October 2016 - ghc-tickets

[GHC] #12606: Linux (ELF) Support for "ghc -static -shared"
by GHC 19 Jul '18

19 Jul '18

#12606: Linux (ELF) Support for "ghc -static -shared" ----------------------------------------+--------------------------------- Reporter: tmobile | Owner: Type: feature request | Status: new Priority: normal | Milestone: ⊥ Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: Unknown/Multiple | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: ----------------------------------------+--------------------------------- On some platforms it is possible to create a stand-alone shared library out of some object `Foo.o` including foreign exports with something like: `ghc -static -shared Foo.o ...` The resulting library will have `Foo`'s dependencies including `base` as well as the RTS statically linked in. On BSDs this is possible because generating position independent code is the default there, and on Windows this is possible because the dynamic loader uses some magic that doesn't require position independent code. Unfortunately on Linux this is not possible because `libHSbase`, `libHSrts`, etc. are not built with `-fPIC` unless one specifically asks for it when building GHC. As far as I know this is the only way to get `-static -shared` to work on this platform. There are use cases for compiling Haskell code into a single shared object that does not have additional runtime dependencies. Suppose one wishes to write a plugin for a program in Haskell. It would be nice to provide the plugin as a single shared object, rather than requiring GHC's libraries to be present at runtime as well. An easy solution might be adding a switch to the configure script or `build.mk` to make it easier to build GHC+libs with `-fPIC` on Linux. Ideally it would be possible to build and install both PIC and non-PIC libraries and have GHC choose the correct one at link time based on the flags and target platform, since PIC code carries a small space/performance overhead and isn't necessary on Linux in most cases. I don't know my way around GHC(outside of the GHC API) or the build system that well yet, but if adding extra build system flags and GHC link-time flag handling code to make this work is an acceptable solution I'm happy to spend the time to get this working. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12606> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #8581: Add support for explicitly-bidirectional pattern synonyms
by GHC 12 Jul '18

12 Jul '18

#8581: Add support for explicitly-bidirectional pattern synonyms ------------------------------------+------------------------------------- Reporter: cactus | Owner: cactus Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: 5144 | Blocking: Related Tickets: | ------------------------------------+------------------------------------- Some patterns cannot, by themselves, be turned over into an expression, so they have to be defined as unidirectional. Maybe the most trivial example would be {{{ pattern P -> _ }}} Sometimes, however, it would be desirable to give an explicit way of turning these pattern synonyms into expressions. The PatternSynonyms wiki page has this example: {{{ import qualified Data.Sequence as Seq pattern Empty -> (Seq.viewl -> Seq.EmptyL) pattern x :< xs -> (Seq.viewl -> x Seq.:< xs) pattern xs :> x -> (Seq.viewr -> xs Seq.:> x) }}} It would make a ton of sense to be able to use this cons/snoc notation as "constructors" for `Seq`s. The proposed syntax for this {{{ pattern x :< xs -> (Seq.viewl -> x Seq.:< xs) where x :< xs = x Seq.<| xs }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8581> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 52

[GHC] #12050: Allow haddock comments on non-record types
by GHC 10 Jul '18

10 Jul '18

#12050: Allow haddock comments on non-record types -------------------------------------+------------------------------------- Reporter: bgamari | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 7.10.3 (Parser) | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- There are a variety of cases I have found myself wanting to place haddock documentation on fields of non-record data types rather often in the past. The syntax for this might look like, {{{#!hs data Foo = Foo -- ^ A foo Int -- ^ how big is the foo? String -- ^ the name of the foo }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12050> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #8822: Allow -- ^ Haddock syntax on record constructors
by GHC 10 Jul '18

10 Jul '18

#8822: Allow -- ^ Haddock syntax on record constructors -------------------------------------+------------------------------------- Reporter: Fuuzetsu | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler (Parser) | Version: 7.9 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | -------------------------------------+------------------------------------- Original ticket is on Haddock Trac [http://trac.haskell.org/haddock/ticket/144 here] which I'm opening here as it requires lexer/parser changes. The idea is that given something like {{{ -- | A tree with labeled internal nodes data Tree = Empty -- ^ Leaf | Branch -- ^ Internal node { label :: a -- ^ Node label , subtrees :: [Tree] -- ^ List of subtrees } }}} we'd like it to parse as the user would expect instead of failing. We're currently forced to do something like {{{ -- | A tree with labeled internal nodes data Tree = Empty -- ^ Leaf | -- | Internal node Branch { label :: a -- ^ Node label , subtrees :: [Tree] -- ^ List of subtrees } }}} which is quite ugly. I can't think of a reason why the first form wouldn't be allowed. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8822> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #11070: Type-level arithmetic of sized-types has weaker inference power than in 7.8
by GHC 10 Jul '18

10 Jul '18

#11070: Type-level arithmetic of sized-types has weaker inference power than in 7.8 -------------------------------------+------------------------------------- Reporter: cactus | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: GHC rejects Unknown/Multiple | valid program Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Given the following definitions (just a copy of the relevant bits from sized-types 0.3.5.1, to keep this example self-contained): {{{ -- Copied from sized-types 0.3.5.1 data X0 = X0 data N1 = N1 data X0_ a = X0_ Integer data X1_ a = X1_ Integer type X1 = X1_ X0 type X2 = X0_ (X1_ X0) class Size ix instance Size X0 instance Size a => Size (X1_ a) instance Size a => Size (X0_ a) type family APP0 a type instance APP0 X0 = X0 type instance APP0 N1 = X0_ N1 type instance APP0 (X1_ a) = X0_ (X1_ a) type instance APP0 (X0_ a) = X0_ (X0_ a) type family APP1 a type instance APP1 X0 = X1_ X0 type instance APP1 N1 = N1 type instance APP1 (X1_ a) = X1_ (X1_ a) type instance APP1 (X0_ a) = X1_ (X0_ a) type family SUCC a type instance SUCC X0 = X1_ X0 type instance SUCC N1 = X0 type instance SUCC (X1_ a) = APP0 (SUCC a) type instance SUCC (X0_ a) = APP1 a type family ADD a b type instance ADD a X0 = a type instance ADD X0 a = a type instance ADD X0 N1 = N1 type instance ADD N1 N1 = APP0 N1 type instance ADD N1 (X1_ b) = APP0 b type instance ADD N1 (X0_ b) = APP1 (ADD N1 b) type instance ADD (X1_ a) N1 = APP0 a type instance ADD (X0_ a) N1 = APP1 (ADD a N1) type instance ADD (X1_ a) (X1_ b) = APP0 (SUCC (ADD a b)) type instance ADD (X1_ a) (X0_ b) = APP1 (ADD a b) type instance ADD (X0_ a) (X1_ b) = APP1 (ADD a b) type instance ADD (X0_ a) (X0_ b) = APP0 (ADD a b) type family NOT a type instance NOT X0 = N1 type instance NOT N1 = X0 type instance NOT (X1_ a) = APP0 (NOT a) type instance NOT (X0_ a) = APP1 (NOT a) type SUB a b = ADD a (SUCC (NOT b)) }}} The following module typechecks with GHC 7.8.3: {{{ data B w = B (&*) :: (Size n, Size n', Size (ADD n' n), n ~ SUB (ADD n' n) n', n' ~ SUB (ADD n' n) n) => [(a, B n)] -> [(b, B n')] -> [((a, b), B (ADD n' n))] mks &* args = undefined foo :: [((), B X1)] foo = [((), B)] bar :: [(((), ()), B X2)] bar = [((), B)] &* foo }}} However, it fails with GHC 7.10.2, with {{{ /tmp/GHCBug.hs:70:7: Couldn't match type ‘ADD X1 n0’ with ‘X0_ (X1_ X0)’ The type variable ‘n0’ is ambiguous Expected type: [(((), ()), B X2)] Actual type: [(((), ()), B (ADD X1 n0))] In the expression: [((), B)] &* foo In an equation for ‘bar’: bar = [((), B)] &* foo /tmp/GHCBug.hs:70:17: Occurs check: cannot construct the infinite type: n0 ~ ADD (ADD X1 n0) N1 The type variable ‘n0’ is ambiguous Expected type: SUB (ADD X1 n0) X1 Actual type: n0 In the expression: [((), B)] &* foo In an equation for ‘bar’: bar = [((), B)] &* foo Failed, modules loaded: none. }}} The workaround/solution is to change the definition of `bar`: {{{ bar :: [(((), ()), B X2)] bar = [((), B :: B X1)] &* foo }}} This second version typechecks with GHC 7.10.2. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11070> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #11282: Injectivity annotation fails in the presence of higher-rank use of constraint family
by GHC 09 Jul '18

09 Jul '18

#11282: Injectivity annotation fails in the presence of higher-rank use of constraint family -------------------------------------+------------------------------------- Reporter: goldfire | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.11 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- When I say {{{ {-# LANGUAGE TypeFamilies, RankNTypes #-} module Bug where import GHC.Exts type family F a = r | r -> a type family G a :: Constraint meth :: (G a => F a) -> () meth = undefined }}} I get {{{ Bug.hs:10:9: error: • Could not deduce: F a ~ F a0 from the context: G a0 bound by the type signature for: meth :: G a0 => F a0 at Bug.hs:10:9-26 NB: ‘F’ is a type function, and may not be injective The type variable ‘a0’ is ambiguous • In the ambiguity check for ‘meth’ To defer the ambiguity check to use sites, enable AllowAmbiguousTypes In the type signature: meth :: (G a => F a) -> () }}} The presence of `G a` is critical for exhibiting this bug. If I replace `G a` with `Show a` the problem disappears. This one was actually encountered in an attempt to do Useful Work, not just idle typecheckery. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11282> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #9123: Need for higher kinded roles
by GHC 08 Jul '18

08 Jul '18

#9123: Need for higher kinded roles ------------------------------------+------------------------------------- Reporter: simonpj | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- This [http://www.haskell.org/pipermail/ghc-devs/2014-May/004964.html thread] on ghc-devs identifies a real shortcoming of the new roles system. Here's a compact example, from the thread {{{ class C m where ret :: a -> m a bind :: m a -> (a -> m b) -> m b join :: m (m a) -> m a newtype T m a = MkT (m a) deriving( C ) }}} The `deriving(C)` is accepted without `join` in the class, but rejected when `join` is added. And the AMP proposal adds `join` to class `Monad`! In one way it is rightly rejected: it really would be unsound to derive an instance for `C (T K)` where `K`'s argument had nominal (but not representational) role. But we have no way to limit the type constructors at which `T` can be used. This deficiency is noted in the [http://research.microsoft.com/en- us/um/people/simonpj/papers/ext-f/ Safe Coercions paper], but this seems to be the first occasion on which it has bitten us badly. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9123> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 66

[GHC] #9672: Error message too long (full case statement printed)
by GHC 06 Jul '18

06 Jul '18

#9672: Error message too long (full case statement printed) -------------------------------------+------------------------------------- Reporter: andreas.abel | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler (Type | Version: 7.8.3 checker) | Operating System: Keywords: | Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: Other Difficulty: Easy (less than 1 | Test Case: hour) | Blocking: Blocked By: | Differential Revisions: Related Tickets: | -------------------------------------+------------------------------------- Imho, it does not make sense to print the other branches of a case when a type error is discovered in one branch. When you see the following example, you will agree with me. {{{ src/full/Agda/TypeChecking/Errors.hs:540:35: Couldn't match type ‘[Char]’ with ‘TCMT IO Doc’ Expected type: TCM Doc Actual type: [Char] In the expression: "Module cannot be imported since it has open interaction points" In a case alternative: SolvedButOpenHoles -> "Module cannot be imported since it has open interaction points" In a stmt of a 'do' block: }}} {{{#!hs case err of { InternalError s -> panic s NotImplemented s -> fwords $ "Not implemented: " ++ s NotSupported s -> fwords $ "Not supported: " ++ s CompilationError s -> sep [fwords "Compilation error:", text s] GenericError s -> fwords s GenericDocError d -> return d TerminationCheckFailed because -> fwords "Termination checking failed for the following functions:" $$ (nest 2 $ fsep $ punctuate comma $ map (pretty . dropTopLevelModule) $ concatMap termErrFunctions because) $$ fwords "Problematic calls:" $$ (nest 2 $ fmap (P.vcat . nub) $ mapM prettyTCM $ sortBy (compare `on` callInfoRange) $ concatMap termErrCalls because) PropMustBeSingleton -> fwords "Datatypes in Prop must have at most one constructor when proof irrelevance is enabled" DataMustEndInSort t -> fsep $ pwords "The type of a datatype must end in a sort." ++ [prettyTCM t] ++ pwords "isn't a sort." ShouldEndInApplicationOfTheDatatype t -> fsep $ pwords "The target of a constructor must be the datatype applied to its parameters," ++ [prettyTCM t] ++ pwords "isn't" ShouldBeAppliedToTheDatatypeParameters s t -> fsep $ pwords "The target of the constructor should be" ++ [prettyTCM s] ++ pwords "instead of" ++ [prettyTCM t] ShouldBeApplicationOf t q -> fsep $ pwords "The pattern constructs an element of" ++ [prettyTCM q] ++ pwords "which is not the right datatype" ShouldBeRecordType t -> fsep $ pwords "Expected non-abstract record type, found " ++ [prettyTCM t] ShouldBeRecordPattern p -> fsep $ pwords "Expected record pattern" NotAProjectionPattern p -> fsep $ pwords "Not a valid projection for a copattern: " ++ [prettyA p] DifferentArities -> fwords "The number of arguments in the defining equations differ" WrongHidingInLHS -> do { fwords "Unexpected implicit argument" } WrongHidingInLambda t -> do { fwords "Found an implicit lambda where an explicit lambda was expected" } WrongIrrelevanceInLambda t -> do { fwords "Found an irrelevant lambda where a relevant lambda was expected" } WrongNamedArgument a -> fsep $ pwords "Function does not accept argument " ++ [prettyTCM a] WrongHidingInApplication t -> do { fwords "Found an implicit application where an explicit application was expected" } HidingMismatch h h' -> fwords $ "Expected " ++ verbalize (Indefinite h') ++ " argument, but found " ++ verbalize (Indefinite h) ++ " argument" RelevanceMismatch r r' -> fwords $ "Expected " ++ verbalize (Indefinite r') ++ " argument, but found " ++ verbalize (Indefinite r) ++ " argument" ColorMismatch c c' -> fsep $ pwords "Expected argument color to be" ++ [prettyTCM c'] ++ pwords "but found color" ++ [prettyTCM c] NotInductive t -> fsep $ [prettyTCM t] ++ pwords "is not an inductive data type" UninstantiatedDotPattern e -> fsep $ pwords "Failed to infer the value of dotted pattern" IlltypedPattern p a -> fsep $ pwords "Type mismatch" IllformedProjectionPattern p -> fsep $ pwords "Ill-formed projection pattern " ++ [prettyA p] CannotEliminateWithPattern p a -> do { let ...; fsep $ pwords "Cannot eliminate type" ++ prettyTCM a : if isProj then pwords "with projection pattern" ++ ... else pwords "with pattern" ++ prettyA p : pwords "(did you supply too many arguments?)" } TooManyArgumentsInLHS a -> fsep $ pwords "Left hand side gives too many arguments to a function of type" ++ [prettyTCM a] WrongNumberOfConstructorArguments c expect given -> fsep $ pwords "The constructor" ++ [prettyTCM c] ++ pwords "expects" ++ [text (show expect)] ++ pwords "arguments (including hidden ones), but has been given" ++ [text (show given)] ++ pwords "(including hidden ones)" CantResolveOverloadedConstructorsTargetingSameDatatype d cs -> fsep $ pwords ("Can't resolve overloaded constructors targeting the same datatype (" ++ show (qnameToConcrete d) ++ "):") ++ map pretty cs DoesNotConstructAnElementOf c t -> fsep $ pwords "The constructor" ++ [prettyTCM c] ++ pwords "does not construct an element of" ++ [prettyTCM t] ConstructorPatternInWrongDatatype c d -> fsep $ [prettyTCM c] ++ pwords "is not a constructor of the datatype" ++ [prettyTCM d] IndicesNotConstructorApplications [i] -> fwords "The index" $$ nest 2 (prettyTCM i) $$ fsep (pwords "is not a constructor (or literal) applied to variables" ++ pwords "(note that parameters count as constructor arguments)") IndicesNotConstructorApplications is -> fwords "The indices" $$ nest 2 (vcat $ map prettyTCM is) $$ fsep (pwords "are not constructors (or literals) applied to variables" ++ pwords "(note that parameters count as constructor arguments)") IndexVariablesNotDistinct vs is -> fwords "The variables" $$ nest 2 (vcat $ map (\ v -> prettyTCM (I.Var v [])) vs) $$ fwords "in the indices" $$ nest 2 (vcat $ map prettyTCM is) $$ fwords "are not distinct (note that parameters count as constructor arguments)" IndicesFreeInParameters vs indices pars -> fwords "The variables" $$ nest 2 (vcat $ map (\ v -> prettyTCM (I.Var v [])) vs) $$ fwords "which are used (perhaps as constructor parameters) in the index expressions" $$ nest 2 (vcat $ map prettyTCM indices) $$ fwords "are free in the parameters" $$ nest 2 (vcat $ map prettyTCM pars) ShadowedModule x [] -> (throwImpossible (Impossible "src/full/Agda/TypeChecking/Errors.hs" 404)) ShadowedModule x ms@(m : _) -> fsep $ pwords "Duplicate definition of module" ++ [prettyTCM x <> text "."] ++ pwords "Previous definition of" ++ [help m] ++ pwords "module" ++ [prettyTCM x] ++ pwords "at" ++ [prettyTCM r] where help m = do { ... } r = case ... of { [] -> ... r : _ -> ... } defSiteOfLast [] = noRange defSiteOfLast ns = nameBindingSite (last ns) ModuleArityMismatch m EmptyTel args -> fsep $ pwords "The module" ++ [prettyTCM m] ++ pwords "is not parameterized, but is being applied to arguments" ModuleArityMismatch m tel@(ExtendTel _ _) args -> fsep $ pwords "The arguments to " ++ [prettyTCM m] ++ pwords "does not fit the telescope" ++ [prettyTCM tel] ShouldBeEmpty t [] -> fsep $ [prettyTCM t] ++ pwords "should be empty, but that's not obvious to me" ShouldBeEmpty t ps -> fsep ([prettyTCM t] ++ pwords "should be empty, but the following constructor patterns are valid:") $$ nest 2 (vcat $ map (showPat 0) ps) ShouldBeASort t -> fsep $ [prettyTCM t] ++ pwords "should be a sort, but it isn't" ShouldBePi t -> fsep $ [prettyTCM t] ++ pwords "should be a function type, but it isn't" NotAProperTerm -> fwords "Found a malformed term" SetOmegaNotValidType -> fwords "Set\969 is not a valid type" InvalidType v -> fsep $ [prettyTCM v] ++ pwords "is not a valid type" SplitOnIrrelevant p t -> fsep $ pwords "Cannot pattern match" ++ [prettyA p] ++ pwords "against irrelevant type" ++ [prettyTCM t] DefinitionIsIrrelevant x -> fsep $ text "Identifier" : prettyTCM x : pwords "is declared irrelevant, so it cannot be used here" VariableIsIrrelevant x -> fsep $ text "Variable" : prettyTCM x : pwords "is declared irrelevant, so it cannot be used here" UnequalBecauseOfUniverseConflict cmp s t -> fsep $ [prettyTCM s, notCmp cmp, ....] UnequalTerms cmp s t a -> do { (d1, d2, d) <- prettyInEqual s t; fsep $ [...] ++ pwords "of type" ++ [...] ++ [...] } UnequalTypes cmp a b -> prettyUnequal a (notCmp cmp) b UnequalColors a b -> error "TODO guilhem 4" HeterogeneousEquality u a v b -> fsep $ pwords "Refuse to solve heterogeneous constraint" ++ [prettyTCM u] ++ pwords ":" ++ [prettyTCM a] ++ pwords "=?=" ++ [prettyTCM v] ++ pwords ":" ++ [prettyTCM b] UnequalRelevance cmp a b -> fsep $ [prettyTCM a, notCmp cmp, ....] ++ pwords "because one is a relevant function type and the other is an irrelevant function type" UnequalHiding a b -> fsep $ [prettyTCM a, text "!=", ....] ++ pwords "because one is an implicit function type and the other is an explicit function type" UnequalSorts s1 s2 -> fsep $ [prettyTCM s1, text "!=", ....] NotLeqSort s1 s2 -> fsep $ pwords "The type of the constructor does not fit in the sort of the datatype, since" ++ [prettyTCM s1] ++ pwords "is not less or equal than" ++ [prettyTCM s2] TooFewFields r xs -> fsep $ pwords "Missing fields" ++ punctuate comma (map pretty xs) ++ pwords "in an element of the record" ++ [prettyTCM r] TooManyFields r xs -> fsep $ pwords "The record type" ++ [prettyTCM r] ++ pwords "does not have the fields" ++ punctuate comma (map pretty xs) DuplicateConstructors xs -> fsep $ pwords "Duplicate constructors" ++ punctuate comma (map pretty xs) ++ pwords "in datatype" DuplicateFields xs -> fsep $ pwords "Duplicate fields" ++ punctuate comma (map pretty xs) ++ pwords "in record" UnexpectedWithPatterns ps -> fsep $ pwords "Unexpected with patterns" ++ (punctuate (text " |") $ map prettyA ps) WithClausePatternMismatch p q -> fsep $ pwords "With clause pattern " ++ [prettyA p] ++ pwords " is not an instance of its parent pattern " ++ [prettyTCM q] MetaCannotDependOn m ps i -> fsep $ pwords "The metavariable" ++ [prettyTCM $ MetaV m []] ++ pwords "cannot depend on" ++ [pvar i] ++ pwords "because it" ++ deps where pvar = prettyTCM . var deps = case map pvar ps of { [] -> ... [x] -> ... xs -> ... } MetaOccursInItself m -> fsep $ pwords "Cannot construct infinite solution of metavariable" ++ [prettyTCM $ MetaV m []] BuiltinMustBeConstructor s e -> fsep $ [prettyA e] ++ pwords "must be a constructor in the binding to builtin" ++ [text s] NoSuchBuiltinName s -> fsep $ pwords "There is no built-in thing called" ++ [text s] DuplicateBuiltinBinding b x y -> fsep $ pwords "Duplicate binding for built-in thing" ++ [text b <> comma] ++ pwords "previous binding to" ++ [prettyTCM x] NoBindingForBuiltin x -> fsep $ pwords "No binding for builtin thing" ++ [text x <> comma] ++ pwords ("use {-# BUILTIN " ++ x ++ " name #-} to bind it to 'name'") NoSuchPrimitiveFunction x -> fsep $ pwords "There is no primitive function called" ++ [text x] BuiltinInParameterisedModule x -> fwords $ "The BUILTIN pragma cannot appear inside a bound context " ++ "(for instance, in a parameterised module or as a local declaration)" IllegalLetInTelescope tb -> fsep $ [pretty tb] ++ pwords " is not allowed in a telescope here." NoRHSRequiresAbsurdPattern ps -> fwords $ "The right-hand side can only be omitted if there " ++ "is an absurd pattern, () or {}, in the left-hand side." AbsurdPatternRequiresNoRHS ps -> fwords $ "The right-hand side must be omitted if there " ++ "is an absurd pattern, () or {}, in the left-hand side." LocalVsImportedModuleClash m -> fsep $ pwords "The module" ++ [text $ show m] ++ pwords "can refer to either a local module or an imported module" SolvedButOpenHoles -> "Module cannot be imported since it has open interaction points" UnsolvedMetas rs -> fsep (pwords "Unsolved metas at the following locations:") $$ nest 2 (vcat $ map prettyTCM rs) UnsolvedConstraints cs -> fsep (pwords "Failed to solve the following constraints:") $$ nest 2 (vcat $ map prettyConstraint cs) where prettyConstraint :: ProblemConstraint -> TCM Doc prettyConstraint c = f (prettyTCM c) where r = ... .... CyclicModuleDependency ms -> fsep (pwords "cyclic module dependency:") $$ nest 2 (vcat $ map pretty ms) FileNotFound x files -> fsep (pwords "Failed to find source of module" ++ [pretty x] ++ pwords "in any of the following locations:") $$ nest 2 (vcat $ map (text . filePath) files) OverlappingProjects f m1 m2 -> fsep (pwords "The file" ++ [text (filePath f)] ++ pwords "can be accessed via several project roots. Both" ++ [pretty m1] ++ pwords "and" ++ [pretty m2] ++ pwords "point to this file.") AmbiguousTopLevelModuleName x files -> fsep (pwords "Ambiguous module name. The module name" ++ [pretty x] ++ pwords "could refer to any of the following files:") $$ nest 2 (vcat $ map (text . filePath) files) ClashingFileNamesFor x files -> fsep (pwords "Multiple possible sources for module" ++ [text $ show x] ++ pwords "found:") $$ nest 2 (vcat $ map (text . filePath) files) ModuleDefinedInOtherFile mod file file' -> fsep $ pwords "You tried to load" ++ [text (filePath file)] ++ pwords "which defines the module" ++ [pretty mod <> text "."] ++ pwords "However, according to the include path this module should" ++ pwords "be defined in" ++ [text (filePath file') <> text "."] ModuleNameDoesntMatchFileName given files -> fsep (pwords "The name of the top level module does not match the file name. The module" ++ [pretty given] ++ pwords "should be defined in one of the following files:") $$ nest 2 (vcat $ map (text . filePath) files) BothWithAndRHS -> fsep $ pwords "Unexpected right hand side" NotInScope xs -> fsep (pwords "Not in scope:") $$ nest 2 (vcat $ map name xs) where name x = fsep [...] suggestion s | elem ':' s = parens $ text "did you forget space around the ':'?" | elem "->" two = parens $ text "did you forget space around the '->'?" | otherwise = empty where two = ... NoSuchModule x -> fsep $ pwords "No such module" ++ [pretty x] AmbiguousName x ys -> vcat [fsep $ pwords "Ambiguous name" ++ [...] ++ pwords "It could refer to any one of", nest 2 $ vcat $ map nameWithBinding ys, ....] AmbiguousModule x ys -> vcat [fsep $ pwords "Ambiguous module name" ++ [...] ++ pwords "It could refer to any one of", nest 2 $ vcat $ map help ys, ....] where help :: ModuleName -> TCM Doc help m = do { ... } UninstantiatedModule x -> fsep (pwords "Cannot access the contents of the parameterised module" ++ [pretty x <> text "."] ++ pwords "To do this the module first has to be instantiated. For instance:") $$ nest 2 (hsep [text "module", pretty x <> text "'", ....]) ClashingDefinition x y -> fsep $ pwords "Multiple definitions of" ++ [pretty x <> text "."] ++ pwords "Previous definition at" ++ [prettyTCM $ nameBindingSite $ qnameName y] ClashingModule m1 m2 -> fsep $ pwords "The modules" ++ [prettyTCM m1, text "and", ....] ++ pwords "clash." ClashingImport x y -> fsep $ pwords "Import clash between" ++ [pretty x, text "and", ....] ClashingModuleImport x y -> fsep $ pwords "Module import clash between" ++ [pretty x, text "and", ....] PatternShadowsConstructor x c -> fsep $ pwords "The pattern variable" ++ [prettyTCM x] ++ pwords "has the same name as the constructor" ++ [prettyTCM c] DuplicateImports m xs -> fsep $ pwords "Ambiguous imports from module" ++ [pretty m] ++ pwords "for" ++ punctuate comma (map pretty xs) ModuleDoesntExport m xs -> fsep $ pwords "The module" ++ [pretty m] ++ pwords "doesn't export the following:" ++ punctuate comma (map pretty xs) NotAModuleExpr e -> fsep $ pwords "The right-hand side of a module definition must have the form 'M e1 .. en'" ++ pwords "where M is a module name. The expression" ++ [pretty e, text "doesn't."] FieldOutsideRecord -> fsep $ pwords "Field appearing outside record declaration." InvalidPattern p -> fsep $ pretty p : pwords "is not a valid pattern" RepeatedVariablesInPattern xs -> fsep $ pwords "Repeated variables in left hand side:" ++ map pretty xs NotAnExpression e -> fsep $ [pretty e] ++ pwords "is not a valid expression." NotAValidLetBinding nd -> fwords $ "Not a valid let-declaration" NothingAppliedToHiddenArg e -> fsep $ [pretty e] ++ pwords "cannot appear by itself. It needs to be the argument to" ++ pwords "a function expecting an implicit argument." NothingAppliedToInstanceArg e -> fsep $ [pretty e] ++ pwords "cannot appear by itself. It needs to be the argument to" ++ pwords "a function expecting an instance argument." NoParseForApplication es -> fsep $ pwords "Could not parse the application" ++ [pretty $ C.RawApp noRange es] AmbiguousParseForApplication es es' -> fsep (pwords "Don't know how to parse" ++ [pretty_es <> (text ".")] ++ pwords "Could mean any one of:") $$ nest 2 (vcat $ map pretty' es') where pretty_es :: TCM Doc pretty_es = pretty $ C.RawApp noRange es pretty' :: C.Expr -> TCM Doc .... BadArgumentsToPatternSynonym x -> fsep $ pwords "Bad arguments to pattern synonym " ++ [prettyTCM x] TooFewArgumentsToPatternSynonym x -> fsep $ pwords "Too few arguments to pattern synonym " ++ [prettyTCM x] UnusedVariableInPatternSynonym -> fsep $ pwords "Unused variable in pattern synonym." NoParseForLHS IsLHS p -> fsep $ pwords "Could not parse the left-hand side" ++ [pretty p] NoParseForLHS IsPatSyn p -> fsep $ pwords "Could not parse the pattern synonym" ++ [pretty p] AmbiguousParseForLHS lhsOrPatSyn p ps -> fsep (pwords "Don't know how to parse" ++ [pretty_p <> text "."] ++ pwords "Could mean any one of:") $$ nest 2 (vcat $ map pretty' ps) where pretty_p :: TCM Doc pretty_p = pretty p pretty' :: C.Pattern -> TCM Doc .... IncompletePatternMatching v args -> fsep $ pwords "Incomplete pattern matching for" ++ [prettyTCM v <> text "."] ++ pwords "No match for" ++ map prettyTCM args UnreachableClauses f pss -> fsep $ pwords "Unreachable" ++ pwords (plural (length pss) "clause") where plural 1 thing = thing plural n thing = thing ++ "s" CoverageFailure f pss -> fsep (pwords "Incomplete pattern matching for" ++ [prettyTCM f <> text "."] ++ pwords "Missing cases:") $$ nest 2 (vcat $ map display pss) where display ps = do { ... } nicify f ps = do { ... } CoverageCantSplitOn c tel cIxs gIxs | length cIxs /= length gIxs -> (throwImpossible (Impossible "src/full/Agda/TypeChecking/Errors.hs" 750)) | otherwise -> addCtxTel tel $ vcat ([fsep $ pwords "I'm not sure if there should be a case for the constructor" ++ [...] ++ pwords "because I get stuck when trying to solve the following" ++ pwords "unification problems (inferred index \8799 expected index):"] ++ zipWith (\ c g -> nest 2 $ prettyTCM c <+> text "\8799" <+> prettyTCM g) cIxs gIxs) CoverageCantSplitIrrelevantType a -> fsep $ pwords "Cannot split on argument of irrelevant datatype" ++ [prettyTCM a] CoverageCantSplitType a -> fsep $ pwords "Cannot split on argument of non-datatype" ++ [prettyTCM a] SplitError e -> prettyTCM e WithoutKError a u v -> fsep $ pwords "Cannot eliminate reflexive equation" ++ [prettyTCM u] ++ pwords "=" ++ [prettyTCM v] ++ pwords "of type" ++ [prettyTCM a] ++ pwords "because K has been disabled." NotStrictlyPositive d ocs -> fsep $ pwords "The datatype" ++ [prettyTCM d] ++ pwords "is not strictly positive, because" ++ prettyOcc "it" ocs where prettyOcc _ [] = [] prettyOcc it (OccCon d c r : ocs) = concat [...] prettyOcc it (OccClause f n r : ocs) = concat [...] prettyR NonPositively = pwords "negatively" prettyR (ArgumentTo i q) = pwords "as the" ++ [...] ++ pwords "argument to" ++ [...] th 0 = text "first" th 1 = text "second" th 2 = text "third" th n = text (show $ n - 1) <> text "th" .... IFSNoCandidateInScope t -> fsep $ pwords "No variable of type" ++ [prettyTCM t] ++ pwords "was found in scope." SafeFlagPostulate e -> fsep $ pwords "Cannot postulate" ++ [pretty e] ++ pwords "with safe flag" SafeFlagPragma xs -> let plural | length xs == 1 = ... | otherwise = ... in fsep $ [fwords ("Cannot set OPTION pragma" ++ plural)] ++ map text xs ++ [fwords "with safe flag."] SafeFlagNoTerminationCheck -> fsep (pwords "Cannot use NO_TERMINATION_CHECK pragma with safe flag.") SafeFlagNonTerminating -> fsep (pwords "Cannot use NON_TERMINATING pragma with safe flag.") SafeFlagTerminating -> fsep (pwords "Cannot use TERMINATING pragma with safe flag.") SafeFlagPrimTrustMe -> fsep (pwords "Cannot use primTrustMe with safe flag") NeedOptionCopatterns -> fsep (pwords "Option --copatterns needed to enable destructor patterns") } }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9672> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #12073: Missing instance of MonadFix for Q
by GHC 06 Jul '18

06 Jul '18

#12073: Missing instance of MonadFix for Q -------------------------------------+------------------------------------- Reporter: jophish | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Template | Version: Haskell | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- `Q` is capable of being an instance of `MonadFix`. The definition of `mfix` is the same as `fixIO` except that the `MVar` handling is wrapped in `runIO`. {{{#!hs instance MonadFix Q where mfix k = do m <- runIO newEmptyMVar ans <- runIO (unsafeInterleaveIO (takeMVar m)) result <- k ans runIO (putMVar m result) pure result }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12073> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #11523: Infinite Loop when mixing UndecidableSuperClasses and the class/instance constraint synonym trick.
by GHC 05 Jul '18

05 Jul '18

#11523: Infinite Loop when mixing UndecidableSuperClasses and the class/instance constraint synonym trick. -------------------------------------+------------------------------------- Reporter: ekmett | Owner: Type: bug | Status: new Priority: highest | Milestone: 8.0.1 Component: Compiler | Version: 7.10.3 (Type checker) | Keywords: | Operating System: Unknown/Multiple UndecidableSuperClasses | Architecture: | Type of failure: GHC rejects Unknown/Multiple | valid program Test Case: | Blocked By: Blocking: | Related Tickets: #11480 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- There seems to be a bad interaction between `UndecidableSuperClasses` and the common trick of using a cyclic definition of a class and instance to make an alias at the constraint level: {{{#!hs class (Foo f, Bar f) => Baz f instance (Foo f, Bar f) => Baz f }}} Unfortunately, there are circumstances in which you can't eliminate it, such as {{{#!hs class (p, q) => p & q instance (p, q) => p & q }}} There we can't partially apply (,) at the constraint level, but we can talk about `(&) :: Constraint -> Constraint -> Constraint` and `(&) (Eq a) :: Constraint -> Constraint`. This doesn't seem to happen on simpler examples like the above, but once I modify the categories example from #11480 to move the domain and codomain of a functor into class associated types, so that they don't infect every single subclass of functor, we run into a problem. The following stripped down version of the code seems to send the compiler into an infinite loop: {{{#!hs {-# language KindSignatures #-} {-# language PolyKinds #-} {-# language DataKinds #-} {-# language TypeFamilies #-} {-# language RankNTypes #-} {-# language NoImplicitPrelude #-} {-# language FlexibleContexts #-} {-# language MultiParamTypeClasses #-} {-# language GADTs #-} {-# language ConstraintKinds #-} {-# language FlexibleInstances #-} {-# language TypeOperators #-} {-# language ScopedTypeVariables #-} {-# language DefaultSignatures #-} {-# language FunctionalDependencies #-} {-# language UndecidableSuperClasses #-} {-# language UndecidableInstances #-} {-# language TypeInType #-} import GHC.Types (Constraint, Type) import qualified Prelude type Cat i = i -> i -> Type newtype Y (p :: i -> j -> Type) (a :: j) (b :: i) = Y { getY :: p b a } class Vacuous (a :: i) instance Vacuous a class (Functor p, Dom p ~ Op p, Cod p ~ Nat p (->)) => Category (p :: Cat i) where type Op p :: Cat i type Op p = Y p type Ob p :: i -> Constraint type Ob p = Vacuous class (Category (Dom f), Category (Cod f)) => Functor (f :: i -> j) where type Dom f :: Cat i type Cod f :: Cat j class (Functor f, Dom f ~ p, Cod f ~ q) => Fun (p :: Cat i) (q :: Cat j) (f :: i -> j) | f -> p q instance (Functor f, Dom f ~ p, Cod f ~ q) => Fun (p :: Cat i) (q :: Cat j) (f :: i -> j) data Nat (p :: Cat i) (q :: Cat j) (f :: i -> j) (g :: i -> j) instance (Category p, Category q) => Category (Nat p q) where type Ob (Nat p q) = Fun p q instance (Category p, Category q) => Functor (Nat p q) where type Dom (Nat p q) = Y (Nat p q) type Cod (Nat p q) = Nat (Nat p q) (->) instance (Category p, Category q) => Functor (Nat p q f) where type Dom (Nat p q f) = Nat p q type Cod (Nat p q f) = (->) instance Category (->) instance Functor ((->) e) where type Dom ((->) e) = (->) type Cod ((->) e) = (->) instance Functor (->) where type Dom (->) = Y (->) type Cod (->) = Nat (->) (->) instance (Category p, Op p ~ Y p) => Category (Y p) where type Op (Y p) = p instance (Category p, Op p ~ Y p) => Functor (Y p a) where type Dom (Y p a) = Y p type Cod (Y p a) = (->) instance (Category p, Op p ~ Y p) => Functor (Y p) where type Dom (Y p) = p type Cod (Y p) = Nat (Y p) (->) }}} Here I need the circular definition of `Fun` to talk about the fact that the objects in the category of natural transformations from a category p to a category q are functors with domain p and codomain q, so to give the definition of the class-associated `type Ob (Nat p q)` I need such a cyclic definition. I can't leave the domain and codomain of `Functor` in fundeps, otherwise if I go to define a subclass of Functor I'd have to include the arguments, and I have a ''lot'' of those subclasses! -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11523> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 33