt_SuppressStgExts
+   | Opt_SuppressStgReps
    | Opt_SuppressTicks     -- Replaces Opt_PprShowTicks
    | Opt_SuppressTimestamps -- ^ Suppress timestamps in dumps
    | Opt_SuppressCoreSizes  -- ^ Suppress per binding Core size stats in dumps
diff --git a/compiler/GHC/Driver/Session.hs b/compiler/GHC/Driver/Session.hs
index 168a204fbc..b717feebaa 100644
--- a/compiler/GHC/Driver/Session.hs
+++ b/compiler/GHC/Driver/Session.hs
@@ -2395,6 +2395,7 @@ dynamic_flags_deps = [
                   setGeneralFlag Opt_SuppressIdInfo
                   setGeneralFlag Opt_SuppressTicks
                   setGeneralFlag Opt_SuppressStgExts
+                  setGeneralFlag Opt_SuppressStgReps
                   setGeneralFlag Opt_SuppressTypeSignatures
                   setGeneralFlag Opt_SuppressCoreSizes
                   setGeneralFlag Opt_SuppressTimestamps)
@@ -3347,6 +3348,7 @@ dFlagsDeps = [
   depFlagSpec' "suppress-stg-free-vars" Opt_SuppressStgExts
      (useInstead "-d" "suppress-stg-exts"),
   flagSpec "suppress-stg-exts"          Opt_SuppressStgExts,
+  flagSpec "suppress-stg-reps"          Opt_SuppressStgReps,
   flagSpec "suppress-coercions"         Opt_SuppressCoercions,
   flagSpec "suppress-coercion-types"    Opt_SuppressCoercionTypes,
   flagSpec "suppress-idinfo"            Opt_SuppressIdInfo,
@@ -3806,8 +3808,8 @@ defaultFlags settings
       Opt_RPath,
       Opt_DumpWithWays,
       Opt_CompactUnwind,
-      Opt_ShowErrorContext
-
+      Opt_ShowErrorContext,
+      Opt_SuppressStgReps
     ]
 
     ++ [f | (ns,f) <- optLevelFlags, 0 `elem` ns]
@@ -5032,6 +5034,7 @@ initSDocContext dflags style = SDC
   , sdocSuppressUniques             = gopt Opt_SuppressUniques dflags
   , sdocSuppressModulePrefixes      = gopt Opt_SuppressModulePrefixes dflags
   , sdocSuppressStgExts             = gopt Opt_SuppressStgExts dflags
+  , sdocSuppressStgReps             = gopt Opt_SuppressStgReps dflags
   , sdocErrorSpans                  = gopt Opt_ErrorSpans dflags
   , sdocStarIsType                  = xopt LangExt.StarIsType dflags
   , sdocLinearTypes                 = xopt LangExt.LinearTypes dflags
diff --git a/compiler/GHC/Stg/Syntax.hs b/compiler/GHC/Stg/Syntax.hs
index 4956920fb1..07fc35b5eb 100644
--- a/compiler/GHC/Stg/Syntax.hs
+++ b/compiler/GHC/Stg/Syntax.hs
@@ -87,7 +87,7 @@ import GHC.Core.Ppr( {- instances -} )
 import GHC.Builtin.PrimOps ( PrimOp, PrimCall )
 import GHC.Core.TyCon    ( PrimRep(..), TyCon )
 import GHC.Core.Type     ( Type )
-import GHC.Types.RepType ( typePrimRep1 )
+import GHC.Types.RepType ( typePrimRep1, typePrimRep )
 import GHC.Utils.Panic.Plain
 
 {-
@@ -740,12 +740,23 @@ pprStgTopBinding = pprGenStgTopBinding
 pprStgTopBindings :: OutputablePass pass => StgPprOpts -> [GenStgTopBinding pass] -> SDoc
 pprStgTopBindings = pprGenStgTopBindings
 
+pprIdWithRep :: Id -> SDoc
+pprIdWithRep v = ppr v <> pprTypeRep (idType v)
+
+pprTypeRep :: Type -> SDoc
+pprTypeRep ty =
+    ppUnlessOption sdocSuppressStgReps $
+    char ':' <> case typePrimRep ty of
+                  [r] -> ppr r
+                  r -> ppr r
+
+
 instance Outputable StgArg where
   ppr = pprStgArg
 
 pprStgArg :: StgArg -> SDoc
-pprStgArg (StgVarArg var) = ppr var
-pprStgArg (StgLitArg con) = ppr con
+pprStgArg (StgVarArg var) = pprIdWithRep var
+pprStgArg (StgLitArg con) = ppr con <> pprTypeRep (literalType con)
 
 instance OutputablePass pass => Outputable  (GenStgExpr pass) where
   ppr = pprStgExpr panicStgPprOpts
diff --git a/compiler/GHC/Stg/Unarise.hs b/compiler/GHC/Stg/Unarise.hs
index 35eb37b17f..c7d7ebcc6a 100644
--- a/compiler/GHC/Stg/Unarise.hs
+++ b/compiler/GHC/Stg/Unarise.hs
@@ -184,6 +184,129 @@ So we pass type arguments of the DataCon's TyCon in StgConApp to decide what
 layout to use. Note that unlifted values can't be let-bound, so we don't need
 types in StgRhsCon.
 
+Note [Casting slot arguments]
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+Consider this function which selects between Int32# and Int64# from a unboxed sum.
+
+    foo ::  (# Int32# | Int64#  #) -> FD
+    foo x = case x of
+        (# x1 | #) -> F x1
+        (# | x2 #) -> D x2
+
+Naturally we would expect x1 to have a PrimRep of Int32Rep and x2 of DoubleRep.
+However we used to generate this (bogus) code after Unarise giving rise to #22208:
+
+    M.foo :: (# GHC.Prim.Int32# | GHC.Prim.Int64# #) -> M.FD
+    [GblId, Arity=1, Unf=OtherCon []] =
+        {} \r [sum_tag sum_field]
+            case sum_tag of tag_gsc {
+              __DEFAULT -> M.F [sum_field];
+              2# -> M.D [sum_field];
+            };
+
+Where sum_field is used both as Int32# and Int64# depending on the branch
+because they share the same SlotTy.
+This usually works out since we put all int's in the same sort of register.
+So even if the reps where wrong (x :: bits32) = (y :: bits64) would produce
+correct code in the most cases.
+However there are cases where this goes wrong, causing lint errors,in the case of #22208
+compiler panics or in some cases incorrect results in the C backend.
+For now our solution is to construct proper casts between the PrimRep of the slot and
+the variables we want to store in, or read out of these slots.
+
+This means when we have a sum (# Int32# | Int64# #) if we want to store a Int32
+we convert it to a Int64 on construction of the tuple value, and convert it back
+to a Int32 once when want to use the field. On most backends these coversions should
+be no-ops at runtime so this seems reasonable.
+
+Conversion for values coming out of a strict field happen in mapSumIdBinders. While
+conversion during the construction of sums happen inside mkUbxSum.
+
+------------- A full example of casting during sum construction ----------------
+
+To compile a constructor application of a unboxed sum of type (# Int32# | Int64# )
+in an expression like  `let sum = (# x | #)` we will call mkUbxSum to determine
+which binders we have to replace sum with at use sites during unarise.
+See also Note [Translating unboxed sums to unboxed tuples].
+
+Int32# and Int64# in this case will share the same slot in the unboxed sum. This means
+the sum after unarise will be represented by two binders. One for the tag and one for
+the field. The later having Int64Rep.
+However our input for the field is of Int32Rep. So in order to soundly construct
+`(# x | #) :: (# Int32# | Int64# )` we must upcast `x` to Int64#.
+To do this mkUbxSum will produce an expression with a hole for constructor application
+to go into. That is the call to mkUbxSum and it's result will look something like:
+
+  >>> mkUbxSum (#|#) [Int32#, Int64#] (x::Int32#) us (x')
+  ([1#::Int#, x'::Int64#], \rhs -> case int32ToInt# x of x' -> rhs )
+
+We will use the returned arguments to construct an application to an unboxed tuple:
+
+  >>> mkTuple [tag::Int#, x'::Int64#]
+  (# tag, x' #)
+
+Which we will then use as the rhs to pass into the casting wrapper to
+construct an expression that casts `x` to the right type before constructing the
+tuple
+
+  >>> (\rhs -> case int32ToInt# x of x' -> rhs ) (# tag, x' #)
+  case int32ToInt# x of x' -> (# #) 1# x'
+
+Which results in the this definition for `sum` after all is said and done:
+
+  let sum = case int32ToInt# x of { x' -> (# #) 1# x' }
+
+Not that the renaming is not optional. Cmm requires binders of different uniques
+to have at least different types. See Note [CorePrep Overview]: 6. Clone all local Ids
+
+------------- A full example of casting during sum matching --------------------
+
+When matching on an unboxed sum constructor we start out with
+something like this the pre-unarise:
+
+    f :: (# Int32 | Int64# ) -> ...
+    f sum = case sum of
+        (# x |#) -> alt_rhs
+        ...
+
+We unarise the function arguments and get:
+
+    f sum_tag sum_slot1 = case sum_tag of
+        1# -> ???
+
+Now we need to match up the original alternative binders with the sum slots passed
+to the function. This is done by mapSumIdBinders which we we call for our
+example alternative like this:
+
+    >>> mapSumIdBinders [x] [sum_slot1] alt_rhs env
+    (env', alt_rhs')
+
+mapSumIdBinders first matches up the list of binders with the slots passed to
+the function which is trivial in this case. Then we check if the slot and the
+variable residing inside it agree on their Rep. If alternative binders and
+the function arguments agree in their slot reps we we just extend the environment
+with a mapping from `x` to `sum_slot1` and we return the rhs as is.
+
+If the reps of the sum_slots do not agree with alternative binders they represent
+then we need to wrap the whole RHS in nested cases which cast the sum_slot<n>
+variables to the correct rep. Here `x` is of Int32Rep while `sum_slot1` will be
+of Int64Rep. This means instead of retuning the original alt_rhs we will return:
+
+  >>> mapSumIdBinders [x] [sum_slot1] alt_rhs env
+  ( env'[x=x']
+  , case int64ToInt32# (sum_slot1 :: Int64#) of
+      (x' :: Int32#) -> alt_rhs
+  )
+
+We then run unarise on alt_rhs within that expression, which will replace the first occurence
+of `x` with sum_slot_arg_1 giving us post-unarise:
+
+    f sum_tag sum_slot1 =
+      case sum_tag of
+        1# -> case int64ToInt32# sum_slot1 of
+          x' -> ... x' ...
+        ...
+
 Note [UnariseEnv]
 ~~~~~~~~~~~~~~~~~~
 At any variable occurrence 'v',
@@ -256,8 +379,8 @@ import GHC.Prelude
 import GHC.Types.Basic
 import GHC.Core
 import GHC.Core.DataCon
-import GHC.Core.TyCon ( isVoidRep )
-import GHC.Data.FastString (FastString, mkFastString)
+import GHC.Core.TyCon
+import GHC.Data.FastString (FastString, mkFastString, fsLit)
 import GHC.Types.Id
 import GHC.Types.Literal
 import GHC.Core.Make (aBSENT_SUM_FIELD_ERROR_ID)
@@ -273,6 +396,7 @@ import GHC.Core.Type
 import GHC.Builtin.Types.Prim (intPrimTy)
 import GHC.Builtin.Types
 import GHC.Types.Unique.Supply
+import GHC.Types.Unique
 import GHC.Utils.Misc
 import GHC.Types.Var.Env
 
@@ -280,7 +404,11 @@ import Data.Bifunctor (second)
 import Data.List.NonEmpty (NonEmpty (..))
 import Data.Maybe (mapMaybe)
 import qualified Data.IntMap as IM
+import GHC.Builtin.PrimOps
+import GHC.Builtin.PrimOps.Casts
+import Data.List (mapAccumL)
 
+-- import GHC.Utils.Trace
 --------------------------------------------------------------------------------
 
 -- | A mapping from binders to the Ids they were expanded/renamed to.
@@ -305,8 +433,10 @@ import qualified Data.IntMap as IM
 -- INVARIANT: OutStgArgs in the range only have NvUnaryTypes
 --            (i.e. no unboxed tuples, sums or voids)
 --
-type UnariseEnv = VarEnv UnariseVal
+newtype UnariseEnv = UnariseEnv  { ue_rho :: (VarEnv UnariseVal) }
 
+initUnariseEnv :: VarEnv UnariseVal -> UnariseEnv
+initUnariseEnv = UnariseEnv
 data UnariseVal
   = MultiVal [OutStgArg] -- MultiVal to tuple. Can be empty list (void).
   | UnaryVal OutStgArg   -- See Note [Renaming during unarisation].
@@ -319,25 +449,27 @@ instance Outputable UnariseVal where
 -- The id is mapped to one or more things.
 -- See Note [UnariseEnv]
 extendRho :: UnariseEnv -> Id -> UnariseVal -> UnariseEnv
-extendRho rho x (MultiVal args)
+extendRho env x (MultiVal args)
   = assert (all (isNvUnaryType . stgArgType) args)
-    extendVarEnv rho x (MultiVal args)
-extendRho rho x (UnaryVal val)
+    env { ue_rho = extendVarEnv (ue_rho env) x (MultiVal args) }
+extendRho env x (UnaryVal val)
   = assert (isNvUnaryType (stgArgType val))
-    extendVarEnv rho x (UnaryVal val)
+    env { ue_rho = extendVarEnv (ue_rho env) x (UnaryVal val) }
 -- Properly shadow things from an outer scope.
 -- See Note [UnariseEnv]
 
 -- The id stands for itself so we don't record a mapping.
 -- See Note [UnariseEnv]
 extendRhoWithoutValue :: UnariseEnv -> Id -> UnariseEnv
-extendRhoWithoutValue rho x = delVarEnv rho x
+extendRhoWithoutValue env x = env { ue_rho = delVarEnv (ue_rho env) x }
 
+lookupRho :: UnariseEnv -> Id -> Maybe UnariseVal
+lookupRho env v = lookupVarEnv (ue_rho env) v
 
 --------------------------------------------------------------------------------
 
 unarise :: UniqSupply -> [StgTopBinding] -> [StgTopBinding]
-unarise us binds = initUs_ us (mapM (unariseTopBinding emptyVarEnv) binds)
+unarise us binds = initUs_ us (mapM (unariseTopBinding (initUnariseEnv emptyVarEnv)) binds)
 
 unariseTopBinding :: UnariseEnv -> StgTopBinding -> UniqSM StgTopBinding
 unariseTopBinding rho (StgTopLifted bind)
@@ -365,7 +497,7 @@ unariseRhs rho (StgRhsCon ccs con mu ts args)
 unariseExpr :: UnariseEnv -> StgExpr -> UniqSM StgExpr
 
 unariseExpr rho e@(StgApp f [])
-  = case lookupVarEnv rho f of
+  = case lookupRho rho f of
       Just (MultiVal args)  -- Including empty tuples
         -> return (mkTuple args)
       Just (UnaryVal (StgVarArg f'))
@@ -378,7 +510,7 @@ unariseExpr rho e@(StgApp f [])
 unariseExpr rho e@(StgApp f args)
   = return (StgApp f' (unariseFunArgs rho args))
   where
-    f' = case lookupVarEnv rho f of
+    f' = case lookupRho rho f of
            Just (UnaryVal (StgVarArg f')) -> f'
            Nothing -> f
            err -> pprPanic "unariseExpr - app2" (pprStgExpr panicStgPprOpts e $$ ppr err)
@@ -389,12 +521,17 @@ unariseExpr _ (StgLit l)
   = return (StgLit l)
 
 unariseExpr rho (StgConApp dc n args ty_args)
-  | Just args' <- unariseMulti_maybe rho dc args ty_args
-  = return (mkTuple args')
-
-  | otherwise
-  , let args' = unariseConArgs rho args
-  = return (StgConApp dc n args' (map stgArgType args'))
+  | isUnboxedSumDataCon dc || isUnboxedTupleDataCon dc
+  = do
+      us <- getUniqueSupplyM
+      case unariseUbxSumOrTupleArgs rho us dc args ty_args of
+        (args', Just cast_wrapper)
+          -> return $ cast_wrapper (mkTuple args')
+        (args', Nothing)
+          -> return $ (mkTuple args')
+  | otherwise =
+      let args' = unariseConArgs rho args in
+      return $ (StgConApp dc n args' (map stgArgType args'))
 
 unariseExpr rho (StgOpApp op args ty)
   = return (StgOpApp op (unariseFunArgs rho args) ty)
@@ -402,15 +539,19 @@ unariseExpr rho (StgOpApp op args ty)
 unariseExpr rho (StgCase scrut bndr alt_ty alts)
   -- tuple/sum binders in the scrutinee can always be eliminated
   | StgApp v [] <- scrut
-  , Just (MultiVal xs) <- lookupVarEnv rho v
+  , Just (MultiVal xs) <- lookupRho rho v
   = elimCase rho xs bndr alt_ty alts
 
   -- Handle strict lets for tuples and sums:
   --   case (# a,b #) of r -> rhs
   -- and analogously for sums
   | StgConApp dc _n args ty_args <- scrut
-  , Just args' <- unariseMulti_maybe rho dc args ty_args
-  = elimCase rho args' bndr alt_ty alts
+  , isUnboxedSumDataCon dc || isUnboxedTupleDataCon dc
+  = do
+    us <- getUniqueSupplyM
+    case unariseUbxSumOrTupleArgs rho us dc args ty_args of
+      (args',Just wrapper) -> wrapper <$> elimCase rho args' bndr alt_ty alts
+      (args',Nothing) -> elimCase rho args' bndr alt_ty alts
 
   -- See (3) of Note [Rubbish literals] in GHC.Types.Literal
   | StgLit lit <- scrut
@@ -435,17 +576,21 @@ unariseExpr rho (StgTick tick e)
   = StgTick tick <$> unariseExpr rho e
 
 -- Doesn't return void args.
-unariseMulti_maybe :: UnariseEnv -> DataCon -> [InStgArg] -> [Type] -> Maybe [OutStgArg]
-unariseMulti_maybe rho dc args ty_args
+unariseUbxSumOrTupleArgs :: UnariseEnv -> UniqSupply -> DataCon -> [InStgArg] -> [Type]
+                   -> ( [OutStgArg]           -- Arguments representing the unboxed sum
+                      , Maybe (StgExpr -> StgExpr)) -- Transformation to apply to the arguments, to bring them
+                                                    -- into the right Rep
+unariseUbxSumOrTupleArgs rho us dc args ty_args
   | isUnboxedTupleDataCon dc
-  = Just (unariseConArgs rho args)
+  = (unariseConArgs rho args, Nothing)
 
   | isUnboxedSumDataCon dc
   , let args1 = assert (isSingleton args) (unariseConArgs rho args)
-  = Just (mkUbxSum dc ty_args args1)
+  = let (args2, cast_wrapper) = mkUbxSum dc ty_args args1 us
+    in (args2, Just cast_wrapper)
 
   | otherwise
-  = Nothing
+  = panic "unariseUbxSumOrTupleArgs: Constructor not a unboxed sum or tuple"
 
 -- Doesn't return void args.
 unariseRubbish_maybe :: Literal -> Maybe [OutStgArg]
@@ -472,15 +617,19 @@ elimCase rho args bndr (MultiValAlt _) [GenStgAlt{ alt_con   = _
                                                  , alt_bndrs = bndrs
                                                  , alt_rhs   = rhs}]
   = do let rho1 = extendRho rho bndr (MultiVal args)
-           rho2
+       (rho2, rhs') <- case () of
+           _
              | isUnboxedTupleBndr bndr
-             = mapTupleIdBinders bndrs args rho1
+             -> return (mapTupleIdBinders bndrs args rho1, rhs)
              | otherwise
-             = assert (isUnboxedSumBndr bndr) $
-               if null bndrs then rho1
-                             else mapSumIdBinders bndrs args rho1
+             -> assert (isUnboxedSumBndr bndr) $
+               case bndrs of
+                -- Sum with a void-type binder?
+                [] -> return (rho1, rhs)
+                [alt_bndr] -> mapSumIdBinders alt_bndr args rhs rho1
+                _ -> pprPanic "mapSumIdBinders" (ppr bndrs $$ ppr args)
 
-       unariseExpr rho2 rhs
+       unariseExpr rho2 rhs'
 
 elimCase rho args@(tag_arg : real_args) bndr (MultiValAlt _) alts
   | isUnboxedSumBndr bndr
@@ -570,18 +719,23 @@ unariseSumAlt :: UnariseEnv
 unariseSumAlt rho _ GenStgAlt{alt_con=DEFAULT,alt_bndrs=_,alt_rhs=e}
   = GenStgAlt DEFAULT mempty <$> unariseExpr rho e
 
-unariseSumAlt rho args GenStgAlt{ alt_con   = DataAlt sumCon
+unariseSumAlt rho args alt@GenStgAlt{ alt_con   = DataAlt sumCon
                                 , alt_bndrs = bs
                                 , alt_rhs   = e
                                 }
-  = do let rho'     = mapSumIdBinders bs args rho
-           lit_case = LitAlt (LitNumber LitNumInt (fromIntegral (dataConTag sumCon)))
-       GenStgAlt lit_case mempty <$> unariseExpr rho' e
+
+  = do (rho',e') <- case bs of
+              [b] -> mapSumIdBinders b args e rho
+              -- Sums must have one binder
+              _ -> pprPanic "unariseSumAlt2" (ppr args $$ pprPanicAlt alt)
+       let lit_case   = LitAlt (LitNumber LitNumInt (fromIntegral (dataConTag sumCon)))
+       GenStgAlt lit_case mempty <$> unariseExpr rho' e'
 
 unariseSumAlt _ scrt alt
-  = pprPanic "unariseSumAlt" (ppr scrt $$ pprPanicAlt alt)
+  = pprPanic "unariseSumAlt3" (ppr scrt $$ pprPanicAlt alt)
 
 --------------------------------------------------------------------------------
+-- Mapping binders when matching und a unboxed sum/tuple
 
 mapTupleIdBinders
   :: [InId]       -- Un-processed binders of a tuple alternative.
@@ -617,27 +771,90 @@ mapTupleIdBinders ids args0 rho0
       map_ids rho0 ids_unarised args0
 
 mapSumIdBinders
-  :: [InId]      -- Binder of a sum alternative (remember that sum patterns
-                 -- only have one binder, so this list should be a singleton)
+  :: InId        -- Binder (in the case alternative).
   -> [OutStgArg] -- Arguments that form the sum (NOT including the tag).
                  -- Can't have void args.
+  -> InStgExpr
   -> UnariseEnv
-  -> UnariseEnv
+  -> UniqSM (UnariseEnv, OutStgExpr)
 
-mapSumIdBinders [id] args rho0
-  = assert (not (any (isZeroBitTy . stgArgType) args)) $
+mapSumIdBinders alt_bndr args rhs rho0
+  = assert (not (any (isZeroBitTy . stgArgType) args)) $ do
+    uss <- listSplitUniqSupply <$> getUniqueSupplyM
     let
+      fld_reps = typePrimRep (idType alt_bndr)
+
+      -- Slots representing the whole sum
       arg_slots = map primRepSlot $ concatMap (typePrimRep . stgArgType) args
-      id_slots  = map primRepSlot $ typePrimRep (idType id)
+      -- The slots representing the field of the sum we bind.
+      id_slots  = map primRepSlot $ fld_reps
       layout1   = layoutUbxSum arg_slots id_slots
-    in
-      if isMultiValBndr id
-        then extendRho rho0 id (MultiVal [ args !! i | i <- layout1 ])
-        else assert (layout1 `lengthIs` 1)
-             extendRho rho0 id (UnaryVal (args !! head layout1))
 
-mapSumIdBinders ids sum_args _
-  = pprPanic "mapSumIdBinders" (ppr ids $$ ppr sum_args)
+      -- See Note [Casting slot arguments]
+      -- Most of the code here is just to make sure our binders are of the
+      -- right type.
+      -- Select only the args which contain parts of the current field.
+      id_arg_exprs   = [ args !! i | i <- layout1 ]
+      id_vars   = [v | StgVarArg v <- id_arg_exprs]
+      -- Output types for the field binders based on their rep
+      id_tys    = map primRepToType fld_reps
+
+      typed_id_arg_input = assert (equalLength id_vars id_tys) $
+                           zip3 id_vars id_tys uss
+
+      mkCastInput :: (Id,Type,UniqSupply) -> ([(PrimOp,Type,Unique)],Id,Id)
+      mkCastInput (id,tar_type,bndr_us) =
+        let (ops,types) = unzip $ getCasts (typePrimRep1 $ idType id) (typePrimRep1 tar_type)
+            cst_opts = zip3 ops types $ uniqsFromSupply bndr_us
+            out_id = case cst_opts of
+              [] -> id
+              _ ->  let (_,ty,uq) = last cst_opts
+                    in mkCastVar uq ty
+        in (cst_opts,id,out_id)
+
+      cast_inputs = map mkCastInput typed_id_arg_input
+      (rhs_with_casts,typed_ids) = mapAccumL cast_arg (\x->x) cast_inputs
+        where
+          cast_arg rhs_in (cast_ops,in_id,out_id) =
+            let rhs_out = castArgRename cast_ops (StgVarArg in_id)
+            in (rhs_in . rhs_out, out_id)
+
+      typed_id_args = map StgVarArg typed_ids
+
+      -- pprTrace "mapSumIdBinders"
+      --           (text "id_tys" <+> ppr id_tys $$
+      --           text "id_args" <+> ppr id_arg_exprs $$
+      --           text "rhs" <+> ppr rhs $$
+      --           text "rhs_with_casts" <+> ppr rhs_with_casts
+      --           ) $
+    if isMultiValBndr alt_bndr
+      then return (extendRho rho0 alt_bndr (MultiVal typed_id_args), rhs_with_casts rhs)
+      else assert (typed_id_args `lengthIs` 1) $
+            return (extendRho rho0 alt_bndr (UnaryVal (head typed_id_args)), rhs_with_casts rhs)
+
+-- Convert the argument to the given type, and wrap the conversion
+-- around the given expression. Use the given Id as a name for the
+-- converted value.
+castArgRename :: [(PrimOp,Type,Unique)] -> StgArg -> StgExpr -> StgExpr
+castArgRename ops in_arg rhs =
+  case ops of
+    [] -> rhs
+    ((op,ty,uq):rest_ops) ->
+      let out_id' = mkCastVar uq ty -- out_name `setIdUnique` uq `setIdType` ty
+          sub_cast = castArgRename rest_ops (StgVarArg out_id')
+      in mkCast in_arg op out_id' ty $ sub_cast rhs
+
+-- Construct a case binder used when casting sums, of a given type and unique.
+mkCastVar :: Unique -> Type -> Id
+mkCastVar uq ty = mkSysLocal (fsLit "cst_sum") uq ManyTy ty
+
+mkCast :: StgArg -> PrimOp -> OutId -> Type -> StgExpr -> StgExpr
+mkCast arg_in cast_op out_id out_ty in_rhs =
+  let r2 = typePrimRep1 out_ty
+      scrut = StgOpApp (StgPrimOp cast_op) [arg_in] out_ty
+      alt = GenStgAlt { alt_con = DEFAULT, alt_bndrs = [], alt_rhs = in_rhs}
+      alt_ty = PrimAlt r2
+  in (StgCase scrut out_id alt_ty [alt])
 
 -- | Build a unboxed sum term from arguments of an alternative.
 --
@@ -650,31 +867,72 @@ mapSumIdBinders ids sum_args _
 --   [ 1#, rubbish ]
 --
 mkUbxSum
-  :: DataCon      -- Sum data con
+  :: HasDebugCallStack
+  => DataCon      -- Sum data con
   -> [Type]       -- Type arguments of the sum data con
   -> [OutStgArg]  -- Actual arguments of the alternative.
-  -> [OutStgArg]  -- Final tuple arguments
-mkUbxSum dc ty_args args0
+  -> UniqSupply
+  -> ([OutStgArg] -- Final tuple arguments
+     ,(StgExpr->StgExpr) -- We might need to cast the args first
+     )
+mkUbxSum dc ty_args args0 us
   = let
       _ :| sum_slots = ubxSumRepType (map typePrimRep ty_args)
-        -- drop tag slot
-
+      -- drop tag slot
+      field_slots = (mapMaybe (typeSlotTy . stgArgType) args0)
       tag = dataConTag dc
+      layout'  = layoutUbxSum sum_slots field_slots
 
-      layout'  = layoutUbxSum sum_slots (mapMaybe (typeSlotTy . stgArgType) args0)
       tag_arg  = StgLitArg (LitNumber LitNumInt (fromIntegral tag))
       arg_idxs = IM.fromList (zipEqual "mkUbxSum" layout' args0)
 
-      mkTupArgs :: Int -> [SlotTy] -> IM.IntMap StgArg -> [StgArg]
-      mkTupArgs _ [] _
-        = []
-      mkTupArgs arg_idx (slot : slots_left) arg_map
-        | Just stg_arg <- IM.lookup arg_idx arg_map
-        = stg_arg : mkTupArgs (arg_idx + 1) slots_left arg_map
-        | otherwise
-        = ubxSumRubbishArg slot : mkTupArgs (arg_idx + 1) slots_left arg_map
+      ((_idx,_idx_map,_us,wrapper),slot_args)
+        = assert (length arg_idxs <= length sum_slots ) $
+          mapAccumL mkTupArg (0,arg_idxs,us,id) sum_slots
+
+      mkTupArg  :: (Int, IM.IntMap StgArg,UniqSupply,StgExpr->StgExpr)
+                -> SlotTy
+                -> ((Int,IM.IntMap StgArg,UniqSupply,StgExpr->StgExpr), StgArg)
+      mkTupArg (arg_idx, arg_map, us, wrapper) slot
+         | Just stg_arg <- IM.lookup arg_idx arg_map
+         =  case castArg us slot stg_arg of
+              -- Slot and arg type missmatched, do a cast
+              Just (casted_arg,us',wrapper') ->
+                ( (arg_idx+1, arg_map, us', wrapper . wrapper')
+                , casted_arg)
+              -- Use the arg as-is
+              Nothing ->
+                ( (arg_idx+1, arg_map, us, wrapper)
+                , stg_arg)
+         -- Garbage slot, fill with rubbish
+         | otherwise
+         =  ( (arg_idx+1, arg_map, us, wrapper)
+            , ubxSumRubbishArg slot)
+
+      castArg :: UniqSupply -> SlotTy -> StgArg -> Maybe (StgArg,UniqSupply,StgExpr -> StgExpr)
+      castArg us slot_ty arg
+        -- Cast the argument to the type of the slot if required
+        | slotPrimRep slot_ty /= typePrimRep1 (stgArgType arg)
+        , out_ty <- primRepToType $ slotPrimRep slot_ty
+        , (ops,types) <- unzip $ getCasts (typePrimRep1 $ stgArgType arg) $ typePrimRep1 out_ty
+        , not . null $ ops
+        = let (us1,us2) = splitUniqSupply us
+              cast_uqs = uniqsFromSupply us1
+              cast_opts = zip3 ops types cast_uqs
+              (_op,out_ty,out_uq) = last cast_opts
+              casts = castArgRename cast_opts arg :: StgExpr -> StgExpr
+          in Just (StgVarArg (mkCastVar out_uq out_ty),us2,casts)
+        -- No need for casting
+        | otherwise = Nothing
+
+      tup_args = tag_arg : slot_args
     in
-      tag_arg : mkTupArgs 0 sum_slots arg_idxs
+      -- pprTrace "mkUbxSum" (
+      --   text "ty_args (slots)" <+> ppr ty_args $$
+      --   text "args0" <+> ppr args0 $$
+      --   text "wrapper" <+>
+      --       (ppr $ wrapper $ StgLit $ LitChar '_'))
+      (tup_args, wrapper)
 
 
 -- | Return a rubbish value for the given slot type.
@@ -787,7 +1045,7 @@ unariseArgBinder is_con_arg rho x =
 -- | MultiVal a function argument. Never returns an empty list.
 unariseFunArg :: UnariseEnv -> StgArg -> [StgArg]
 unariseFunArg rho (StgVarArg x) =
-  case lookupVarEnv rho x of
+  case lookupRho rho x of
     Just (MultiVal [])  -> [voidArg]   -- NB: do not remove void args
     Just (MultiVal as)  -> as
     Just (UnaryVal arg) -> [arg]
@@ -809,7 +1067,7 @@ unariseFunArgBinder = unariseArgBinder False
 -- | MultiVal a DataCon argument. Returns an empty list when argument is void.
 unariseConArg :: UnariseEnv -> InStgArg -> [OutStgArg]
 unariseConArg rho (StgVarArg x) =
-  case lookupVarEnv rho x of
+  case lookupRho rho x of
     Just (UnaryVal arg) -> [arg]
     Just (MultiVal as) -> as      -- 'as' can be empty
     Nothing
diff --git a/compiler/GHC/Types/RepType.hs b/compiler/GHC/Types/RepType.hs
index a5e39f5c63..57e1c6099b 100644
--- a/compiler/GHC/Types/RepType.hs
+++ b/compiler/GHC/Types/RepType.hs
@@ -244,7 +244,8 @@ ubxSumRepType constrs0
     in
       sumRep
 
-layoutUbxSum :: SortedSlotTys -- Layout of sum. Does not include tag.
+layoutUbxSum :: HasDebugCallStack
+             => SortedSlotTys -- Layout of sum. Does not include tag.
                               -- We assume that they are in increasing order
              -> [SlotTy]      -- Slot types of things we want to map to locations in the
                               -- sum layout
@@ -267,7 +268,8 @@ layoutUbxSum sum_slots0 arg_slots0 =
       | otherwise
       = findSlot arg (slot_idx + 1) slots useds
     findSlot _ _ [] _
-      = pprPanic "findSlot" (text "Can't find slot" $$ ppr sum_slots0 $$ ppr arg_slots0)
+      = pprPanic "findSlot" (text "Can't find slot" $$ text "sum_slots:" <> ppr sum_slots0
+                                                    $$ text "arg_slots:" <> ppr arg_slots0 )
 
 --------------------------------------------------------------------------------
 
@@ -347,18 +349,17 @@ fitsIn ty1 ty2
   = Just ty1
   | isWordSlot ty1 && isWordSlot ty2
   = Just (max ty1 ty2)
-  | isFloatSlot ty1 && isFloatSlot ty2
-  = Just (max ty1 ty2)
   | otherwise
   = Nothing
+  -- We used to share slots between Float/Double but currently we can't easily
+  -- covert between float/double in a way that is both work free and safe.
+  -- So we put them in different slots.
+  -- See Note [Casting slot arguments]
   where
     isWordSlot Word64Slot = True
     isWordSlot WordSlot   = True
     isWordSlot _          = False
 
-    isFloatSlot DoubleSlot = True
-    isFloatSlot FloatSlot  = True
-    isFloatSlot _          = False
 
 
 {- **********************************************************************
diff --git a/compiler/GHC/Utils/Outputable.hs b/compiler/GHC/Utils/Outputable.hs
index 812edf15cd..74f77af445 100644
--- a/compiler/GHC/Utils/Outputable.hs
+++ b/compiler/GHC/Utils/Outputable.hs
@@ -414,6 +414,7 @@ data SDocContext = SDC
   , sdocSuppressUniques             :: !Bool
   , sdocSuppressModulePrefixes      :: !Bool
   , sdocSuppressStgExts             :: !Bool
+  , sdocSuppressStgReps             :: !Bool
   , sdocErrorSpans                  :: !Bool
   , sdocStarIsType                  :: !Bool
   , sdocLinearTypes                 :: !Bool
@@ -474,6 +475,7 @@ defaultSDocContext = SDC
   , sdocSuppressUniques             = False
   , sdocSuppressModulePrefixes      = False
   , sdocSuppressStgExts             = False
+  , sdocSuppressStgReps             = True
   , sdocErrorSpans                  = False
   , sdocStarIsType                  = False
   , sdocLinearTypes                 = False
diff --git a/compiler/ghc.cabal.in b/compiler/ghc.cabal.in
index a867801951..f3c4be57e3 100644
--- a/compiler/ghc.cabal.in
+++ b/compiler/ghc.cabal.in
@@ -168,6 +168,7 @@ Library
         GHC.Builtin.Names
         GHC.Builtin.Names.TH
         GHC.Builtin.PrimOps
+        GHC.Builtin.PrimOps.Casts
         GHC.Builtin.PrimOps.Ids
         GHC.Builtin.Types
         GHC.Builtin.Types.Literals
diff --git a/docs/users_guide/debugging.rst b/docs/users_guide/debugging.rst
index 83d093cd06..d4fe59b2d1 100644
--- a/docs/users_guide/debugging.rst
+++ b/docs/users_guide/debugging.rst
@@ -946,6 +946,16 @@ parts that you are not interested in.
 
     Suppress the printing of core size stats per binding
 
+.. ghc-flag:: -dsuppress-stg-reps
+    :shortdesc: Suppress rep annotations on STG args.
+    :type: dynamic
+
+    :since: 9.6.1
+
+    default: enabled
+
+    Disabling this will annoate certain stg arguments with their prim rep.
+
 
 .. _checking-consistency:
 
diff --git a/testsuite/driver/testlib.py b/testsuite/driver/testlib.py
index a0f97098ab..947d4c1f81 100644
--- a/testsuite/driver/testlib.py
+++ b/testsuite/driver/testlib.py
@@ -1465,7 +1465,8 @@ def compile_cmp_asm(name: TestName,
                     ext: str,
                     extra_hc_opts: str
                     ) -> PassFail:
-    print('Compile only, extra args = ', extra_hc_opts)
+    if extra_hc_opts:
+        print('Compile only, extra args = ', extra_hc_opts)
     result = simple_build(name + '.' + ext, way, '-keep-s-files -O ' + extra_hc_opts, False, None, [], False, False)
 
     if badResult(result):
@@ -1492,7 +1493,8 @@ def compile_grep_asm(name: TestName,
                      is_substring: bool,
                      extra_hc_opts: str
                      ) -> PassFail:
-    print('Compile only, extra args = ', extra_hc_opts)
+    if extra_hc_opts:
+        print('Compile and grep asm, extra args = ', extra_hc_opts)
     result = simple_build(name + '.' + ext, way, '-keep-s-files -O ' + extra_hc_opts, False, None, [], False, False)
 
     if badResult(result):
@@ -1513,7 +1515,8 @@ def compile_grep_core(name: TestName,
                       way: WayName,
                       extra_hc_opts: str
                       ) -> PassFail:
-    print('Compile only, extra args = ', extra_hc_opts)
+    if extra_hc_opts:
+        print('Compile only, extra args = ', extra_hc_opts)
     result = simple_build(name + '.hs', way, '-ddump-to-file -dsuppress-all -ddump-simpl -O ' + extra_hc_opts, False, None, [], False, False)
 
     if badResult(result):
diff --git a/testsuite/tests/unboxedsums/GenManyUbxSums.hs b/testsuite/tests/unboxedsums/GenManyUbxSums.hs
new file mode 100755
index 0000000000..5d38f10b5b
--- /dev/null
+++ b/testsuite/tests/unboxedsums/GenManyUbxSums.hs
@@ -0,0 +1,109 @@
+#!/usr/bin/env runghc
+{-# LANGUAGE MagicHash #-}
+{-# LANGUAGE UnboxedTuples #-}
+{-# LANGUAGE UnboxedSums #-}
+
+-- This little piece of code constructs a large set of functions
+-- constructing and deconstructing unboxed tuples of various types.
+module Main where
+
+import GHC.Exts
+import System.IO
+import Data.List (intersperse)
+inputs = ["Int", "Word"]
+sizes = ["","8","16","32","64"]
+
+-- ["Addr#","Int#","Int8#","Int16#","Int32#","Int64#","Word#","Word8#","Word16#","Word32#","Word64#"]
+types = "Addr#" : do
+    r <- inputs
+    s <- sizes
+    return $ r++s++"#"
+
+-- We eventually build two sums, one of type (# t1 | t2 #) and one of (# t1 | t3).
+-- So we build all possible combinations of three types here.
+combos = do
+    t1 <- types
+    t2 <- types
+    t3 <- types
+    return (t1,t2,t3)
+
+mkCon ty = case ty of
+    "Addr#" -> "Addr"
+    "Int#" -> "I#"
+    "Int8#" -> "I8#"
+    "Int16#" -> "I16#"
+    "Int32#" -> "I32#"
+    "Int64#" -> "I64#"
+    "Word#" -> "W#"
+    "Word8#" -> "W8#"
+    "Word16#" -> "W16#"
+    "Word32#" -> "W32#"
+    "Word64#" -> "W64#"
+
+-- Construct a function like the one below, varying the types in the sums based on the
+-- given type tuples.
+-- We need to NOINLINE or the function will be constant folded away.
+-- {-# NOINLINE fun0 #-}
+-- fun0 :: (# Addr# | I16# #) -> (# Addr# | I# #)
+-- fun0 x = case x of
+--   (# x1 | #) -> (# x1 | #) :: (# Addr# | I# #)
+mkFun n (t1,t2,t3) =
+    "{-# NOINLINE fun" ++ show n ++ " #-}\n" ++
+    "fun" ++ show n ++ " :: (# " ++ t1 ++" | " ++ t2 ++ " #) -> (# " ++ t1 ++" | " ++ t3 ++ " #)\n" ++
+    "fun" ++ show n ++ " x = case x of\n" ++
+    "  (# x1 | #) -> (# x1 | #) :: (# " ++ t1 ++ " | " ++ t3 ++ " #)"
+
+-- Generate functions for all the tuple combinations.
+mkFuns _ [] = ""
+mkFuns n (combo:combos) =
+    mkFun n combo ++ "\n" ++ mkFuns (n+1) combos
+
+-- generate a test that will put a value into a unboxed sum and then retrieve it later on.
+-- It generates code like the one below:
+-- test0 =
+--     let in_val = maxBound
+--         out_val = case in_val of I# x -> case fun0 (# x | #) of (# y | #) -> I# y
+--     in in_val == out_val
+mkTest n (t1,_,_)=
+    let test_name = "test" ++ show n
+        test_code = test_name ++ " =\n" ++
+                    "  let in_val = (maxBound)\n" ++
+                    "      out_val = case in_val of " ++ mkCon t1 ++ " x -> case fun" ++ show n ++ " (# x | #) of (# y | #) -> " ++ mkCon t1 ++ " y\n" ++
+                    "  in in_val == out_val"
+    in (test_code,test_name)
+
+-- Test all the tuples
+mkTests n combos =
+    let (defs, names) = unzip $ zipWith mkTest [0..] combos
+        assert_results = "\nassert_results = and [" ++ (concat $ intersperse "," names) ++ "]\n" :: String
+    in unlines defs ++ assert_results
+
+header =
+    "{-# LANGUAGE MagicHash #-}\n\
+    \{-# LANGUAGE UnboxedTuples #-}\n\
+    \{-# LANGUAGE UnboxedSums #-}\n\
+    \module Main where\n\
+    \import GHC.Exts\n\
+    \import GHC.Word\n\
+    \import GHC.Int\n\
+    \import ManyUbxSums_Addr\n"
+main = do
+    out <- openFile "ManyUbxSums.hs" WriteMode
+    hPutStrLn out header
+
+    let combo:_ = combos
+    -- putStrLn $ mkFun 1 combo
+    hPutStrLn out $ mkFuns 0 combos
+
+    hPutStrLn out $ mkTests 0 combos
+    hPutStrLn out "main = do"
+
+    hPutStrLn out $ "  putStrLn . show $ assert_results"
+
+    -- The snippet below would print all individual test results.
+    -- But for CI really just check if all results match the input
+    -- let runTest n =
+    --         hPutStrLn out $ "  putStrLn $ \"test" ++ show n ++ " \" ++ (show test" ++ show n ++ ")"
+    --         mapM runTest [0 .. length combos - 1]
+
+    hClose out
diff --git a/testsuite/tests/unboxedsums/ManyUbxSums.stdout b/testsuite/tests/unboxedsums/ManyUbxSums.stdout
new file mode 100644
index 0000000000..0ca95142bb
--- /dev/null
+++ b/testsuite/tests/unboxedsums/ManyUbxSums.stdout
@@ -0,0 +1 @@
+True
diff --git a/testsuite/tests/unboxedsums/ManyUbxSums_Addr.hs b/testsuite/tests/unboxedsums/ManyUbxSums_Addr.hs
new file mode 100644
index 0000000000..6e718168e8
--- /dev/null
+++ b/testsuite/tests/unboxedsums/ManyUbxSums_Addr.hs
@@ -0,0 +1,26 @@
+{-# LANGUAGE MagicHash #-}
+{-# LANGUAGE UnboxedTuples #-}
+{-# LANGUAGE UnboxedSums #-}
+
+{-# OPTIONS_GHC -Wno-missing-methods #-}
+
+module ManyUbxSums_Addr where
+
+import GHC.Exts
+-- import GHC.Word
+-- import GHC.Int
+--import GHC.Utils.Misc
+
+data Addr = Addr Addr#
+
+instance Eq Addr where
+    (Addr x) == (Addr y) = case (eqAddr# x y) of
+      1# -> True
+      0# -> False
+
+instance Num Addr where
+  fromInteger x = case fromIntegral x of I# x1 -> Addr (int2Addr# x1)
+
+instance Bounded Addr where
+  maxBound = fromIntegral (maxBound :: Word)
+  minBound = 0
\ No newline at end of file
diff --git a/testsuite/tests/unboxedsums/T22208.hs b/testsuite/tests/unboxedsums/T22208.hs
new file mode 100644
index 0000000000..cc85eafcef
--- /dev/null
+++ b/testsuite/tests/unboxedsums/T22208.hs
@@ -0,0 +1,41 @@
+{-# LANGUAGE MagicHash                 #-}
+{-# LANGUAGE UnboxedSums               #-}
+{-# LANGUAGE UnboxedTuples             #-}
+module M where
+
+import GHC.Base
+
+-- Reproducer from #22208
+foo :: (# Float# | Double# #) -> (# Float# | Float #)
+foo (# x | #) = (# x | #)
+bar :: (# Word# | Int64# #) -> (# Double# | Word# #)
+bar (# y | #) = let x = y in (# | x #)
+baz :: (# Word# | Word64# #) -> (# Word# | (##) #)
+baz (# x | #) = (# x | #)
+
+foo1 :: (# Float# | Double# #) -> (# Float# | Float #)
+foo1 (# x | #) = (# x | #)
+bar1 :: (# Word# | Int64# #) -> (# Double# | Word# #)
+bar1 (# y | #) = let x = y in (# | x #)
+baz1 :: (# Word# | Word64# #) -> (# Word# | (##) #)
+baz1 (# x | #) = (# x | #)
+
+-- i8 value from w64 slot
+baz2 :: (# Int8# | Word64# #) -> (# Int8# | (##) #)
+baz2 (# x | #) = (# x | #)
+
+-- w8 value from w64 slot
+baz3 :: (# Word8# | Word64# #) -> (# Word8# | (##) #)
+baz3 (# x | #) = (# x | #)
+
+-- w8 from w slot
+baz4 :: (# Word8# | Word# #) -> (# Word8# | (##) #)
+baz4 (# x | #) = (# x | #)
+
+-- w from w slot
+baz5 :: (# Word8# | Word# #) -> (# Word# | (##) #)
+baz5 (# | x #) = (# x | #)
+
+-- addr from w slot
+baz6 :: (# Addr# | Word# #) -> (# Addr# | (##) #)
+baz6 (# x | #) = (# x | #)
\ No newline at end of file
diff --git a/testsuite/tests/unboxedsums/all.T b/testsuite/tests/unboxedsums/all.T
index 0d887c60ed..dc078fc8e6 100644
--- a/testsuite/tests/unboxedsums/all.T
+++ b/testsuite/tests/unboxedsums/all.T
@@ -55,3 +55,12 @@ test('unpack_sums_9', normal, compile, [""])
 #      [only_ways(['normal']),
 #       extra_files([ "unboxedsums" + str(i) + ".hs" for i in range(1, 12) ])],
 #      makefile_test, [])
+test('T22208', normal, compile, ['-dstg-lint -dcmm-lint'])
+test('ManyUbxSums',
+     [ pre_cmd('{compiler} --run ./GenManyUbxSums.hs'),
+       extra_files(['GenManyUbxSums.hs', 'ManyUbxSums_Addr.hs']),
+     ],
+     multi_compile_and_run,
+     ['ManyUbxSums',
+        [('ManyUbxSums_Addr.hs','')]
+        , '-v0 -dstg-lint -dcmm-lint'])