March 2016 - ghc-tickets

[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] #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] #11727: Allow one type signature for multiple pattern synonyms
by GHC 01 Jul '16

01 Jul '16

#11727: Allow one type signature for multiple pattern synonyms -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: feature | Status: new request | Priority: normal | Milestone: Component: Compiler | Version: 7.10.3 Keywords: | Operating System: Unknown/Multiple PatternSynonyms | Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- There's no reason why we shouldn't allow the same type signature for multiple pattern synonyms just like ordinary functions. For example, {{{ pattern A,B,C,D :: Int pattern A = 5 pattern B = 6 pattern C = 7 pattern D = 8 }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11727> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #11165: Testsuite framework failures are too easy to ignore and too hard to find
by GHC 01 Jul '16

01 Jul '16

#11165: Testsuite framework failures are too easy to ignore and too hard to find -------------------------------------+------------------------------------- Reporter: bgamari | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Test Suite | 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 Rev(s): | Wiki Page: -------------------------------------+------------------------------------- The testsuite driver really should do a better job of drawing the user's attention to framework failures. The only evidence in the testsuite summary that something went wrong is a non-zero "framework failures" count. If you notice that this count is non-zero then you need to sift manually through the initial log output produced by the driver and find the culprit. It's just generally terrible. Testsuite failures are failures, we should at least treat them as such and include them in the `TEST=...` message in the testsuite summary so they can be easily reproduced. Ideally we would also save the exception that caused the failure and show it in the summary as well. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11165> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #11471: Note [The kind invariant] in TyCoRep needs to be updated
by GHC 28 Jun '16

28 Jun '16

#11471: Note [The kind invariant] in TyCoRep needs to be updated -------------------------------------+------------------------------------- Reporter: bgamari | Owner: Type: task | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.3 (Type checker) | Keywords: TypeInType | Operating System: Unknown/Multiple Architecture: | Type of failure: None/Unknown Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- This Note, found in TyCoRep, needs to be updated to reflect `TypeInType`. It should not reference `OpenKind`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/11471> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 30

[GHC] #9078: Segfault with makeStableName
by GHC 28 Jun '16

28 Jun '16

#9078: Segfault with makeStableName ------------------------------------+------------------------------------- Reporter: robertce | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: Unknown/Multiple Architecture: Unknown/Multiple | Type of failure: None/Unknown Difficulty: Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | ------------------------------------+------------------------------------- Calling `makeStableName` on the same object many times causes a segfault. The following triggers this behaviour on both OSX and Linux-x86_64, provided it is compiled (it doesn't seem to occur if the call to `makeStableName` is interpreted). {{{ module Main where import Control.Monad import System.Mem.StableName main :: IO () main = replicateM_ 100000 (makeStableName foo) foo :: Int foo = 1 }}} It seems to happen largely at random. The 100000 calls to `makeStableName` in the above example essentially guarantee its occurrence, but I have observed it happen with far fewer. The reason this bug was discovered is it is causing frequent failures in Accelerate on 7.8.2. See [https://github.com/AccelerateHS/accelerate/issues/162] In that case it seems to be causing more `strange closure type` errors than segfaults, but I assume it is the same root cause. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9078> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 24

[GHC] #10686: Process stops responding to sigINT
by GHC 22 Jun '16

22 Jun '16

#10686: Process stops responding to sigINT -------------------------------------+------------------------------------- Reporter: hamish | 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 | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- Every now and then the leksah-server process has required a kill -9 it to get rid of it. Now Leksah is using the threaded RTS it seems to be happening in Leksah too. It is worse in Leksah because child processes inherit the behaviour and do not terminate when sent Leksah sends them sigINT (which Leksah does very frequently). Once the problem arrises it continues until the Leksah is restarted. Both leksah-server and Leksah start child processes using the process package. I think process package and the GHC GC both call blockUserSignals and it only has one place to store savedSignals. My guess is that two calls go into blockUserSignals from different threads before unblockUserSignals is called on either. The result is that when unblockUserSignals is called it does nothing (because savedSignals was overwritten). Perhaps savedSignals could use some kind of thread local storage? Also docs seem to indicate sigprocmask should not be used at all in multithreaded processes. Perhaps the GHC RTS should use pthread_sigprocmask instead? If the calling process delegates Ctrl+C the problem might be largely hidden. The sigINT signal is enabled again in the child fork and the parent will get an exception if the child is terminated with a sigINT. However I believe the parent is still left in a state where it will no longer respond to sigINT if the child process terminates normally (assuming two calls in a row were made to blockUserSignals at some point). The work around for now in Leksah is to reenable sigINT before each createProcess call. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10686> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2