- ghc-tickets - Haskell.org

[GHC] #13882: Template variable unbound in rewrite rule
by GHC 28 Jun '17

28 Jun '17

#13882: Template variable unbound in rewrite rule --------------------------------------+--------------------------------- Reporter: achirkin | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1-rc2 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: --------------------------------------+--------------------------------- I'm getting GHC panic on `GHC 8.0.1`, `GHC 8.2.1-rc1`, and `GHC 8.2.1-rc2`. I think this error can somehow be related to https://ghc.haskell.org/trac/ghc/ticket/10924 or https://ghc.haskell.org/trac/ghc/ticket/13410 . On all three versions of compiler it builds fine with `-O1` or `-O0`. With `-O2` it crashes with the following error: {{{ [1 of 1] Compiling Numeric.Dimensions.Dim ( Dim.hs, Dim.o ) ghc: panic! (the 'impossible' happened) (GHC version 8.2.0.20170507 for x86_64-unknown-linux): Template variable unbound in rewrite rule Variable: ipv_s1e7 Rule "SC:concatDim0" Rule bndrs: [ipv_s1ee, ipv_s1ed, ipv_s1ef, ys_a19R, ipv_s1e4, ipv_s1e6, sc_s1fF, sc_s1fG, sc_s1fH, sc_s1fI, sc_s1fE, ipv_s1e7] LHS args: [TYPE: (ipv1_s1e6 |> ([Nth:0 (Sym ipv2)])_N), TYPE: ys_a19R, sc_s1fE, :* @ [Nat] @ ys_a19R @ ipv_s1ed @ ipv_s1ee @ ipv1_s1ef @~ (sc_s1fF :: ([Nat] :: *) ~# ([ipv_s1ed] :: *)) @~ (sc_s1fG :: (ys_a19R :: [Nat]) ~# (ConsDim ipv_s1ee ipv2_s1ef :: [ipv1_s1ed])) sc_s1fH sc_s1fI] Actual args: [TYPE: (ipv1_X1eZ |> ([Nth:0 (Sym ipv2)])_N), TYPE: ys_a19R, ipv_s1ea `cast` ((Dim ([Nth:0 (Sym ipv)])_N (Sym (Coh <ipv>_N ([Nth:0 (Sym ipv)])_N)))_R :: (Dim ipv1_X1eZ :: *) ~R# (Dim (ipv1_X1eZ |> ([Nth:0 (Sym ipv2)])_N) :: *)), :* @ [Nat] @ ys_a19R @ ipv_s1ed @ ipv_s1ee @ ipv1_s1ef @~ (sc_s1fF :: ([Nat] :: *) ~# ([ipv_s1ed] :: *)) @~ (sc_s1fG :: (ys_a19R :: [Nat]) ~# (ConsDim ipv_s1ee ipv2_s1ef :: [ipv1_s1ed])) sc_s1fH sc_s1fI] Call stack: CallStack (from HasCallStack): prettyCurrentCallStack, called at compiler/utils/Outputable.hs:1134:58 in ghc:Outputable callStackDoc, called at compiler/utils/Outputable.hs:1138:37 in ghc:Outputable pprPanic, called at compiler/specialise/Rules.hs:584:19 in ghc:Rules Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} Here is a minimal example I could come up with. Function `concatDim` is what causes the panic. {{{#!hs {-# LANGUAGE DataKinds #-} {-# LANGUAGE GADTs #-} {-# LANGUAGE KindSignatures #-} {-# LANGUAGE PolyKinds #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TypeFamilyDependencies #-} {-# LANGUAGE TypeOperators #-} {-# LANGUAGE UndecidableInstances #-} module Numeric.Dimensions.Dim (Dim (..), concatDim) where import Data.Type.Equality ((:~:)(..)) import GHC.TypeLits (Nat, TypeError, ErrorMessage (..)) import Unsafe.Coerce (unsafeCoerce) -- | Type-level dimensionality data Dim (ns :: k) where D :: Dim '[] (:*) :: forall (n::Nat) (ns::[k]) . Dim n -> Dim ns -> Dim (ConsDim n ns) Dn :: forall (n :: Nat) . Dim (n :: Nat) infixr 5 :* -- | Either known or unknown at compile-time natural number data XNat = XN Nat | N Nat -- | Unknown natural number, known to be not smaller than the given Nat type XN (n::Nat) = 'XN n -- | Known natural number type N (n::Nat) = 'N n -- | List concatenation type family (as :: [k]) ++ (bs :: [k]) :: [k] where (++) '[] bs = bs (++) as '[] = as (++) (a ': as) bs = a ': (as ++ bs) infixr 5 ++ type family Head (xs :: [k]) :: k where Head (x ': xs) = x Head '[] = TypeError ( 'Text "Head -- empty type-level list." ) type family Tail (xs :: [k]) :: [k] where Tail (x ': xs) = xs Tail '[] = TypeError ( 'Text "Tail -- empty type-level list." ) -- | Unify usage of XNat and Nat. -- This is useful in function and type definitions. -- Mainly used in the definition of Dim. type family ConsDim (x :: l) (xs :: [k]) = (ys :: [k]) | ys -> x xs l where ConsDim (x :: Nat) (xs :: [Nat]) = x ': xs ConsDim (x :: Nat) (xs :: [XNat]) = N x ': xs ConsDim (XN m) (xs :: [XNat]) = XN m ': xs concatDim :: forall (xs :: [Nat]) (ys :: [Nat]) . Dim xs -> Dim ys -> Dim (xs ++ ys) concatDim D ys = ys concatDim xs D = xs concatDim (x :* xs) ys = case unsafeCoerce Refl :: (xs ++ ys) :~: (Head xs ': (Tail xs ++ ys)) of Refl -> x :* concatDim xs ys }}} I am a little bit suspicious about `unsafeCoerce Refl`, but it worked in all other places so far. So if you find it not a bug but my incorrect usage of `unsafeCoerce`, please explain me why :) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13882> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #11990: Custom Type Error not getting triggered in the nested Type function call
by GHC 27 Jun '17

27 Jun '17

#11990: Custom Type Error not getting triggered in the nested Type function call -------------------------------------+------------------------------------- Reporter: magesh.b | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.0.1 Component: Compiler | Version: 8.0.1-rc3 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 have partial type function which is invoked by another type function. When the inner type function fails with TypeError, outer type function is not been able to propagate that type error to its caller. As a result of it, I'm getting following error • No instance for (KnownSymbol (NestedPartialTF (TypeError ...))) instead of • Unexpected type @ NestedPartialTF: Char {{{#!hs {-# LANGUAGE DataKinds, TypeOperators, TypeFamilies, UndecidableInstances, ScopedTypeVariables, FlexibleContexts #-} -- | module CErrs where import GHC.TypeLits import Data.Proxy type family PartialTF t :: Symbol where PartialTF Int = "Int" PartialTF Bool = "Bool" PartialTF a = TypeError (Text "Unexpected type @ PartialTF: " :<>: ShowType a) type family NestedPartialTF (tsym :: Symbol) :: Symbol where NestedPartialTF "Int" = "int" NestedPartialTF "Bool" = "bool" NestedPartialTF a = TypeError (Text "Unexpected type @ NestedPartialTF: " :<>: ShowType a) testPartialTF :: forall a.(KnownSymbol (PartialTF a)) => a -> String testPartialTF t = symbolVal (Proxy :: Proxy (PartialTF a)) --t1 = testPartialTF 'a' {- Above code rightly fails with the following error: • Unexpected type: Char • In the expression: testPartialTF 'a' In an equation for ‘t1’: t1 = testPartialTF 'a' -} -- Bug? testNesPartialTF :: forall a.(KnownSymbol (NestedPartialTF (PartialTF a))) => a -> String testNesPartialTF t = symbolVal (Proxy :: Proxy (NestedPartialTF (PartialTF a))) t2 = testNesPartialTF 'a' {- Above code fails with the following error: • No instance for (KnownSymbol (NestedPartialTF (TypeError ...))) arising from a use of ‘testNesPartialTF’ • In the expression: testNesPartialTF 'a' In an equation for ‘t2’: t2 = testNesPartialTF 'a' -} }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11990> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 12

[GHC] #13880: panic initTc: unsolved constraints
by GHC 27 Jun '17

27 Jun '17

#13880: panic initTc: unsolved constraints -------------------------------------+------------------------------------- Reporter: vanto | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Compile-time Unknown/Multiple | crash or panic Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I'm trying to play with some piece of code in GHCi. Sorry if it's a duplicate bug.\\ {{{ {-# LANGUAGE TemplateHaskell #-} module Testbar where printf :: String -> String printf s = gen (parse s) $(printf "Error: %s at line %d") }}} {{{ Prelude> :l testbar [1 of 1] Compiling Testbar ( testbar.hs, interpreted ) ghc.exe: panic! (the 'impossible' happened) (GHC version 8.0.2 for i386-unknown-mingw32): initTc: unsolved constraints WC {wc_insol = [W] gen_a10Q :: t_a10P[tau:1] (CHoleCan: gen) [W] parse_a10Z :: t_a10Y[tau:1] (CHoleCan: parse)} Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13880> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

Re: [GHC] #602: Warning Suppression
by GHC 27 Jun '17

27 Jun '17

#602: Warning Suppression -------------------------------------+------------------------------------- Reporter: simonmar | Owner: (none) Type: task | Status: new Priority: lowest | Milestone: Component: Compiler | Version: None Resolution: None | Keywords: warnings Operating System: Unknown/Multiple | Architecture: Type of failure: Incorrect | Unknown/Multiple warning at compile-time | Test Case: N/A Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | Design/LocalWarningPragmas | -------------------------------------+------------------------------------- Changes (by ntc2): * cc: ntc2 (added) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/602#comment:31> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #8033: add AVX register support to llvm calling convention
by GHC 27 Jun '17

27 Jun '17

#8033: add AVX register support to llvm calling convention -------------------------------------+------------------------------------- Reporter: carter | Owner: (none) Type: task | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.7 Resolution: | Keywords: SIMD Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by carter): sounds good to me :) might be worth doing 0-7 to have 8 rather than only 6, but that would require ABI changes on the ghc side, though even this patch does too :) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8033#comment:33> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #8033: add AVX register support to llvm calling convention
by GHC 27 Jun '17

27 Jun '17

#8033: add AVX register support to llvm calling convention -------------------------------------+------------------------------------- Reporter: carter | Owner: (none) Type: task | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.7 Resolution: | Keywords: SIMD Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by bgamari): In the interest of unsticking this, I'm going to submit this patch upstream to LLVM tomorrow if there is no objection, {{{#!patch diff --git a/lib/Target/X86/X86CallingConv.td b/lib/Target/X86/X86CallingConv.td index 7d146d050a5..cb6ed24a155 100644 --- a/lib/Target/X86/X86CallingConv.td +++ b/lib/Target/X86/X86CallingConv.td @@ -650,8 +650,16 @@ def CC_X86_64_GHC : CallingConv<[ // Pass in STG registers: F1, F2, F3, F4, D1, D2 CCIfType<[f32, f64, v16i8, v8i16, v4i32, v2i64, v4f32, v2f64], - CCIfSubtarget<"hasSSE1()", - CCAssignToReg<[XMM1, XMM2, XMM3, XMM4, XMM5, XMM6]>>> + CCIfSubtarget<"hasSSE1()", + CCAssignToReg<[XMM1, XMM2, XMM3, XMM4, XMM5, XMM6]>>>, + // AVX + CCIfType<[v32i8, v16i16, v8i32, v4i64, v8f32, v4f64], + CCIfSubtarget<"hasAVX()", + CCAssignToReg<[YMM1, YMM2, YMM3, YMM4, YMM5, YMM6]>>>, + // AVX-512 + CCIfType<[v64i8, v32i16, v16i32, v8i64, v16f32, v8f64], + CCIfSubtarget<"hasAVX512()", + CCAssignToReg<[ZMM1, ZMM2, ZMM3, ZMM4, ZMM5, ZMM6]>>> ]>; def CC_X86_64_HiPE : CallingConv<[ }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8033#comment:32> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #13337: GHC doesn't think a type is of kind *, despite having evidence for it
by GHC 27 Jun '17

27 Jun '17

#13337: GHC doesn't think a type is of kind *, despite having evidence for it -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.2 (Type checker) | 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: -------------------------------------+------------------------------------- The easiest way to illustrate this bug is this: {{{#!hs {-# LANGUAGE PolyKinds #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TypeFamilies #-} module Foo where import Data.Kind (Type) blah :: forall (a :: k). k ~ Type => a -> a blah x = x }}} {{{ $ ghci Foo.hs GHCi, version 8.0.2: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 1] Compiling Foo ( Foo.hs, interpreted ) Foo.hs:8:43: error: • Expected a type, but ‘a’ has kind ‘k’ • In the type signature: blah :: forall (a :: k). k ~ Type => a -> a }}} I find this behavior quite surprising, especially since we have evidence that `k ~ Type` in scope! If the program above looks too contrived, consider a similar program that uses `Typeable`. I want to write something like this: {{{#!hs {-# LANGUAGE RankNTypes #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE TypeInType #-} {-# LANGUAGE TypeOperators #-} module Foo where import Data.Kind (Type) import Data.Typeable foo :: forall k (a :: k) proxy. (Typeable k, Typeable a) => proxy a -> String foo _ = case eqT :: Maybe (k :~: Type) of Nothing -> "Non-Type kind" Just Refl -> case eqT :: Maybe (a :~: Int) of Nothing -> "Non-Int type" Just Refl -> "It's an Int!" }}} This exhibits the same problem. Despite the fact that we pattern-matched on a `Refl` of type `k :~: Type`, GHC refuses to consider the possibility that `a :~: Int` is well kinded, even though `(a :: k)`, and we learned from the first `Refl` that `k ~ Type`! -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13337> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 18

[GHC] #12912: IO library should not use select()
by GHC 27 Jun '17

27 Jun '17

#12912: IO library should not use select() -------------------------------------+------------------------------------- Reporter: simonmar | Owner: simonmar Type: bug | Status: new Priority: normal | Milestone: 8.0.2 Component: | Version: 8.0.1 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: -------------------------------------+------------------------------------- The IO library uses `select()` in `fdReady()`, which fails in processes with more than 1024 file descriptors. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12912> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #13855: Syntactic sugar to write the recursion in GHC
by GHC 27 Jun '17

27 Jun '17

#13855: Syntactic sugar to write the recursion 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: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The recursion is defined by two case that are a base case and a recursive case.\\ Instead of writing two functions to describe the recursion, I propose to write a single function that makes both.\\ {{{<op>}}} should be set to tell the compiler how to compile the function.\\ syntaxe : {{{ exp = <base_case_exp1;...;base_case_expn> <op> <recursive_case_exp1;...;recursive_case_expn> <recursive_case_guards> }}} examples: {{{ fac :: Int -> Int fac n = 0 = 1 <op> n * fac (n-1) }}} {{{ product :: Num a => [a] -> a product (n:ns) = [] = 1 <op> n * product ns }}} {{{ reverse :: [a] -> [a] reverse (x:xs) = [] = [] <op> reverse xs ++ [x] }}} {{{ insert :: Ord a => a -> [a] -> [a] insert x (y:ys) = x [] = [x] <op> | x <= y = x:y:ys | otherwise = y:insert x ys }}} {{{ drop :: Int -> [a] -> [a] drop n (_:xs) = 0 xs = xs; _ [] = [] <op> drop (n-1) xs }}} {{{ fib :: Int -> Int fib n = 0 = 0; 1 = 1 <op> fib (n-2) + fib (n-1) }}} If someone is interested in this, I encourage him to write a proposal. I don't know how to do it. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13855> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

Re: [GHC] #7080: Make RULES and SPECIALISE more consistent
by GHC 27 Jun '17

27 Jun '17

#7080: Make RULES and SPECIALISE more consistent -------------------------------------+------------------------------------- Reporter: simonpj | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: ⊥ Component: Compiler | Version: 7.4.2 Resolution: | Keywords: Specialise 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 simonpj): * keywords: => Specialise -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7080#comment:3> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0