{- (c) The University of Glasgow 2006 (c) The GRASP/AQUA Project, Glasgow University, 1992-1998 @Uniques@ are used to distinguish entities in the compiler (@Ids@, @Classes@, etc.) from each other. Thus, @Uniques@ are the basic comparison key in the compiler. If there is any single operation that needs to be fast, it is @Unique@ comparison. Unsurprisingly, there is quite a bit of huff-and-puff directed to that end. Some of the other hair in this code is to be able to use a ``splittable @UniqueSupply@'' if requested/possible (not standard Haskell). -} {-# LANGUAGE CPP, BangPatterns, MagicHash #-} module Unique ( -- * Main data types Unique, Uniquable(..), uNIQUE_BITS, -- ** Constructors, destructors and operations on 'Unique's hasKey, pprUniqueAlways, mkUniqueGrimily, -- Used in UniqSupply only! getKey, -- Used in Var, UniqFM, Name only! mkUnique, unpkUnique, -- Used in BinIface only eqUnique, ltUnique, deriveUnique, -- Ditto newTagUnique, -- Used in CgCase initTyVarUnique, initExitJoinUnique, nonDetCmpUnique, isValidKnownKeyUnique, -- Used in PrelInfo.knownKeyNamesOkay -- ** Making built-in uniques -- now all the built-in Uniques (and functions to make them) -- [the Oh-So-Wonderful Haskell module system wins again...] mkAlphaTyVarUnique, mkPrimOpIdUnique, mkPrimOpWrapperUnique, mkPreludeMiscIdUnique, mkPreludeDataConUnique, mkPreludeTyConUnique, mkPreludeClassUnique, mkCoVarUnique, mkVarOccUnique, mkDataOccUnique, mkTvOccUnique, mkTcOccUnique, mkRegSingleUnique, mkRegPairUnique, mkRegClassUnique, mkRegSubUnique, mkCostCentreUnique, mkBuiltinUnique, mkPseudoUniqueD, mkPseudoUniqueE, mkPseudoUniqueH, -- ** Deriving uniques -- *** From TyCon name uniques tyConRepNameUnique, -- *** From DataCon name uniques dataConWorkerUnique, dataConTyRepNameUnique ) where #include "HsVersions.h" #include "Unique.h" import GhcPrelude import BasicTypes import FastString import Outputable import Util -- just for implementing a fast [0,61) -> Char function import GHC.Exts (indexCharOffAddr#, Char(..), Int(..)) import Data.Char ( chr, ord ) import Data.Bits {- ************************************************************************ * * \subsection[Unique-type]{@Unique@ type and operations} * * ************************************************************************ The @Chars@ are ``tag letters'' that identify the @UniqueSupply@. Fast comparison is everything on @Uniques@: -} -- | Unique identifier. -- -- The type of unique identifiers that are used in many places in GHC -- for fast ordering and equality tests. You should generate these with -- the functions from the 'UniqSupply' module -- -- These are sometimes also referred to as \"keys\" in comments in GHC. newtype Unique = MkUnique Int {-# INLINE uNIQUE_BITS #-} uNIQUE_BITS :: Int uNIQUE_BITS = finiteBitSize (0 :: Int) - UNIQUE_TAG_BITS {- Now come the functions which construct uniques from their pieces, and vice versa. The stuff about unique *supplies* is handled further down this module. -} unpkUnique :: Unique -> (Char, Int) -- The reverse mkUniqueGrimily :: Int -> Unique -- A trap-door for UniqSupply getKey :: Unique -> Int -- for Var incrUnique :: Unique -> Unique stepUnique :: Unique -> Int -> Unique deriveUnique :: Unique -> Int -> Unique newTagUnique :: Unique -> Char -> Unique mkUniqueGrimily = MkUnique {-# INLINE getKey #-} getKey (MkUnique x) = x incrUnique (MkUnique i) = MkUnique (i + 1) stepUnique (MkUnique i) n = MkUnique (i + n) -- deriveUnique uses an 'X' tag so that it won't clash with -- any of the uniques produced any other way -- SPJ says: this looks terribly smelly to me! deriveUnique (MkUnique i) delta = mkUnique 'X' (i + delta) -- newTagUnique changes the "domain" of a unique to a different char newTagUnique u c = mkUnique c i where (_,i) = unpkUnique u -- | How many bits are devoted to the unique index (as opposed to the class -- character). uniqueMask :: Int uniqueMask = (1 `shiftL` uNIQUE_BITS) - 1 -- pop the Char in the top 8 bits of the Unique(Supply) -- No 64-bit bugs here, as long as we have at least 32 bits. --JSM -- and as long as the Char fits in 8 bits, which we assume anyway! mkUnique :: Char -> Int -> Unique -- Builds a unique from pieces -- NOT EXPORTED, so that we can see all the Chars that -- are used in this one module mkUnique c i = MkUnique (tag .|. bits) where tag = ord c `shiftL` uNIQUE_BITS bits = i .&. uniqueMask unpkUnique (MkUnique u) = let -- as long as the Char may have its eighth bit set, we -- really do need the logical right-shift here! tag = chr (u `shiftR` uNIQUE_BITS) i = u .&. uniqueMask in (tag, i) -- | The interface file symbol-table encoding assumes that known-key uniques fit -- in 30-bits; verify this. -- -- See Note [Symbol table representation of names] in BinIface for details. isValidKnownKeyUnique :: Unique -> Bool isValidKnownKeyUnique u = case unpkUnique u of (c, x) -> ord c < 0xff && x <= (1 `shiftL` 22) {- ************************************************************************ * * \subsection[Uniquable-class]{The @Uniquable@ class} * * ************************************************************************ -} -- | Class of things that we can obtain a 'Unique' from class Uniquable a where getUnique :: a -> Unique hasKey :: Uniquable a => a -> Unique -> Bool x `hasKey` k = getUnique x == k instance Uniquable FastString where getUnique fs = mkUniqueGrimily (uniqueOfFS fs) instance Uniquable Int where getUnique i = mkUniqueGrimily i {- ************************************************************************ * * \subsection[Unique-instances]{Instance declarations for @Unique@} * * ************************************************************************ And the whole point (besides uniqueness) is fast equality. We don't use `deriving' because we want {\em precise} control of ordering (equality on @Uniques@ is v common). -} -- Note [Unique Determinism] -- ~~~~~~~~~~~~~~~~~~~~~~~~~ -- The order of allocated @Uniques@ is not stable across rebuilds. -- The main reason for that is that typechecking interface files pulls -- @Uniques@ from @UniqSupply@ and the interface file for the module being -- currently compiled can, but doesn't have to exist. -- -- It gets more complicated if you take into account that the interface -- files are loaded lazily and that building multiple files at once has to -- work for any subset of interface files present. When you add parallelism -- this makes @Uniques@ hopelessly random. -- -- As such, to get deterministic builds, the order of the allocated -- @Uniques@ should not affect the final result. -- see also wiki/deterministic-builds -- -- Note [Unique Determinism and code generation] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- The goal of the deterministic builds (wiki/deterministic-builds, #4012) -- is to get ABI compatible binaries given the same inputs and environment. -- The motivation behind that is that if the ABI doesn't change the -- binaries can be safely reused. -- Note that this is weaker than bit-for-bit identical binaries and getting -- bit-for-bit identical binaries is not a goal for now. -- This means that we don't care about nondeterminism that happens after -- the interface files are created, in particular we don't care about -- register allocation and code generation. -- To track progress on bit-for-bit determinism see #12262. eqUnique :: Unique -> Unique -> Bool eqUnique (MkUnique u1) (MkUnique u2) = u1 == u2 ltUnique :: Unique -> Unique -> Bool ltUnique (MkUnique u1) (MkUnique u2) = u1 < u2 -- Provided here to make it explicit at the call-site that it can -- introduce non-determinism. -- See Note [Unique Determinism] -- See Note [No Ord for Unique] nonDetCmpUnique :: Unique -> Unique -> Ordering nonDetCmpUnique (MkUnique u1) (MkUnique u2) = if u1 == u2 then EQ else if u1 < u2 then LT else GT {- Note [No Ord for Unique] ~~~~~~~~~~~~~~~~~~~~~~~~~~ As explained in Note [Unique Determinism] the relative order of Uniques is nondeterministic. To prevent from accidental use the Ord Unique instance has been removed. This makes it easier to maintain deterministic builds, but comes with some drawbacks. The biggest drawback is that Maps keyed by Uniques can't directly be used. The alternatives are: 1) Use UniqFM or UniqDFM, see Note [Deterministic UniqFM] to decide which 2) Create a newtype wrapper based on Unique ordering where nondeterminism is controlled. See Module.ModuleEnv 3) Change the algorithm to use nonDetCmpUnique and document why it's still deterministic 4) Use TrieMap as done in CmmCommonBlockElim.groupByLabel -} instance Eq Unique where a == b = eqUnique a b a /= b = not (eqUnique a b) instance Uniquable Unique where getUnique u = u -- We do sometimes make strings with @Uniques@ in them: showUnique :: Unique -> String showUnique uniq = case unpkUnique uniq of (tag, u) -> finish_show tag u (iToBase62 u) finish_show :: Char -> Int -> String -> String finish_show 't' u _pp_u | u < 26 = -- Special case to make v common tyvars, t1, t2, ... -- come out as a, b, ... (shorter, easier to read) [chr (ord 'a' + u)] finish_show tag _ pp_u = tag : pp_u pprUniqueAlways :: Unique -> SDoc -- The "always" means regardless of -dsuppress-uniques -- It replaces the old pprUnique to remind callers that -- they should consider whether they want to consult -- Opt_SuppressUniques pprUniqueAlways u = text (showUnique u) instance Outputable Unique where ppr = pprUniqueAlways instance Show Unique where show uniq = showUnique uniq {- ************************************************************************ * * \subsection[Utils-base62]{Base-62 numbers} * * ************************************************************************ A character-stingy way to read/write numbers (notably Uniques). The ``62-its'' are \tr{[0-9a-zA-Z]}. We don't handle negative Ints. Code stolen from Lennart. -} iToBase62 :: Int -> String iToBase62 n_ = ASSERT(n_ >= 0) go n_ "" where go n cs | n < 62 = let !c = chooseChar62 n in c : cs | otherwise = go q (c : cs) where (!q, r) = quotRem n 62 !c = chooseChar62 r chooseChar62 :: Int -> Char {-# INLINE chooseChar62 #-} chooseChar62 (I# n) = C# (indexCharOffAddr# chars62 n) chars62 = "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"# {- ************************************************************************ * * \subsection[Uniques-prelude]{@Uniques@ for wired-in Prelude things} * * ************************************************************************ Allocation of unique supply characters: v,t,u : for renumbering value-, type- and usage- vars. B: builtin C-E: pseudo uniques (used in native-code generator) X: uniques derived by deriveUnique _: unifiable tyvars (above) 0-9: prelude things below (no numbers left any more..) :: (prelude) parallel array data constructors other a-z: lower case chars for unique supplies. Used so far: d desugarer f AbsC flattener g SimplStg k constraint tuple tycons m constraint tuple datacons n Native codegen r Hsc name cache s simplifier z anonymous sums -} mkAlphaTyVarUnique :: Int -> Unique mkPreludeClassUnique :: Int -> Unique mkPreludeTyConUnique :: Int -> Unique mkPreludeDataConUnique :: Arity -> Unique mkPrimOpIdUnique :: Int -> Unique -- See Note [Primop wrappers] in PrimOp.hs. mkPrimOpWrapperUnique :: Int -> Unique mkPreludeMiscIdUnique :: Int -> Unique mkCoVarUnique :: Int -> Unique mkAlphaTyVarUnique i = mkUnique '1' i mkCoVarUnique i = mkUnique 'g' i mkPreludeClassUnique i = mkUnique '2' i -------------------------------------------------- -- Wired-in type constructor keys occupy *two* slots: -- * u: the TyCon itself -- * u+1: the TyConRepName of the TyCon mkPreludeTyConUnique i = mkUnique '3' (2*i) tyConRepNameUnique :: Unique -> Unique tyConRepNameUnique u = incrUnique u -- Data constructor keys occupy *two* slots. The first is used for the -- data constructor itself and its wrapper function (the function that -- evaluates arguments as necessary and calls the worker). The second is -- used for the worker function (the function that builds the constructor -- representation). -------------------------------------------------- -- Wired-in data constructor keys occupy *three* slots: -- * u: the DataCon itself -- * u+1: its worker Id -- * u+2: the TyConRepName of the promoted TyCon -- Prelude data constructors are too simple to need wrappers. mkPreludeDataConUnique i = mkUnique '6' (3*i) -- Must be alphabetic -------------------------------------------------- dataConTyRepNameUnique, dataConWorkerUnique :: Unique -> Unique dataConWorkerUnique u = incrUnique u dataConTyRepNameUnique u = stepUnique u 2 -------------------------------------------------- mkPrimOpIdUnique op = mkUnique '9' (2*op) mkPrimOpWrapperUnique op = mkUnique '9' (2*op+1) mkPreludeMiscIdUnique i = mkUnique '0' i -- The "tyvar uniques" print specially nicely: a, b, c, etc. -- See pprUnique for details initTyVarUnique :: Unique initTyVarUnique = mkUnique 't' 0 mkPseudoUniqueD, mkPseudoUniqueE, mkPseudoUniqueH, mkBuiltinUnique :: Int -> Unique mkBuiltinUnique i = mkUnique 'B' i mkPseudoUniqueD i = mkUnique 'D' i -- used in NCG for getUnique on RealRegs mkPseudoUniqueE i = mkUnique 'E' i -- used in NCG spiller to create spill VirtualRegs mkPseudoUniqueH i = mkUnique 'H' i -- used in NCG spiller to create spill VirtualRegs mkRegSingleUnique, mkRegPairUnique, mkRegSubUnique, mkRegClassUnique :: Int -> Unique mkRegSingleUnique = mkUnique 'R' mkRegSubUnique = mkUnique 'S' mkRegPairUnique = mkUnique 'P' mkRegClassUnique = mkUnique 'L' mkCostCentreUnique :: Int -> Unique mkCostCentreUnique = mkUnique 'C' mkVarOccUnique, mkDataOccUnique, mkTvOccUnique, mkTcOccUnique :: FastString -> Unique -- See Note [The Unique of an OccName] in OccName mkVarOccUnique fs = mkUnique 'i' (uniqueOfFS fs) mkDataOccUnique fs = mkUnique 'd' (uniqueOfFS fs) mkTvOccUnique fs = mkUnique 'v' (uniqueOfFS fs) mkTcOccUnique fs = mkUnique 'c' (uniqueOfFS fs) initExitJoinUnique :: Unique initExitJoinUnique = mkUnique 's' 0