September 2017 - ghc-tickets

[GHC] #11959: Importing doubly exported pattern synonym and associated pattern synonym panics
by GHC 04 Oct '17

04 Oct '17

#11959: Importing doubly exported pattern synonym and associated pattern synonym panics -------------------------------------+------------------------------------- Reporter: darchon | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.0.1 Component: Compiler | Version: 8.0.1-rc3 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: -------------------------------------+------------------------------------- Given {{{#!hs {-# LANGUAGE PatternSynonyms, ViewPatterns #-} module Pat (Vec2(Nil,(:>)), pattern (:>)) where newtype Vec2 a = Vec2 {unvec2 :: [a]} pattern Nil :: Vec2 a pattern Nil = Vec2 [] pattern (:>) x xs <- ((\ys -> (head $ unvec2 ys,Vec2 . tail $ unvec2 ys)) -> (x,xs)) where (:>) x xs = Vec2 (x:unvec2 xs) }}} and {{{#!hs {-# LANGUAGE PatternSynonyms #-} module Main where import Pat (Vec2(..),pattern (:>)) }}} I get: {{{ $ ghci Main.hs GHCi, version 8.0.0.20160411: http://www.haskell.org/ghc/ :? for help [1 of 2] Compiling Pat ( Pat.hs, interpreted ) Pat.hs:2:29: warning: [-Wduplicate-exports] ‘:>’ is exported by ‘(:>)’ and ‘Vec2(Nil, type (:>))’ [2 of 2] Compiling Main ( Main.hs, interpreted ) ghc: panic! (the 'impossible' happened) (GHC version 8.0.0.20160411 for x86_64-unknown-linux): filterImports/combine (:>, :>, Nothing) (:>, Vec2{Vec2, :>, Nil}, Nothing) Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} Now, the warning about the duplicate export is of course correct. But the panic shouldn't happen when I try to import both the associated pattern synonym and the normal pattern synonym. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11959> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 17

[GHC] #12747: INLINE vs NOINLINE vs <nothing> give three different results; two would be better
by GHC 04 Oct '17

04 Oct '17

#12747: INLINE vs NOINLINE vs <nothing> give three different results; two would be better -------------------------------------+------------------------------------- Reporter: | Owner: MikolajKonarski | Type: feature | Status: new request | 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: -------------------------------------+------------------------------------- INLINE vs NOINLE vs <nothing> for this function https://github.com/LambdaHack/LambdaHack/blob/master/Game/LambdaHack/Client… produces distinct figures in all three cases. See https://ghc.haskell.org/trac/ghc/ticket/12603?replyto=8#comment:7 and some later comments, but I'd like to move here the discussion of the specific feature request that <nothing> should at any given point be equivalent to either INLINE or NOINLINE. Three different results are very unwieldy while optimizing code, because one cannot tweak the INLINE pragma for functions in isolation, via fixing INLINE or NOINLINE for all other functions and varying only one, because the optimal setting for some functions may be <nothing>, and it naturally varies (as it should; that is fine; but in each particular codebase/compilation flags state it should be equal to either of INLINE or NOINLINE). It seems almost as if GHC considers a function for inlining in some stages, optimizes accordingly, but changes its mind in other places, not being bound by either INLINE nor NOINLINE, and so the resulting stack of optimizations is different than in either case. IMHO that's too complex behaviour for the programmer to follow, it should irreversibly decide whether to inline at exactly the same place it first takes into account the NO/INLINE pragmas, if present. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12747> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #13561: Remove unsafe Chan combinators
by GHC 03 Oct '17

03 Oct '17

#13561: Remove unsafe Chan combinators -------------------------------------+------------------------------------- Reporter: bgamari | Owner: (none) Type: task | Status: new Priority: high | Milestone: 8.4.1 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: #4154 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Namely `isEmptyChan` and `unGetChan`. See #4154. Let's delete them for 8.4. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13561> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #13848: Unexpected order of variable quantification with GADT constructor
by GHC 03 Oct '17

03 Oct '17

#13848: Unexpected order of variable quantification with GADT constructor -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 (Type checker) | Keywords: | Operating System: Unknown/Multiple TypeApplications | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Here's some code: {{{#!hs {-# LANGUAGE DataKinds #-} {-# LANGUAGE ExistentialQuantification #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE KindSignatures #-} {-# LANGUAGE ScopedTypeVariables #-} module Foo where data N = Z | S N data Vec (n :: N) a where VNil :: forall a. Vec Z a VCons :: forall n a. a -> Vec n a -> Vec (S n) a }}} I want to use `TypeApplications` on `VCons`. I tried doing so in GHCi: {{{ GHCi, version 8.2.0.20170523: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 1] Compiling Foo ( Foo.hs, interpreted ) Ok, modules loaded: Foo. λ> :set -XTypeApplications -XDataKinds λ> :t VCons @Z @Int 1 VNil <interactive>:1:8: error: • Expected a type, but ‘'Z’ has kind ‘N’ • In the type ‘Z’ In the expression: VCons @Z @Int 1 VNil <interactive>:1:11: error: • Expected kind ‘N’, but ‘Int’ has kind ‘*’ • In the type ‘Int’ In the expression: VCons @Z @Int 1 VNil <interactive>:1:17: error: • Couldn't match type ‘'Z’ with ‘Int’ Expected type: Vec Int 'Z Actual type: Vec 'Z 'Z • In the fourth argument of ‘VCons’, namely ‘VNil’ In the expression: VCons @Z @Int 1 VNil }}} Huh? That's strange, I would have expected the first type application to be of kind `N`, and the second to be of kind `*`. But GHC disagrees! {{{ λ> :set -fprint-explicit-foralls λ> :type +v VCons VCons :: forall a (n :: N). a -> Vec n a -> Vec ('S n) a }}} That's downright unintuitive to me, since I explicitly specified the order in which the quantified variables should appear in the type signature for `VCons`. Similarly, if you leave out the `forall`s: {{{#!hs data Vec (n :: N) a where VNil :: Vec Z a VCons :: a -> Vec n a -> Vec (S n) a }}} Then `:type +v` will also report the same quantified variable order for `VCons`. This is perhaps less surprising, since the `n` and `a` in `data Vec (n :: N) a` don't scope over the constructors, so GHC must infer the topological order in which the variables appear in each constructor. But I would certainly not expect GHC to do this if I manually specify the order with `forall`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13848> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #9725: Constraint deduction failure
by GHC 03 Oct '17

03 Oct '17

#9725: Constraint deduction failure -------------------------------------+------------------------------------- Reporter: heisenbug | 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: GHC Difficulty: Unknown | rejects valid program Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- I get a strange error message like this {{{ [1 of 1] Compiling Main ( 08-10.hs, interpreted ) 08-10.hs:254:33: Could not deduce (KnownLoc (ent par1 a)) arising from a use of ‘Hidden'’ from the context (List ConfFamily ts) bound by the type signature for locAware :: List ConfFamily ts => (forall (i :: k). KnownLoc (ent par i) => ConfFamily (Facility (ent par i)) ts) -> Con ConfFamily (On ent par) at 08-10.hs:251:13-160 or from (ent par a ~ 'Monitor par1 mp, KnownLoc ('Monitor par1 mp)) bound by a pattern with constructor MonitorF :: forall (par :: Entity) (mp :: MonitorPoint). KnownLoc ('Monitor par mp) => String -> String -> Observer -> Bool -> Facility ('Monitor par mp), in an equation for ‘h’ at 08-10.hs:254:19-28 In the expression: Hidden' (d f) In an equation for ‘h’: h (Hide f@(MonitorF {})) = Hidden' (d f) In an equation for ‘locAware’: locAware descr = Abstr h where h :: (ent `On` par) -> ConfFamily (ent `On` par) ts h (Hide f@(MonitorF {})) = Hidden' (d f) d :: KnownLoc (ent par i) => Facility (ent par i) -> ConfFamily (Facility (ent par i)) ts d _ = descr Failed, modules loaded: none. }}} I have many language extensions active, such as, `DataKinds`, `ScopedTypeVariables`, etc. If somebody says this indeed looks like a bug, then I'll try to post a minimal reproduction case. ** My reasoning goes like this ** Here are the relevant facts that GHC figured out from a GADT pattern match: `(ent par a ~ 'Monitor par1 mp, KnownLoc ('Monitor par1 mp))` From this I discover these equalities: {{{ ent === 'Monitor (*) par === par1 (**) a === mp (***) }}} Substituring `(*)`^-1^ and `(***)`^-1^ into the given `KnownLoc ('Monitor par1 mp)` results in the desired `KnownLoc (ent par1 a)` Looks like a bug, right? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9725> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 21

[GHC] #13819: GHC Panics
by GHC 03 Oct '17

03 Oct '17

#13819: GHC Panics -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) 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: -------------------------------------+------------------------------------- {{{#!hs {-# LANGUAGE DeriveFunctor, TypeApplications #-} -- {-# Language ViewPatterns #-} -- {-# Language InstanceSigs, ViewPatterns, TupleSections, GeneralizedNewtypeDeriving, TemplateHaskell, LambdaCase #-} module D where import Data.Coerce import Control.Applicative newtype A a = A (IO a) deriving Functor instance Applicative A where pure = pure @(_ -> WrappedMonad A _) @(_ -> A _) pure instance Monad A where }}} causes a panic {{{ $ ghci -ignore-dot-ghci /tmp/a.hs GHCi, version 8.3.20170605: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling D ( /tmp/a.hs, interpreted ) ghc: panic! (the 'impossible' happened) (GHC version 8.3.20170605 for x86_64-unknown-linux): repSplitAppTys w0_a1xc[tau:4] WrappedMonad A w0_a1xe[tau:4] [] 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/types/Type.hs:809:9 in ghc:Type Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug > }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13819> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 17

[GHC] #14110: GHC Panic on over-saturated associated type family
by GHC 03 Oct '17

03 Oct '17

#14110: GHC Panic on over-saturated associated type family -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new 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: -------------------------------------+------------------------------------- If we over-saturate a type family {{{#!hs {-# Language TypeFamilies, ScopedTypeVariables, PolyKinds, DataKinds #-} import Data.Kind class R (c :: k -> Constraint) where type R_ (c :: k -> Constraint) :: k -> Type instance R Eq where type R_ Eq a = a -> a -> Bool }}} we get {{{ $ ghci -ignore-dot-ghci test.hs GHCi, version 8.3.20170605: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Main ( test.hs, interpreted ) ghc: panic! (the 'impossible' happened) (GHC version 8.3.20170605 for x86_64-unknown-linux): repSplitAppTys a_a2qu[sk:1] a_a2qu[sk:1] -> Bool [] 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/types/Type.hs:809:9 in ghc:Type Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug > }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14110> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #14232: Invalid type signature prokoves error in GHCi
by GHC 03 Oct '17

03 Oct '17

#14232: Invalid type signature prokoves error in GHCi -------------------------------------+------------------------------------- Reporter: walling | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: GHCi | Version: 8.2.1 Keywords: | Operating System: MacOS X Architecture: x86_64 | Type of failure: Poor/confusing (amd64) | error message Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- When I create and save a file with the following content (notice the missing arrow `->`): {{{#!hs f :: (String -> a) String -> a f g s = g s }}} ... it outputs this snippet in GHCi (copy-pasted from terminal): {{{ $ ghci bug.hs GHCi, version 8.2.1: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Main ( Playground/bug.hs, interpreted ) ghc: panic! (the 'impossible' happened) (GHC version 8.2.1 for x86_64-apple-darwin): repSplitAppTys String a_a1pl[sk:1] [] 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/types/Type.hs:808:9 in ghc:Type Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug > Leaving GHCi. }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14232> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #12759: Latest Debian GCC breajs GHC
by GHC 02 Oct '17

02 Oct '17

#12759: Latest Debian GCC breajs GHC ----------------------------------------+-------------------------- Reporter: erikd | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.0.2 Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: Unknown/Multiple | Type of failure: Other Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: ----------------------------------------+-------------------------- Just got `gcc (Debian 6.2.0-9) 6.2.0 20161019` in a recent upgrade (Debian Testing) and now: {{{ > cat hello.hs main :: IO () main = putStrLn "Hello" > ghc hello.hs -o hello [1 of 1] Compiling Main ( hello.hs, hello.o ) Linking hello ... /usr/bin/ld: hello.o: relocation R_X86_64_32S against symbol `stg_bh_upd_frame_info' \ can not be used when making a shared object; recompile with -fPIC /usr/bin/ld: /usr/lib/ghc-8.0/lib/base-4.9.0.0/libHSbase-4.9.0.0.a(Base__166.o): \ relocation R_X86_64_32S against symbol `stg_bh_upd_frame_info' can not be used when \ making a shared object; recompile with -fPIC .... }}} The previous version of gcc in Debian Testing (gcc (Debian 6.2.0-6) 6.2.0 20161010) is fine. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12759> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 27

[GHC] #13884: To an automatic compilation of file-header LANGUAGE pragmas in GHC.
by GHC 02 Oct '17

02 Oct '17

#13884: To an automatic compilation of file-header LANGUAGE pragmas in GHC. -------------------------------------+------------------------------------- Reporter: vanto | Owner: (none) Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 8.0.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Other Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I wrote this little piece of code to show that the compiler could very well execute the pragma by itself without being ordered to do so.\\ {{{ module Testfoo where import Language.Haskell.TH.Lib import Data.Kind tup = $(tupE $ take 4 $ cycle [ [| "hi" |] , [| 5 |] ]) data T = MkT (forall a. [a] -> [a]) answer_read = show (read @Int "3") data Foo (a :: Type) (b :: Type) where MkFoo :: (a ~ Int, b ~ Char) => Foo a b data family Song (a :: k) }}} By compiling the code without the Pragmas, GHC responds:\\ * 1) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) testfoo.hs:13:7: error: parse error on input `$' Perhaps you intended to use TemplateHaskell\\ * 2) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) testfoo.hs:15:23: error: Illegal symbol '.' in type Perhaps you intended to use {{{RankNTypes}}} or a similar language extension to enable explicit-forall syntax: forall <tvs>.<type>\\ * 3) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) testfoo.hs:19:16: error: Illegal kind signature: `Type' Perhaps you intended to use KindSignatures In the data type declaration for `Foo' testfoo.hs:19:28: error: Illegal kind signature: `Type' Perhaps you intended to use KindSignatures In the data type declaration for `Foo' testfoo.hs:22:1: error: Unexpected kind variable `k' Perhaps you intended to use {{{PolyKinds}}} In the declaration for type family `Song' testfoo.hs:22:24: error: Illegal kind signature: `k' Perhaps you intended to use KindSignatures In the declaration for type family `Song'\\ * 4) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) testfoo.hs:18:21: error: Pattern syntax in expression context: read@Int Did you mean to enable {{{TypeApplications}}}?\\ * 5) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) testfoo.hs:20:1: error: * Illegal generalised algebraic data declaration for `Foo' (Use {{{GADTs}}} to allow GADTs) * In the data declaration for `Foo'\\ * 6) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) testfoo.hs:23:1: error: * Illegal family declaration for `Song' Use {{{TypeFamilies}}} to allow indexed type families * In the data family declaration for `Song'\\ * 7) c:\Testghc>ghc testfoo.hs [1 of 1] Compiling Testfoo ( testfoo.hs, testfoo.o ) c:\Testghc>\\ Here is the code once completed. {{{ {-# LANGUAGE TemplateHaskell #-} {-# LANGUAGE RankNTypes #-} {-# LANGUAGE PolyKinds #-} {-# LANGUAGE TypeApplications #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE TypeFamilies #-} module Testfoo where import Language.Haskell.TH.Lib import Data.Kind tup = $(tupE $ take 4 $ cycle [ [| "hi" |] , [| 5 |] ]) data T = MkT (forall a. [a] -> [a]) answer_read = show (read @Int "3") data Foo (a :: Type) (b :: Type) where MkFoo :: (a ~ Int, b ~ Char) => Foo a b data family Song (a :: k) }}} As you may notice, GHC suggests the appropriate pragma.\\ If we add manually in the code the Pragmas one after the other and we arrive at the end, the code is fully compiled without error.\\ The compiler could do this alone.\\ We could test it using a "-auto" option on the compiler command line.\\ This is a start to the compiler automation technology, what do you think of that? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13884> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9