September 2018 - ghc-tickets

[GHC] #15657: Support promotion of pattern synonyms to kinds
by GHC 19 Sep '18

19 Sep '18

#15657: Support promotion of pattern synonyms to kinds -------------------------------------+------------------------------------- Reporter: infinity0 | 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: -------------------------------------+------------------------------------- Suppose we define a heterogeneous binary tree: {{{#!hs {-# LANGUAGE DataKinds , GADTs , PatternSynonyms #-} data Tree a = TLeaf | TNode a (Tree a) (Tree a) data HTree xs where HLeaf :: HTree 'TLeaf HNode :: x -> HTree ls -> HTree rs -> HTree ('TNode x ls rs) }}} A tree representation is chosen because it's pretty general, and easy to combine - just `HNode` a bunch of them together. With a `HList` for example, it's harder to do this in nested fashion, and we want to be able to do that for the `$BigRealWorld` things we're writing. However in the majority of cases, the `$BigRealWorld` things defined by actual clients of the API don't need the full power of the HTree, and so pattern synonyms potentially allow them to easily define a tree of one item, or a small unnested list of items. {{{#!hs -- as above, then: pattern TPure :: a -> Tree a pattern TPure a = TNode a TLeaf TLeaf pattern TCons :: a -> Tree a -> Tree a pattern TCons x y = TNode x TLeaf y pattern HTPure :: x -> HTree ('TPure x) -- error, has to be ('TNode x 'TLeaf 'TLeaf) pattern HTPure a = HNode a HLeaf HLeaf clientThing :: HTree ('TPure Int) -- error, has to be ('TNode Int 'TLeaf 'TLeaf) clientThing = HTPure 3 }}} Oh no! GHC fails with: {{{ • Pattern synonym ‘TPure’ cannot be used here (Pattern synonyms cannot be promoted) • In the first argument of ‘HTree’, namely ‘( 'TPure x)’ In the type ‘x -> HTree ( 'TPure x)’ | 20 | pattern HTPure :: x -> HTree ('TPure x) }}} Actually the first one is not a big deal, we only write that once so it doesn't matter if we need to expand it fully. But things like `clientAPI` might be defined several times and then it's annoying to have to write the synonym out in full every time. I appreciate `ViewPatterns` make it hard to do this and would be totally happy with a solution that only works to promote non-ViewPattern pattern synonyms. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15657> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #15364: Replace the atomicModifyMutVar# primop
by GHC 18 Sep '18

18 Sep '18

#15364: Replace the atomicModifyMutVar# primop -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: (none) Type: feature | Status: new request | Priority: normal | Milestone: 8.8.1 Component: Runtime | Version: 8.5 System | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): Phab:D4884 | Wiki Page: -------------------------------------+------------------------------------- This was [https://github.com/ghc-proposals/ghc- proposals/blob/master/proposals/0027-atomicModifyMutVar.rst accepted proposal #27]. Specifically: 1. Replace `atomicModifyMutVar#` with `atomicModifyMutVar2#`. 2. Add `atomicModifyMutVar_#` and `atomicSwapMutVar#`. 3. Add a function imitating `atomicModifyMutVar#` to `GHC.Exts` for backwards compatibility. 4. Use the new primops where appropriate throughout `base`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15364> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #15530: Type applications in patterns
by GHC 18 Sep '18

18 Sep '18

#15530: Type applications in patterns -------------------------------------+------------------------------------- Reporter: nomeata | Owner: (none) Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 8.5 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 Type applications in pattern proposal was accepted: https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0031 -type-applications-in-patterns.rst This ticket tracks its implementation. I am tempted to give it a shot, to learn more about the type checker. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15530> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #15655: Simpliify tcTyConScopedTyVars
by GHC 18 Sep '18

18 Sep '18

#15655: Simpliify tcTyConScopedTyVars -------------------------------------+------------------------------------- Reporter: simonpj | 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: -------------------------------------+------------------------------------- Currently we have {{{ tcTyConScopedTyVars :: [(Name,TyVar)], }}} But actually every call of `mkTcTyCon` passes something like `(mkTyVarNamePairs tvs)` to it. So we could noticeably simplify this to {{{ tcTyConScopedTyVars :: [TyVar], }}} Less fiddling around. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15655> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #15651: Check if some auto apply code is dead and remove if appropriate.
by GHC 18 Sep '18

18 Sep '18

#15651: Check if some auto apply code is dead and remove if appropriate. -------------------------------------+------------------------------------- Reporter: AndreasK | Owner: (none) Type: task | Status: new Priority: normal | Milestone: 8.6.1 Component: Runtime | Version: 8.4.3 System | Keywords: Newcomer | 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 is a whole family of stg_ap_stk_* and stg_stk_save_* functions generated in AutoApply.cmm which as far as I can tell is not used anywhere in the compiler and can likely be removed. In particular this would involve: * Stop the code in question from being generated. (utils/genapply) * Make sure that doesn't break things. (Validate/Run the testsuite). * Grep for stk in the compiler to make sure we don't build calls to these in rare circumstances with string concatenation for extra safety. * Document the change in patch notes just in case. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15651> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #11284: Lambda-lifting fails in simple Text example
by GHC 18 Sep '18

18 Sep '18

#11284: Lambda-lifting fails in simple Text example -------------------------------------+------------------------------------- Reporter: bgamari | Owner: Type: bug | Status: new 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: -------------------------------------+------------------------------------- Consider the example, {{{#!hs import Data.Char (isSpace) import Data.List (foldl') import GHC.Exts (build) import qualified Data.Text as T longestWord :: T.Text -> Int longestWord t = foldl' max 0 $ map T.length $ fusedWords t fusedWords :: T.Text -> [T.Text] fusedWords t0 = build $ \cons nil -> let go !t | T.null t = nil | otherwise = let (w, rest) = T.span (not . isSpace) t in cons w (go $ T.dropWhile isSpace rest) in go t0 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11284> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 21

[GHC] #15502: -ddump-splices truncates Integer literals to Int literals
by GHC 18 Sep '18

18 Sep '18

#15502: -ddump-splices truncates Integer literals to Int literals -------------------------------------+------------------------------------- Reporter: ChaiTRex | 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: -------------------------------------+------------------------------------- I trusted that the splice results shown by `-ddump-splices` were correct. They weren't, which caused me to waste a lot of time debugging my Template Haskell expressions when they were already correct. {{{ $ ghc --version The Glorious Glasgow Haskell Compilation System, version 8.4.3 }}} == Example program == {{{#!hs {-# OPTIONS_GHC -ddump-splices #-} {-# LANGUAGE TemplateHaskell #-} module Main where import Language.Haskell.TH.Syntax (Lift(lift)) main = print ( $( lift (toInteger (maxBound :: Int) + 1) ) , $( lift (minBound :: Int) ) ) }}} == Output of `runghc` == Note that the output of the program on the bottom line below is correct. The two splice results shown by `-ddump-splices` incorrectly match each other: {{{ Example.hs:8:19-56: Splicing expression lift (toInteger (maxBound :: Int) + 1) ======> -9223372036854775808 Example.hs:9:19-40: Splicing expression lift (minBound :: Int) ======> (-9223372036854775808) (9223372036854775808,-9223372036854775808) }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15502> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #15509: `showEFloat` inconsistency introduced in base-4.12
by GHC 18 Sep '18

18 Sep '18

#15509: `showEFloat` inconsistency introduced in base-4.12 -------------------------------------+------------------------------------- Reporter: hvr | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 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 base-4.11 had a `showEFloat` that behaved like {{{#!hs showEFloat (Just n) | n <= 0 == showEFloat (Just 1) }}} For non-positive values, base-4.12 introduced an unfortunate inconsistency: {{{#!hs showEFloat (Just n) | n < 0 == showEFloat (Just 1) -- NB: *not* equivalent to showEFloat (Just 0) }}} e.g. now we have {{{#!hs showEFloat (Just 2) 1.0 "" == "1.00e0" showEFloat (Just 1) 1.0 "" == "1.0e0" showEFloat (Just 0) 1.0 "" == "1e0" showEFloat (Just (-1)) 1.0 "" == "1.0e0" showEFloat (Just (-2)) 1.0 "" == "1.0e0" }}} iow, negative `n`s in the precision argument are now treated differently from `(Just 0)` which is a source of subtle bugs for API consumers which assume `showEFloat` to have "continuous total" semantics for the precision argument. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15509> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #15550: Names of RULES aren't quoted in -ddump-splices
by GHC 18 Sep '18

18 Sep '18

#15550: Names of RULES aren't quoted in -ddump-splices -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Template | Version: 8.4.3 Haskell | 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: -------------------------------------+------------------------------------- Compile the following program: {{{#!hs {-# LANGUAGE TemplateHaskell #-} {-# OPTIONS_GHC -ddump-splices #-} module Bug where $([d| myId :: a -> a myId x = x {-# NOINLINE [1] myId #-} {-# RULES "myId" forall x. myId x = x #-} |]) }}} {{{ $ /opt/ghc/8.4.3/bin/ghci Bug.hs GHCi, version 8.4.3: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/ryanglscott/.ghci [1 of 1] Compiling Bug ( Bug.hs, interpreted ) Bug.hs:(5,3)-(9,6): Splicing declarations [d| {-# RULES "myId" forall x_a1xu. myId_a1xr x_a1xu = x_a1xu #-} myId_a1xr :: a_a1xs -> a_a1xs myId_a1xr x_a1xt = x_a1xt {-# NOINLINE [1] myId_a1xr #-} |] ======> myId_a49f :: a_a49e -> a_a49e myId_a49f x_a49g = x_a49g {-# NOINLINE [1] myId_a49f #-} {-# RULES myId forall x_a49h. myId_a49f x_a49h = x_a49h #-} Ok, one module loaded. }}} Notice how in the bottom of the `-ddump-splices` output, the name of the rewrite rule for `myId` isn't surrounded by double quotes, thus making it syntactically invalid. That is, it's printed as: {{{#!hs {-# RULES myId forall x_a49h. myId_a49f x_a49h = x_a49h #-} }}} Whereas it should be: {{{#!hs {-# RULES "myId" forall x_a49h. myId_a49f x_a49h = x_a49h #-} }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15550> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #15572: TH improperly converts promoted data cons in ConT
by GHC 18 Sep '18

18 Sep '18

#15572: TH improperly converts promoted data cons in ConT -------------------------------------+------------------------------------- Reporter: RyanGlScott | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Template | Version: 8.4.3 Haskell | 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 you compile the following program: {{{#!hs {-# LANGUAGE DataKinds #-} {-# LANGUAGE TemplateHaskell #-} {-# OPTIONS_GHC -ddump-splices #-} module Bug where import Language.Haskell.TH $([d| type AbsoluteUnit1 = '() |]) $(pure [TySynD (mkName "AbsoluteUnit2") [] (ConT '())]) }}} {{{ $ /opt/ghc/8.6.1/bin/ghci Bug.hs GHCi, version 8.6.0.20180810: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/rgscott/.ghci [1 of 1] Compiling Bug ( Bug.hs, interpreted ) Bug.hs:8:3-33: Splicing declarations [d| type AbsoluteUnit1_a1HN = '() |] ======> type AbsoluteUnit1_a4qs = '() Bug.hs:9:3-54: Splicing declarations pure [TySynD (mkName "AbsoluteUnit2") [] (ConT '())] ======> type AbsoluteUnit2 = () }}} You'll notice an unusual discrepancy between the two `-ddump-splices` logs. In the first one: {{{#!hs type AbsoluteUnit1_a4qs = '() }}} The `'()` constructor is properly preceded with a single quote. In the second one, however: {{{#!hs type AbsoluteUnit2 = () }}} `'()` incorrectly appears without a single quote! The culprit is in the way `Convert` [http://git.haskell.org/ghc.git/blob/154d4e219cc0cebbef8a845609bd63ec45fdbea… handles] `ConT`: {{{#!hs ConT nm -> do { nm' <- tconName nm ; mk_apps (HsTyVar noExt NotPromoted (noLoc nm')) tys'} }}} This code naïvely assumes that `ConT` will never contain a promoted data constructor name by hardcoding `NotPromoted`. We really ought to be checking if `nm'` is a data con `RdrName` here and using `Promoted` if so, and `NotPromoted` otherwise. Patch incoming. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15572> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4