{-# LANGUAGE CPP, BangPatterns #-} ----------------------------------------------------------------------------- -- -- Stg to C-- code generation: expressions -- -- (c) The University of Glasgow 2004-2006 -- ----------------------------------------------------------------------------- module GHC.StgToCmm.Expr ( cgExpr ) where #include "HsVersions.h" import GhcPrelude hiding ((<*>)) import {-# SOURCE #-} GHC.StgToCmm.Bind ( cgBind ) import GHC.StgToCmm.Monad import GHC.StgToCmm.Heap import GHC.StgToCmm.Env import GHC.StgToCmm.DataCon import GHC.StgToCmm.Prof (saveCurrentCostCentre, restoreCurrentCostCentre, emitSetCCC) import GHC.StgToCmm.Layout import GHC.StgToCmm.Prim import GHC.StgToCmm.Hpc import GHC.StgToCmm.Ticky import GHC.StgToCmm.Utils import GHC.StgToCmm.Closure import StgSyn import MkGraph import BlockId import Cmm hiding ( succ ) import CmmInfo import CoreSyn import DataCon import DynFlags ( mAX_PTR_TAG ) import ForeignCall import Id import PrimOp import TyCon import Type ( isUnliftedType ) import RepType ( isVoidTy, countConRepArgs ) import CostCentre ( CostCentreStack, currentCCS ) import Maybes import Util import FastString import Outputable import Control.Monad ( unless, void ) import Control.Arrow ( first ) import Data.List ( partition ) ------------------------------------------------------------------------ -- cgExpr: the main function ------------------------------------------------------------------------ cgExpr :: CgStgExpr -> FCode ReturnKind cgExpr (StgApp fun args) = cgIdApp fun args -- seq# a s ==> a -- See Note [seq# magic] in PrelRules cgExpr (StgOpApp (StgPrimOp SeqOp) [StgVarArg a, _] _res_ty) = cgIdApp a [] -- dataToTag# :: a -> Int# -- See Note [dataToTag#] in primops.txt.pp cgExpr (StgOpApp (StgPrimOp DataToTagOp) [StgVarArg a] _res_ty) = do dflags <- getDynFlags emitComment (mkFastString "dataToTag#") tmp <- newTemp (bWord dflags) _ <- withSequel (AssignTo [tmp] False) (cgIdApp a []) -- TODO: For small types look at the tag bits instead of reading info table emitReturn [getConstrTag dflags (cmmUntag dflags (CmmReg (CmmLocal tmp)))] cgExpr (StgOpApp op args ty) = cgOpApp op args ty cgExpr (StgConApp con args _)= cgConApp con args cgExpr (StgTick t e) = cgTick t >> cgExpr e cgExpr (StgLit lit) = do cmm_lit <- cgLit lit emitReturn [CmmLit cmm_lit] cgExpr (StgLet _ binds expr) = do { cgBind binds; cgExpr expr } cgExpr (StgLetNoEscape _ binds expr) = do { u <- newUnique ; let join_id = mkBlockId u ; cgLneBinds join_id binds ; r <- cgExpr expr ; emitLabel join_id ; return r } cgExpr (StgCase expr bndr alt_type alts) = cgCase expr bndr alt_type alts cgExpr (StgLam {}) = panic "cgExpr: StgLam" ------------------------------------------------------------------------ -- Let no escape ------------------------------------------------------------------------ {- Generating code for a let-no-escape binding, aka join point is very very similar to what we do for a case expression. The duality is between let-no-escape x = b in e and case e of ... -> b That is, the RHS of 'x' (ie 'b') will execute *later*, just like the alternative of the case; it needs to be compiled in an environment in which all volatile bindings are forgotten, and the free vars are bound only to stable things like stack locations.. The 'e' part will execute *next*, just like the scrutinee of a case. -} ------------------------- cgLneBinds :: BlockId -> CgStgBinding -> FCode () cgLneBinds join_id (StgNonRec bndr rhs) = do { local_cc <- saveCurrentCostCentre -- See Note [Saving the current cost centre] ; (info, fcode) <- cgLetNoEscapeRhs join_id local_cc bndr rhs ; fcode ; addBindC info } cgLneBinds join_id (StgRec pairs) = do { local_cc <- saveCurrentCostCentre ; r <- sequence $ unzipWith (cgLetNoEscapeRhs join_id local_cc) pairs ; let (infos, fcodes) = unzip r ; addBindsC infos ; sequence_ fcodes } ------------------------- cgLetNoEscapeRhs :: BlockId -- join point for successor of let-no-escape -> Maybe LocalReg -- Saved cost centre -> Id -> CgStgRhs -> FCode (CgIdInfo, FCode ()) cgLetNoEscapeRhs join_id local_cc bndr rhs = do { (info, rhs_code) <- cgLetNoEscapeRhsBody local_cc bndr rhs ; let (bid, _) = expectJust "cgLetNoEscapeRhs" $ maybeLetNoEscape info ; let code = do { (_, body) <- getCodeScoped rhs_code ; emitOutOfLine bid (first (<*> mkBranch join_id) body) } ; return (info, code) } cgLetNoEscapeRhsBody :: Maybe LocalReg -- Saved cost centre -> Id -> CgStgRhs -> FCode (CgIdInfo, FCode ()) cgLetNoEscapeRhsBody local_cc bndr (StgRhsClosure _ cc _upd args body) = cgLetNoEscapeClosure bndr local_cc cc (nonVoidIds args) body cgLetNoEscapeRhsBody local_cc bndr (StgRhsCon cc con args) = cgLetNoEscapeClosure bndr local_cc cc [] (StgConApp con args (pprPanic "cgLetNoEscapeRhsBody" $ text "StgRhsCon doesn't have type args")) -- For a constructor RHS we want to generate a single chunk of -- code which can be jumped to from many places, which will -- return the constructor. It's easy; just behave as if it -- was an StgRhsClosure with a ConApp inside! ------------------------- cgLetNoEscapeClosure :: Id -- binder -> Maybe LocalReg -- Slot for saved current cost centre -> CostCentreStack -- XXX: *** NOT USED *** why not? -> [NonVoid Id] -- Args (as in \ args -> body) -> CgStgExpr -- Body (as in above) -> FCode (CgIdInfo, FCode ()) cgLetNoEscapeClosure bndr cc_slot _unused_cc args body = do dflags <- getDynFlags return ( lneIdInfo dflags bndr args , code ) where code = forkLneBody $ do { ; withNewTickyCounterLNE (idName bndr) args $ do ; restoreCurrentCostCentre cc_slot ; arg_regs <- bindArgsToRegs args ; void $ noEscapeHeapCheck arg_regs (tickyEnterLNE >> cgExpr body) } ------------------------------------------------------------------------ -- Case expressions ------------------------------------------------------------------------ {- Note [Compiling case expressions] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ It is quite interesting to decide whether to put a heap-check at the start of each alternative. Of course we certainly have to do so if the case forces an evaluation, or if there is a primitive op which can trigger GC. A more interesting situation is this (a Plan-B situation) !P!; ...P... case x# of 0# -> !Q!; ...Q... default -> !R!; ...R... where !x! indicates a possible heap-check point. The heap checks in the alternatives *can* be omitted, in which case the topmost heapcheck will take their worst case into account. In favour of omitting !Q!, !R!: - *May* save a heap overflow test, if ...P... allocates anything. - We can use relative addressing from a single Hp to get at all the closures so allocated. - No need to save volatile vars etc across heap checks in !Q!, !R! Against omitting !Q!, !R! - May put a heap-check into the inner loop. Suppose the main loop is P -> R -> P -> R... Q is the loop exit, and only it does allocation. This only hurts us if P does no allocation. If P allocates, then there is a heap check in the inner loop anyway. - May do more allocation than reqd. This sometimes bites us badly. For example, nfib (ha!) allocates about 30\% more space if the worst-casing is done, because many many calls to nfib are leaf calls which don't need to allocate anything. We can un-allocate, but that costs an instruction Neither problem hurts us if there is only one alternative. Suppose the inner loop is P->R->P->R etc. Then here is how many heap checks we get in the *inner loop* under various conditions Alloc Heap check in branches (!Q!, !R!)? P Q R yes no (absorb to !P!) -------------------------------------- n n n 0 0 n y n 0 1 n . y 1 1 y . y 2 1 y . n 1 1 Best choices: absorb heap checks from Q and R into !P! iff a) P itself does some allocation or b) P does allocation, or there is exactly one alternative We adopt (b) because that is more likely to put the heap check at the entry to a function, when not many things are live. After a bunch of single-branch cases, we may have lots of things live Hence: two basic plans for case e of r { alts } ------ Plan A: the general case --------- ...save current cost centre... ...code for e, with sequel (SetLocals r) ...restore current cost centre... ...code for alts... ...alts do their own heap checks ------ Plan B: special case when --------- (i) e does not allocate or call GC (ii) either upstream code performs allocation or there is just one alternative Then heap allocation in the (single) case branch is absorbed by the upstream check. Very common example: primops on unboxed values ...code for e, with sequel (SetLocals r)... ...code for alts... ...no heap check... -} ------------------------------------- data GcPlan = GcInAlts -- Put a GC check at the start the case alternatives, [LocalReg] -- which binds these registers | NoGcInAlts -- The scrutinee is a primitive value, or a call to a -- primitive op which does no GC. Absorb the allocation -- of the case alternative(s) into the upstream check ------------------------------------- cgCase :: CgStgExpr -> Id -> AltType -> [CgStgAlt] -> FCode ReturnKind cgCase (StgOpApp (StgPrimOp op) args _) bndr (AlgAlt tycon) alts | isEnumerationTyCon tycon -- Note [case on bool] = do { tag_expr <- do_enum_primop op args -- If the binder is not dead, convert the tag to a constructor -- and assign it. See Note [Dead-binder optimisation] ; unless (isDeadBinder bndr) $ do { dflags <- getDynFlags ; tmp_reg <- bindArgToReg (NonVoid bndr) ; emitAssign (CmmLocal tmp_reg) (tagToClosure dflags tycon tag_expr) } ; (mb_deflt, branches) <- cgAlgAltRhss (NoGcInAlts,AssignedDirectly) (NonVoid bndr) alts -- See Note [GC for conditionals] ; emitSwitch tag_expr branches mb_deflt 0 (tyConFamilySize tycon - 1) ; return AssignedDirectly } where do_enum_primop :: PrimOp -> [StgArg] -> FCode CmmExpr do_enum_primop TagToEnumOp [arg] -- No code! = getArgAmode (NonVoid arg) do_enum_primop primop args = do dflags <- getDynFlags tmp <- newTemp (bWord dflags) cgPrimOp [tmp] primop args return (CmmReg (CmmLocal tmp)) {- Note [case on bool] ~~~~~~~~~~~~~~~~~~~ This special case handles code like case a <# b of True -> False -> --> case tagToEnum# (a <$# b) of True -> .. ; False -> ... --> case (a <$# b) of r -> case tagToEnum# r of True -> .. ; False -> ... If we let the ordinary case code handle it, we'll get something like tmp1 = a < b tmp2 = Bool_closure_tbl[tmp1] if (tmp2 & 7 != 0) then ... // normal tagged case but this junk won't optimise away. What we really want is just an inline comparison: if (a < b) then ... So we add a special case to generate tmp1 = a < b if (tmp1 == 0) then ... and later optimisations will further improve this. Now that #6135 has been resolved it should be possible to remove that special case. The idea behind this special case and pre-6135 implementation of Bool-returning primops was that tagToEnum# was added implicitly in the codegen and then optimized away. Now the call to tagToEnum# is explicit in the source code, which allows to optimize it away at the earlier stages of compilation (i.e. at the Core level). Note [Scrutinising VoidRep] ~~~~~~~~~~~~~~~~~~~~~~~~~~~ Suppose we have this STG code: f = \[s : State# RealWorld] -> case s of _ -> blah This is very odd. Why are we scrutinising a state token? But it can arise with bizarre NOINLINE pragmas (#9964) crash :: IO () crash = IO (\s -> let {-# NOINLINE s' #-} s' = s in (# s', () #)) Now the trouble is that 's' has VoidRep, and we do not bind void arguments in the environment; they don't live anywhere. See the calls to nonVoidIds in various places. So we must not look up 's' in the environment. Instead, just evaluate the RHS! Simple. Note [Dead-binder optimisation] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ A case-binder, or data-constructor argument, may be marked as dead, because we preserve occurrence-info on binders in CoreTidy (see CoreTidy.tidyIdBndr). If the binder is dead, we can sometimes eliminate a load. While CmmSink will eliminate that load, it's very easy to kill it at source (giving CmmSink less work to do), and in any case CmmSink only runs with -O. Since the majority of case binders are dead, this optimisation probably still has a great benefit-cost ratio and we want to keep it for -O0. See also Phab:D5358. This probably also was the reason for occurrence hack in Phab:D5339 to exist, perhaps because the occurrence information preserved by 'CoreTidy.tidyIdBndr' was insufficient. But now that CmmSink does the job we deleted the hacks. -} cgCase (StgApp v []) _ (PrimAlt _) alts | isVoidRep (idPrimRep v) -- See Note [Scrutinising VoidRep] , [(DEFAULT, _, rhs)] <- alts = cgExpr rhs {- Note [Dodgy unsafeCoerce 1] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Consider case (x :: HValue) |> co of (y :: MutVar# Int) DEFAULT -> ... We want to generate an assignment y := x We want to allow this assignment to be generated in the case when the types are compatible, because this allows some slightly-dodgy but occasionally-useful casts to be used, such as in RtClosureInspect where we cast an HValue to a MutVar# so we can print out the contents of the MutVar#. If instead we generate code that enters the HValue, then we'll get a runtime panic, because the HValue really is a MutVar#. The types are compatible though, so we can just generate an assignment. -} cgCase (StgApp v []) bndr alt_type@(PrimAlt _) alts | isUnliftedType (idType v) -- Note [Dodgy unsafeCoerce 1] = -- assignment suffices for unlifted types do { dflags <- getDynFlags ; unless (reps_compatible dflags) $ pprPanic "cgCase: reps do not match, perhaps a dodgy unsafeCoerce?" (pp_bndr v $$ pp_bndr bndr) ; v_info <- getCgIdInfo v ; emitAssign (CmmLocal (idToReg dflags (NonVoid bndr))) (idInfoToAmode v_info) -- Add bndr to the environment ; _ <- bindArgToReg (NonVoid bndr) ; cgAlts (NoGcInAlts,AssignedDirectly) (NonVoid bndr) alt_type alts } where reps_compatible dflags = primRepCompatible dflags (idPrimRep v) (idPrimRep bndr) pp_bndr id = ppr id <+> dcolon <+> ppr (idType id) <+> parens (ppr (idPrimRep id)) {- Note [Dodgy unsafeCoerce 2, #3132] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ In all other cases of a lifted Id being cast to an unlifted type, the Id should be bound to bottom, otherwise this is an unsafe use of unsafeCoerce. We can generate code to enter the Id and assume that it will never return. Hence, we emit the usual enter/return code, and because bottom must be untagged, it will be entered. The Sequel is a type-correct assignment, albeit bogus. The (dead) continuation loops; it would be better to invoke some kind of panic function here. -} cgCase scrut@(StgApp v []) _ (PrimAlt _) _ = do { dflags <- getDynFlags ; mb_cc <- maybeSaveCostCentre True ; _ <- withSequel (AssignTo [idToReg dflags (NonVoid v)] False) (cgExpr scrut) ; restoreCurrentCostCentre mb_cc ; emitComment $ mkFastString "should be unreachable code" ; l <- newBlockId ; emitLabel l ; emit (mkBranch l) -- an infinite loop ; return AssignedDirectly } {- Note [Handle seq#] ~~~~~~~~~~~~~~~~~~~~~ See Note [seq# magic] in PrelRules. The special case for seq# in cgCase does this: case seq# a s of v (# s', a' #) -> e ==> case a of v (# s', a' #) -> e (taking advantage of the fact that the return convention for (# State#, a #) is the same as the return convention for just 'a') -} cgCase (StgOpApp (StgPrimOp SeqOp) [StgVarArg a, _] _) bndr alt_type alts = -- Note [Handle seq#] -- And see Note [seq# magic] in PrelRules -- Use the same return convention as vanilla 'a'. cgCase (StgApp a []) bndr alt_type alts cgCase scrut bndr alt_type alts = -- the general case do { dflags <- getDynFlags ; up_hp_usg <- getVirtHp -- Upstream heap usage ; let ret_bndrs = chooseReturnBndrs bndr alt_type alts alt_regs = map (idToReg dflags) ret_bndrs ; simple_scrut <- isSimpleScrut scrut alt_type ; let do_gc | is_cmp_op scrut = False -- See Note [GC for conditionals] | not simple_scrut = True | isSingleton alts = False | up_hp_usg > 0 = False | otherwise = True -- cf Note [Compiling case expressions] gc_plan = if do_gc then GcInAlts alt_regs else NoGcInAlts ; mb_cc <- maybeSaveCostCentre simple_scrut ; let sequel = AssignTo alt_regs do_gc{- Note [scrut sequel] -} ; ret_kind <- withSequel sequel (cgExpr scrut) ; restoreCurrentCostCentre mb_cc ; _ <- bindArgsToRegs ret_bndrs ; cgAlts (gc_plan,ret_kind) (NonVoid bndr) alt_type alts } where is_cmp_op (StgOpApp (StgPrimOp op) _ _) = isComparisonPrimOp op is_cmp_op _ = False {- Note [GC for conditionals] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For boolean conditionals it seems that we have always done NoGcInAlts. That is, we have always done the GC check before the conditional. This is enshrined in the special case for case tagToEnum# (a>b) of ... See Note [case on bool] It's odd, and it's flagrantly inconsistent with the rules described Note [Compiling case expressions]. However, after eliminating the tagToEnum# (#13397) we will have: case (a>b) of ... Rather than make it behave quite differently, I am testing for a comparison operator here in in the general case as well. ToDo: figure out what the Right Rule should be. Note [scrut sequel] ~~~~~~~~~~~~~~~~~~~ The job of the scrutinee is to assign its value(s) to alt_regs. Additionally, if we plan to do a heap-check in the alternatives (see Note [Compiling case expressions]), then we *must* retreat Hp to recover any unused heap before passing control to the sequel. If we don't do this, then any unused heap will become slop because the heap check will reset the heap usage. Slop in the heap breaks LDV profiling (+RTS -hb) which needs to do a linear sweep through the nursery. Note [Inlining out-of-line primops and heap checks] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ If shouldInlinePrimOp returns True when called from GHC.StgToCmm.Expr for the purpose of heap check placement, we *must* inline the primop later in GHC.StgToCmm.Prim. If we don't things will go wrong. -} ----------------- maybeSaveCostCentre :: Bool -> FCode (Maybe LocalReg) maybeSaveCostCentre simple_scrut | simple_scrut = return Nothing | otherwise = saveCurrentCostCentre ----------------- isSimpleScrut :: CgStgExpr -> AltType -> FCode Bool -- Simple scrutinee, does not block or allocate; hence safe to amalgamate -- heap usage from alternatives into the stuff before the case -- NB: if you get this wrong, and claim that the expression doesn't allocate -- when it does, you'll deeply mess up allocation isSimpleScrut (StgOpApp op args _) _ = isSimpleOp op args isSimpleScrut (StgLit _) _ = return True -- case 1# of { 0# -> ..; ... } isSimpleScrut (StgApp _ []) (PrimAlt _) = return True -- case x# of { 0# -> ..; ... } isSimpleScrut _ _ = return False isSimpleOp :: StgOp -> [StgArg] -> FCode Bool -- True iff the op cannot block or allocate isSimpleOp (StgFCallOp (CCall (CCallSpec _ _ safe _ _)) _) _ = return $! not (playSafe safe) -- dataToTag# evalautes its argument, see Note [dataToTag#] in primops.txt.pp isSimpleOp (StgPrimOp DataToTagOp) _ = return False isSimpleOp (StgPrimOp op) stg_args = do arg_exprs <- getNonVoidArgAmodes stg_args dflags <- getDynFlags -- See Note [Inlining out-of-line primops and heap checks] return $! shouldInlinePrimOp dflags op arg_exprs isSimpleOp (StgPrimCallOp _) _ = return False ----------------- chooseReturnBndrs :: Id -> AltType -> [CgStgAlt] -> [NonVoid Id] -- These are the binders of a case that are assigned by the evaluation of the -- scrutinee. -- They're non-void, see Note [Post-unarisation invariants] in UnariseStg. chooseReturnBndrs bndr (PrimAlt _) _alts = assertNonVoidIds [bndr] chooseReturnBndrs _bndr (MultiValAlt n) [(_, ids, _)] = ASSERT2(ids `lengthIs` n, ppr n $$ ppr ids $$ ppr _bndr) assertNonVoidIds ids -- 'bndr' is not assigned! chooseReturnBndrs bndr (AlgAlt _) _alts = assertNonVoidIds [bndr] -- Only 'bndr' is assigned chooseReturnBndrs bndr PolyAlt _alts = assertNonVoidIds [bndr] -- Only 'bndr' is assigned chooseReturnBndrs _ _ _ = panic "chooseReturnBndrs" -- MultiValAlt has only one alternative ------------------------------------- cgAlts :: (GcPlan,ReturnKind) -> NonVoid Id -> AltType -> [CgStgAlt] -> FCode ReturnKind -- At this point the result of the case are in the binders cgAlts gc_plan _bndr PolyAlt [(_, _, rhs)] = maybeAltHeapCheck gc_plan (cgExpr rhs) cgAlts gc_plan _bndr (MultiValAlt _) [(_, _, rhs)] = maybeAltHeapCheck gc_plan (cgExpr rhs) -- Here bndrs are *already* in scope, so don't rebind them cgAlts gc_plan bndr (PrimAlt _) alts = do { dflags <- getDynFlags ; tagged_cmms <- cgAltRhss gc_plan bndr alts ; let bndr_reg = CmmLocal (idToReg dflags bndr) (DEFAULT,deflt) = head tagged_cmms -- PrimAlts always have a DEFAULT case -- and it always comes first tagged_cmms' = [(lit,code) | (LitAlt lit, code) <- tagged_cmms] ; emitCmmLitSwitch (CmmReg bndr_reg) tagged_cmms' deflt ; return AssignedDirectly } cgAlts gc_plan bndr (AlgAlt tycon) alts = do { dflags <- getDynFlags ; (mb_deflt, branches) <- cgAlgAltRhss gc_plan bndr alts ; let !fam_sz = tyConFamilySize tycon !bndr_reg = CmmLocal (idToReg dflags bndr) !ptag_expr = cmmConstrTag1 dflags (CmmReg bndr_reg) !branches' = first succ <$> branches !maxpt = mAX_PTR_TAG dflags (!via_ptr, !via_info) = partition ((< maxpt) . fst) branches' !small = isSmallFamily dflags fam_sz -- Is the constructor tag in the node reg? -- See Note [Tagging big families] ; if small || null via_info then -- Yes, bndr_reg has constructor tag in ls bits emitSwitch ptag_expr branches' mb_deflt 1 (if small then fam_sz else maxpt) else -- No, the get exact tag from info table when mAX_PTR_TAG -- See Note [Double switching for big families] do let !untagged_ptr = cmmUntag dflags (CmmReg bndr_reg) !itag_expr = getConstrTag dflags untagged_ptr !info0 = first pred <$> via_info if null via_ptr then emitSwitch itag_expr info0 mb_deflt 0 (fam_sz - 1) else do infos_lbl <- newBlockId infos_scp <- getTickScope let spillover = (maxpt, (mkBranch infos_lbl, infos_scp)) (mb_shared_deflt, mb_shared_branch) <- case mb_deflt of (Just (stmts, scp)) -> do lbl <- newBlockId return ( Just (mkLabel lbl scp <*> stmts, scp) , Just (mkBranch lbl, scp)) _ -> return (Nothing, Nothing) -- Switch on pointer tag emitSwitch ptag_expr (spillover : via_ptr) mb_shared_deflt 1 maxpt join_lbl <- newBlockId emit (mkBranch join_lbl) -- Switch on info table tag emitLabel infos_lbl emitSwitch itag_expr info0 mb_shared_branch (maxpt - 1) (fam_sz - 1) emitLabel join_lbl ; return AssignedDirectly } cgAlts _ _ _ _ = panic "cgAlts" -- UbxTupAlt and PolyAlt have only one alternative -- Note [Double switching for big families] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- -- An algebraic data type can have a n >= 0 summands -- (or alternatives), which are identified (labeled) by -- constructors. In memory they are kept apart by tags -- (see Note [Data constructor dynamic tags] in GHC.StgToCmm.Closure). -- Due to the characteristics of the platform that -- contribute to the alignment of memory objects, there -- is a natural limit of information about constructors -- that can be encoded in the pointer tag. When the mapping -- of constructors to the pointer tag range 1..mAX_PTR_TAG -- is not injective, then we have a "big data type", also -- called a "big (constructor) family" in the literature. -- Constructor tags residing in the info table are injective, -- but considerably more expensive to obtain, due to additional -- memory access(es). -- -- When doing case analysis on a value of a "big data type" -- we need two nested switch statements to make up for the lack -- of injectivity of pointer tagging, also taking the info -- table tag into account. The exact mechanism is described next. -- -- In the general case, switching on big family alternatives -- is done by two nested switch statements. According to -- Note [Tagging big families], the outer switch -- looks at the pointer tag and the inner dereferences the -- pointer and switches on the info table tag. -- -- We can handle a simple case first, namely when none -- of the case alternatives mention a constructor having -- a pointer tag of 1..mAX_PTR_TAG-1. In this case we -- simply emit a switch on the info table tag. -- Note that the other simple case is when all mentioned -- alternatives lie in 1..mAX_PTR_TAG-1, in which case we can -- switch on the ptr tag only, just like in the small family case. -- -- There is a single intricacy with a nested switch: -- Both should branch to the same default alternative, and as such -- avoid duplicate codegen of potentially heavy code. The outer -- switch generates the actual code with a prepended fresh label, -- while the inner one only generates a jump to that label. -- -- For example, let's assume a 64-bit architecture, so that all -- heap objects are 8-byte aligned, and hence the address of a -- heap object ends in `000` (three zero bits). -- -- Then consider the following data type -- -- > data Big = T0 | T1 | T2 | T3 | T4 | T5 | T6 | T7 | T8 -- Ptr tag: 1 2 3 4 5 6 7 7 7 -- As bits: 001 010 011 100 101 110 111 111 111 -- Info pointer tag (zero based): -- 0 1 2 3 4 5 6 7 8 -- -- Then \case T2 -> True; T8 -> True; _ -> False -- will result in following code (slightly cleaned-up and -- commented -ddump-cmm-from-stg): {- R1 = _sqI::P64; -- scrutinee if (R1 & 7 != 0) goto cqO; else goto cqP; cqP: // global -- enter call (I64[R1])(R1) returns to cqO, args: 8, res: 8, upd: 8; cqO: // global -- already WHNF _sqJ::P64 = R1; _cqX::P64 = _sqJ::P64 & 7; -- extract pointer tag switch [1 .. 7] _cqX::P64 { case 3 : goto cqW; case 7 : goto cqR; default: {goto cqS;} } cqR: // global _cr2 = I32[I64[_sqJ::P64 & (-8)] - 4]; -- tag from info pointer switch [6 .. 8] _cr2::I64 { case 8 : goto cr1; default: {goto cr0;} } cr1: // global R1 = GHC.Types.True_closure+2; call (P64[(old + 8)])(R1) args: 8, res: 0, upd: 8; cr0: // global -- technically necessary label goto cqS; cqW: // global R1 = GHC.Types.True_closure+2; call (P64[(old + 8)])(R1) args: 8, res: 0, upd: 8; cqS: // global R1 = GHC.Types.False_closure+1; call (P64[(old + 8)])(R1) args: 8, res: 0, upd: 8; -} -- -- For 32-bit systems we only have 2 tag bits in the pointers at our disposal, -- so the performance win is dubious, especially in face of the increased code -- size due to double switching. But we can take the viewpoint that 32-bit -- architectures are not relevant for performance any more, so this can be -- considered as moot. -- Note [alg-alt heap check] -- -- In an algebraic case with more than one alternative, we will have -- code like -- -- L0: -- x = R1 -- goto L1 -- L1: -- if (x & 7 >= 2) then goto L2 else goto L3 -- L2: -- Hp = Hp + 16 -- if (Hp > HpLim) then goto L4 -- ... -- L4: -- call gc() returns to L5 -- L5: -- x = R1 -- goto L1 -- Note [Tagging big families] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- -- Both the big and the small constructor families are tagged, -- that is, greater unions which overflow the tag space of TAG_BITS -- (i.e. 3 on 32 resp. 7 constructors on 64 bit archs). -- -- For example, let's assume a 64-bit architecture, so that all -- heap objects are 8-byte aligned, and hence the address of a -- heap object ends in `000` (three zero bits). Then consider -- > data Maybe a = Nothing | Just a -- > data Day a = Mon | Tue | Wed | Thu | Fri | Sat | Sun -- > data Grade = G1 | G2 | G3 | G4 | G5 | G6 | G7 | G8 | G9 | G10 -- -- Since `Grade` has more than 7 constructors, it counts as a -- "big data type" (also referred to as "big constructor family" in papers). -- On the other hand, `Maybe` and `Day` have 7 constructors or fewer, so they -- are "small data types". -- -- Then -- * A pointer to an unevaluated thunk of type `Maybe Int`, `Day` or `Grade` will end in `000` -- * A tagged pointer to a `Nothing`, `Mon` or `G1` will end in `001` -- * A tagged pointer to a `Just x`, `Tue` or `G2` will end in `010` -- * A tagged pointer to `Wed` or `G3` will end in `011` -- ... -- * A tagged pointer to `Sat` or `G6` will end in `110` -- * A tagged pointer to `Sun` or `G7` or `G8` or `G9` or `G10` will end in `111` -- -- For big families we employ a mildly clever way of combining pointer and -- info-table tagging. We use 1..MAX_PTR_TAG-1 as pointer-resident tags where -- the tags in the pointer and the info table are in a one-to-one -- relation, whereas tag MAX_PTR_TAG is used as "spill over", signifying -- we have to fall back and get the precise constructor tag from the -- info-table. -- -- Consequently we now cascade switches, because we have to check -- the pointer tag first, and when it is MAX_PTR_TAG, fetch the precise -- tag from the info table, and switch on that. The only technically -- tricky part is that the default case needs (logical) duplication. -- To do this we emit an extra label for it and branch to that from -- the second switch. This avoids duplicated codegen. See Trac #14373. -- See note [Double switching for big families] for the mechanics -- involved. -- -- Also see note [Data constructor dynamic tags] -- and the wiki https://gitlab.haskell.org/ghc/ghc/wikis/commentary/rts/haskell-execution/pointer-tagging -- ------------------- cgAlgAltRhss :: (GcPlan,ReturnKind) -> NonVoid Id -> [CgStgAlt] -> FCode ( Maybe CmmAGraphScoped , [(ConTagZ, CmmAGraphScoped)] ) cgAlgAltRhss gc_plan bndr alts = do { tagged_cmms <- cgAltRhss gc_plan bndr alts ; let { mb_deflt = case tagged_cmms of ((DEFAULT,rhs) : _) -> Just rhs _other -> Nothing -- DEFAULT is always first, if present ; branches = [ (dataConTagZ con, cmm) | (DataAlt con, cmm) <- tagged_cmms ] } ; return (mb_deflt, branches) } ------------------- cgAltRhss :: (GcPlan,ReturnKind) -> NonVoid Id -> [CgStgAlt] -> FCode [(AltCon, CmmAGraphScoped)] cgAltRhss gc_plan bndr alts = do dflags <- getDynFlags let base_reg = idToReg dflags bndr cg_alt :: CgStgAlt -> FCode (AltCon, CmmAGraphScoped) cg_alt (con, bndrs, rhs) = getCodeScoped $ maybeAltHeapCheck gc_plan $ do { _ <- bindConArgs con base_reg (assertNonVoidIds bndrs) -- alt binders are always non-void, -- see Note [Post-unarisation invariants] in UnariseStg ; _ <- cgExpr rhs ; return con } forkAlts (map cg_alt alts) maybeAltHeapCheck :: (GcPlan,ReturnKind) -> FCode a -> FCode a maybeAltHeapCheck (NoGcInAlts,_) code = code maybeAltHeapCheck (GcInAlts regs, AssignedDirectly) code = altHeapCheck regs code maybeAltHeapCheck (GcInAlts regs, ReturnedTo lret off) code = altHeapCheckReturnsTo regs lret off code ----------------------------------------------------------------------------- -- Tail calls ----------------------------------------------------------------------------- cgConApp :: DataCon -> [StgArg] -> FCode ReturnKind cgConApp con stg_args | isUnboxedTupleCon con -- Unboxed tuple: assign and return = do { arg_exprs <- getNonVoidArgAmodes stg_args ; tickyUnboxedTupleReturn (length arg_exprs) ; emitReturn arg_exprs } | otherwise -- Boxed constructors; allocate and return = ASSERT2( stg_args `lengthIs` countConRepArgs con, ppr con <> parens (ppr (countConRepArgs con)) <+> ppr stg_args ) do { (idinfo, fcode_init) <- buildDynCon (dataConWorkId con) False currentCCS con (assertNonVoidStgArgs stg_args) -- con args are always non-void, -- see Note [Post-unarisation invariants] in UnariseStg -- The first "con" says that the name bound to this -- closure is "con", which is a bit of a fudge, but -- it only affects profiling (hence the False) ; emit =<< fcode_init ; tickyReturnNewCon (length stg_args) ; emitReturn [idInfoToAmode idinfo] } cgIdApp :: Id -> [StgArg] -> FCode ReturnKind cgIdApp fun_id args = do dflags <- getDynFlags fun_info <- getCgIdInfo fun_id self_loop_info <- getSelfLoop let fun_arg = StgVarArg fun_id fun_name = idName fun_id fun = idInfoToAmode fun_info lf_info = cg_lf fun_info n_args = length args v_args = length $ filter (isVoidTy . stgArgType) args node_points dflags = nodeMustPointToIt dflags lf_info case getCallMethod dflags fun_name fun_id lf_info n_args v_args (cg_loc fun_info) self_loop_info of -- A value in WHNF, so we can just return it. ReturnIt | isVoidTy (idType fun_id) -> emitReturn [] | otherwise -> emitReturn [fun] -- ToDo: does ReturnIt guarantee tagged? EnterIt -> ASSERT( null args ) -- Discarding arguments emitEnter fun SlowCall -> do -- A slow function call via the RTS apply routines { tickySlowCall lf_info args ; emitComment $ mkFastString "slowCall" ; slowCall fun args } -- A direct function call (possibly with some left-over arguments) DirectEntry lbl arity -> do { tickyDirectCall arity args ; if node_points dflags then directCall NativeNodeCall lbl arity (fun_arg:args) else directCall NativeDirectCall lbl arity args } -- Let-no-escape call or self-recursive tail-call JumpToIt blk_id lne_regs -> do { adjustHpBackwards -- always do this before a tail-call ; cmm_args <- getNonVoidArgAmodes args ; emitMultiAssign lne_regs cmm_args ; emit (mkBranch blk_id) ; return AssignedDirectly } -- Note [Self-recursive tail calls] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- -- Self-recursive tail calls can be optimized into a local jump in the same -- way as let-no-escape bindings (see Note [What is a non-escaping let] in -- stgSyn/CoreToStg.hs). Consider this: -- -- foo.info: -- a = R1 // calling convention -- b = R2 -- goto L1 -- L1: ... -- ... -- ... -- L2: R1 = x -- R2 = y -- call foo(R1,R2) -- -- Instead of putting x and y into registers (or other locations required by the -- calling convention) and performing a call we can put them into local -- variables a and b and perform jump to L1: -- -- foo.info: -- a = R1 -- b = R2 -- goto L1 -- L1: ... -- ... -- ... -- L2: a = x -- b = y -- goto L1 -- -- This can be done only when function is calling itself in a tail position -- and only if the call passes number of parameters equal to function's arity. -- Note that this cannot be performed if a function calls itself with a -- continuation. -- -- This in fact implements optimization known as "loopification". It was -- described in "Low-level code optimizations in the Glasgow Haskell Compiler" -- by Krzysztof Woś, though we use different approach. Krzysztof performed his -- optimization at the Cmm level, whereas we perform ours during code generation -- (Stg-to-Cmm pass) essentially making sure that optimized Cmm code is -- generated in the first place. -- -- Implementation is spread across a couple of places in the code: -- -- * FCode monad stores additional information in its reader environment -- (cgd_self_loop field). This information tells us which function can -- tail call itself in an optimized way (it is the function currently -- being compiled), what is the label of a loop header (L1 in example above) -- and information about local registers in which we should arguments -- before making a call (this would be a and b in example above). -- -- * Whenever we are compiling a function, we set that information to reflect -- the fact that function currently being compiled can be jumped to, instead -- of called. This is done in closureCodyBody in GHC.StgToCmm.Bind. -- -- * We also have to emit a label to which we will be jumping. We make sure -- that the label is placed after a stack check but before the heap -- check. The reason is that making a recursive tail-call does not increase -- the stack so we only need to check once. But it may grow the heap, so we -- have to repeat the heap check in every self-call. This is done in -- do_checks in GHC.StgToCmm.Heap. -- -- * When we begin compilation of another closure we remove the additional -- information from the environment. This is done by forkClosureBody -- in GHC.StgToCmm.Monad. Other functions that duplicate the environment - -- forkLneBody, forkAlts, codeOnly - duplicate that information. In other -- words, we only need to clean the environment of the self-loop information -- when compiling right hand side of a closure (binding). -- -- * When compiling a call (cgIdApp) we use getCallMethod to decide what kind -- of call will be generated. getCallMethod decides to generate a self -- recursive tail call when (a) environment stores information about -- possible self tail-call; (b) that tail call is to a function currently -- being compiled; (c) number of passed non-void arguments is equal to -- function's arity. (d) loopification is turned on via -floopification -- command-line option. -- -- * Command line option to turn loopification on and off is implemented in -- DynFlags. -- -- -- Note [Void arguments in self-recursive tail calls] -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -- -- State# tokens can get in the way of the loopification optimization as seen in -- #11372. Consider this: -- -- foo :: [a] -- -> (a -> State# s -> (# State s, Bool #)) -- -> State# s -- -> (# State# s, Maybe a #) -- foo [] f s = (# s, Nothing #) -- foo (x:xs) f s = case f x s of -- (# s', b #) -> case b of -- True -> (# s', Just x #) -- False -> foo xs f s' -- -- We would like to compile the call to foo as a local jump instead of a call -- (see Note [Self-recursive tail calls]). However, the generated function has -- an arity of 2 while we apply it to 3 arguments, one of them being of void -- type. Thus, we mustn't count arguments of void type when checking whether -- we can turn a call into a self-recursive jump. -- emitEnter :: CmmExpr -> FCode ReturnKind emitEnter fun = do { dflags <- getDynFlags ; adjustHpBackwards ; sequel <- getSequel ; updfr_off <- getUpdFrameOff ; case sequel of -- For a return, we have the option of generating a tag-test or -- not. If the value is tagged, we can return directly, which -- is quicker than entering the value. This is a code -- size/speed trade-off: when optimising for speed rather than -- size we could generate the tag test. -- -- Right now, we do what the old codegen did, and omit the tag -- test, just generating an enter. Return -> do { let entry = entryCode dflags $ closureInfoPtr dflags $ CmmReg nodeReg ; emit $ mkJump dflags NativeNodeCall entry [cmmUntag dflags fun] updfr_off ; return AssignedDirectly } -- The result will be scrutinised in the sequel. This is where -- we generate a tag-test to avoid entering the closure if -- possible. -- -- The generated code will be something like this: -- -- R1 = fun -- copyout -- if (fun & 7 != 0) goto Lret else goto Lcall -- Lcall: -- call [fun] returns to Lret -- Lret: -- fun' = R1 -- copyin -- ... -- -- Note in particular that the label Lret is used as a -- destination by both the tag-test and the call. This is -- because Lret will necessarily be a proc-point, and we want to -- ensure that we generate only one proc-point for this -- sequence. -- -- Furthermore, we tell the caller that we generated a native -- return continuation by returning (ReturnedTo Lret off), so -- that the continuation can be reused by the heap-check failure -- code in the enclosing case expression. -- AssignTo res_regs _ -> do { lret <- newBlockId ; let (off, _, copyin) = copyInOflow dflags NativeReturn (Young lret) res_regs [] ; lcall <- newBlockId ; updfr_off <- getUpdFrameOff ; let area = Young lret ; let (outArgs, regs, copyout) = copyOutOflow dflags NativeNodeCall Call area [fun] updfr_off [] -- refer to fun via nodeReg after the copyout, to avoid having -- both live simultaneously; this sometimes enables fun to be -- inlined in the RHS of the R1 assignment. ; let entry = entryCode dflags (closureInfoPtr dflags (CmmReg nodeReg)) the_call = toCall entry (Just lret) updfr_off off outArgs regs ; tscope <- getTickScope ; emit $ copyout <*> mkCbranch (cmmIsTagged dflags (CmmReg nodeReg)) lret lcall Nothing <*> outOfLine lcall (the_call,tscope) <*> mkLabel lret tscope <*> copyin ; return (ReturnedTo lret off) } } ------------------------------------------------------------------------ -- Ticks ------------------------------------------------------------------------ -- | Generate Cmm code for a tick. Depending on the type of Tickish, -- this will either generate actual Cmm instrumentation code, or -- simply pass on the annotation as a @CmmTickish@. cgTick :: Tickish Id -> FCode () cgTick tick = do { dflags <- getDynFlags ; case tick of ProfNote cc t p -> emitSetCCC cc t p HpcTick m n -> emit (mkTickBox dflags m n) SourceNote s n -> emitTick $ SourceNote s n _other -> return () -- ignore }