December 2016 - ghc-tickets

[GHC] #8684: hWaitForInput cannot be interrupted by async exceptions on unix
by GHC 05 Feb '18

05 Feb '18

#8684: hWaitForInput cannot be interrupted by async exceptions on unix ------------------------------------+------------------------------------- Reporter: nh2 | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: libraries/base | Version: 7.6.3 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- http://hackage.haskell.org/package/base-4.6.0.1/docs/System-Timeout.html claims that {{{timeout}}} can interrupt {{{hWaitForInput}}}, but in fact that's false (e.g. mentioned in https://ghc.haskell.org/trac/ghc/ticket/7353#comment:4) {{{ -- import Control.Concurrent import System.IO import System.Timeout main = timeout (1 * 1000000) $ hWaitForInput stdin (5 * 1000) }}} will not be killed after 1 second, but instead wait for the full 5 seconds timeout passed to {{{hWaitForInput}}}. The implementation is {{{ready}}} at http://www.haskell.org/ghc/docs/latest/html/libraries/base/src/GHC-IO- FD.html, where we have two foreign calls: {{{safe fdReady}}} and {{{unsafe unsafe_fdReady}}}. The actual C implementation is at https://github.com/haskell- suite/base/blob/master/cbits/inputReady.c#L16. It uses {{{select}}} on Unix, and does check for {{{EINTR}}}, so I believe that according to http://www.haskell.org/ghc/docs/7.6.3/html/users_guide/ffi.html#ffi- interruptible both foreign calls can be replaced by a single {{{interruptible}}} one. Is that true? If not, it's a documentation bug in {{{timeout}}} at least. Also, does {{{interruptible}}}, apart from allowing the function to be interrupted, behave more like {{{safe}}} or {{{unsafe}}}? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8684> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 32

[GHC] #8228: GHC built under Windows does not generate dyn_hi files
by GHC 04 Feb '18

04 Feb '18

#8228: GHC built under Windows does not generate dyn_hi files ------------------------------+-------------------------------------------- Reporter: ezyang | Owner: Type: bug | Status: new Priority: high | Milestone: 7.8.1 Component: Compiler | Version: 7.7 Keywords: | Operating System: Windows Architecture: | Type of failure: GHC doesn't work at all Unknown/Multiple | Test Case: cabal01 Difficulty: Unknown | Blocking: Blocked By: | Related Tickets: | ------------------------------+-------------------------------------------- I found cabal01 failing with: {{{ =====> cabal01(normal) 1 of 1 [0, 0, 0] cd . && $MAKE -s --no-print-directory cabal01 VANILLA=--enable-library- vanilla PROF=--disable-l ibrary-profiling DYN=--enable-shared </dev/null >cabal01.run.stdout 2>cabal01.run.stderr Wrong exit code (expected 0 , actual 2 ) Stdout: Stderr: Creating library file: dist\build\libHStest-1.0-ghc7.7.20130904.dll.a setup.exe: Error: Could not find module: A with any suffix: ["dyn_hi"] in the search path: ["dist\\build"] make[2]: *** [cabal01] Error 1 }}} but I bet some of the other failures are due to this. Indeed, in dist/build, there are no dyn_hi files When I look more closely at the commands being run, I see: {{{ bash.exe-3.1$ c:/Users/ezyang/Dev/ghc-validate/inplace/bin/ghc-stage2.exe --make -fbuilding-ca bal-package -O -static -dynamic-too -dynosuf dyn_o -dynhisuf dyn_hi -outputdir dist/build -odi r dist/build -hidir dist/build -stubdir dist/build -i -idist/build -i. -idist/build/autogen -I dist/build/autogen -Idist/build -optP-include -optPdist/build/autogen/cabal_macros.h -package- name test-1.0 -hide-all- packages -no-user-package-db -package-db local.db -package-db dist/pac kage.conf.inplace -package-id base-4.7.0.0-inplace -XHaskell98 -XForeignFunctionInterface A B. A -fforce-recomp -dcore-lint -dcmm-lint -dno-debug-output -no-user-package-db [1 of 2] Compiling B.A ( B\A.hs, dist\build\B\A.o ) [2 of 2] Compiling A ( A.hs, dist\build\A.o ) }}} which should result in dyn_hi files but does not. I haven't traced it any further yet. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8228> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 17

[GHC] #11706: Increase precedence of statements (if-then-else, case-of, do)
by GHC 03 Feb '18

03 Feb '18

#11706: Increase precedence of statements (if-then-else, case-of, do) -------------------------------------+------------------------------------- Reporter: YoYoYonnY | Owner: Type: feature | Status: new request | Priority: lowest | Milestone: ⊥ Component: Compiler | Version: (Parser) | 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: -------------------------------------+------------------------------------- Warning: Skip to the {{{TL; DR}}}, and maybe the {{{How it works}}} section; You don't want to read this. (I've warned you.) == The problem == We do it all the time: Writing that one dollar sign just before our expressions, just because Haskell can't parse it: {{{#!hs f x = g $ if x then {- ... -} else {- ... -} f x = g $ case x of {- ... -} f x = g $ do {- ... -} }}} However, we programmers are lazy and want to type as few characters as possible. Think about the precious characters we would save if we could write code this way: {{{#!hs f x = g if x then {- ... -} else {- ... -} f x = g case x of {- ... -} f x = g do {- ... -} }}} Wow, our productivity just shot up by a whopping 6%! However, this code looks ugly, noone would ever write this... Right? There has to be some way to catagorize this feature with others people would never use... == Pragmas to the rescue! == The {{{LANGUAGE}}} pragma! Of course! We can add a syntactic extension to enable above syntax! But how should we name it... {{{InlineStatements}}}? {{{InfixStatements}}}? {{{ParenthesizedStatements}}}? {{{IncreasedStatementPrecedence}}}? {{{StatementsWithoutParenthesis}}}? I don't know. I am [1] not the parents who might one day give birth to this feature. I am merely the one who conceptualized it. [1] (probably, unless this request remains unnoticed for another 20 years or so) == How it works == * Implicit parenthesis will be put around ''all'' statements, according to the following rules: 1. Parenthesis will be opened at the start of the statement. 2. Parenthesis will be closed at the end of the statement. The end of the statement is determined by 1. The curly brackets around the statement, or; 2. The indentation of the statement Basically, these are the rule Haskell already uses. Sorry if this is obvious, but some people seemed to get confused, so here's some examples: === (Correct) Examples === Examples are in the form: {{{#!hs implicitParenthesis -- New syntax explicitParenthesis -- Old syntax }}} Call {{{idM :: Monad m => m a -> m a}}} as {{{idM :: IO String -> IO String}}}. {{{#!hs idM do putStrLn "What's my name?" getLine idM (do { putStrLn "What's my name?"; getLine; }) }}} Create function {{{whatsMyName :: Maybe String -> String}}}. {{{#!hs whatsMyName x = id case x of Just name = "My name is " ++ name Nothing = "What's my name?" whatsMyName x = id (case x of { Just name = "My name is " ++ name; Nothing = "What's my name?"; }) }}} Another example using {{{let}}} and {{{if-then-else}}}. {{{#!hs main = putStrLn (++) "I've been tryna work it out... The square root of 69 is 8 something, right? " let eight_something = 8.306623862918075 in if sqrt 69 == eight_something then "Yeah" else "Oh na na" main = putStrLn ((++) "I've been tryna work it out... The square root of 69 is 8 something, right? " (let eight_something = 8.306623862918075 in (if sqrt 69 == eight_something then "Yeah" else "Oh na na")) }}} === Incorrect examples === {{{#!hs f do putStrLn "This won't work..." True f (do { putStrLn "This won't work..."; True; }) }}} {{{#!hs f do putStrLn "This won't work..." True -- Indented by extra space f (do { putStrLn "This won't work..."; True; }) }}} == TL; DR == What there is: {{{#!hs main = when $ do putStrLn "Parenthesis..." main = when True (do putStrLn "Parenthesis...") }}} What I want: {{{#!hs when (doBlocksNeedParenthesis) do putStrLn "This code is invalid." when (doBlocksNeedParenthesis) $ do putStrLn "This code is valid." when (doBlocksNeedParenthesis) (do putStrLn "This code is valid aswell.") when (thisIsImplemented) do putStrLn "These are all equal" when (thisIsImplemented) (do putStrLn "These are all equal") when (thisIsImplemented) $ do putStrLn "These are all equal" }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11706> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #13032: Redundant forcing of Given dictionaries
by GHC 01 Feb '18

01 Feb '18

#13032: Redundant forcing of Given dictionaries -------------------------------------+------------------------------------- Reporter: simonpj | 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: -------------------------------------+------------------------------------- If you look at the test case for #13025 you'll see code like {{{ f :: (a~~b) => a -> b -> Bool f d x y = case HEq_sc d of (cobox :: a ~# b) -> True }}} From the Given dictionary `(a~~b)`, the contraint solver generates a given binding, just in case {{{ (cobox :: a~# b) = HEq_sc d }}} But because `cobox` is a ''coercion'' we evaluate this binding strictly, and so the desugarer produces the case expression above. Most redundant Given bindings are let-bound and hence discarded quickly by the occurrence analyser, but GHC doesn't discard cases (termination worries). So the task here is to remove the redundant Given in some way. (Of course, if the equality is actually used, we need to keep it.) NB: See `Note [The equality types story]` in `TysPrim` for the meanagerie of equality types. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13032> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 12

[GHC] #11195: New pattern-match check can be non-performant
by GHC 01 Feb '18

01 Feb '18

#11195: New pattern-match check can be non-performant -------------------------------------+------------------------------------- Reporter: goldfire | Owner: Type: bug | Status: new Priority: highest | Milestone: Component: Compiler | Version: 7.11 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: -------------------------------------+------------------------------------- In my work in merging Phab:D808 (and originally pointed out by thomie), I ran into a peculiar bug: the new pattern-match check took gobs of memory (~10GB) to check !OptCoercion in the stage-2 compiler. My analysis got only as far as nailing the problem down to `pmTraverse`, which took 80% of the time, and got called roughly a gajillion times. To reproduce, simply compile !OptCoercion (with `-package ghc`) and you'll see your memory usage explode. So that you'll be able to build GHC until this is fixed, I've disable the pattern-match check in that file. There is a good chance that this is caused by an interaction between the pattern-match check and the type=kind code. I'm happy to probe further, but I'd want George's help. NB: This applies only after my patch is merged. Which should be today. For some value of today. Hopefully today's value of today. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11195> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 21

[GHC] #12665: Make Read instances faster, and make them fail fast
by GHC 01 Feb '18

01 Feb '18

#12665: Make Read instances faster, and make them fail fast -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: Type: feature | Status: new request | Priority: high | Milestone: 8.2.1 Component: Core | Version: 8.0.1 Libraries | 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: -------------------------------------+------------------------------------- At present, the `Read` instances for standard types are generally written ''as though'' `Read` were a typical parser for a programming language. However, it most assuredly is not. Specifically, we currently follow the pattern of lexing and then parsing. Lexing looks at the string and attempts to identify the next token, whatever it may be. But for `Read`, we don't need that. Thanks to the types, we have a very clear sense of what characters we expect to encounter. I've started to sketch out an improvement, which reads `Int` and `Word` around seven times as fast (9.5 times as fast with parentheses and negation), and will fail immediately on something like `read (fix ('a':)) :: Int` rather than going into an infinite loop. To begin with, modify the definition of `paren` thus: {{{#!hs expectP' :: Char -> ReadPrec () expectP' c = lift (expect c) {-# INLINE expectP' #-} paren :: ReadPrec a -> ReadPrec a -- ^ @(paren p)@ parses \"(P0)\" -- where @p@ parses \"P0\" in precedence context zero paren p = do expectP' '(' x <- reset p expectP' ')' return x }}} This allows fast failure when looking for parentheses, so we don't have to scan to the end of the first token (whatever it may be) before concluding that it is not `'('`. Now we can parse `Word` and `Int` very efficiently. I had to specialize earlier than I wanted to convince GHC that I don't want to convert through `Integer`. I'm not sure why the `fromIntegral` rule doesn't fire reliably around here. The code below (temporarily) uses the current definition for base 16, because that's a bit fussy; I'll rewrite it soon. {{{#!hs charDiff :: Char -> Char -> Word charDiff c1 c2 = fromIntegral (ord c1 - ord c2) {-# INLINE charDiff #-} readHexOct :: ReadP Word readHexOct = do _ <- ReadP.char '0' baseId <- lexBaseChar case baseId of Oct -> readBaseP 8 Hex -> L.readHexP {-# INLINE readHexOct #-} data BaseId = Oct | Hex lexBaseChar :: ReadP BaseId lexBaseChar = do { c <- ReadP.get; case c of 'o' -> pure Oct 'O' -> pure Oct 'x' -> pure Hex 'X' -> pure Hex _ -> ReadP.pfail } readWord :: ReadP Word readWord = readNumber (readHexOct ReadP.<++ readBaseP 10) readInt :: ReadP Int readInt = fromIntegral <$> readWord readBaseP :: Integral a => Word -> ReadP a readBaseP !base = do c <- ReadP.get let diff = charDiff c '0' if diff < base then readBaseP' base (fromIntegral diff) else ReadP.pfail {-# INLINE readBaseP #-} readBaseP' :: Integral a => Word -> a -> ReadP a readBaseP' !base !acc0 = ReadP.look >>= go acc0 where go !acc (c:cs) | diff < base = ReadP.get *> go (fromIntegral base * acc + fromIntegral diff) cs where diff = charDiff c '0' go !acc _ = pure acc {-# INLINE readBaseP' #-} readNumber :: Num a => ReadP.ReadP a -> ReadPrec a readNumber p = parens $ do cs <- lift skipSpaces *> look case cs of ('-': _) -> get *> lift (skipSpaces *> (negate <$> p)) _ -> lift $ skipSpaces *> p {-# INLINE readNumber #-} }}} Other `Word` and `Int`-like types can be built on top of that foundation. I haven't yet attempted to deal with other instances, but I think there are probably a lot of opportunities for similar improvements. WARNING: I know very little about parsing, and less about `ReadPrec`. It's conceivable that I've made some semantic errors here, although I don't think I have. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12665> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 21

[GHC] #10855: GHC rejects code that Haskell 2010 report accepts
by GHC 01 Feb '18

01 Feb '18

#10855: GHC rejects code that Haskell 2010 report accepts -------------------------------------+------------------------------------- Reporter: goldfire | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.2 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- According to [https://www.haskell.org/onlinereport/haskell2010/haskellch3.html#x8-220003 The Haskell 2010 Report], the following two modules should compile. But they don't, failing with parse errors: {{{ {-# LANGUAGE Haskell2010 #-} module Bug where bool = - case b of False -> 0; True -> (-1) }}} {{{ {-# LANGUAGE Haskell2010 #-} instance Num (IO a) where negate = id main = - do putStrLn "hi!" }}} Note the unary `-` signs. I'm not sure whether we should complicate the parser to fix these, but we should document the compliance-failure in the manual. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10855> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #10074: Implement the 'Improved LLVM Backend' proposal
by GHC 01 Feb '18

01 Feb '18

#10074: Implement the 'Improved LLVM Backend' proposal -------------------------------------+------------------------------------- Reporter: | Owner: thoughtpolice thoughtpolice | Status: new Type: task | Milestone: 7.12.1 Priority: high | Version: Component: Compiler | Operating System: Unknown/Multiple (LLVM) | Type of failure: None/Unknown Keywords: llvm, | Blocked By: codegen | Related Tickets: Architecture: | Unknown/Multiple | Test Case: | Blocking: | Differential Revisions: Phab:D530 | -------------------------------------+------------------------------------- This is a meta ticket designed to reflect the current implementation status of the 'Improved LLVM Backend' proposal, documented here: https://ghc.haskell.org/trac/ghc/wiki/ImprovedLLVMBackend -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10074> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 34

[GHC] #12245: Deriving Data at higher kinds
by GHC 01 Feb '18

01 Feb '18

#12245: Deriving Data at higher kinds -------------------------------------+------------------------------------- Reporter: simonpj | 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: -------------------------------------+------------------------------------- Here's what [https://mail.haskell.org/pipermail/generics/2016-June/000564.html Lennart Spitzner wanted to do]: {{{ > {-# LANGUAGE StandaloneDeriving #-} > {-# LANGUAGE DeriveDataTypeable #-} > {-# LANGUAGE FlexibleInstances #-} > > import Data.Data ( Data ) > > data Foo f = Foo (f Bool) (f Int) > > deriving instance Data (Foo []) > deriving instance Data (Foo Maybe) }}} Of course you can't derive `Data` for `Foo` because we don't know what `f` is, so Lennart is making multiple instances, one for each instance of `f`. It's a bit clumsy. What we would really like is {{{ deriving instance (forall a. Data => Data (f a)) => Data (Foo f) }}} but we don't have higher order instances yet! So Lennart is manually making two instances. This should work, but he gets {{{ > Main.hs: line 45, column 1: > Multiple declarations of ‘$cr2C’ > Declared at: Main.hs:44:1 > Main.hs:45:1 > Main.hs: line 45, column 1: > Multiple declarations of ‘$tr2D’ > Declared at: Main.hs:44:1 > Main.hs:45:1 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12245> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #10754: truncate /= double2Int
by GHC 28 Jan '18

28 Jan '18

#10754: truncate /= double2Int -------------------------------------+------------------------------------- Reporter: cblp | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: | Version: 7.10.2 libraries/base | Keywords: truncate, | Operating System: Unknown/Multiple double2Int, rewrite, rule | Architecture: | Type of failure: Incorrect result Unknown/Multiple | at runtime Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- There is a rule in libraries/base/GHC/Float.hs: {{{#!hs "truncate/Double->Int" truncate = double2Int }}} But it is not true for some values. Particularly, {{{#!hs let infinity = 1/0 :: Double truncate infinity :: Int == 0 double2Int infinity == -9223372036854775808 -- minBound }}} As a result, the value of {{{truncate (1/0 :: Double) :: Int}}} depends on optimization level. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10754> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7