August 2018 - ghc-tickets

[GHC] #13296: stat() calls can block Haskell runtime
by GHC 10 Aug '18

10 Aug '18

#13296: stat() calls can block Haskell runtime -------------------------------------+------------------------------------- Reporter: nh2 | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Runtime | Version: 8.0.2 System | Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Runtime Unknown/Multiple | performance bug Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- `getFileStatus` (`stat()` syscall) is marked as `unsafe`, which means that if I have e.g. `+RTS -N4`, I can't stat more than 4 files at the same time without completely stopping the Haskell world. This is an issue on network file systems, where a single `stat()` can easily take 2 milliseconds, so you typically want to do them in parallel (but due to the above you can't). The underlying problem is that there are some Linux syscalls you typically need on networked file systems that have no asynchronous equivalent; according to http://blog.libtorrent.org/2012/10/asynchronous-disk-io/ these are at least: * `stat()` * `open()` * `fallocate()` * `rename()` A quick skim through `libraries/base/System/Posix/Internals.hs` reveals the situation: * `stat()` only exists as `unsafe` * `open()` has both `safe` and `unsafe` variants (I haven't checked which one is used in practice) The remaining ones are in the `unix` package * `fallocate()` is `safe` * `rename()` is `unsafe` It seems to me that there are two issues here: 1) None of these calls should be `unsafe` because they may block for a very long time (e.g. > 0.5 ms even on the fastest LANs). 2) We need to answer the question: If we marked them as `safe`, how many of them would the RTS execute in parallel? To my current knowledge (thanks `rwbarton` and `slyfox` on #ghc), there's a pool of Haskell executing threads (the usual `-RTS -N`), and a pool of FFI threads. Are there any restrictions on the size of that latter pool? The docs https://downloads.haskell.org/~ghc/latest/docs/html/users_guide/ffi- chap.html#foreign-imports-and-multi-threading are not specific on that topic, simply mentioning "but there may be an arbitrary number of foreign calls in progress at any one time, regardless of the +RTS -N value". Does that mean the amount of FFI threads is truly unbounded? -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/13296> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

[GHC] #15478: ghc-pkg package config validation too strict
by GHC 10 Aug '18

10 Aug '18

#15478: ghc-pkg package config validation too strict -------------------------------------+------------------------------------- Reporter: mboes | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 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: -------------------------------------+------------------------------------- Registration of a package using the following config works: {{{ name: frob version: 0.1.0.0 id: frob-0.1.0.0-inplace key: frob-0.1.0.0-inplace exposed: True exposed-modules: Baz from base-4.11.1.0:Data.Bool, Foo from base-4.11.1.0:Data.List, Frob depends: base-4.11.1.0 }}} But while GHC is happy to work with the following variant of the above configuration added by hand to the package database, it's not possible to add it via `ghc-pkg register`, because it's rejected at validation phase: {{{ name: frob version: 0.1.0.0 id: frob-0.1.0.0-inplace key: frob-0.1.0.0-inplace exposed: True exposed-modules: Baz from base:Data.Bool, Foo from base:Data.List, Frob depends: base }}} We get the following error: {{{ $ ghc-pkg register frob.conf frob-0.1.0.0: module reexport refers to a non-existent defining package: base (use --force to override) }}} This is unfortunate, because in general unit-id's don't just have version numbers but also hashes, and in some settings these can be hard to guess. The general point is that ghc-pkg should be exactly as permissive as GHC itself, not less or more. One way to achieve that is to make the `checkModule` function in `utils/ghc-pkg/Main.hs` (called by `validatePackageConfig`) use `lookupPackageName` as a fallback to `lookupUnitId`, and say that if `lookupPackageName` returns exactly one result (i.e. the name is unambiguous) then that should be considered okay. Perhaps not A-okay, because the name could become ambiguous in the future when more entries are added to the database, but that should at most be a warning, not an error, I think. Note: in principle we should be able to use the `--force` flag here to bypass the results of configuration validation. But oddly enough, `--force` ''filters out'' entries that it objects about, rather than just letting them pass through. So `--force` is not a workable workaround. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15478> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #15498: HPC: do notation marks () as non-covered
by GHC 10 Aug '18

10 Aug '18

#15498: HPC: do notation marks () as non-covered -------------------------------------+------------------------------------- Reporter: tom-bop | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Code Coverage | Version: 8.2.1 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Incorrect Unknown/Multiple | error/warning at compile-time Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- With this Main.hs, the "()" in "pure ()" is marked by HPC as a non-covered expression: {{{#!hs foo :: IO () foo = do _ <- putStrLn "Unimportant string" pure () main :: IO () main = do foo putStrLn "Unimportant #2" }}} Repro: - `ghc -fhpc Main.hs` - `./Main` - `hpc markup Main.tix` This is an admittedly small issue, but it can be valuable for a project to aspire to 100% code coverage and false negatives prevent us from getting there. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15498> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #15500: internal error: Unable to commit 1048576 bytes of memory. Deepseq
by GHC 09 Aug '18

09 Aug '18

#15500: internal error: Unable to commit 1048576 bytes of memory. Deepseq --------------------------------------+---------------------------------- Reporter: alxdb | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 Keywords: memory deepseq | Operating System: Linux Architecture: x86_64 (amd64) | Type of failure: Runtime crash Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: --------------------------------------+---------------------------------- the following code is a (nearly) minimal working example that causes the aforementioned internal error. {{{#!hs module Main where import Control.DeepSeq myfunc :: Int -> Int myfunc x = sum . take x $ [0..] logiter :: (NFData a) => Int -> (a -> a) -> a -> IO a logiter iter f x | iter >= 0 = do let y = f x deepseq y print $ "iter " ++ show iter if iter == 0 then return y else logiter (iter - 1) f y | otherwise = error "no negative iter!" main :: IO () main = do print "start" print . show $ myfunc 2000000 print "done" print "start" res <- logiter 5 myfunc 2000000 print "done" print . show $ res }}} Perhaps this is an issue with deepseq, however it the message does say to come and report the bug, so that's what I'm doing. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15500> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #15496: if a variable that isn't in scope is used and main is not defined ghc freaks out
by GHC 09 Aug '18

09 Aug '18

#15496: if a variable that isn't in scope is used and main is not defined ghc freaks out ----------------------------------------+--------------------------------- Reporter: kuhnsB | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Compiler | Version: 8.4.3 Keywords: | Operating System: Linux Architecture: Unknown/Multiple | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: ----------------------------------------+--------------------------------- --y = 5::Double f::Double ->Double f x = x+y --main = print 5 produces: ghc: panic! (the 'impossible' happened) (GHC version 8.0.2 for x86_64-unknown-linux): initTc: unsolved constraints WC {wc_insol = [W] y_axB :: t_axA[tau:1] (CHoleCan: y)} Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug un-commenting either line 1 or 4 causes ghc to identify the other problem correctly -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15496> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #9334: Implement "instance chains"
by GHC 09 Aug '18

09 Aug '18

#9334: Implement "instance chains" -------------------------------------+------------------------------------- Reporter: diatchki | Owner: diatchki Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler (Type | Version: 7.9 checker) | Differential Revisions: Keywords: | Architecture: Operating System: Unknown/Multiple | Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------- It would be useful to implement a version of "instance chains" [1] in GHC, which would eliminate the need for OVERLAPPING_INSTANCES in most (all practcial?) programs. The idea is that programmers can explicitly group and order instances into an "instance chain". For example: {{{ instance (Monad m) => StateM (StateT s m) s where ... else (MonadTrans t, StateM m s) => StateM (t m) s where ... }}} When GHC searches for instances, the instances in a chain are considered together and in order, starting with the first one: 1. If the goal matches the current instance's head, then this instance is selected and the rest are ignored, as if they were not there; 2. If the goal does not match the current instance's head, AND it does not unify with the current instance's head, then we skip the instance and proceed to the next member of the chain; 3. If the goal does not match the current instance's head, but it does unify with it, then we cannot use this chain to solve the goal. In summary: earlier instances in a chain "hide" later instances, and later instances can be reached only if we are sure that none of the previous instance will match. [1] http://web.cecs.pdx.edu/~mpj/pubs/instancechains.pdf -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9334> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 19

[GHC] #15491: No control over synopsis in Library Documentation
by GHC 08 Aug '18

08 Aug '18

#15491: No control over synopsis in Library Documentation -------------------------------------+------------------------------------- Reporter: h001 | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: 8.6.1 Component: Documentation | Version: 8.4.3 Keywords: | Operating System: Windows Architecture: x86_64 | Type of failure: Documentation (amd64) | bug Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- You can not hide synopsis of Library Documentation (Microsoft Edge 41.16299.402.0, Microsoft EdgeHTML 16.16299) so that reading the documentation is impossible. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15491> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #12002: Pragmas after a module declaration are ignored without warning.
by GHC 08 Aug '18

08 Aug '18

#12002: Pragmas after a module declaration are ignored without warning. -------------------------------------+------------------------------------- Reporter: seanparsons | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.10.3 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Other Unknown/Multiple | Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Normally pragmas are placed above the module declaration like so: {{{#!hs {-# LANGUAGE OverloadedStrings #-} module Main where }}} However if the above 2 lines are flipped around like this: {{{#!hs module Main where {-# LANGUAGE OverloadedStrings #-} }}} Then the behaviour seen by a user is that equivalent to the pragma not being present, because it would appear GHC ignores the multi-line comment it looks like. For a novice (I'm not one but it just tripped me up) this would be incredibly baffling (it was for me), it would be nice to see a warning for this at the least. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12002> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

[GHC] #12043: internal error: evacuate: strange closure type
by GHC 08 Aug '18

08 Aug '18

#12043: internal error: evacuate: strange closure type -------------------------------------+------------------------------------- Reporter: mattchan | Owner: Type: bug | Status: new Priority: normal | Milestone: Research needed Component: Compiler | Version: 7.10.3 Keywords: internal | Operating System: Unknown/Multiple error, strange closuren | Architecture: ia64 | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- Scary error while building Liquid Haskell from source: {{{ $ stack install [truncated] text-1.2.2.1: copy/register Progress: 40/61 -- While building package Cabal-1.22.8.0 using: /Users/matt/.stack/setup-exe-cache/x86_64-osx/setup-Simple- Cabal-1.22.5.0-ghc-7.10.3 --builddir=.stack- work/dist/x86_64-osx/Cabal-1.22.5.0 build --ghc-options " -ddump-hi -ddump-to-file" Process exited with code: ExitFailure (-6) Logs have been written to: /Users/matt/liquid/liquidhaskell/.stack- work/logs/Cabal-1.22.8.0.log Configuring Cabal-1.22.8.0... Building Cabal-1.22.8.0... Preprocessing library Cabal-1.22.8.0... [ 1 of 81] Compiling Paths_Cabal ( .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/autogen/Paths_Cabal.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Paths_Cabal.o ) [ 2 of 81] Compiling Distribution.TestSuite ( Distribution/TestSuite.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/TestSuite.o ) [ 3 of 81] Compiling Distribution.Simple.PreProcess.Unlit ( Distribution/Simple/PreProcess/Unlit.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Simple/PreProcess/Unlit.o ) [ 4 of 81] Compiling Distribution.GetOpt ( Distribution/GetOpt.hs, .stack-work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/GetOpt.o ) [ 5 of 81] Compiling Distribution.PackageDescription.Utils ( Distribution/PackageDescription/Utils.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/PackageDescription/Utils.o ) [ 6 of 81] Compiling Distribution.Simple.CCompiler ( Distribution/Simple/CCompiler.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Simple/CCompiler.o ) [ 7 of 81] Compiling Distribution.Compat.Binary ( Distribution/Compat/Binary.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Compat/Binary.o ) [ 8 of 81] Compiling Distribution.Compat.ReadP ( Distribution/Compat/ReadP.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Compat/ReadP.o ) [ 9 of 81] Compiling Distribution.Text ( Distribution/Text.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Text.o ) [10 of 81] Compiling Distribution.Version ( Distribution/Version.hs, .stack-work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Version.o ) [11 of 81] Compiling Language.Haskell.Extension ( Language/Haskell/Extension.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Language/Haskell/Extension.o ) [12 of 81] Compiling Distribution.Compiler ( Distribution/Compiler.hs, .stack-work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Compiler.o ) [13 of 81] Compiling Distribution.Simple.Compiler ( Distribution/Simple/Compiler.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Simple/Compiler.o ) [14 of 81] Compiling Distribution.Simple.GHC.ImplInfo ( Distribution/Simple/GHC/ImplInfo.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Simple/GHC/ImplInfo.o ) [15 of 81] Compiling Distribution.License ( Distribution/License.hs, .stack-work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/License.o ) [16 of 81] Compiling Distribution.ModuleName ( Distribution/ModuleName.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/ModuleName.o ) [17 of 81] Compiling Distribution.Package ( Distribution/Package.hs, .stack-work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/Package.o ) [18 of 81] Compiling Distribution.System ( Distribution/System.hs, .stack-work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/System.o ) [19 of 81] Compiling Distribution.PackageDescription ( Distribution/PackageDescription.hs, .stack- work/dist/x86_64-osx/Cabal-1.22.5.0/build/Distribution/PackageDescription.o ) ghc: internal error: evacuate: strange closure type 41832616 (GHC version 7.10.3 for x86_64_apple_darwin) Please report this as a GHC bug: http://www.haskell.org/ghc/reportabug }}} [[Image(http://i.imgur.com/T44UPg0.png)]] -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/12043> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 6

Re: [GHC] #4442: Add unaligned version of indexWordArray#
by GHC 08 Aug '18

08 Aug '18

#4442: Add unaligned version of indexWordArray# -------------------------------------+------------------------------------- Reporter: tibbe | Owner: reinerp Type: feature request | Status: closed Priority: low | Milestone: Component: Compiler | Version: 7.1 Resolution: fixed | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: #14447 | Differential Rev(s): Phab:D4488 Wiki Page: | -------------------------------------+------------------------------------- Changes (by thomie): * status: new => closed * differential: => Phab:D4488 * resolution: => fixed -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/4442#comment:20> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0