June 2016 - ghc-tickets

[GHC] #12094: Unlifted types and pattern synonym signatures
by GHC 12 Jul '16

12 Jul '16

#12094: Unlifted types and pattern synonym signatures -------------------------------------+------------------------------------- Reporter: akio | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 (Type checker) | Keywords: pattern | Operating System: Unknown/Multiple synonyms | Architecture: | Type of failure: GHC rejects Unknown/Multiple | valid program Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The following program is rejected by ghc-8.0.1: {{{#!hs {-# LANGUAGE MagicHash #-} {-# LANGUAGE PatternSynonyms #-} module Foo where import GHC.Exts (Int#) pattern Zero :: Int# -- commenting out this line works pattern Zero <- 0# }}} {{{ % ghc-stage2 unlifted-pattern-synonym.hs [1 of 1] Compiling Foo ( unlifted-pattern-synonym.hs, unlifted-pattern-synonym.o ) unlifted-pattern-synonym.hs:7:17: error: • Expecting a lifted type, but ‘Int#’ is unlifted • In the type ‘Int#’ }}} Commenting out the type signature removes the error. GHC 7.10.3 accepts this program, with and without the type signature. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12094> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #11690: stage 1 compiler silently ignores plugins
by GHC 12 Jul '16

12 Jul '16

#11690: stage 1 compiler silently ignores plugins -------------------------------------+------------------------------------- Reporter: osa1 | Owner: osa1 Type: feature | Status: new request | Priority: lowest | 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: -------------------------------------+------------------------------------- It can't load plugins but it shouldn't just silently ignore IMHO. At least a warning would be nice. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11690> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #12212: GHC 8.0.1 crash
by GHC 12 Jul '16

12 Jul '16

#12212: GHC 8.0.1 crash -------------------------------------+------------------------------------- Reporter: dibblego | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: | Type of failure: Compile-time Unknown/Multiple | crash Test Case: | Blocked By: Blocking: | Related Tickets: 9160, 10602 Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Attempting to install the geodetics package results in a GHC 8.0.1 crash. {{{ % ghc --version The Glorious Glasgow Haskell Compilation System, version 8.0.1 % uname -a Linux moron 4.6.2-1-ARCH #1 SMP PREEMPT Wed Jun 8 08:40:59 CEST 2016 x86_64 GNU/Linux % cabal install geodetics-0.0.4 ...omitted [2 of 9] Compiling Geodetics.Ellipsoids ( src/Geodetics/Ellipsoids.hs, dist/dist-sandbox-49e45c7e/build/Geodetics/Ellipsoids.o ) ghc: panic! (the 'impossible' happened) (GHC version 8.0.1 for x86_64-unknown-linux): Template variable unbound in rewrite rule Variable: cobox_scSX Rule "SPEC t @ a @ 'DQuantity @ 'DQuantity @ d @ d' @ 'DQuantity @ 'DQuantity @ d @ d' @ 'DQuantity @ 'DQuantity @ d @ d'" Rule bndrs: [cobox_scSX, cobox_scSY, cobox_scSZ, cobox_scT0, cobox_scT1, $dKnownVariant_scT2, $dKnownVariant_scT3, $dKnownVariant_scT4, $dKnownVariant_scT5, $dKnownVariant_scT6, $dKnownVariant_scT7, $dNum_scT8] LHS args: [TYPE: a_ac0a, TYPE: 'DQuantity, TYPE: 'DQuantity, TYPE: d_ac0b, TYPE: d'_ac0c, TYPE: 'DQuantity, TYPE: 'DQuantity, TYPE: d_ac0b, TYPE: d'_ac0c, TYPE: 'DQuantity, TYPE: 'DQuantity, TYPE: d_ac0b, TYPE: d'_ac0c, CO: <d_ac0b * d'_ac0c>_N, CO: cobox_scSY, CO: <d_ac0b * d'_ac0c>_N, CO: cobox_scT0, CO: cobox_scT1, $dKnownVariant_scT2, $dKnownVariant_scT3, $dKnownVariant_scT4, $dKnownVariant_scT5, $dKnownVariant_scT6, $dKnownVariant_scT7, $dNum_scT8] Actual args: [TYPE: a_ac0a, TYPE: 'DQuantity, TYPE: 'DQuantity, TYPE: d_ac0b, TYPE: d'_ac0c, TYPE: 'DQuantity, TYPE: 'DQuantity, TYPE: d_ac0b, TYPE: d'_ac0c, TYPE: 'DQuantity, TYPE: 'DQuantity, TYPE: d_ac0b, TYPE: d'_ac0c, CO: <d_ac0b * d'_ac0c>_N, CO: D:R:*[1], CO: <d_ac0b * d'_ac0c>_N, CO: D:R:*[1], CO: D:R:*[1], $fKnownVariantDQuantity, $fKnownVariantDQuantity, $fKnownVariantDQuantity, $fKnownVariantDQuantity, $fKnownVariantDQuantity, $fKnownVariantDQuantity, $dNum_acqm, tx_aaQW, v_aaQZ] Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug cabal: Error: some packages failed to install: geodetics-0.0.4 failed during the building phase. The exception was: ExitFailure 1 }}} Similarity to #10602 and #9160 has been pointed out. I also tried: {{{ % cabal --ghc-options="-O2" install geodetics-0.0.4 }}} with no difference in the crash outcome. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12212> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #12168: panic! (the 'impossible' happened) with gi-gtk 3.0.4
by GHC 12 Jul '16

12 Jul '16

#12168: panic! (the 'impossible' happened) with gi-gtk 3.0.4 -------------------------------------+------------------------------------- Reporter: drb226 | 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: -------------------------------------+------------------------------------- I'm able to get two different panics, depending on how I try to build gi- gtk. Firstly, on my mac: {{{ brew install gtk+3 stack unpack gi-gtk-3.0.4 cd gi-gtk-3.0.4 stack init --resolver nightly-2016-06-08 PKG_CONFIG_PATH=/usr/local/opt/libffi/lib/pkgconfig/ stack build }}} This results in {{{ [650 of 650] Compiling GI.Gtk ( GI/Gtk.hs, .stack- work/dist/x86_64-osx/Cabal-1.24.0.0/build/GI/Gtk.o ) ghc: panic! (the 'impossible' happened) (GHC version 8.0.1 for x86_64-apple-darwin): RnEnv.plusAvail bindingSignalWriteNext CheckButtonK{CheckButtonK} }}} Secondly, also on my mac, but within an ubuntu-based docker image: {{{ docker run -it snoyberg/stackage:nightly cd ~ stack unpack gi-gtk-3.0.4 cd gi-gtk-3.0.4 stack init --resolver nightly-2016-06-08 stack build }}} This results in {{{ [599 of 654] Compiling GI.Gtk.Objects.Widget ( GI/Gtk/Objects/Widget.hs, .stack-work/dist/x86_64-linux/Cabal-1.24.0.0/build/GI/Gtk/Objects/Widget.o ) ghc: panic! (the 'impossible' happened) (GHC version 8.0.1 for x86_64-unknown-linux): urk! lookup local fingerprint $fAttrInfoWidgetMarginTopPropertyInfo4 }}} The latter panic was first observed on the stackage build server, during the June 8th nightly build. That's why, to use the same dependencies, I specified the nightly-2016-06-08 stack resolver in order to reproduce it. I originally recorded it here: https://github.com/fpco/stackage/issues/1569 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12168> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #10886: Remove the magic from `Any`
by GHC 11 Jul '16

11 Jul '16

#10886: Remove the magic from `Any` -------------------------------------+------------------------------------- Reporter: goldfire | Owner: goldfire Type: task | Status: new Priority: normal | Milestone: 8.0.1 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: | -------------------------------------+------------------------------------- The `Any` type used to be very highly magical: a poly-kinded type in a mono-kinded world. Then it became only normally magical: a datatype with a return kind of `k` (instead of `*`). Then it became only a bit magical: a closed type family with no equations. Now that we have non-magical empty closed type families, I believe that we can remove the last bit of magic and just define {{{ type family Any :: k where { } }}} somewhere (`GHC.Types`?), wire it in (like, e.g., `Bool`), and be done with it. Objections? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10886> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 14

[GHC] #10311: package name returned from tyConPackage is garbled
by GHC 08 Jul '16

08 Jul '16

#10311: package name returned from tyConPackage is garbled -------------------------------------+------------------------------------- Reporter: j.waldmann | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.1 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- I am using tyConPackage (etc.) to programmatically generate links to haddock documentation (e.g., from a typed edit widget). This is now broken: {{{ import Data.Map import Data.Typeable tyConPackage $ typeRepTyCon $ typeOf $ fromList [(1,2)] in ghc-7.8.4: "containers-0.5.5.1" in ghc-7.10.1: "conta_47ajk3tbda43DFWyeF3oHQ" }}} and I don't see how to get the actual package name. Cf. http://stackoverflow.com/questions/29638177/tyconpackage-changed-from- ghc-7-8-to-ghc-7-10 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10311> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 11

[GHC] #9625: ghc: panic using --enable-executable-dynamic
by GHC 08 Jul '16

08 Jul '16

#9625: ghc: panic using --enable-executable-dynamic -------------------------------------+------------------------------------- Reporter: CoreyOConnor | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.3 Keywords: | Operating System: Architecture: Unknown/Multiple | Unknown/Multiple Difficulty: Unknown | Type of failure: Compile- Blocked By: | time crash Related Tickets: | Test Case: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- (from: https://github.com/haskell/cabal/issues/2039) Using --enable-executable-dynamic leads to a ghc panic. Reproduction steps: 1. clone https://github.com/coreyoconnor/executable-dynamic-issue 2. cabal configure --enable-tests --enable-executable-dynamic 3. cabal test Expected results: Test executes and passes. Actual results: {{{ ghc: panic! (the 'impossible' happened) (GHC version 7.8.2 for x86_64 -unknown-linux): Don't understand library name verify-foo }}} Actual results varies based on the name of "verify-foo" test suite. Other names tried: "verifyFoo", "VerifyFoo". All of which also fail. Removing "--enable-executable-dynamic" results in a pass regardless of test suite name. Reproduced on GHC 7.8.2, GHC 7.8.3, Cabal 1.18, Cabal 1.20. Ubuntu. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9625> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #11617: DYLD_LIBRARY_PATH ignored on Mac OS X 10.11.x
by GHC 07 Jul '16

07 Jul '16

#11617: DYLD_LIBRARY_PATH ignored on Mac OS X 10.11.x --------------------------------------+--------------------------------- Reporter: borsboom | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: hsc2hs | Version: 7.10.3 Keywords: | Operating System: MacOS X Architecture: x86_64 (amd64) | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: --------------------------------------+--------------------------------- This was first reported on the Stack issue tracker at https://github.com/commercialhaskell/stack/issues/1161, but I've traced it to the fact that GHC uses shell script wrappers for many common commands, including `ghc`, `ghci`, and `hsc2hs`. Background: Mac OS X 10.11 introduced System Integrity Protection, which treats parts of the filesystem specially to prevent tampering with the system. Among those protections is that executables in `/bin`, such as `/bin/sh`, have the DYLD_LIBRARY_PATH variable stripped from the environment before being invoked. Since `ghc`, `hsc2hs`, and others are shell scripts that use `#!/bin/sh`, it means DYLD_LIBRARY_PATH is stripped from the environment when invoking them. I was able to work around this issue by copying `/bin/sh` to my home directory and then modifying the `#!` path in the `hsc2hs` to use the copy. I'm not sure what the best way to move forward is. Some possibilities I can think of: * Have installer tools for OS X (such as Stack, ghcformacosx, and Homebrew) apply the workaround themselves, although that leaves out anyone who installs directly from an official bindist or builds from source, and is a duplication of effort. * Have `make install` on OS X copy `/bin/sh` and "fix" the wrappers itself. * Replace the shell script wrappers with something written in Haskell that reads the paths from a configuration file. Note: I set the component to `hsc2hs` since that's where I encountered the problem, but I wouldn't be surprised if it also effected other cases (like `ghci`, and `ghc` with Template Haskell). I also hesitated to categorize this a "bug" since it's OS X's weird behaviour, and not actually something I feel GHC is doing wrong, causing this problem. But I doubt OS X is going to change, although it is possible (but not recommended) to disable SIP by booting into recovery mode and running `csrutil disable`, after which these protections are no longer in place. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11617> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #12228: Make newtype polykinded
by GHC 06 Jul '16

06 Jul '16

#12228: Make newtype polykinded -------------------------------------+------------------------------------- Reporter: dfeuer | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 8.0.1 (Type checker) | 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 hate having to give up so much top type safety when I use unlifted types. It seems to me that I should be able to use `newtype` to create new types of kind `#` as well as `*`: {{{#!hs newtype MyWord# = MyWord# Word# newtype GoodStuff# a = GoodStuff (# Int#, a #) }}} In an ideal world, such types would be in Core (so RULES could see them), but I'd be very happy even to just have them in Haskell. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12228> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #12222: ghc panic kindFunResult with template haskell 'isInstance'
by GHC 06 Jul '16

06 Jul '16

#12222: ghc panic kindFunResult with template haskell 'isInstance' -------------------------------------+------------------------------------- Reporter: ghorn | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Template | Version: 7.10.3 Haskell | Keywords: kindFunResult | Operating System: Unknown/Multiple Architecture: | Type of failure: Compile-time Unknown/Multiple | crash Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- {{{#!hs --TH.hs {-# OPTIONS_GHC -Wall #-} {-# LANGUAGE TemplateHaskell #-} module TH ( SomeClass , doThStuff ) where import Control.Monad ( void ) import Language.Haskell.TH class SomeClass a where doThStuff :: Name -> Q [Dec] doThStuff name = reify name >>= go go :: Info -> Q [Dec] go (TyConI (DataD [] _ _ [(NormalC _ [(_,typ)])] _)) = do void (isInstance ''SomeClass [typ]) -- THIS LINE CRASHES GHC return [] go _ = fail "wrong info" }}} {{{#!hs -- Bug.hs {-# OPTIONS_GHC -Wall #-} {-# LANGUAGE TemplateHaskell #-} module Bug ( Bar(..) ) where import TH data Foo a = Foo a instance SomeClass (Foo a) data Bar f = Bar (Foo (f Int)) doThStuff ''Bar }}} I get this error: {{{ $ runghc -ddump-splices Bug.hs Bug.hs:1:1: Exception when trying to run compile-time code: ghc: panic! (the 'impossible' happened) (GHC version 7.10.3 for x86_64-unknown-linux): kindFunResult Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug Code: doThStuff ''Bar }}} If I comment out either `instance SomeClass (Foo a)` or `void (isInstance ''SomeClass [typ])`, the crash does not occur. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12222> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1