September 2018 - ghc-tickets

[GHC] #11498: GHC requires kind-polymorphic signatures on class head
by GHC 01 Oct '18

01 Oct '18

#11498: GHC requires kind-polymorphic signatures on class head -------------------------------------+------------------------------------- Reporter: crockeea | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.2-rc2 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: -------------------------------------+------------------------------------- I have the following compiling example: {{{ {-# LANGUAGE DataKinds, FlexibleContexts, FlexibleInstances, KindSignatures, MultiParamTypeClasses, PolyKinds, RankNTypes, ScopedTypeVariables, TypeFamilies, TypeOperators #-} import Data.Proxy import GHC.Prim type family CtxOf d (args :: k) :: Constraint class Run ctx (params :: [k]) where runAll :: Proxy params -> Proxy ctx -> (forall (args :: k) . (CtxOf ctx args) => Proxy args -> Bool) -> [Bool] data BasicCtxD type instance CtxOf BasicCtxD '(a,b) = (a ~ b) instance (Run BasicCtxD params, a ~ b) => Run BasicCtxD ( '(a,b) ': params) where runAll _ pctx f = (f (Proxy::Proxy '(a,b))) : (runAll (Proxy::Proxy params) pctx f) }}} But if I move the kind signature of `params` to its occurrence in the signature of `runAll`: {{{ class Run ctx params where runAll :: Proxy (params :: [k]) -> Proxy ctx -> (forall (args :: k) . (CtxOf ctx args) => Proxy args -> Bool) -> [Bool] }}} I get a couple of nasty (and confusing) compile errors. {{{ Main.hs:20:25: Couldn't match kind ‘k1’ with ‘(,) k k’ ‘k1’ is a rigid type variable bound by the type signature for runAll :: Proxy ('(a, b) : params) -> Proxy BasicCtxD -> (forall (args :: k1). CtxOf BasicCtxD args => Proxy args -> Bool) -> [Bool] at Main.hs:20:3 Expected type: Proxy args0 Actual type: Proxy '(a, b) Relevant bindings include f :: forall (args :: k1). CtxOf BasicCtxD args => Proxy args -> Bool (bound at Main.hs:20:17) runAll :: Proxy ('(a, b) : params) -> Proxy BasicCtxD -> (forall (args :: k1). CtxOf BasicCtxD args => Proxy args -> Bool) -> [Bool] (bound at Main.hs:20:3) In the first argument of ‘f’, namely ‘(Proxy :: Proxy '(a, b))’ In the first argument of ‘(:)’, namely ‘(f (Proxy :: Proxy '(a, b)))’ In the expression: (f (Proxy :: Proxy '(a, b))) : (runAll (Proxy :: Proxy params) pctx f) Main.hs:21:42: Could not deduce (k1 ~ (,) k k) from the context (CtxOf BasicCtxD args) bound by a type expected by the context: CtxOf BasicCtxD args => Proxy args -> Bool at Main.hs:21:8-42 ‘k1’ is a rigid type variable bound by the type signature for runAll :: Proxy ('(a, b) : params) -> Proxy BasicCtxD -> (forall (args :: k1). CtxOf BasicCtxD args => Proxy args -> Bool) -> [Bool] at Main.hs:20:3 Expected type: Proxy args -> Bool Actual type: Proxy args1 -> Bool Relevant bindings include f :: forall (args :: k1). CtxOf BasicCtxD args => Proxy args -> Bool (bound at Main.hs:20:17) runAll :: Proxy ('(a, b) : params) -> Proxy BasicCtxD -> (forall (args :: k1). CtxOf BasicCtxD args => Proxy args -> Bool) -> [Bool] (bound at Main.hs:20:3) In the third argument of ‘runAll’, namely ‘f’ In the second argument of ‘(:)’, namely ‘(runAll (Proxy :: Proxy params) pctx f)’ }}} It seems to me that the two definitions of the class are equivalent, so it would be nice if they both compiled. It wasn't obvious to me that the kind signature had to go on the class parameter rather than somewhere else. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11498> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #15637: Ambiguous type variables in GeneralisedNewtypeDeriving
by GHC 01 Oct '18

01 Oct '18

#15637: Ambiguous type variables in GeneralisedNewtypeDeriving -------------------------------------+------------------------------------- Reporter: i-am-tom | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 Keywords: | Operating System: Unknown/Multiple GeneralisedNewtypeDeriving, | GeneralizedNewtypeDeriving | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- When deriving the `C` instance in the following code: {{{#!hs {-# LANGUAGE AllowAmbiguousTypes #-} {-# LANGUAGE GeneralisedNewtypeDeriving #-} module Test where class C a where f :: String instance C () where f = "foo" newtype T = T () deriving C }}} The following error occurs: {{{ Test.hs:10:27: error: • Ambiguous type variable ‘a0’ arising from a use of ‘f’ prevents the constraint ‘(C a0)’ from being solved. Probable fix: use a type annotation to specify what ‘a0’ should be. These potential instances exist: instance C T -- Defined at Test.hs:10:27 instance C () -- Defined at Test.hs:8:10 • In the third argument of ‘GHC.Prim.coerce’, namely ‘f’ In the expression: GHC.Prim.coerce @String @String f :: String In an equation for ‘f’: f = GHC.Prim.coerce @String @String f :: String When typechecking the code for ‘f’ in a derived instance for ‘C T’: To see the code I am typechecking, use -ddump-deriv | 10 | newtype T = T () deriving C | ^ }}} ... and the following core is produced: {{{ ==================== Derived instances ==================== Derived class instances: instance Test.C Test.T where Test.f = GHC.Prim.coerce @GHC.Base.String @GHC.Base.String Test.f :: GHC.Base.String Derived type family instances: }}} The problem seems to be that the `a` should have been set to `()` within the coerced instance. I've been working round this with a `newtype X a = X String` as the result value so that the `a` is present in the signature, but I think this is a bug; should a more specialised instance be generated? I hope this is enough of an explanation! Thanks, Tom -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15637> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #15663: T9675 inexplicably regressed in allocations due to text submodule bump
by GHC 01 Oct '18

01 Oct '18

#15663: T9675 inexplicably regressed in allocations due to text submodule bump -------------------------------------+------------------------------------- Reporter: bgamari | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 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: -------------------------------------+------------------------------------- Harbormaster somehow started showing a significant (~20%) regression in `T9675`'s allocations starting in 989dca6cbd93. Strangely, this is a bump of the `text` submodule. Something is extremely fishy here. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15663> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #15661: Nullary constraint in GHCi breaks `:t` command
by GHC 01 Oct '18

01 Oct '18

#15661: Nullary constraint in GHCi breaks `:t` command -------------------------------------+------------------------------------- Reporter: taktoa | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: GHCi | Version: 8.4.3 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: -------------------------------------+------------------------------------- If you create a value whose type has a nullary constraint (i.e.: a constraint that does not reference any of the type variables in scope) and then try to run `:t` on it, GHCi attempts to run instance resolution and fails before printing the type. Expected output: {{{ GHCi, version 8.4.3: http://www.haskell.org/ghc/ :? for help Prelude> :set -XFlexibleContexts Prelude> data Foo = Foo Prelude> let x :: (Show Foo) => () ; x = () Prelude> :t x x :: Show Foo => () }}} Actual output: {{{ GHCi, version 8.4.3: http://www.haskell.org/ghc/ :? for help Prelude> :set -XFlexibleContexts Prelude> data Foo = Foo Prelude> let x :: (Show Foo) => () ; x = () Prelude> :t x <interactive>:1:1: error: No instance for (Show Foo) arising from a use of ‘x’ }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15661> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #15630: panic! Simplifier ticks exhausted
by GHC 01 Oct '18

01 Oct '18

#15630: panic! Simplifier ticks exhausted --------------------------------------+--------------------------------- Reporter: micahshahn | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.2.2 Keywords: | Operating System: Windows Architecture: x86_64 (amd64) | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: --------------------------------------+--------------------------------- Compiling the following with -O2 causes a panic. {{{#!hs module GHCPanic where data IValue = IDefault | IInt Int | IBlob String (?) :: Applicative m => (IValue -> m a) -> IValue -> m (Maybe a) (?) _ IDefault = pure Nothing (?) p x = Just <$> p x getInt :: IValue -> Either () Int getInt (IInt i) = Right i getInt v = Left () getString :: IValue -> Either () String getString (IBlob b) = Right $ b getString v = Left () (<+>) :: Applicative m => (m (a -> b), [IValue]) -> (IValue -> m a) -> (m b, [IValue]) (<+>) (f, (v:vs)) p = (f <*> (p v), vs) data TestStructure = TestStructure { _param1 :: Int , _param2 :: Maybe String , _param3 :: Maybe Int , _param4 :: Maybe String , _param5 :: Maybe Int , _param6 :: Maybe Int , _param7 :: Maybe String , _param8 :: Maybe String , _param9 :: Maybe Int , _param10 :: Maybe Int , _param11 :: Maybe String , _param12 :: Maybe String , _param13 :: Maybe Int , _param14 :: Maybe Int , _param15 :: Maybe String } getMenuItem :: [IValue] -> Either () TestStructure getMenuItem vs = fst $ (pure TestStructure, vs) <+> getInt <+> (getString ?) <+> (getInt ?) <+> (getString ?) <+> (getInt ?) <+> (getInt ?) <+> (getString ?) <+> (getString ?) <+> (getInt ?) <+> (getInt ?) <+> (getString ?) <+> (getString ?) <+> (getInt ?) <+> (getInt ?) <+> (getString ?) }}} {{{ ghc.exe: panic! (the 'impossible' happened) (GHC version 8.2.2 for x86_64-unknown-mingw32): Simplifier ticks exhausted When trying UnfoldingDone $j_s1y9 To increase the limit, use -fsimpl-tick-factor=N (default 100) If you need to do this, let GHC HQ know, and what factor you needed To see detailed counts use -ddump-simpl-stats Total ticks: 71323 Call stack: CallStack (from HasCallStack): prettyCurrentCallStack, called at compiler\utils\Outputable.hs:1133:58 in ghc:Outputable callStackDoc, called at compiler\utils\Outputable.hs:1137:37 in ghc:Outputable pprPanic, called at compiler\simplCore\SimplMonad.hs:199:31 in ghc:SimplMonad }}} This seems similar to #8319 which was marked as being fixed. It compiles (albeit very very slowly!) if I remove the last parameter and the last application of (<+>). -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15630> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #15691: Marking Pred(S n) = n as injective
by GHC 30 Sep '18

30 Sep '18

#15691: Marking Pred(S n) = n as injective -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.6.1 (Type checker) | Keywords: TypeFamilies, | Operating System: Unknown/Multiple InjectiveFamilies | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Should `Pred` be injective? Please close the ticket if this is a known limitation {{{#!hs {-# Language DataKinds, TypeFamilyDependencies #-} data N = O | S N type family Pred n = res | res -> n where Pred(S n) = n }}} fails with {{{ • Type family equation violates injectivity annotation. RHS of injective type family equation is a bare type variable but these LHS type and kind patterns are not bare variables: ‘'S n’ Pred ('S n) = n -- Defined at 462.hs:7:2 • In the equations for closed type family ‘Pred’ In the type family declaration for ‘Pred’ | 7 | Pred(S n) = n | ^^^^^^^^^^^^^ }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15691> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

Re: [GHC] #3073: Avoid reconstructing dictionaries in recursive instance methods
by GHC 30 Sep '18

30 Sep '18

#3073: Avoid reconstructing dictionaries in recursive instance methods -------------------------------------+------------------------------------- Reporter: simonpj | Owner: (none) Type: bug | Status: new Priority: lowest | Milestone: Component: Compiler | Version: 6.10.1 Resolution: | Keywords: Inlining Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Changes (by nh2): * cc: nh2 (added) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/3073#comment:16> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #10986: GHC should delete all temporary files it creates in /tmp
by GHC 30 Sep '18

30 Sep '18

#10986: GHC should delete all temporary files it creates in /tmp -------------------------------------+------------------------------------- Reporter: erikd | Owner: Type: feature | Status: new request | Priority: normal | Milestone: 8.0.1 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: -------------------------------------+------------------------------------- GHC creates large numbers of temporary files in /tmp. These would normally get cleaned up when the machine is rebooted, but for things like Jenkins build machines that don't get rebooted often they buld up pretty quickly. The /tmp on my Jenkins build machine currently has over 200k files and directories, but only about 40 of them are non-GHC related files. Under normal operation, GHC should delete all temporary files it creates. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10986> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 26

[GHC] #15623: Wrong Endian, ghc-pwd and ghc-cabal on ghc 8.0.1 for powerpc64le (IBM Power)
by GHC 30 Sep '18

30 Sep '18

#15623: Wrong Endian, ghc-pwd and ghc-cabal on ghc 8.0.1 for powerpc64le (IBM Power) -------------------------------------+------------------------------------- Reporter: | Owner: (none) francescantoncastro | Type: bug | Status: new Priority: highest | Milestone: 8.6.1 Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: powerpc64 | Type of failure: Building GHC | failed Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I tried to compile ghc-8.0.1 from source using the powerpc64/unknown linux .tar.xz file (since the other file was for AIX). The ghc-pwd and ghc-cabal files won't execute even after stripping them with strip. The libraries included are Big Endian instead of Little Endian. I replaced the libffi library by the local one on my system (IBM Power8 with redhatenterpriseserver). ghc complains there is no cabal file after naming a symlink to cabal (on /bin) ghc-cabal (instead of the faulty ghc-cabal file). See attached file. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15623> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #15690: Add Eq1/Ord1/Read1/Show1 instances to newtypes in Data.Semigroup
by GHC 29 Sep '18

29 Sep '18

#15690: Add Eq1/Ord1/Read1/Show1 instances to newtypes in Data.Semigroup -------------------------------------+------------------------------------- Reporter: Bodigrim | Owner: (none) Type: feature | Status: new request | Priority: normal | Milestone: 8.8.1 Component: | Version: 8.6.1 libraries/base | 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: -------------------------------------+------------------------------------- I'd like to have `Eq1`/`Ord1`/`Read1`/`Show1` instances for newtypes in `Data.Semigroup`, at least for `Min`, `Max` and `Option`. Is there a specific reason for their absence? If no, I'll be happy to prepare a PR. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15690> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0