February 2015 - ghc-tickets

[GHC] #9965: getChanContents BlockedIndefinitelyOnMVar on 7.10 RC1
by GHC 08 Apr '15

08 Apr '15

#9965: getChanContents BlockedIndefinitelyOnMVar on 7.10 RC1 -------------------------------------+------------------------------------- Reporter: jberryman | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: | Version: 7.10.1-rc1 libraries/base | Operating System: Unknown/Multiple Keywords: | Type of failure: Incorrect result Architecture: | at runtime Unknown/Multiple | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- All of the stanzas work on 7.8.3, but the middle one works incorrectly on 7.10: {{{#!hs module Main where import Control.Concurrent.Chan import Control.Monad main = do io <- newChan let inp = [0] :: [Int] mapM_ (writeChan io) inp {- -- WORKS: outp <- getChanContents io if 0 == head outp then putStrLn "OK" else error $ "Smoke test failed" -} -- BROKEN ON 7.10 (BlockedIndefinitelyOnMVar): outp <- getChanContents io if and (zipWith (==) inp outp) then putStrLn "OK" else error $ "Smoke test failed" {- -- WORKS: forM_ inp $ \xIn-> do xOut <- readChan io unless (xIn == xOut) $ error $ "Smoke test failed" -} }}} I first noticed the behavior in a Chan-like library I wrote, whose internals are completely different from Chan, but where the getChanContents is a copy-paste from Chan. So I assume it has something to do with `unsafeInterleaveIO`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9965> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #10114: Kind mismatches with AnyK in rank-2 types
by GHC 08 Apr '15

08 Apr '15

#10114: Kind mismatches with AnyK in rank-2 types -------------------------------------+------------------------------------- Reporter: cam | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.4 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: GHC rejects Unknown/Multiple | valid program Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- The following module: {{{#!hs {-# LANGUAGE Rank2Types #-} module KindBug where type T = forall f a. f a foo :: T foo = undefined bar :: forall f a. f a bar = foo }}} ...produces the following error: {{{ [1 of 1] Compiling KindBug ( KindBug.hs, KindBug.o ) KindBug.hs:10:7: Kind incompatibility when matching types: f0 :: AnyK -> * f :: * -> * Expected type: f a Actual type: f0 a0 Relevant bindings include bar :: f a (bound at KindBug.hs:10:1) In the expression: foo In an equation for ‘bar’: bar = foo }}} Note that no extensions are enabled other than rank-2 types. Enabling and using kind annotations fixes the error, as does replacing the type synonym with its definition in the type signature of foo. Similar errors can also be generated with PolyKinds enabled and less plausible code, e.g. {{{ (return undefined :: forall f a. f a) :: f a }}}. The non-PolyKinds example is simplified from some actual code, where the type synonym was similar to the ones used in lens, and is the simplest example I could find (with help from Shachaf on IRC) that resembled real code. I searched for existing tickets and the closest I could find are #7916 and #5935, both of which are already fixed. Also, according to glguy (Eric Mertens) just now in #haskell-lens, the same error occurs in the 7.10 branch and in 7.11.20150120. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10114> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #9160: Panic: Template variable unbound in rewrite rule
by GHC 08 Apr '15

08 Apr '15

#9160: Panic: Template variable unbound in rewrite rule ----------------------------------+--------------------------------------- Reporter: mietek | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: MacOS X Architecture: x86_64 (amd64) | Type of failure: Compile-time crash Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ----------------------------------+--------------------------------------- Installing `singletons-1.0` with GHC 7.8.2 on OS X 10.9.2 fails: {{{ ghc: panic! (the 'impossible' happened) (GHC version 7.8.2 for x86_64-apple-darwin): Template variable unbound in rewrite rule }}} Full build log attached. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9160> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 18

[GHC] #9964: GHC crash with NOINLINE and weird IO stuff
by GHC 07 Apr '15

07 Apr '15

#9964: GHC crash with NOINLINE and weird IO stuff -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.11 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Compile-time Unknown/Multiple | crash Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- {{{#!hs {-# LANGUAGE UnboxedTuples #-} module Crash where import GHC.Base crash :: IO Int crash = IO (\s -> let {-# NOINLINE s' #-} s' = s in (# s', 1::Int #)) }}} This compiles under 7.6.3, but both 7.8.3 and head choke: {{{ ghc: panic! (the 'impossible' happened) (GHC version 7.8.3 for x86_64-unknown-linux): StgCmmEnv: variable not found s{v sLw} [lid] local binds for: Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9964> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 15

[GHC] #9958: System.IO.Error: Fix a documentation link to Control.Exception.Exception
by GHC 07 Apr '15

07 Apr '15

#9958: System.IO.Error: Fix a documentation link to Control.Exception.Exception -------------------------------------+------------------------------------- Reporter: mineo | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: | Version: 7.8.4 libraries/base | Operating System: Unknown/Multiple Keywords: | Type of failure: None/Unknown Architecture: | Blocked By: Unknown/Multiple | Related Tickets: Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- On http://hackage.haskell.org/package/base-4.7.0.2/docs/System-IO- Error.html, the link to Control.Exception.Exception in the sentence "For a more general type of exception, including also those that arise in pure code, see Control.Exception.Exception." is broken because (in the source) it's a link to a module, but C.E.E is a type. Changing the double to single quotes fixes that. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9958> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #9225: Missing syntax error for package qualified import with version number
by GHC 06 Apr '15

06 Apr '15

#9225: Missing syntax error for package qualified import with version number -------------------------+------------------------------------------------- Reporter: | Owner: ezyang ezyang | Status: new Type: bug | Milestone: Priority: | Version: 7.8.2 normal | Operating System: Unknown/Multiple Component: | Type of failure: Incorrect warning at Compiler (Parser) | compile-time Keywords: | Test Case: Architecture: | Blocking: Unknown/Multiple | Difficulty: | Unknown | Blocked By: | Related Tickets: | -------------------------+------------------------------------------------- Consider the following test file: {{{ {-# LANGUAGE PackageImports #-} module Foo where import "containers-0.5.0.0" Data.Map }}} where we have {{{ ezyang@sabre:~$ ghc-pkg list | grep containers containers-0.5.0.0 unordered-containers-0.2.3.0 }}} GHC will fail to compile this, but in an unusual way: {{{ ezyang@sabre:~$ ghc test.hs [1 of 1] Compiling Foo ( test.hs, test.o ) test.hs:3:1: Failed to load interface for `Data.Map' It is not a module in the current program, or in any known package. }}} That's misleading: the real reason the import failed is because this is a package qualified import (not package ID qualified imports). For my purposes, it would be useful to also support package ID imports, but in the meantime, it would be a good idea to run Cabal's syntax check on the package name, which requires every hyphenated component to contain an alphabetic character. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9225> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #8057: PackageImports with versions for ghci
by GHC 06 Apr '15

06 Apr '15

#8057: PackageImports with versions for ghci ------------------------------------+------------------------------------- Reporter: Lemming | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: GHCi | 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: | ------------------------------------+------------------------------------- Imports with explicit package names and versions would be nice to have in GHCi, because you cannot use the -package and -hide-package options from within GHCi. This is what currently happens: {{{ $ ghci-7.6.3 -XPackageImports GHCi, version 7.6.3: http://www.haskell.org/ghc/ :? for help Loading package ghc-prim ... linking ... done. Loading package integer-gmp ... linking ... done. Loading package base ... linking ... done. Prelude> import Control.Monad.Trans.List <no location info>: Ambiguous module name `Control.Monad.Trans.List': it was found in multiple packages: transformers-0.3.0.0 List-0.5.1 Prelude> import "transformers-0.3.0.0" Control.Monad.Trans.List <no location info>: Could not find module `Control.Monad.Trans.List' It is not a module in the current program, or in any known package. Prelude> :q Leaving GHCi. $ ghci-7.6.3 -hide-package List GHCi, version 7.6.3: http://www.haskell.org/ghc/ :? for help Loading package ghc-prim ... linking ... done. Loading package integer-gmp ... linking ... done. Loading package base ... linking ... done. Prelude> import Control.Monad.Trans.List Prelude Control.Monad.Trans.List> }}} However, I now see that the import with explicit package name works, if I omit the package version. I think GHCi should at least warn that providing a version does not work. See also #2362 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8057> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #8811: Profiling output jumbled together
by GHC 06 Apr '15

06 Apr '15

#8811: Profiling output jumbled together ------------------------------------+------------------------------------- Reporter: augustss | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | 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: | ------------------------------------+------------------------------------- In the time profile (-p), if the MODULE column is very wide and the no. column is also very wide they will run together with no space between. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8811> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #9457: hsc2hs breaks with `--cflag=-Werror` in cross-compilation mode
by GHC 05 Apr '15

05 Apr '15

#9457: hsc2hs breaks with `--cflag=-Werror` in cross-compilation mode -------------------------------------+------------------------------------- Reporter: rwbarton | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: hsc2hs | Version: 7.9 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- I would like to be able to do {{{ config_args='--target=i386-unknown-linux --with-gcc=i386-unknown-linux- gcc' ./validate }}} to validate a 32-bit GHC on a 64-bit Linux system. (I don't know why I have to specify `--with-gcc` in addition to `--target`; an unrelated mystery.) However `validate` sets `-Werror` and the C code that `hsc2hs` generates in cross-compilation mode is not always warning-free, e.g., {{{ Files.hsc: In function ‘_hsc2hs_test’: Files.hsc:81:14: error: variable ‘test_array’ set but not used [-Werror =unused-but-set-variable] cc1: all warnings being treated as errors }}} so `hsc2hs` draws incorrect conclusions from its tests. Either the `hsc2hs`-generated C code should be made (and kept) warning- free, or perhaps `hsc2hs` should append `-Wwarn` to the options that it passes to the C compiler. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9457> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #9737: List ANN in pragmas chapter of user manual
by GHC 01 Apr '15

01 Apr '15

#9737: List ANN in pragmas chapter of user manual -------------------------------------+------------------------------------- Reporter: goldfire | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Documentation | Version: 7.8.3 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Blocked By: | None/Unknown Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- That chapter seems to be a comprehensive listing of pragmas. It should fulfill that promise. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9737> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1