November 2017 - ghc-tickets

[GHC] #14333: GHC doesn't use the fact that Coercible is symmetric
by GHC 23 Jul '18

23 Jul '18

#14333: GHC doesn't use the fact that Coercible is symmetric -------------------------------------+------------------------------------- Reporter: Iceland_jack | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 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: -------------------------------------+------------------------------------- {{{#!hs {-# Language ScopedTypeVariables, GADTs, DeriveGeneric #-} import GHC.Generics import Data.Type.Coercion data AB = A | B deriving (Generic) data XY = X | Y deriving (Generic) data This ab xy where At :: This AB XY foo :: This ab xy -> Coercion (Rep ab a) (Rep xy a) foo At = Coercion bar :: forall ab xy a. This ab xy -> Coercion (Rep ab a) (Rep xy a) bar at | Coercion <- foo at :: Coercion (Rep ab a) (Rep xy a) = Coercion }}} This compiles on 8.2.1 and 8.3.20170920 but flipping the arguments to `Coercion` fails {{{#!hs -- ... -- • Could not deduce: Coercible (Rep xy a) (Rep ab a) -- arising from a use of ‘Coercion’ -- from the context: Coercible (Rep ab a) (Rep xy a) -- bound by a pattern with constructor: bar :: forall ab xy a. This ab xy -> Coercion (Rep xy a) (Rep ab a) bar at | Coercion <- foo at :: Coercion (Rep ab a) (Rep xy a) = Coercion }}} Conceptually there should be an (`Undecidable`) superclass constraint for `Coercible` {{{#!hs class Coercible b a => Coercible a b }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14333> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 16

[GHC] #14414: Profiled program runs 2.5x faster than non-profiled
by GHC 23 Jul '18

23 Jul '18

#14414: Profiled program runs 2.5x faster than non-profiled -------------------------------------+------------------------------------- Reporter: Fuuzetsu | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 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 was looking at benchmarks game (attached as fasta.ghc-2.hs). I have found that with the flags give there, this program on GHC 8.2.1 runs in about a second with `-prof -fprof-auto` and 2.5 seconds without! To run without profiling: {{{ ghc --make -Wall -fforce-recomp -fllvm -O2 -XBangPatterns -threaded -rtsopts -XOverloadedStrings fasta.ghc-2.hs -o fasta.ghc-2.ghc_run && ./fasta.ghc-2.ghc_run +RTS -N4 -s -RTS 250000 > /dev/null }}} Same program with profiling {{{ ghc --make -fforce-recomp -fllvm -prof -fprof-auto -O2 -XBangPatterns -threaded -rtsopts -XOverloadedStrings fasta.ghc-2.hs -o fasta.ghc-2.ghc_run && ./fasta.ghc-2.ghc_run +RTS -N4 -p -s -RTS 250000 > /dev/null }}} I also attach Core outputs for both profiled and unprofiled version. To me this seems very strange: profiled version is somehow faster. Perhaps what's worse is that this means that there's some optimisation GHC is performing when profiling is not on that makes the program a lot slower than it could be! This program is not minimised. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14414> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 16

[GHC] #12714: T9405 fails on Windows
by GHC 22 Jul '18

22 Jul '18

#12714: T9405 fails on Windows --------------------------------------+--------------------------------- Reporter: bgamari | Owner: Type: bug | Status: new Priority: normal | Milestone: 8.2.1 Component: Runtime System | Version: 8.0.1 Keywords: | Operating System: Windows Architecture: x86_64 (amd64) | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: --------------------------------------+--------------------------------- The `T9405` testcase currently fails on Windows, {{{ }}} Contrary to what was suggested in ticket:12004#comment:3 it looks like the problem is that an empty `.ticky` file is produced. Even increasing the `sleep` time to 10 seconds (after bumping the sleep in `T9405.hs` accordingly) doesn't change this, so I suspect there is a runtime system bug here. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12714> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #9599: app runs 10 times faster when compiled with profilling information than without it
by GHC 21 Jul '18

21 Jul '18

#9599: app runs 10 times faster when compiled with profilling information than without it -------------------------------------+------------------------------------- Reporter: kgardas | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.8.2 Keywords: | Operating System: Solaris Architecture: x86 | Type of failure: Runtime Difficulty: Unknown | performance bug Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- Hello, I've found interesting application code which when compiled with {{{ -O2 -threaded --make }}} runs 10 times slower than the same code compiled with {{{ -O2 -threaded --make -prof -auto-all -caf-all -fforce-recomp }}} the application is compiled by GHC 7.8.2 on i386-solaris2 platform. The profiling application is then run as normal application, i.e. without +RTS -p. The application itself is my crude beginner Haskell code which takes Wikipedia dump (XML file) and tries wiki pages from there import into XWiki using REST API. It uses Data.Text package for lazy load of Wikipedia data, whole logic of selection of page/title/content is done on Data.Text and later saved into XPage which is using ByteString. The application also uses Simon Marlow's Async/async/wait code which is copied from Simon Marlow's Parallel and Concurrent Programming in Haskell http://chimera.labs.oreilly.com/books/1230000000929 App attached. My question is, what to do with it to help finding the issue? E.g. I assume it's issue when profiling app is that much faster than common optimized app. For your information, optimized app takes 9m50s to upload 1000 pages while profiled optimized app takes only 50s to upload the same amount of pages on my equipped with E5-2620 (2 GHz 6 core/12 threads xeon) Also the difference in time is that big that it's not caused by a noise in import of pages on XWiki side. I've tested several times of course. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9599> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #14242: Ticks and join points don't play well
by GHC 21 Jul '18

21 Jul '18

#14242: Ticks and join points don't play well -------------------------------------+------------------------------------- Reporter: bgamari | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 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: -------------------------------------+------------------------------------- Currently the presence of ticks appear to inhibit our ability to discover join points as described by this comment in `OccurAnal.occAnal`, {{{#!hs usage_lam = markAllNonTailCalled (markAllInsideLam usage) -- TODO There may be ways to make ticks and join points play -- nicer together, but right now there are problems: -- let j x = ... in tick<t> (j 1) -- Making j a join point may cause the simplifier to drop t -- (if the tick is put into the continuation). So we don't -- count j 1 as a tail call. }}} This causes spurious warnings when compiling with a `DEBUG` compiler and `-g` or profiling enabled. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14242> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3

[GHC] #14388: GHC Panic
by GHC 20 Jul '18

20 Jul '18

#14388: GHC Panic -------------------------------------+------------------------------------- Reporter: aferrandi | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.2 Keywords: IfaceTyLit | Operating System: Linux panic | Architecture: | Type of failure: Compile-time Unknown/Multiple | crash or panic Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Building the project exmap I,ve got: Information:stack: Building 'exmap' ... Information:stack: /usr/bin/stack build . --stack-yaml /home/andrea/prj/exmap/server/stack.yaml "--ghc-options=-Wall -Werror" exmap-0.1.0.0: configure (exe) Configuring exmap-0.1.0.0... exmap-0.1.0.0: build (exe) Preprocessing executable 'exmap' for exmap-0.1.0.0... ghc: panic! (the 'impossible' happened) (GHC version 8.0.2 for x86_64-unknown-linux): get IfaceTyLit 255 Information:stack: Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug Information:stack: -- While building package exmap-0.1.0.0 using: /home/andrea/.stack/setup-exe-cache/x86_64-linux/Cabal- simple_mPHDZzAJ_1.24.2.0_ghc-8.0.2 --builddir=.stack- work/dist/x86_64-linux/Cabal-1.24.2.0 build exe:exmap --ghc-options " -ddump-hi -ddump-to-file" Process exited with code: ExitFailure 1 Information:25.10.17 16:19 - Compilation completed with 1 error and 0 warnings in 1s 946ms Error:stack: Stack build failed with nonzero exit status -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/14388> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #12606: Linux (ELF) Support for "ghc -static -shared"
by GHC 19 Jul '18

19 Jul '18

#12606: Linux (ELF) Support for "ghc -static -shared" ----------------------------------------+--------------------------------- Reporter: tmobile | Owner: Type: feature request | Status: new Priority: normal | Milestone: ⊥ Component: Compiler | Version: 8.0.1 Keywords: | Operating System: Linux Architecture: Unknown/Multiple | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: ----------------------------------------+--------------------------------- On some platforms it is possible to create a stand-alone shared library out of some object `Foo.o` including foreign exports with something like: `ghc -static -shared Foo.o ...` The resulting library will have `Foo`'s dependencies including `base` as well as the RTS statically linked in. On BSDs this is possible because generating position independent code is the default there, and on Windows this is possible because the dynamic loader uses some magic that doesn't require position independent code. Unfortunately on Linux this is not possible because `libHSbase`, `libHSrts`, etc. are not built with `-fPIC` unless one specifically asks for it when building GHC. As far as I know this is the only way to get `-static -shared` to work on this platform. There are use cases for compiling Haskell code into a single shared object that does not have additional runtime dependencies. Suppose one wishes to write a plugin for a program in Haskell. It would be nice to provide the plugin as a single shared object, rather than requiring GHC's libraries to be present at runtime as well. An easy solution might be adding a switch to the configure script or `build.mk` to make it easier to build GHC+libs with `-fPIC` on Linux. Ideally it would be possible to build and install both PIC and non-PIC libraries and have GHC choose the correct one at link time based on the flags and target platform, since PIC code carries a small space/performance overhead and isn't necessary on Linux in most cases. I don't know my way around GHC(outside of the GHC API) or the build system that well yet, but if adding extra build system flags and GHC link-time flag handling code to make this work is an acceptable solution I'm happy to spend the time to get this working. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12606> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #13194: Concurrent modification of package.cache is not safe
by GHC 18 Jul '18

18 Jul '18

#13194: Concurrent modification of package.cache is not safe -------------------------------------+------------------------------------- Reporter: arybczak | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: ghc-pkg | 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: -------------------------------------+------------------------------------- There are a couple of different issues here. 1. On Linux, issuing `ghc-pkg register` for multiple packages in parallel might result in lost updates to package database because of how `registerPackage` function works - it reads existing package databases, picks the one to modify, then checks that package info for the package to register is fine and replaces package database with what was read in the beginning + new package info. Therefore, if updates interleave, it might happen that process1 reads the database, then process2 updates it while process1 still has the old version and uses it for its update later, so update made by process2 is lost. 2. On Windows, update to package database might fail - the issue is that GHC attempts to update it using rename trick, which fails whenever any other process has file to be replaced open for reading. Combine that with the fact that GHC reads package database when compiling packages and you get problems in both Stack (https://github.com/commercialhaskell/stack/issues/2617) and Cabal (https://github.com/haskell/cabal/issues/4005) BTW, rename trick (used for atomic database updates) not only doesn't work on Windows, it's also not atomic e.g. on NFS (https://stackoverflow.com/questions/41362016/rename-atomicity-and-nfs) The solution to both problems is to use OS specific features to lock database file (in shared mode when reading and in exclusive mode when writing). This can be done on Windows using LockFileEx. Unfortunately for POSIX things are a bit more complicated. There are two ways to lock a file on Linux: 1. Using fcntl(F_SET_LK) (POSIX API) 2. Using flock (BSD API) However, fcntl locks have a serious limitation: The record locks described above are associated with the process (unlike the open file description locks described below). This has some unfortunate consequences: * If a process closes any file descriptor referring to a file, then all of the process's locks on that file are released, regardless of the file descriptor(s) on which the locks were obtained. This is bad: it means that a process can lose its locks on a file such as /etc/passwd or /etc/mtab when for some reason a library function decides to open, read, and close the same file. * The threads in a process share locks. In other words, a multithreaded program can't use record locking to ensure that threads don't simultaneously access the same region of a file. Whereas flock is not guaranteed to work with NFS, according to https://en.wikipedia.org/wiki/File_locking#Problems: Whether and how flock locks work on network filesystems, such as NFS, is implementation dependent. On BSD systems, flock calls on a file descriptor open to a file on an NFS-mounted partition are successful no-ops. On Linux prior to 2.6.12, flock calls on NFS files would act only locally. Kernel 2.6.12 and above implement flock calls on NFS files using POSIX byte-range locks. These locks will be visible to other NFS clients that implement fcntl-style POSIX locks, but invisible to those that do not.[4] Assuming that the solution would be to go with locking the database, we would need to: 1. In `registerPackage`, lock all read databases in shared mode except for the database that will later be modified, which has to be locked in exclusive mode. The handle also would need to be kept open and passed to `changeDB` later and used for rewriting the database with updated version in `GHC.PackageDb.writePackageDb` instead of `writeFileAtomic` (which is not actually unconditionally atomic, as demonstrated above). 2. `GHC.PackageDb.decodeFromFile` would lock a file in appropriate mode and return the handle to open file if appropriate. 3. Add support for locking a file. This should be fairly easy to do in GHC.IO.Handle.FD by extending function `openFile'` with appropriate parameters and then adding wrapper function `openLockedFile` or something. We can add both blocking and non-blocking locking to make ghc-pkg show information about waiting for locked package database if appropriate. Alternatively we could add a function similar to the following: `hLock :: Handle -> LockMode -> Bool{-block-} -> IO Bool`, but that requires extracting file descriptor from Handle, which as far as I see is problematic. Is going with locking an acceptable solution here? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13194> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 15

[GHC] #13775: Type family expansion is too lazy, allows accepting of ill-typed terms
by GHC 17 Jul '18

17 Jul '18

#13775: Type family expansion is too lazy, allows accepting of ill-typed terms -------------------------------------+------------------------------------- Reporter: fizruk | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.0.2 (Type checker) | Keywords: | Operating System: MacOS X Architecture: x86_64 | Type of failure: GHC accepts (amd64) | invalid program Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- I'm using GHC 8.0.2 and I've just witnessed a weird bug. To reproduce a bug I use this type family, using `TypeError` (this is a minimal type family I could get to keep bug reproducible): {{{ {-# LANGUAGE DataKinds #-} {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE TypeOperators #-} {-# LANGUAGE UndecidableInstances #-} import GHC.TypeLits type family Head xs where Head '[] = TypeError (Text "empty list") Head (x ': xs) = x }}} Then I go to GHCi, load this code and observe this: {{{ >>> show (Proxy @ (Head '[])) "Proxy" }}} This looks like a bug to me! I expect `Head '[]` to produce a type error. And indeed it does, if I ask differently: {{{ >>> Proxy @ (Head '[]) <interactive>:9:1: error: • empty list • When checking the inferred type it :: Proxy (TypeError ...) }}} So far it looks like `show` somehow "lazily" evaluates it's argument type and that's why it's possible to `show Proxy` even when `Proxy` is ill- typed. But if I expand `Head '[]` manually then it all works as expected again: {{{ >>> show $ Proxy @ (TypeError (Text "error")) <interactive>:13:8: error: • error • In the second argument of ‘($)’, namely ‘Proxy @(TypeError (Text "error"))’ In the expression: show $ Proxy @(TypeError (Text "error")) In an equation for ‘it’: it = show $ Proxy @(TypeError (Text "error")) }}} You can remove `TypeError` from the original type family: {{{ type family Head xs where Head (x ': xs) = x }}} And it gets even weirder: {{{ >>> show (Proxy @ (Head '[])) "Proxy" >>> Proxy @ (Head '[]) Proxy }}} I did not test this with GHC 8.2. I think this behaviour is not critical for me, but accepting ill-typed terms looks like a bad sign, especially for type-level-heavy programs. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13775> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 10

[GHC] #13184: :show bindings broken under -fexternal-interpreter
by GHC 17 Jul '18

17 Jul '18

#13184: :show bindings broken under -fexternal-interpreter -------------------------------------+------------------------------------- Reporter: niteria | Owner: Type: bug | Status: new Priority: low | Milestone: Component: GHCi | 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: -------------------------------------+------------------------------------- This is straight from GHC HEAD: {{{ $ ./inplace/bin/ghc-stage2 --interactive -fexternal-interpreter GHCi, version 8.1.20170123: http://www.haskell.org/ghc/ :? for help Loaded GHCi configuration from /home/bnitka/.ghc/ghci.conf package flags have changed, resetting and loading new packages... Loaded GHCi configuration from /home/bnitka/.ghci Prelude> let a = 1 Prelude> :show bindings a :: Num p => p = *** Exception: ghc-stage2: this operation requires -fno-external- interpreter Prelude> }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13184> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 3