January 2018 - ghc-tickets

[GHC] #14642: ‘class C using’ fails to parse with MonadComprehensions
by GHC 07 Jan '18

07 Jan '18

#14642: ‘class C using’ fails to parse with MonadComprehensions -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.2 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: -------------------------------------+------------------------------------- `TransformListComp` / `MonadComprehensions` reserve `using` but should this not work? {{{#!hs {-# Language TransformListComp #-} class C using }}} gives {{{ $ ghci -ignore-dot-ghci /tmp/X.hs GHCi, version 8.5.20180105: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Main ( /tmp/X.hs, interpreted ) /tmp/X.hs:3:9: error: parse error on input ‘using’ | 3 | class C using | ^^^^^ Failed, no modules loaded. Prelude> }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14642> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #14639: "unresolved constraints" when import and main missing
by GHC 06 Jan '18

06 Jan '18

#14639: "unresolved constraints" when import and main missing -------------------------------------+------------------------------------- Reporter: bainewedlock | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: GHCi | Version: 8.0.2 Keywords: | Operating System: Windows Architecture: | Type of failure: Compile-time Unknown/Multiple | crash or panic Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I gut a weird error because I forgot an import and the main method. (the commented lines in the code below) GHCi told me to post a bug about this, so... :-) == Steps to reproduce: {{{ > ghci GHCi, version 8.0.2: http://www.haskell.org/ghc/ :? for help Prelude> :l bug.hs [1 of 1] Compiling Main ( bug.hs, interpreted ) ghc.exe: panic! (the 'impossible' happened) (GHC version 8.0.2 for x86_64-unknown-mingw32): initTc: unsolved constraints WC {wc_insol = [W] printf_a2MG :: t_a2MF[tau:1] (CHoleCan: printf)} Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} == File bug.hs: {{{#!hs module Main where import Test.Hspec import Test.QuickCheck -- import Text.Printf testMovie :: Int -> Int -> Double -> Int -> Spec testMovie card ticket perc s = it (printf "should return Movie for card: %d, ticket: %d perc %s " card ticket (show perc)) $ movie card ticket perc `shouldBe` s movie :: Int -> Int -> Double -> Int movie card ticket perc = 4 -- main :: IO () -- main = undefined }}} == Expected behavior: Useful error message. == Actual behavior: "panic" error message. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14639> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #14638: Simplifier ticks exhausted on Ω lambda term
by GHC 05 Jan '18

05 Jan '18

#14638: Simplifier ticks exhausted on Ω lambda term -------------------------------------+------------------------------------- Reporter: merlynfry | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.2 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 panics when trying to compile the following program: {{{#!hs module Omega where newtype Mu a = Roll { unRoll :: Mu a -> a } omega = (\h -> $ Roll h) (\x -> unRoll x $ x) }}} However, GHC has no complaints about interpreting the module. ---- I've attached the most verbose output of GHC, should you so need it. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14638> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #14592: Totality checking
by GHC 05 Jan '18

05 Jan '18

#14592: Totality checking -------------------------------------+------------------------------------- Reporter: varosi | Owner: (none) Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 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: -------------------------------------+------------------------------------- Example of that is Idris totality checking (http://docs.idris- lang.org/en/latest/tutorial/theorems.html#sect-totality) This feature could be started as a very specific types of pure functions which could be found to be total for sure and improve it in time. I'm putting a ticket for discussion here, because I didn't find any ticket related to that in GHC Trac. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14592> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #14507: Core Lint error with Type.Reflection and pattern synonyms
by GHC 05 Jan '18

05 Jan '18

#14507: Core Lint error with Type.Reflection and pattern synonyms -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 Keywords: | Operating System: Unknown/Multiple PatternSynonyms | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- {{{#!hs {-# Language PatternSynonyms, ViewPatterns, GADTs, ConstraintKinds, RankNTypes, KindSignatures, PolyKinds, ScopedTypeVariables, DataKinds, TypeInType, TypeOperators, TypeApplications, TypeFamilies, TypeFamilyDependencies #-} import Type.Reflection import Prelude import Data.Kind data Dict c where Dict :: c => Dict c data N = O | S N foo :: forall (kind::k') (a::k). TypeRep kind -> TypeRep a -> Maybe (kind :~~: k, TypeRep a) foo krep rep = do HRefl <- eqTypeRep krep (typeRepKind rep) pure (HRefl, rep) pattern Bloop' :: forall k' (a::k). Typeable k' => k':~~:k -> TypeRep (a::k) -> TypeRep (a::k) pattern Bloop' hrefl rep <- (foo (typeRep @k') -> Just (hrefl, rep)) where Bloop' HRefl rep = rep type family SING = (res :: k -> Type) | res -> k where -- Core Lint error disappears with this line: SING = (TypeRep :: N -> Type) pattern RepN :: forall (a::kk). Typeable a => N~kk => SING a -> TypeRep (a::kk) pattern RepN tr <- Bloop' (HRefl::N:~~:kk) (tr :: TypeRep (a::N)) where RepN tr = tr asTypeable :: TypeRep a -> Dict (Typeable a) asTypeable rep = withTypeable rep Dict pattern TypeRep :: () => Typeable a => TypeRep a pattern TypeRep <- (asTypeable -> Dict) where TypeRep = typeRep -- error pattern SO = RepN TypeRep }}} triggers “Core Lint errors” on 8.2.1, 8.3.20170920 when run with `ghci -ignore-dot-ghci -dcore-lint`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14507> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #14608: Different GHCi error messages for similar scenarios
by GHC 04 Jan '18

04 Jan '18

#14608: Different GHCi error messages for similar scenarios -------------------------------------+------------------------------------- Reporter: mb64 | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: GHCi | Version: 8.2.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Poor/confusing Unknown/Multiple | error message Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- With the following code: {{{#!hs {-# LANGUAGE UnboxedTuples #-} module Test where data UnboxedTupleData = MkUTD (# (),() #) doThings :: UnboxedTupleData -> () doThings (MkUTD t) = () }}} This is accepted and compiled with a simple `ghc --make Test.hs`. However, with `ghci Test.hs`, you get an ugly error message: {{{#!hs GHCi, version 8.2.2: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Test ( Test.hs, interpreted ) ghc: panic! (the 'impossible' happened) (GHC version 8.2.2 for x86_64-unknown-linux): bcIdPrimRep t_s1ro :: (# (), () #) 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/ghci/ByteCodeGen.hs:1582:5 in ghc:ByteCodeGen Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} Replacing `(MkUTD t)` with just `t` obtains a different error, with a nicer message: {{{#!hs GHCi, version 8.2.2: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Test ( Test.hs, interpreted ) Error: bytecode compiler can't handle unboxed tuples and sums. Possibly due to foreign import/export decls in source. Workaround: use -fobject-code, or compile this module to .o separately. }}} True to its word, GHCi accepts it with `-fobject-code`, but it also accepts the unwrapping one, with the worse error message, with `-fobject- code`. It only happens with unboxed tuples in GHCi: if you replace the unboxed tuple with an unboxed int, GHCi will happily accept it. GHC cheerfully compiles everything. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14608> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #10731: System.IO.openTempFile is not thread safe on Windows
by GHC 04 Jan '18

04 Jan '18

#10731: System.IO.openTempFile is not thread safe on Windows -------------------------------------+------------------------------------- Reporter: | Owner: NeilMitchell | Type: bug | Status: new Priority: normal | Milestone: Component: | Version: 7.10.1 libraries/base | Keywords: | Operating System: Windows Architecture: | Type of failure: Runtime crash Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- Given the test program: {{{#!hs import Control.Concurrent import System.IO import Control.Monad import System.Directory main = do putStrLn "Starting" var <- newEmptyMVar tdir <- getTemporaryDirectory xs <- replicateM 10 $ do var <- newEmptyMVar flip forkFinally (\s -> do print s; putMVar var ()) $ do replicateM_ 100000 $ do (file, h) <- openTempFile tdir "test.txt" hClose h removeFile file return var mapM_ takeMVar xs }}} If I compile and run that with {{{ghc --make TmpFile.hs -threaded && tmpfile +RTS -N5}}} I get: {{{ Starting Left C:\Users\NDMIT_~1\AppData\Local\Temp\: openTempFile: permission denied (Permission denied) Left C:\Users\NDMIT_~1\AppData\Local\Temp\: openTempFile: permission denied (Permission denied) ... }}} I've reproduced this on GHC 7.8 and 7.10.1. We hit this in production about twice a day. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10731> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #14636: GHC Panic
by GHC 04 Jan '18

04 Jan '18

#14636: GHC Panic -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.2 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: -------------------------------------+------------------------------------- {{{ $ ghci -ignore-dot-ghci -XPolyKinds -XDataKinds GHCi, version 8.3.20171208: http://www.haskell.org/ghc/ :? for help Prelude> import Data.Proxy Prelude Data.Proxy> data ID :: a -> a -> * Prelude Data.Proxy> :kind (Proxy :: _ ID) ghc-stage2: panic! (the 'impossible' happened) (GHC version 8.3.20171208 for x86_64-unknown-linux): piResultTy k_a1yY[tau:1] ID Call stack: CallStack (from HasCallStack): callStackDoc, called at compiler/utils/Outputable.hs:1144:37 in ghc:Outputable pprPanic, called at compiler/types/Type.hs:951:35 in ghc:Type Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug Prelude Data.Proxy> }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14636> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #14633: -fwarn-redundant-constraints false positive
by GHC 04 Jan '18

04 Jan '18

#14633: -fwarn-redundant-constraints false positive -------------------------------------+------------------------------------- Reporter: ghorn | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Incorrect Unknown/Multiple | error/warning at compile-time Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I had code which compiled cleanly on GHC 8.0.2 with -fwarn-redundant- constraints which now gives a warning on GHC 8.2.2. Here is the code, and my workaround: {{{#!haskell {-# OPTIONS_GHC -Wall -Werror -fwarn-redundant-constraints #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE DataKinds #-} {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE FlexibleContexts #-} module Bug ( bug , workaround ) where import GHC.Generics ( D1, Datatype, Meta, Rep, datatypeName ) import Data.Proxy ( Proxy ) -- /home/greghorn/hslibs/ghc82_bug_maybe/Bug.hs:17:1: warning: [-Wredundant-constraints] -- • Redundant constraint: Rep a ~ D1 d p -- • In the type signature for: -- bug :: forall a (d :: Meta) (p :: * -> *). -- (Datatype d, Rep a ~ D1 d p) => -- Proxy a -> String -- | -- 25 | bug :: forall a d p . (Datatype d, Rep a ~ D1 d p) => Proxy a -> String -- | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ bug :: forall a d p . (Datatype d, Rep a ~ D1 d p) => Proxy a -> String bug = const name where name = datatypeName (undefined :: D1 d p b) type family GetD a :: Meta where GetD (D1 d p) = d workaround :: forall a d p . (Datatype (GetD (Rep a)), Rep a ~ D1 d p) => Proxy a -> String workaround = const name where name = datatypeName (undefined :: D1 d p b) }}} I suspect it is a bug because if I remove the "redundant" constraint it no longer typechecks. Here is a minimal setup to reproduce with stack: {{{ name: bug version: 0.0.0.2 license: AllRightsReserved author: Greg Horn maintainer: gregmainland(a)gmail.com build-type: Simple cabal-version: >=1.10 library exposed-modules: Bug build-depends: base >= 4.7 && < 5 default-language: Haskell2010 }}} {{{ resolver: lts-10.2 compiler-check: newer-minor # Local packages, usually specified by relative directory name packages: - . }}} Alternatively: {{{ git clone https://github.com/ghorn/ghc-redundant-constraint-bug cd ghc-redundant-constraint-bug stack build }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14633> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #14607: Core Lint error
by GHC 03 Jan '18

03 Jan '18

#14607: Core Lint error -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.2 Keywords: TypeInType, | Operating System: Unknown/Multiple DeferredTypeErrors | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: #14605 #14584 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- This produces long {{{#!hs {-# Language DerivingStrategies #-} {-# Language GADTs #-} {-# Language GeneralizedNewtypeDeriving #-} {-# Language InstanceSigs #-} {-# Language KindSignatures #-} {-# Language TypeFamilies #-} {-# Language TypeInType #-} {-# Language TypeOperators #-} import Data.Kind data DEFUNC :: Type -> Type -> Type where (:~>) :: a -> b -> DEFUNC a b type a ~> b = DEFUNC a b -> Type data LamCons a :: Type ~> Type where LamCons :: a -> LamCons a ([a] :~> [a]) class Mk (app :: Type ~> Type) where type Arg app :: Type mk :: Arg app -> app (a :~> b) instance Mk (LamCons a :: Type ~> Type) where type Arg (LamCons a) = a mk :: a -> LamCons (a :~> b) mk = LamCons }}} erros with `ghci -ignore-dot-ghci -fdefer-type-errors -dcore-lint bug.hs` on GHC 8.2.1 and 8.3.20171208. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14607> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5