November 2017 - ghc-tickets

[GHC] #14427: GHCi output change breaks tooling users
by GHC 08 Nov '17

08 Nov '17

#14427: GHCi output change breaks tooling users -------------------------------------+------------------------------------- Reporter: bgamari | Owner: (none) Type: bug | Status: new Priority: high | Milestone: 8.2.2 Component: GHCi | 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: -------------------------------------+------------------------------------- In Phab:D3651 we changed the output produced by GHCi when loading a module. Unfortunately this broke `haskell-mode`, which relies on this output. Upstream tickets: * https://github.com/haskell/haskell-mode/issues/1553 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14427> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #14339: GHC 8.2.1 regression when combining GND with TypeError (solveDerivEqns: probable loop)
by GHC 08 Nov '17

08 Nov '17

#14339: GHC 8.2.1 regression when combining GND with TypeError (solveDerivEqns: probable loop) -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: high | Milestone: 8.4.1 Component: Compiler | Version: 8.2.1 (Type checker) | Keywords: deriving, | Operating System: Unknown/Multiple CustomTypeErrors | Architecture: | Type of failure: Compile-time Unknown/Multiple | crash or panic Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- This code panics on GHC 8.2.1 and later: {{{#!hs {-# LANGUAGE DataKinds #-} {-# LANGUAGE GeneralizedNewtypeDeriving #-} {-# LANGUAGE UndecidableInstances #-} module Bug where import GHC.TypeLits newtype Baz = Baz Foo deriving Bar newtype Foo = Foo Int class Bar a where bar :: a instance (TypeError (Text "Boo")) => Bar Foo where bar = undefined }}} {{{ $ /opt/ghc/8.2.1/bin/ghci Bug.hs GHCi, version 8.2.1: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 1] Compiling Bug ( Bug.hs, interpreted ) ghc: panic! (the 'impossible' happened) (GHC version 8.2.1 for x86_64-unknown-linux): solveDerivEqns: probable loop DerivSpec ds_loc = Bug.hs:9:12-14 ds_name = $fBarBaz ds_tvs = [] ds_cls = Bar ds_tys = [Baz] ds_theta = [ThetaOrigin to_tvs = [] to_givens = [] to_wanted_origins = [Bar Foo, (Foo :: *) ~R# (Baz :: *)]] ds_mechanism = newtype [[s0_a1D7[fuv:0]]] 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/typecheck/TcDerivInfer.hs:515:9 in ghc:TcDerivInfer }}} This is a regression since GHC 8.0.2, in which it does compile successfully. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14339> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #14430: lintUnfolding does not allow unfoldings to jump to enclosing join points
by GHC 08 Nov '17

08 Nov '17

#14430: lintUnfolding does not allow unfoldings to jump to enclosing join points -------------------------------------+------------------------------------- Reporter: nomeata | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.3 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: -------------------------------------+------------------------------------- I am not sure why this does not show up in HEAD but it does show up in my loopification branch. It does look like a bug in GHC to me. With loopification, I get the following code in the interface of `Foreign.C.String`: {{{ d5fd65f7834390bebd0e80edc7b8f627 withCAStringLen1 :: String -> (CStringLen -> IO a) -> State# RealWorld -> (# State# RealWorld, a #) {- Arity: 3, HasNoCafRefs, Strictness: <S,U><C(C(S(SL))),C(C1(U(U,U)))><S,U>, Unfolding: (\ @ a (str :: String) (f :: CStringLen -> IO a) (eta :: State# RealWorld) -> case lenAcc @ Char str newCAString2 of wild { I# x -> case newAlignedPinnedByteArray# @ RealWorld x 1# eta of ds2 { (#,#) ipv ipv1 -> case unsafeFreezeByteArray# @ RealWorld ipv1 ipv of ds3 { (#,#) ipv2 ipv3 -> let { ptr :: Addr# = byteArrayContents# ipv3 } in let { $j :: State# RealWorld -> () -> (# State# RealWorld, a #) <join 2> {- Arity: 2, Strictness: <S,U><L,A>, Inline: [0], Unfolding: InlineRule (2, True, False) (\ (w :: State# RealWorld) (w1 :: ()) -> case (f (Ptr @ CChar ptr, wild)) `cast` (N:IO[0] <a>_R) w of ds4 { (#,#) ipv4 ipv5 -> case touch# @ 'UnliftedRep @ ByteArray# ipv3 ipv4 of s4 { DEFAULT -> (# s4, ipv5 #) } }) -} = \ (w :: State# RealWorld)[OneShot] (w1 :: ())[OneShot] -> case (f (Ptr @ CChar ptr, wild)) `cast` (N:IO[0] <a>_R) w of ds4 { (#,#) ipv4 ipv5 -> case touch# @ 'UnliftedRep @ ByteArray# ipv3 ipv4 of s4 { DEFAULT -> (# s4, ipv5 #) } } } in let { go :: [Char] -> Int -> State# RealWorld -> (# State# RealWorld, a #) <join 3> {- Arity: 3, Strictness: <S,1*U><S(S),1*U(U)><S,U>, Inline: [~], Unfolding: InlineRule (3, True, False) (\ (ds :: [Char])[OneShot] (n :: Int)[OneShot] (eta1 :: State# RealWorld)[OneShot] -> letrec { go :: [Char] -> Int -> State# RealWorld -> (# State# RealWorld, a #) <join 3> {- Arity: 3 -} = \ (ds1 :: [Char]) (n1 :: Int) (eta2 :: State# RealWorld) -> case ds1 of wild1 { [] -> case n1 of n2 { I# ipv4 -> $j eta2 () } : c cs -> case n1 of wild2 { I# i -> case c of wild3 { C# c# -> case writeInt8OffAddr# @ RealWorld ptr i (narrow8Int# (ord# c#)) eta2 of s2 { DEFAULT -> go cs (I# (+# i 1#)) s2 } } } } } in go ds n eta1) -} = \ (ds :: [Char])[OneShot] (n :: Int)[OneShot] (eta1 :: State# RealWorld)[OneShot] -> case n of ww { I# ww1 -> let { exit :: State# RealWorld -> (# State# RealWorld, a #) <join 1> {- Arity: 1, Strictness: <S,U> -} = \ (w :: State# RealWorld)[OneShot] -> case (f (Ptr @ CChar ptr, wild)) `cast` (N:IO[0] <a>_R) w of ds4 { (#,#) ipv4 ipv5 -> case touch# @ 'UnliftedRep @ ByteArray# ipv3 ipv4 of s4 { DEFAULT -> (# s4, ipv5 #) } } } in letrec { $wgo :: [Char] -> Int# -> State# RealWorld -> (# State# RealWorld, a #) <join 3> {- Arity: 3, Strictness: <S,1*U><L,U><S,U>, Inline: [0] -} = \ (w :: [Char]) (ww2 :: Int#) (w1 :: State# RealWorld) -> case w of wild1 { [] -> exit w1 : c cs -> case c of wild2 { C# c# -> case writeInt8OffAddr# @ RealWorld ptr ww2 (narrow8Int# (ord# c#)) w1 of s2 { DEFAULT -> $wgo cs (+# ww2 1#) s2 } } } } in $wgo ds ww1 eta1 } } in go str newCAString2 ipv2 } } }) -} }}} Note how `go` references `j` not only in its RHS, but also in its unfolding. When loading this interface, I get this error: {{{ HC [stage 1] libraries/base/dist-install/build/System/Posix/Internals.o ghc-stage1: panic! (the 'impossible' happened) (GHC version 8.3.20171101 for x86_64-unknown-linux): Iface Lint failure In interface for Foreign.C.String Unfolding of go_a6h7 <no location info>: warning: In the expression: jump $j_a6gX eta2_a6hS () Invalid occurrence of a join variable: $j_a6gX The binder is either not a join point, or not valid here go_a6h7 = \ (ds_a6hM [OS=OneShot] :: [Char]) (n_a6hN [OS=OneShot] :: Int) (eta1_a6hO [OS=OneShot] :: State# RealWorld) -> joinrec { go_a6hP :: [Char] -> Int -> State# RealWorld -> (# State# RealWorld, a_a6gD #) [LclId[JoinId(3)], Arity=3] go_a6hP (ds1_a6hQ :: [Char]) (n1_a6hR :: Int) (eta2_a6hS :: State# RealWorld) = case ds1_a6hQ of wild1_a6hT { [] -> case n1_a6hR of n2_a6hW { I# ipv4_a6hY -> jump $j_a6gX eta2_a6hS () }; : c_a6i1 cs_a6i2 -> case n1_a6hR of wild2_a6i4 { I# i_a6i6 -> case c_a6i1 of wild3_a6i8 { C# c#_a6ia -> case writeInt8OffAddr# @ RealWorld ptr_a6gT i_a6i6 (narrow8Int# (ord# c#_a6ia)) eta2_a6hS of s2_a6ic { __DEFAULT -> jump go_a6hP cs_a6i2 (I# (+# i_a6i6 1#)) s2_a6ic } } } }; } in jump go_a6hP ds_a6hM n_a6hN eta1_a6hO Iface expr = \ (ds :: [Char])[OneShot] (n :: Int)[OneShot] (eta1 :: State# RealWorld)[OneShot] -> letrec { go :: [Char] -> Int -> State# RealWorld -> (# State# RealWorld, a #) <join 3> {- Arity: 3 -} = \ (ds1 :: [Char]) (n1 :: Int) (eta2 :: State# RealWorld) -> case ds1 of wild1 { [] -> case n1 of n2 { I# ipv4 -> $j eta2 () } : c cs -> case n1 of wild2 { I# i -> case c of wild3 { C# c# -> case writeInt8OffAddr# @ RealWorld ptr i (narrow8Int# (ord# c#)) eta2 of s2 { DEFAULT -> go cs (I# (+# i 1#)) s2 } } } } } in go ds n eta1 Call stack: CallStack (from HasCallStack): callStackDoc, called at compiler/utils/Outputable.hs:1147:37 in ghc:Outputable pprPanic, called at compiler/iface/TcIface.hs:1664:33 in ghc:TcIface }}} When reading the code in `tcPragExpr` I see that it lints the unfolding of `go` using `lintUnfolding`, which initializes the `LintM` with an empty `le_joins`. But when linting `go`, the join `j` is valid, even in the unfolding, isn't it? Unless of course I just don’t see the problem with the above unfolding, and am barking up the wrong tree. Anyways, either there is a problem in how we lint the unfoldings, or there is a secret invariant that I violated with my patch -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14430> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #13795: :kind! is not expanding type synonyms anymore
by GHC 08 Nov '17

08 Nov '17

#13795: :kind! is not expanding type synonyms anymore --------------------------------------+--------------------------------- Reporter: Hjulle | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.2 Keywords: | Operating System: Linux Architecture: x86_64 (amd64) | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: --------------------------------------+--------------------------------- Given {{{#!hs type A = () :kind! A }}} Expected result: {{{#!hs A :: * = () }}} Actual result: {{{#!hs A :: * = A }}} ---- Some IRC conversation on the topic (on #ghc): {{{ 23:37 < hjulle> :kind! does not seem to expand type synonyms for me (in GHC 8.0.2), it just prints them verbatim. Does anyone else have this problem? Example: "type A = ()" ":kind! A" will print out "A :: * = A" (which is not very helpful). 23:40 < hjulle> Is this a bug? The documentation (https://downloads.haskell.org/~ghc/latest/docs/html/users_guide/ghci.html #ghci-cmd-:kind) explicitly states that :kind! should expand type synonyms, so I think yes? 23:57 < RyanGlScott> hjulle: That's absolute a bug. File a ticket! 23:57 < RyanGlScott> *absolutely 23:58 < RyanGlScott> Moreover, I know why that's happening 23:59 < RyanGlScott> Internally, :kind! uses the normalise_type function to reduce type families: http://git.haskell.org/ghc.git/blob/e77b9a2069bca9018f989d7c4f54da099e3ab21… 23:59 < RyanGlScott> But see the comment there Day changed to 07 jun 2017 00:00 < RyanGlScott> -- Try to not to disturb type synonyms if possible 00:01 < RyanGlScott> So fixing this would just be a matter of calling coreView afterwards (which expands type synonyms) 00:02 < RyanGlScott> er, actually, expandTypeSynonyms is even better: http://git.haskell.org/ghc.git/blob/e77b9a2069bca9018f989d7c4f54da099e3ab21… }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13795> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 12

[GHC] #14440: Duplicate type family instances are permitted
by GHC 08 Nov '17

08 Nov '17

#14440: Duplicate type family instances are permitted -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 (Type checker) | Keywords: TypeFamilies | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- This threw me for a loop recently. To my surprise, GHC is quite happy to allow duplicate type family instances, provided that their RHSes are the same: {{{#!hs {-# LANGUAGE TypeFamilies #-} module Lib where type family Foo b }}} {{{#!hs {-# LANGUAGE TypeFamilies #-} module A where import Lib type instance Foo Bool = Bool }}} {{{#!hs {-# LANGUAGE TypeFamilies #-} module B where import Lib type instance Foo Bool = Bool }}} {{{#!hs module C where import A import B import Lib f :: Bool -> Foo Bool f x = not x }}} {{{ $ /opt/ghc/8.2.1/bin/ghci C.hs GHCi, version 8.2.1: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 4] Compiling Lib ( Lib.hs, interpreted ) [2 of 4] Compiling B ( B.hs, interpreted ) [3 of 4] Compiling A ( A.hs, interpreted ) [4 of 4] Compiling C ( C.hs, interpreted ) Ok, 4 modules loaded. λ> :i Foo type family Foo b :: * -- Defined at Lib.hs:4:1 type instance Foo Bool = Bool -- Defined at A.hs:6:15 type instance Foo Bool = Bool -- Defined at B.hs:6:15 }}} Is this intended? My intuition screams "no", since if we offer //class// instance coherence, it seems like one ought to offer //type family// instance coherence as well. At the same time, I can't think of any threat to type soundness imposed by this (although it's quite strange to see two duplicate type family instances in the output of `:i` with two completely different provenances). -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14440> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #14394: Inferred type for pattern synonym has redundant equality constraint
by GHC 08 Nov '17

08 Nov '17

#14394: Inferred type for pattern synonym has redundant equality constraint -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 (Type checker) | Keywords: | Operating System: Unknown/Multiple PatternSynonyms, TypeInType | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Load this file into GHCi: {{{#!hs {-# LANGUAGE GADTs #-} {-# LANGUAGE PatternSynonyms #-} module Bug where import Data.Type.Equality pattern Foo = HRefl }}} {{{ $ /opt/ghc/8.2.1/bin/ghci Bug.hs GHCi, version 8.2.1: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 1] Compiling Bug ( Bug.hs, interpreted ) Ok, 1 module loaded. λ> :i Foo pattern Foo :: () => (* ~ *, b ~ a) => a :~~: b }}} Notice that the type signature for `Foo` has an entirely redundant `* ~ *` constraint. The same does not happen if `TypeInType` is enabled. {{{#!hs {-# LANGUAGE GADTs #-} {-# LANGUAGE PatternSynonyms #-} {-# LANGUAGE TypeInType #-} module Works where import Data.Type.Equality pattern Foo = HRefl }}} {{{ λ> :i Foo pattern Foo :: forall k2 k1 (a :: k1) (b :: k2). () => (k2 ~ k1, (b :: k2) ~~ (a :: k1)) => a :~~: b }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14394> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #14408: error in refineFromInScope
by GHC 08 Nov '17

08 Nov '17

#14408: error in refineFromInScope -------------------------------------+------------------------------------- Reporter: duog | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.3 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: -------------------------------------+------------------------------------- There are several instances of the following line in the output in a validate build: {{{ WARNING: file compiler/simplCore/SimplEnv.hs, line 678 $j_setr }}} For example: https://phabricator.haskell.org/harbormaster/build/36855/?l=0, the harbourmaster build of 609f2844b92d5aa474f34b989c6ec5ad9fdb2ce3. That line indicates that said warning "is an error!" -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14408> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #12791: Superclass methods could be more aggressively specialised.
by GHC 08 Nov '17

08 Nov '17

#12791: Superclass methods could be more aggressively specialised. -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 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: -------------------------------------+------------------------------------- Say `R` is a superclass of `MR` but only uses one of the type variables. If this type variable is fixed then we know that we are going to use the specific dictionary. Thus, the optimiser *could* specialise all methods from the superclass at this point leading to better code. {{{#!hs {-# LANGUAGE ConstraintKinds #-} {-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE FlexibleContexts #-} module Foo where class R t => MR t m where push :: t -> m -> Int class R t where pull :: t -> Int --type MR2 t m = (R t, MR t m) instance MR Int Int where push = max instance R Int where pull = negate myf :: (MR Int a) => a -> Int -> Int myf _ = pull }}} To give a concrete example, `R` is a super class of `MR` but only mentions the first type variable. Thus when we fix it in `myf`, we could optimise the definition to `myf _ = negate` by inlining the class method. Reid points out that if you have a definition like {{{ data X = X f :: R X => Int -> Int f = pull }}} then the instance for `R X` could be provided by another module. However it is common to structure large applications with super class constraints so it would be desirable to do better. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12791> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 35

[GHC] #14433: Windows10 GHCI Crash, it can't run correctly
by GHC 08 Nov '17

08 Nov '17

#14433: Windows10 GHCI Crash, it can't run correctly --------------------------------------+---------------------------------- Reporter: Shizumi | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: GHCi | Version: 8.2.1 Keywords: | Operating System: Windows Architecture: x86_64 (amd64) | Type of failure: Runtime crash Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: --------------------------------------+---------------------------------- I followed the Haskel Platform website download the version 8.2.1 and installed it sucessfully, i add the extra path too.But when I open GHCI in start menu or in PowerShell,it cant run sucessfully, below are the error message, ps: here is Windows 10 Pro Insider Preview, does it matter or what? PS C:\Users\SHIZU-NOTEBOOK\Desktop> ghci GHCI, version 8.2.1: http://www.haskell.org/ghc/ :? for help ghc.exe: internal error: mkPath failed converting char* to wchar_t* (GHC version 8.2.1 for x86_64_unknown_mingw32) Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14433> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

Re: [GHC] #83: typo: %lt; should be <
by GHC 08 Nov '17

08 Nov '17

#83: typo: %lt; should be < ---------------------------------+-------------------- Reporter: nobody | Owner: nobody Type: bug | Status: closed Priority: normal | Milestone: Component: Documentation | Version: 5.04.1 Resolution: Fixed | Keywords: Type of failure: None/Unknown | ---------------------------------+-------------------- Changes (by Mateusz Kowalczyk <fuuzetsu@…>): * failure: => None/Unknown Comment: In [changeset:"ed18f47f931361a9adbb109085c6feb432ec41aa/ghc" ed18f47/ghc]: {{{ #!CommitTicketReference repository="ghc" revision="ed18f47f931361a9adbb109085c6feb432ec41aa" Factor out builds into steps. Address ghc/ghc#83 comments. This should greatly improve log output. }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/83#comment:2> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0