August 2018 - ghc-tickets

[GHC] #14259: Worker/Wrapper for sum return
by GHC 26 Jan '19

26 Jan '19

#14259: Worker/Wrapper for sum return -------------------------------------+------------------------------------- Reporter: jheek | 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: -------------------------------------+------------------------------------- GHC version 8.2 introduces Unboxed Sum types. It would be great if the worker/wrapper transformation could make use of this new functionality. For clarification I would expect a function like this: {{{#!haskell maybeEven :: Int -> Maybe Int maybeEven n = case even n of True -> Just n False -> Nothing }}} to be transformed into (the core equivalent) of {{{#!haskell maybeEven :: Int -> Maybe Int MaybeEven (I# n#) = case workerMaybeEven n# of (# | x #) -> Just x (# (# #) | #) -> Nothing workerMaybeEven :: Int# -> (# (# #) | Int #) {-# NOINLINE workerMaybeEven #-} workerMaybeEven n# = case even (I# n#) of True -> (# | I# n# #) False -> (# (# #) | #) }}} Currently the core output for the maybeEven worker is: {{{#!haskell Main.$wmaybeEven :: Int# -> Maybe Int Main.$wmaybeEven = \ (ww_s4WH :: Int#) -> case remInt# ww_s4WH 2# of { __DEFAULT -> GHC.Base.Nothing @ Int; 0# -> GHC.Base.Just @ Int (GHC.Types.I# ww_s4WH) } }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14259> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #12364: Demand analysis for sum types
by GHC 25 Jan '19

25 Jan '19

#12364: Demand analysis for sum types -------------------------------------+------------------------------------- Reporter: nomeata | Owner: Type: task | 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: -------------------------------------+------------------------------------- While working on #10613 it crossed my mind that it might be worthwhile to expand the demand type also onto sum types. So instead of {{{ | UProd [ArgUse] -- Product }}} we would have {{{ | UData [[ArgUse]] -- Data type }}} and a function like {{{ fromMaybe :: a -> Maybe a -> b }}} would have a signature of {{{ <1*U><1*U(;1*U)> }}} which indicates that the second argument of `fromMaybe` is evaluated at most once; the first constructor of the result has no arguments, the second (separated by `;`) has one argument which is also used at most once. I could imagine that this gives a lot of useful information with parsers and other code that repeatedly retuns stuff wrapped in a `Maybe` or similar type. Now, sum types are often recursive (`[]`…), and we probably want to be smarter about recursion here. But note that this is not a new problem. If you write {{{ data Stream = Stream Int Stream Stream Stream foo (Stream 0 x y z) = 0 foo (Stream 1 x y z) = foo x foo (Stream 2 x y z) = foo y foo (Stream _ x y z) = foo z }}} you already get huge demand signatures. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12364> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 13

[GHC] #12599: Add Hadrian build artifacts to gitignore
by GHC 25 Jan '19

25 Jan '19

#12599: Add Hadrian build artifacts to gitignore -------------------------------------+------------------------------------- Reporter: bgamari | Owner: Type: task | 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: -------------------------------------+------------------------------------- Andrey wrote requesting the following, > For top-level `.gitignore`: > {{{ > # ------------------------------------------------------------------------------------- > # Hadrian's stuff. Note, **/stageN/ folders will be relocated to _build/ after > # resolving https://github.com/snowleopard/hadrian/issues/113 > # ------------------------------------------------------------------------------------- > _build/ > hadrian/ > **/stage?/ > }}} > > For `.gitignore` in affected submodules (array, deepseq, directory, filepath, > hoopl, hpc, process, utils/hsc2hs): > > {{{ > # ------------------------------------------------------------------------------------- > # Hadrian's stuff. Note, stageN/ folders will be relocated to top-level _build/ > # after resolving https://github.com/snowleopard/hadrian/issues/113 > # ------------------------------------------------------------------------------------- > stage?/ > }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12599> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 9

[GHC] #9476: Implement late lambda-lifting
by GHC 25 Jan '19

25 Jan '19

#9476: Implement late lambda-lifting -------------------------------------+------------------------------------- Reporter: simonpj | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- This ticket exists to coordinate work on Nick Frisby's '''late lambda- lifting''' optimisation. * Wiki page: [wiki:LateLamLift]. All the details are here. * Branch in GHC repo: `wip/llf` * Related tickets * #9279: local closure remains in final program * #9374: static argument transformation -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9476> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 82

[GHC] #15437: Internal error when applying a scoped type variable inside a typed expression quotation
by GHC 24 Jan '19

24 Jan '19

#15437: Internal error when applying a scoped type variable inside a typed expression quotation -------------------------------------+------------------------------------- Reporter: dminuoso | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 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: -------------------------------------+------------------------------------- {{{#!hs {-# LANGUAGE TemplateHaskell #-} import TestMod f :: Int f = $$(foo) main :: IO () main = main }}} {{{#!hs {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TemplateHaskell #-} {-# LANGUAGE TypeApplications #-} module TestMod where import Language.Haskell.TH.Syntax (Q, TExp) get :: forall a. Int get = 1 foo :: forall a. Q (TExp Int) foo = [|| get @a ||] }}} {{{ Test.hs:6:8: error: • The exact Name ‘a’ is not in scope Probable cause: you used a unique Template Haskell name (NameU), perhaps via newName, but did not bind it If that's it, then -ddump-splices might be useful • In the result of the splice: $foo To see what the splice expanded to, use -ddump-splices In the Template Haskell splice $$(foo) In the expression: $$(foo) | 6 | f = $$(foo) | ^^^ Test.hs:6:8: error: • GHC internal error: ‘a’ is not in scope during type checking, but it passed the renamer tcl_env of environment: [r3Kl :-> Identifier[f::Int, TopLevelLet [] True], r3PI :-> Identifier[main::IO (), TopLevelLet [r3PI :-> main] True]] • In the type ‘a’ In the expression: get @a In the result of the splice: $foo To see what the splice expanded to, use -ddump-splices | 6 | f = $$(foo) | }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15437> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #13253: Exponential compilation time with RWST & ReaderT stack with `-02`
by GHC 24 Jan '19

24 Jan '19

#13253: Exponential compilation time with RWST & ReaderT stack with `-02` -------------------------------------+------------------------------------- Reporter: phadej | 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: -------------------------------------+------------------------------------- See the attached file. The compilation time grows exponentially with additional fields in `HuteStruct`. Tried with GHC-8.0.1 and quite recent HEAD (8.2) adding a field doubles the compilation time with `-O2`. - 9: 17s - 10: 34s - 11: 83s on my machine. With `-01` around 3sec, independently of the struct size. The original issue: https://github.com/yesodweb/yesod/issues/1348 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13253> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 45

[GHC] #12920: Overzealous unused-top-binds
by GHC 23 Jan '19

23 Jan '19

#12920: Overzealous unused-top-binds -------------------------------------+------------------------------------- Reporter: crockeea | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.2-rc1 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: -------------------------------------+------------------------------------- The following example produces three warnings: {{{ module Foo () where type A = Int type B = A foo :: IO () foo = print (bar 3 :: B) bar :: a -> a bar = id }}} `ghci Foo -Wall` produces {{{ GHCi, version 8.0.1.20161117: http://www.haskell.org/ghc/ :? for help [1 of 1] Compiling Foo ( Foo.hs, interpreted ) Foo.hs:5:1: warning: [-Wunused-top-binds] Defined but not used: type constructor or class ‘B’ Foo.hs:8:1: warning: [-Wunused-top-binds] Defined but not used: ‘foo’ Foo.hs:11:1: warning: [-Wunused-top-binds] Defined but not used: ‘bar’ Ok, modules loaded: Foo. }}} The three errors reported are inconsistent because GHC reports that `B` is unused, but somehow feels that `A` *is* used. It seems to me that two possibilities could work: 1. GHC reports 1 error: GHC should only report that `foo` is unused, since `A` is used in `B`, `B` is used in `foo`, and `bar` is used in `foo`. This seems like the best approach to me. 2. GHC reports 4 errors: One might choose to apply the definition of "unused" transitively so that, since `foo` is unused, everything used in `foo` is considered unused (unless "used" elsewhere). In this case, `A` should also be reported as unused, but it currently isn't. This definition would be exceedingly useful for removing dead code, but it also produces very noisy output when there is a lot of dead code. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12920> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #15470: Record projections with ambiguous types
by GHC 23 Jan '19

23 Jan '19

#15470: Record projections with ambiguous types -------------------------------------+------------------------------------- Reporter: sweirich | Owner: (none) Type: feature | Status: new request | Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 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: -------------------------------------+------------------------------------- The following code fails {{{ {-# LANGUAGE DataKinds, KindSignatures, PolyKinds, TypeFamilies, RankNTypes, AllowAmbiguousTypes #-} module TypeLambda where data Fun a b type family App (f :: Fun t k) (x :: t) :: k data MONAD m = MONAD { return :: forall a. a -> App m a , (>>=) :: forall a b. App m a -> (a -> App m b) -> App m b } }}} with the error message {{{ TypeLambda.hs:12:5: error: • Couldn't match type ‘App m b0’ with ‘App m b’ Expected type: App m a -> (a -> App m b) -> App m b Actual type: App m a -> (a -> App m b0) -> App m b0 NB: ‘App’ is a non-injective type family The type variable ‘b0’ is ambiguous • In the expression: (>>=) In an equation for ‘TypeLambda.>>=’: (TypeLambda.>>=) MONAD {>>= = (>>=)} = (>>=) • Relevant bindings include (>>=) :: forall a b. App m a -> (a -> App m b) -> App m b (bound at TypeLambda.hs:12:5) | 12 | , (>>=) :: forall a b. App m a -> (a -> App m b) -> App m b | ^^^^^ Failed, no modules loaded. }}} However, with ScopedTypeVariables and TypeApplications, it is possible to define the projections: {{{ data MONAD2 m = MONAD2 (forall a. a -> App m a) (forall a b. App m a -> (a -> App m b) -> App m b) bind :: forall b a m. MONAD2 m -> App m a -> (a -> App m b) -> App m b bind (MONAD2 _ b) = b @a @b }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15470> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #13586: ghc --make seems to leak memory
by GHC 22 Jan '19

22 Jan '19

#13586: ghc --make seems to leak memory -------------------------------------+------------------------------------- Reporter: | Owner: (none) MikolajKonarski | Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: x86_64 | Type of failure: Compile-time (amd64) | performance bug Test Case: | Blocked By: Blocking: | Related Tickets: #13379 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- (This is probably not reproducible with a small example.) When I build this project with `cabal build` https://github.com/LambdaHack/LambdaHack/commit/138123ab13edd4db6c8143720af… the peek memory, as observed with `top`, is 10G*. When I instead interrupt the compilation with `^C` at the following point (compilation of this file take a couple of minutes, so it's easy to interrupt): {{{ [123 of 123] Compiling Game.LambdaHack.SampleImplementation.SampleMonadServer ( Game/LambdaHack/SampleImplementation/SampleMonadServer.hs, dist/build/Game/LambdaHack/SampleImplementation/SampleMonadServer.o ) }}} and then restart and continue to the end, peek memory usage in either of the two compilation parts is 5G*. So it seems `ghc --make` keeps some data that is either eventually not used or could as well be read on demand instead of kept in memory. Confirmed with 8.2.1-rc1 as well, but it's not trivial to compile due to restrictive upper bounds of many packages. *exact numerical values are made up -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13586> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 13

[GHC] #12953: Use computed gotos in the interpreter when the compiler supports it
by GHC 22 Jan '19

22 Jan '19

#12953: Use computed gotos in the interpreter when the compiler supports it -------------------------------------+------------------------------------- Reporter: dobenour | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Runtime | Version: 8.0.1 System | Keywords: interpreter | 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: -------------------------------------+------------------------------------- We can use computed gotos to speed up the interpreter on systems that support them. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12953> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5