January 2016 - ghc-tickets

[GHC] #11439: Request for comments: Allow duplicate type signatures
by GHC 17 May '17

17 May '17

#11439: Request for comments: Allow duplicate type signatures -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 7.10.3 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: -------------------------------------+------------------------------------- Interested in feedback (Code inspired by [http://www.cs.ox.ac.uk/projects/utgp/school/idris-tutorial.pdf Idris tutorial]) {{{#!hs sum :: SBool single -> IsSingleton single -> Natural sum STrue x = x sum SFalse [] = 0 sum SFalse (x : xs) = x + sum SFalse xs }}} Allow duplicate type signatures: {{{#!hs sum :: SBool single -> IsSingleton single -> Natural sum :: SBool True -> Natural -> Natural sum STrue x = x sum :: SBool False -> [Natural] -> Natural sum SFalse [] = 0 sum SFalse (x : xs) = x + sum SFalse xs }}} Where {{{#!hs data SBool b where STrue :: SBool True SFalse :: SBool False type family IsSingleton (b :: Bool) :: Type where IsSingleton True = Natural IsSingleton False = [Natural] }}} ---- == Behaviour == You could do something fancy but I would be happy with a check that any duplicate type signature is a specialization of the ''top'' top-level one, this would allow something like this: {{{#!hs length :: [a] -> Int length :: [()] -> Int length [] = 0 length (_:xs) = 1 + length xs }}} Which isn't very useful, hah what ever. ---- == Benefits == Mainly compiler-checked documentation, for the `sum` function we make it clearer that a `STrue` gives you `Natural -> Natural` while `SFalse` gives you `[Natural] -> Natural`. This is most apparent in the final syntax idea: {{{#!hs sum :: SBool single -> IsSingleton single -> Natural @True :: _ -> Natural -> Natural @False :: _ -> [Natural] -> Natural }}} Reader need not look at definition of `IsSingleton`, it is fully described by the duplicate signatures. See ''lens'' where most functions have some kind of type specialisation declared in comments: [https://hackage.haskell.org/package/lens-4.13/docs /Control-Lens-Fold.html#v:-94--63- (^?)] whose type is `(^?) :: s -> Getting (First a) s a -> Maybe a`: {{{#!hs (^?) :: s -> Getter s a -> Maybe a (^?) :: s -> Fold s a -> Maybe a (^?) :: s -> Lens' s a -> Maybe a (^?) :: s -> Iso' s a -> Maybe a (^?) :: s -> Traversal' s a -> Maybe a }}} ---- == Alternative syntax == Throwing it out there, regardless of ambiguity: {{{#!hs sum :: SBool single -> IsSingleton single -> Natural :: SBool True -> Natural -> Natural sum STrue x = x :: SBool False -> [Natural] -> Natural sum SFalse [] = 0 sum SFalse (x : xs) = x + sum SFalse xs }}} or {{{#!hs sum :: SBool single -> IsSingleton single -> Natural ... :: SBool True -> Natural -> Natural sum STrue x = x ... :: SBool False -> [Natural] -> Natural sum SFalse [] = 0 sum SFalse (x : xs) = x + sum SFalse xs }}} or {{{#!hs sum :: SBool single -> IsSingleton single -> Natural ... @True :: _ -> Natural -> Natural sum STrue x = x ... @False :: _ -> [Natural] -> Natural sum SFalse [] = 0 sum SFalse (x : xs) = x + sum SFalse xs }}} or {{{#!hs sum :: SBool single -> IsSingleton single -> Natural @True :: _ -> Natural -> Natural sum STrue x = x @False :: _ -> [Natural] -> Natural sum SFalse [] = 0 sum SFalse (x : xs) = x + sum SFalse xs }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11439> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #11272: Overloaded state-monadic function is not specialised
by GHC 13 May '17

13 May '17

#11272: Overloaded state-monadic function is not specialised -------------------------------------+------------------------------------- Reporter: NickSmallbone | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.3 Keywords: | 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: -------------------------------------+------------------------------------- I have a simple typeclass-polymorphic function which fails to specialise. Here is module A which defines the function `overloaded`: {{{#!hs module A where import Control.Monad.Trans.State import Control.Monad overloaded :: Ord a => a -> a -> State () () overloaded x y = do () <- get when (x <= y) (overloaded y x) }}} In module B I use `overloaded` on `Int`s: {{{#!hs module B where import A import Control.Monad.Trans.State specialised :: Int -> Int -> () specialised x y = execState (A.overloaded x y) () }}} Unfortunately the generated code is not specialised but passes an `Ord` dictionary around. It doesn't make any difference if I mark `overloaded` as `INLINEABLE` or not. In the core file, `overloaded` has been worker-wrapper transformed but the worker is marked `INLINEABLE[0]` - so I'm not sure why it's not being specialised. Curiously, if I make `overloaded` be a normal function instead of one in the state monad, or if I replace `() <- get` with simply `get`, specialisation goes through fine. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11272> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 11

[GHC] #9450: GHC instantiates Data instances before checking hs-boot files
by GHC 11 May '17

11 May '17

#9450: GHC instantiates Data instances before checking hs-boot files -------------------------------------+------------------------------------- Reporter: alanz | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.3 Keywords: boot,deriving | Operating System: Architecture: x86_64 (amd64) | Unknown/Multiple Difficulty: Unknown | Type of failure: Compile- Blocked By: | time crash Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- When compiling a file with deriving instances in it and making use of boot files, it appears that the derivation is attempted before the boot files are checked for consistency. This can lead to GHC panics, as when compiling GHC after this commit https://github.com/alanz/ghc/commit/c73182ca7345f0debba47d0b17a907bcac27c41f In the attached files, if you try to load HsLit.lhs into ghci it will complain about the missing Data instance, not about the boot file inconsistency. If you remove the Data from the deriving clause of HsOverLit, it complains about the boot file inconsistency. I have not managed to reproduce the original panic in this stripped down environment, but if it is needed to more fully understand the problem I can try again to do that. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9450> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #10584: Installation of SFML failed
by GHC 10 May '17

10 May '17

#10584: Installation of SFML failed -------------------------------------+------------------------------------- Reporter: SoleSoul | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.1 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- This report is bad because I have no idea what I'm reporting but the compiler asked me to so there you go: % cabal install sfml :( Resolving dependencies... Configuring SFML-0.2.0.0... Building SFML-0.2.0.0... Failed to install SFML-0.2.0.0 Build log ( /home/user/.cabal/logs/SFML-0.2.0.0.log ): Configuring SFML-0.2.0.0... Building SFML-0.2.0.0... Preprocessing library SFML-0.2.0.0... [ 1 of 71] Compiling SFML.Window.WindowHandle ( src/SFML/Window/WindowHandle.hs, dist/build/SFML/Window/WindowHandle.o ) [ 2 of 71] Compiling SFML.SFDisplayable ( src/SFML/SFDisplayable.hs, dist/build/SFML/SFDisplayable.o ) [ 3 of 71] Compiling SFML.SFCopyable ( src/SFML/SFCopyable.hs, dist/build/SFML/SFCopyable.o ) [ 4 of 71] Compiling SFML.System.InputStream ( dist/build/SFML/System/InputStream.hs, dist/build/SFML/System/InputStream.o ) [ 5 of 71] Compiling SFML.SFResource ( src/SFML/SFResource.hs, dist/build/SFML/SFResource.o ) [ 6 of 71] Compiling SFML.Utils ( src/SFML/Utils.hs, dist/build/SFML/Utils.o ) [ 7 of 71] Compiling SFML.Window.VideoMode ( dist/build/SFML/Window/VideoMode.hs, dist/build/SFML/Window/VideoMode.o ) [ 8 of 71] Compiling SFML.Window.Types ( src/SFML/Window/Types.hs, dist/build/SFML/Window/Types.o ) [ 9 of 71] Compiling SFML.Window.Keyboard ( dist/build/SFML/Window/Keyboard.hs, dist/build/SFML/Window/Keyboard.o ) [10 of 71] Compiling SFML.Window.Joystick ( dist/build/SFML/Window/Joystick.hs, dist/build/SFML/Window/Joystick.o ) [11 of 71] Compiling SFML.Window.ContextSettings ( dist/build/SFML/Window/ContextSettings.hs, dist/build/SFML/Window/ContextSettings.o ) [12 of 71] Compiling SFML.Window.Context ( dist/build/SFML/Window/Context.hs, dist/build/SFML/Window/Context.o ) [13 of 71] Compiling SFML.System.Vector3 ( dist/build/SFML/System/Vector3.hs, dist/build/SFML/System/Vector3.o ) [14 of 71] Compiling SFML.System.Vector2 ( dist/build/SFML/System/Vector2.hs, dist/build/SFML/System/Vector2.o ) [15 of 71] Compiling SFML.Window.Mouse ( dist/build/SFML/Window/Mouse.hs, dist/build/SFML/Window/Mouse.o ) [16 of 71] Compiling SFML.Window.Event ( dist/build/SFML/Window/Event.hs, dist/build/SFML/Window/Event.o ) [17 of 71] Compiling SFML.Window.SFWindow ( src/SFML/Window/SFWindow.hs, dist/build/SFML/Window/SFWindow.o ) [18 of 71] Compiling SFML.Window.Window ( dist/build/SFML/Window/Window.hs, dist/build/SFML/Window/Window.o ) [19 of 71] Compiling SFML.System.Time ( dist/build/SFML/System/Time.hs, dist/build/SFML/System/Time.o ) [20 of 71] Compiling SFML.Audio.SFSoundBuffer ( src/SFML/Audio/SFSoundBuffer.hs, dist/build/SFML/Audio/SFSoundBuffer.o ) [21 of 71] Compiling SFML.System.Sleep ( dist/build/SFML/System/Sleep.hs, dist/build/SFML/System/Sleep.o ) [22 of 71] Compiling SFML.System.Clock ( src/SFML/System/Clock.hs, dist/build/SFML/System/Clock.o ) [23 of 71] Compiling SFML.System ( src/SFML/System.hs, dist/build/SFML/System.o ) [24 of 71] Compiling SFML.Window ( src/SFML/Window.hs, dist/build/SFML/Window.o ) [25 of 71] Compiling SFML.Graphics.Types ( src/SFML/Graphics/Types.hs, dist/build/SFML/Graphics/Types.o ) [26 of 71] Compiling SFML.Graphics.SFSmoothTexture ( src/SFML/Graphics/SFSmoothTexture.hs, dist/build/SFML/Graphics/SFSmoothTexture.o ) [27 of 71] Compiling SFML.Graphics.SFShapeResizable ( src/SFML/Graphics/SFShapeResizable.hs, dist/build/SFML/Graphics/SFShapeResizable.o ) [28 of 71] Compiling SFML.Graphics.SFCoordSpace ( src/SFML/Graphics/SFCoordSpace.hs, dist/build/SFML/Graphics/SFCoordSpace.o ) [29 of 71] Compiling SFML.Graphics.SFBindable ( src/SFML/Graphics/SFBindable.hs, dist/build/SFML/Graphics/SFBindable.o ) [30 of 71] Compiling SFML.Graphics.Rect ( dist/build/SFML/Graphics/Rect.hs, dist/build/SFML/Graphics/Rect.o ) [31 of 71] Compiling SFML.Graphics.SFBounded ( src/SFML/Graphics/SFBounded.hs, dist/build/SFML/Graphics/SFBounded.o ) [32 of 71] Compiling SFML.Graphics.SFViewable ( src/SFML/Graphics/SFViewable.hs, dist/build/SFML/Graphics/SFViewable.o ) [33 of 71] Compiling SFML.Graphics.Texture ( src/SFML/Graphics/Texture.hs, dist/build/SFML/Graphics/Texture.o ) [34 of 71] Compiling SFML.Graphics.Transform ( dist/build/SFML/Graphics/Transform.hs, dist/build/SFML/Graphics/Transform.o ) ghc: panic! (the 'impossible' happened) (GHC version 7.10.1 for x86_64-unknown-linux): Simplifier ticks exhausted When trying UnfoldingDone $j_sJFj 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: 190129 Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug cabal: Error: some packages failed to install: SFML-0.2.0.0 failed during the building phase. The exception was: ExitFailure 1 cabal install sfml 7.18s user 0.76s system 93% cpu 8.460 total -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10584> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8

[GHC] #10323: Add Android and iOS Operating Systems in Trac
by GHC 10 May '17

10 May '17

#10323: Add Android and iOS Operating Systems in Trac -------------------------------------+------------------------------------- Reporter: rwbarton | Owner: hvr Type: feature | Status: new request | Milestone: Priority: normal | Version: Component: Trac & Git | Operating System: Unknown/Multiple Keywords: | Type of failure: None/Unknown Architecture: | Blocked By: Unknown/Multiple | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- These are sufficiently different from Linux/MacOS X to be distinguished from those Operating System values, and sufficiently mainstream to not fall under Other. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10323> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #9120: Cache intermediate powers
by GHC 10 May '17

10 May '17

#9120: Cache intermediate powers ------------------------------------+------------------------------------- Reporter: basvandijk | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- `GHC.Float` caches powers of 2 and 10. The arrays of powers are currently constructed using: {{{ expts :: Array Int Integer expts = array (minExpt,maxExpt) [(n,2^n) | n <- [minExpt .. maxExpt]] }}} Unfortunately, the intermediate powers in the `2^n` computation are not stored back in the array, only the result is. I propose to use the following scheme that does store the intermediate powers: {{{ -- | Exponentiation with a cache for the most common numbers. expt :: Integer -> Int -> Integer expt _ e | e < 0 = error "Negative exponent" expt 2 e | e <= maxExpt2 = expts2 ! e | otherwise = expts2 ! maxExpt2 * 2^(e-maxExpt2) expt 10 e | e <= maxExpt10 = expts10 ! e | otherwise = expts10 ! maxExpt10 * 10^(e-maxExpt10) expt base e = base^e maxExpt2 :: Int maxExpt2 = 1100 maxExpt10 :: Int maxExpt10 = 324 expts2 :: Array Int Integer expts2 = expts 2 maxExpt2 expts2 expts10 :: Array Int Integer expts10 = expts 10 maxExpt10 expts10 expts :: Integer -> Int -> Array Int Integer -> Array Int Integer expts base hi arr = listArray (0, hi) $ 1 : base : go 2 where go :: Int -> [Integer] go !ix = xx : base*xx : go (ix+2) where xx = x * x x = arr ! half half = ix `unsafeShiftR` 1 }}} I will attach a patch. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9120> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #8875: Track Exceptions
by GHC 10 May '17

10 May '17

#8875: Track Exceptions ----------------------------------------------+---------------------------- Reporter: yokto | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Project (more than a week) | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: ----------------------------------------------+---------------------------- Figure out for each function which exceptions can be thrown by it. This could probably be done by tracking calls to {{{ throw }}} and {{{ catch }}} and their types. The info would probably need to be stored in the .hi files, and could then be displayed by haddock. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8875> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #8396: cleanup / refactor native code gens
by GHC 10 May '17

10 May '17

#8396: cleanup / refactor native code gens ----------------------------------------+---------------------------------- Reporter: carter | Owner: Type: feature request | Status: new Priority: high | Milestone: 7.10.1 Component: Compiler | Version: Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: None/Unknown Blocked By: | Test Case: Related Tickets: #8299 , #8287 ,#8272 | Blocking: ----------------------------------------+---------------------------------- the native code gen could use some cleanup / love. 1. Catchup with feature parity of the llvm a. this includes adding SIMD support on applicable native code gen architectures, starting with x86_64 2. make it easier to extend / modify experiment with the native code gen 3. make it easier to add new primops (maybe even a magic inline asm primop?!) 4. other things i'm overlooking right now this also overlaps with a number of related tickets i've been adding to trac. a more ambitious strategic goal that won't be the initial goal, that simon marlow suggested when i chatted with him at ICFP, is to make the native code gen codebase a sort of "mini-llvm". I think thats a bit ambitious for the near term, but maybe thats the point on the horizon we may want to aim towards. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8396> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #8573: "evacuate: strange closure type 0" when creating large array
by GHC 10 May '17

10 May '17

#8573: "evacuate: strange closure type 0" when creating large array ----------------------------+---------------------------------- Reporter: nad | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Keywords: | Operating System: Linux Architecture: x86 | Type of failure: Runtime crash Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ----------------------------+---------------------------------- Consider the following code: {{{ module Main where import Data.Array xs :: [Int] xs = [0 .. 64988] crash :: Int -> IO () crash m = array (0, m) [ (x, x) | x <- xs ] `seq` return () strangeClosureType = do print (sum xs) crash (maxBound - 1) segFault1 = crash (maxBound - 1) segFault2 = do print (sum xs) crash (maxBound - 2) }}} If I compile the program using `ghc --make Main.hs -O -main-is strangeClosureType`, then I get the following error message: {{{ Main: internal error: evacuate: strange closure type 0 (GHC version 7.6.3 for i386_unknown_linux) Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug Aborted (core dumped) }}} If I don't use `-O`, or if I let `segFault1` or `segFault2` be `main`, then I get the following error message instead: {{{ Segmentation fault (core dumped) }}} If the number `30000` is replaced by some other number, then the strange closure error may be replaced by a segfault, or even no error at all. Perhaps this is another instance of bug #7762; I have only tested using GHC 7.6.3. I am using base 4.6.0.1 and array 0.4.0.1. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8573> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #8082: Ordering of assembly blocks affects performance
by GHC 10 May '17

10 May '17

#8082: Ordering of assembly blocks affects performance ----------------------------------+--------------------------------- Reporter: jstolarek | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler (NCG) | Version: 7.6.3 Keywords: | Operating System: Linux Architecture: x86_64 (amd64) | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ----------------------------------+--------------------------------- During my work on #6135 I noticed that performance of reverse-complem benchmark in nofib depends highly on the order in which assembly block are laid out. With my patches I am consistently getting a 18-20% speed-up. In theory my patches should not impact performance of existing programs, but for some reason they affect the ordering of generated assembly blocks. On of the earlier versions of my patch I noticed that kahan benchmark suffered a 16% performance hit and again the only difference I noticed in the generated assembly was ordering of blocks. I did a more in-depth investigation in case of kahan and it turned out that this difference results from the way Core is generated: the difference between HEAD and my patches was that a worker function had its three parameters passed in different order. I did not investigate this for reverse-complem because Core is considerably larger, but I could spend some time on it if it might be relevant. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8082> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 8