- ghc-tickets - Haskell.org

[GHC] #12775: T3424 hangs with -fregs-graph
by GHC 29 Oct '16

29 Oct '16

#12775: T3424 hangs with -fregs-graph -------------------------------------+------------------------------------- Reporter: tjakway | 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: T3424 | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The -fregs-graph option causes rts/T3424 to hang. I waited 15+ minutes when compiling with inplace/ghc-stage2 before killing it; when run with the test suite after a similarly long time I get a compile failed (exit code 99) message. It passes when run in ghci or when using the default register allocator. To reproduce: make TEST=T3424 EXTRA_HC_OPTIONS=-fregs-graph -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12775> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #12536: Clarify commentary surrounding unboxed tuples and kind invariant
by GHC 29 Oct '16

29 Oct '16

#12536: Clarify commentary surrounding unboxed tuples and kind invariant -------------------------------------+------------------------------------- Reporter: bgamari | Owner: simonpj 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: -------------------------------------+------------------------------------- ticket:12115#comment:8 notes that there are a few inconsistencies in the comments describing these areas. Simon, could you have a look? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12536> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #12657: GHC and GHCi: RWX mmap denied by GrSec, results in a segfault
by GHC 27 Oct '16

27 Oct '16

#12657: GHC and GHCi: RWX mmap denied by GrSec, results in a segfault ----------------------------------------+---------------------------------- Reporter: PoroCYon | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Runtime System | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: Unknown/Multiple | Type of failure: Runtime crash Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: ----------------------------------------+---------------------------------- When compiling a package (cabal or stack doesn't matter) or launching GHCi while running a kernel with GrSec and PaX patches, GHC segfaults because GrSec denies an RWX mmap. Some packages do compile successfully, but others (such as optparse- applicative) don't. Launching GHCi always results in a segfault. Here's the output of ghc when compiling the optparse-applicative package: {{{ Building optparse-applicative-0.13.0.0... Preprocessing library optparse-applicative-0.13.0.0... [ 1 of 15] Compiling Options.Applicative.Help.Pretty ( Options/Applicative/Help/Pretty.hs, dist/build/Options/Applicative/Help/Pretty.o ) [ 2 of 15] Compiling Options.Applicative.Help.Chunk ( Options/Applicative/Help/Chunk.hs, dist/build/Options/Applicative/Help/Chunk.o ) [ 3 of 15] Compiling Options.Applicative.Help.Types ( Options/Applicative/Help/Types.hs, dist/build/Options/Applicative/Help/Types.o ) [ 4 of 15] Compiling Options.Applicative.Types ( Options/Applicative/Types.hs, dist/build/Options/Applicative/Types.o ) [ 5 of 15] Compiling Options.Applicative.Internal ( Options/Applicative/Internal.hs, dist/build/Options/Applicative/Internal.o ) [ 6 of 15] Compiling Options.Applicative.Common ( Options/Applicative/Common.hs, dist/build/Options/Applicative/Common.o ) [ 7 of 15] Compiling Options.Applicative.Help.Core ( Options/Applicative/Help/Core.hs, dist/build/Options/Applicative/Help/Core.o ) ghc: mmap 131072 bytes at (nil): Operation not permitted ghc: Try specifying an address with +RTS -xm<addr> -RTS Segmentation fault (core dumped) }}} GrSec log in dmesg: {{{ [15873.000867] grsec: denied RWX mmap of <anonymous mapping> by /usr/lib/ghc-8.0.1/bin/ghc[ghc_worker:16494] uid/euid:1000/1000 gid/egid:1000/1000, parent /usr/bin/cabal[cabal:16486] uid/euid:1000/1000 gid/egid:1000/1000 [15873.001018] ghc_worker[16494]: segfault at 0 ip 000003b6c3a7167f sp 000002b6c23d8e50 error 6 in libHSrts_thr-ghc8.0.1.so[3b6c3a4f000+62000] }}} And here's the one for ghci: {{{ $ ghci GHCi, version 8.0.1: http://www.haskell.org/ghc/ :? for help ghc: mmap 131072 bytes at (nil): Operation not permitted ghc: Try specifying an address with +RTS -xm<addr> -RTS Segmentation fault (core dumped) $ echo $? 139 }}} GrSec complained here with: {{{ [14756.354725] grsec: denied RWX mmap of <anonymous mapping> by /usr/lib/ghc-8.0.1/bin/ghc[ghc:15883] uid/euid:1000/1000 gid/egid:1000/1000, parent /usr/bin/xonsh[xonsh:15879] uid/euid:1000/1000 gid/egid:1000/1000 [14756.354875] ghc[15883]: segfault at 0 ip 00000379c0e0b67f sp 000003f0b7278ce0 error 6 in libHSrts_thr-ghc8.0.1.so[379c0de9000+62000] }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12657> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 11

[GHC] #12617: Make the interface of traceTc and traceRn the same
by GHC 26 Oct '16

26 Oct '16

#12617: Make the interface of traceTc and traceRn the same -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: task | Status: new Priority: lowest | 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: -------------------------------------+------------------------------------- It always seemed strange to me that the interfaces for `traceTc` and `traceRn` were different. {{{#!hs traceTc :: String -> SDoc -> TcM () traceRn :: SDoc -> RnM () }}} I personally prefer using the `traceTc` interface and have to look up each time which function uses which interface. It seems beneficial but annoying to make `traceRn` have the same interface to improve usability slightly. Are there any objections if someone was to do this? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12617> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

Re: [GHC] #4162: GHC API messes up signal handlers
by GHC 26 Oct '16

26 Oct '16

#4162: GHC API messes up signal handlers -------------------------------------+------------------------------------- Reporter: jcpetruzza | Owner: Type: bug | Status: patch Priority: low | Milestone: Component: GHC API | Version: 6.12.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): phab:D2633 Wiki Page: | -------------------------------------+------------------------------------- Comment (by hsyl20): Yes, I have updated the patch in Phab:D2633 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/4162#comment:17> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #12522: GHC 8.0.1 hangs, looping forever in type-checker
by GHC 26 Oct '16

26 Oct '16

#12522: GHC 8.0.1 hangs, looping forever in type-checker -------------------------------------+------------------------------------- Reporter: clinton | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Other Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I'm not sure if this is a bug or hanging the compiler is expected here. This was the minimal example that causes GHC to hang: {{{#!hs {-# LANGUAGE TypeFamilyDependencies #-} {-# LANGUAGE FlexibleInstances #-} {-# LANGUAGE AllowAmbiguousTypes #-} main = return $ f (Just 'c') data D1 x data D2 type family TF x = t | t -> x type instance TF (D1 x, a) = Maybe (TF (x, a)) type instance TF (D2, ()) = Char class C p where f :: TF (x, a) -> () }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12522> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 21

[GHC] #9858: Typeable instance for datatype and its promoted constructor is the same
by GHC 26 Oct '16

26 Oct '16

#9858: Typeable instance for datatype and its promoted constructor is the same -------------------------------------+------------------------------------- Reporter: dreixel | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.10.1 Component: Compiler | Version: 7.9 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Easy (less than 1 | Type of failure: hour) | None/Unknown Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- {{{ {-# LANGUAGE DataKinds #-} {-# LANGUAGE AutoDeriveTypeable #-} import Data.Typeable data A = A main = print $ typeRep (Proxy :: Proxy A) == typeRep (Proxy :: Proxy 'A) }}} This returns `True`, but it should return `False`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9858> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 119

Re: [GHC] #4162: GHC API messes up signal handlers
by GHC 26 Oct '16

26 Oct '16

#4162: GHC API messes up signal handlers -------------------------------------+------------------------------------- Reporter: jcpetruzza | Owner: Type: bug | Status: patch Priority: low | Milestone: Component: GHC API | Version: 6.12.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): phab:D2633 Wiki Page: | -------------------------------------+------------------------------------- Comment (by simonmar): Good point, maybe use a reference count instead? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/4162#comment:16> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #4162: GHC API messes up signal handlers
by GHC 26 Oct '16

26 Oct '16

#4162: GHC API messes up signal handlers -------------------------------------+------------------------------------- Reporter: jcpetruzza | Owner: Type: bug | Status: patch Priority: low | Milestone: Component: GHC API | Version: 6.12.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): phab:D2633 Wiki Page: | -------------------------------------+------------------------------------- Changes (by hsyl20): * differential: D2633 => phab:D2633 Comment: Note that my patch in comment:14 doesn't make `runGhc` thread-safe if several `runGhc` are executed concurrently. E.g., in the following case: {{{#!haskell runGhc (1) install handlers runGhc (2) install handlers runGhc (1) uninstall handlers runGhc (2) uninstall handlers -- reinstall handlers installed by `runGhc (1)` -- overwriting initial handlers }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/4162#comment:15> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #1687: A faster (^)-function.
by GHC 26 Oct '16

26 Oct '16

#1687: A faster (^)-function. --------------------------------------------+----------------------------- Reporter: moonlite@… | Owner: Type: bug | Status: new Priority: normal | Milestone: ⊥ Component: Prelude | Version: 6.6.1 Resolution: | Keywords: Operating System: Linux | Architecture: x86 Type of failure: Runtime performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | --------------------------------------------+----------------------------- Comment (by siddhanathan): rwbarton explained this earlier, but it took a while to wrap my head around it. I find this a bit clearer to understand: {{{ {-# SPECIALISE [1] (^) :: Integer -> Integer -> Integer, Integer -> Int -> Integer, Int -> Int -> Int #-} {-# INLINABLE [1] (^) #-} -- See Note [Inlining (^)] (^) :: (Num a, Integral b) => a -> b -> a x0 ^ y0 | y0 < 0 = errorWithoutStackTrace "Negative exponent" | y0 == 0 = 1 | otherwise = f x0 y0 1 where f x y z | even y = let k = f x (y `quot` 2) z in k * k | y == 1 = x * z | otherwise = let k = f x ((y - 1) `quot` 2) z in k * k * x }}} This has identical performance to the version in the description above. This makes use of the fact that x^2a^ = x^a+a^ = x^a^ * x^a^ ; and x^2a+1^ = x^a^ * x^a^ * x -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/1687#comment:18> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0