- ghc-tickets - Haskell.org

Re: [GHC] #7961: Remove restrictions on promoting GADT's
by GHC 06 Apr '15

06 Apr '15

#7961: Remove restrictions on promoting GADT's -------------------------------------+------------------------------------- Reporter: danharaj | Owner: Type: feature request | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: None/Unknown | Unknown/Multiple Blocked By: | Test Case: Related Tickets: #6024 | Blocking: | Differential Revisions: Phab:D808 -------------------------------------+------------------------------------- Changes (by goldfire): * differential: => Phab:D808 Comment: I have posted an early version of my patch to implement kind equalities, fixing this ticket. Comments/experimentation welcome. It is easiest to access the branch from https://github.com/goldfirere/ghc, on the `nokinds` branch. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7961#comment:17> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

[GHC] #9457: hsc2hs breaks with `--cflag=-Werror` in cross-compilation mode
by GHC 06 Apr '15

06 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] #10237: compiler/ghci/LibFFI.hsc:42 sizeof(ffi_cif) is not an integer
by GHC 06 Apr '15

06 Apr '15

#10237: compiler/ghci/LibFFI.hsc:42 sizeof(ffi_cif) is not an integer -------------------------------------+------------------------------------- Reporter: erikd | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.12.1 Component: Compiler | Version: 7.11 Keywords: | Operating System: Unknown/Multiple Architecture: | Type of failure: Building GHC Unknown/Multiple | failed Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: | -------------------------------------+------------------------------------- Cross-compiling from linux-amd64 to linux-armhf and I get the above error message. The command (after removing a bunhc of irrelevant command line flags) is: {{{ inplace/bin/hsc2hs --cc=/usr/bin/aarch64-linux-gnu-gcc \ --ld=/usr/bin/aarch64-linux-gnu-gcc \ --cross-safe --cflag=-fno-stack-protector --cross-compile \ --cflag=-Daarch64_HOST_ARCH=1 \ --cflag=-Dlinux_HOST_OS=1 --cflag=-D__GLASGOW_HASKELL__=711 \ --cflag=-Icompiler/../rts/dist/build \ --cflag=-Werror=unused-but-set-variable \ --cflag=-include \ --cflag=compiler/stage2/build/autogen/cabal_macros.h \ compiler/ghci/LibFFI.hsc -o compiler/stage2/build/LibFFI.hs }}} The command `grep 'sizeof(ffi_cif)' compiler/ghci/LibFFI.hsc` finds: {{{#!hs cif <- mallocBytes (#const sizeof(ffi_cif)) }}} Native compile on this system works fine. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10237> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 7

[GHC] #10243: SrcSpan incorrect for non-alphanumeric VarPat
by GHC 05 Apr '15

05 Apr '15

#10243: SrcSpan incorrect for non-alphanumeric VarPat -------------------------------------+------------------------------------- Reporter: mpickering | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.10.2 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: | -------------------------------------+------------------------------------- There is an inconsistency when parsing patterns which contain non- alphanumeric characters. For example, {{{!#hs foo (.) = 5 }}} the pattern parses as the following. {{{ (L {tests/examples/ArgPuncParens.hs:2:5-7} (VarPat (Unqual {OccName: .})))] }}} Then consider the following similar program. {{{#!hs foo (x) = 5 }}} The parse is as follows. {{{ (L {tests/examples/ArgPuncParens.hs:2:5-7} (ParPat (L {tests/examples/ArgPuncParens.hs:2:6} (VarPat (Unqual {OccName: x})))))] }}} Notice two differences. 1. The first parse lacks a surrounding `ParPat`. 2. If the `ParPat` is deliberately omitted then the given location encompasses the surrounding parentheses which is not consistent with the second example where `VarPat` specifically refers to `x`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10243> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #10240: Build flavour 'quick-cross' tries to build integer-gmp
by GHC 03 Apr '15

03 Apr '15

#10240: Build flavour 'quick-cross' tries to build integer-gmp -------------------------------------+------------------------------------- Reporter: erikd | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.12.1 Component: Build | Version: 7.11 System | Operating System: Unknown/Multiple Keywords: cross- | Type of failure: None/Unknown compiling | Blocked By: Architecture: | Related Tickets: Unknown/Multiple | Test Case: | Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- The `BuildFlavour=quick-cross` setting is supposed to build `integer- simple` but some time after commit 5971ad56af it started trying to build `integer-gmp` regardless of the setting of `INTEGER_LIBRARY`. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10240> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 1

[GHC] #10231: TMVar - fmap - orElse clashes in ghc 7.8.4
by GHC 03 Apr '15

03 Apr '15

#10231: TMVar - fmap - orElse clashes in ghc 7.8.4 -------------------------------------+------------------------------------- Reporter: ran | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: libraries | Version: 7.8.4 (other) | Operating System: Linux Keywords: | Type of failure: Incorrect result Architecture: x86_64 | at runtime (amd64) | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- Attached is a tar archive with minimal reproducible example. The problem is that after orElse success I get wrong constructor: I should have timer tick but get reset instead. That is, expected output is {{{ timer thread timer tick timer thread timer tick ... }}} but I get {{{ timer thread reset timer thread reset ... }}} The strange thing is that the problem presents, first, only when the binary is built by cabal (checked only inside sandbox): {{{ $ cabal sandbox init ... $ cabal configure Resolving dependencies... Configuring mvk-0.1.0.0... cabal: At least the following dependencies are missing: stm ==2.4.4 $ cabal install --dependencies-only ... $ cabal build }}} When I run {{{ cabal exec -- ghc --make mvk.hs }}} the resulting binary doesn't show the problem. Also, problem disappears if I: * either comment start of the (long-sleeping) reset thread at all; * or change content type of one of TMVars to, say, Bool But if I change type of content for both to, say, Bool, the problem presents. Furthermore, if with Bool content type I put False in timer thread and True in reset thread (and derive Show and output received value) I get {{{ timer thread received Reset False reset ... }}} That is, underlying Bool is right, but fmapped constructor is wrong. The problem was absent in Debian Wheezy's ghc 7.4.1 (not sure about STM version). Reproduced both in * Debian Jessie on x86_64 with GHC 7.8.4 binary from official site and cabal-install 1.20.0.3 from deb.haskell.org * Debian Wheezy on x86 with GHC 7.8.3 binary from deb.haskell.org and cabal-install 1.20.0.3 built by GHC from dist. In all cases used cabal.config from Stackage's lts-1.4 (included in the attached archive) Output from cabal build -v: {{{ Skipping add-source deps check... Using a sandbox located at /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox The sandbox was created after the package was already configured. Re-configuring with most recently used options. If this fails, please run configure manually. Using a sandbox located at /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox '/opt/ghc/7.8.4/bin/ghc' '--numeric-version' looking for tool ghc-pkg near compiler in /opt/ghc/7.8.4/bin found ghc-pkg in /opt/ghc/7.8.4/bin/ghc-pkg '/opt/ghc/7.8.4/bin/ghc-pkg' '--version' '/opt/ghc/7.8.4/bin/ghc' '--supported-languages' '/opt/ghc/7.8.4/bin/ghc' '--info' Reading available packages... Reading available packages... Choosing modular solver. Resolving dependencies... Configuring mvk-0.1.0.0... Dependency base ==4.7.0.2: using base-4.7.0.2 Dependency stm ==2.4.4: using stm-2.4.4 Using Cabal-1.20.0.0 compiled by ghc-7.4 Using compiler: ghc-7.8.4 Using install prefix: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox Binaries installed in: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/bin Libraries installed in: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/lib/x86_64 -linux-ghc-7.8.4/mvk-0.1.0.0 Private binaries installed in: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/libexec Data files installed in: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/share/x86_64 -linux-ghc-7.8.4/mvk-0.1.0.0 Documentation installed in: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/share/doc/x86_64 -linux-ghc-7.8.4/mvk-0.1.0.0 Configuration files installed in: /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/etc Using alex version 3.1.3 found on system at: /usr/bin/alex Using ar found on system at: /usr/bin/ar No c2hs found No cpphs found No ffihugs found Using gcc version 4.9.2 found on system at: /usr/bin/gcc Using ghc version 7.8.4 found on system at: /opt/ghc/7.8.4/bin/ghc Using ghc-pkg version 7.8.4 found on system at: /opt/ghc/7.8.4/bin/ghc-pkg No greencard found Using haddock version 2.14.3 found on system at: /opt/ghc/7.8.4/bin/haddock Using happy version 1.19.4 found on system at: /usr/bin/happy Using haskell-suite found on system at: haskell-suite-dummy-location Using haskell-suite-pkg found on system at: haskell-suite-pkg-dummy- location No hmake found Using hpc version 0.67 found on system at: /opt/ghc/7.8.4/bin/hpc Using hsc2hs version 0.67 found on system at: /opt/ghc/7.8.4/bin/hsc2hs Using hscolour version 1.20 found on system at: /usr/bin/HsColour No hugs found No jhc found Using ld found on system at: /usr/bin/ld No lhc found No lhc-pkg found No nhc98 found Using pkg-config version 0.28 found on system at: /usr/bin/pkg-config Using strip found on system at: /usr/bin/strip Using tar found on system at: /bin/tar No uhc found Component build order: executable 'mvk' creating dist/build creating dist/build/autogen Building mvk-0.1.0.0... Preprocessing executable 'mvk' for mvk-0.1.0.0... Building executable mvk... creating dist/build/mvk creating dist/build/mvk/mvk-tmp /opt/ghc/7.8.4/bin/ghc --make -no-link -fbuilding-cabal-package -O -j4 -static -outputdir dist/build/mvk/mvk-tmp -odir dist/build/mvk/mvk-tmp -hidir dist/build/mvk/mvk-tmp -stubdir dist/build/mvk/mvk-tmp -i -idist/build/mvk/mvk-tmp -i. -idist/build/autogen -Idist/build/autogen -Idist/build/mvk/mvk-tmp -optP-include -optPdist/build/autogen/cabal_macros.h -hide-all-packages -no-user- package-db -package-db /.main/home/ran/src/transas/knei24/mvk.bug/.cabal- sandbox/x86_64-linux-ghc-7.8.4-packages.conf.d -package-db dist/package.conf.inplace -package-id base-4.7.0.2-bfd89587617e381ae01b8dd7b6c7f1c1 -package-id stm-2.4.4-c36cb8072081a12d13d98a3b4449e55c -XHaskell98 ./mvk.hs -Wall Linking... /opt/ghc/7.8.4/bin/ghc --make -fbuilding-cabal-package -O -static -outputdir dist/build/mvk/mvk-tmp -odir dist/build/mvk/mvk-tmp -hidir dist/build/mvk/mvk-tmp -stubdir dist/build/mvk/mvk-tmp -i -idist/build/mvk /mvk-tmp -i. -idist/build/autogen -Idist/build/autogen -Idist/build/mvk /mvk-tmp -optP-include -optPdist/build/autogen/cabal_macros.h -hide-all- packages -no-user-package-db -package-db /.main/home/ran/src/transas/knei24/mvk.bug/.cabal-sandbox/x86_64-linux- ghc-7.8.4-packages.conf.d -package-db dist/package.conf.inplace -package- id base-4.7.0.2-bfd89587617e381ae01b8dd7b6c7f1c1 -package-id stm-2.4.4-c36cb8072081a12d13d98a3b4449e55c -XHaskell98 ./mvk.hs -o dist/build/mvk/mvk -Wall Linking dist/build/mvk/mvk ... }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10231> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 2

[GHC] #10208: libffi issues executable stacks on i386
by GHC 03 Apr '15

03 Apr '15

#10208: libffi issues executable stacks on i386 -----------------------------------------+--------------------------------- Reporter: slyfox | Owner: simonmar Type: bug | Status: new Priority: normal | Milestone: 7.10.2 Component: Runtime System | Version: 7.10.1 Keywords: | Operating System: Linux Architecture: x86 | Type of failure: None/Unknown Test Case: | Blocked By: Blocking: | Related Tickets: Differential Revisions: D764 | -----------------------------------------+--------------------------------- Noticed today when built binaries for i386-gentoo-linux. Latest build phase checks for executable stacks in final result. {{{ * RWX --- --- usr/lib/ghc-7.10.1/rts/libffi.so.6.0.2 * RWX --- --- usr/lib/ghc-7.10.1/rts/libffi.so * RWX --- --- usr/lib/ghc-7.10.1/rts/libffi.so.6 * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_p.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_l.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_debug.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_thr.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_thr_debug.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_thr_l.a:win32.o * !WX --- --- usr/lib/ghc-7.10.1/rts/libCffi_thr_p.a:win32.o }}} Issue first found and fixed in gentoo: http://bugs.gentoo.org/511634 and then sent upstream: http://sourceware.org/ml/libffi-discuss/2014/msg00058.html -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10208> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 4

[GHC] #10213: GHC User Guide documentation of -fwarn-tabs is now incorrect
by GHC 03 Apr '15

03 Apr '15

#10213: GHC User Guide documentation of -fwarn-tabs is now incorrect -------------------------------------+------------------------------------- Reporter: bmjames | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: | Version: 7.10.1 Documentation | Operating System: Unknown/Multiple Keywords: | Type of failure: Documentation Architecture: | bug Unknown/Multiple | Blocked By: Test Case: | Related Tickets: Blocking: | Differential Revisions: | -------------------------------------+------------------------------------- The [https://downloads.haskell.org/~ghc/7.10.1/docs/users_guide.pdf GHC user guide], section 1.5.2.2: -fwarn-tabs warning flag is turned on by default with this release of GHC. However, section 4.8, regarding -fwarn-tabs, still says: This warning is off by default. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/10213> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 5

Re: [GHC] #2710: -main-is flag in {-# OPTIONS #-} pragma not fully supported
by GHC 03 Apr '15

03 Apr '15

#2710: -main-is flag in {-# OPTIONS #-} pragma not fully supported -------------------------------------+------------------------------------- Reporter: Stephan202 | Owner: Type: bug | Status: infoneeded Priority: lowest | Milestone: 7.12.1 Component: Compiler | Version: 6.8.3 Resolution: | Keywords: -main-is Operating System: Unknown/Multiple | ghc Type of failure: Incorrect | Architecture: warning at compile-time | Unknown/Multiple Blocked By: | Test Case: Related Tickets: #1312 | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- Changes (by thomie): * status: new => infoneeded * related: => #1312 Comment: Please read the following irc discussion. {{{ thomie > Should `-main-is` be a dynamic flag? That is, should the following be allowed: {-# OPTIONS_GHC -main-is Foo.bar -#} thomie > It's currently broken #2710 thomie > But what should happen when it is specified in multiple modules? rwbarton > yeah, or if a module specifies that Main is another module rwbarton > it's really a link-time setting isn't it? rwbarton > Probably the only thing that makes sense to put in an OPTIONS pragma is for a module to redefine which of its own symbols is main, but I'm not sure even that is worth supporting rwbarton > for that matter, what does it mean to mean to specify in an OPTIONS pragma something like -rtsopts/-with-rtopts/-debug/etc. rwbarton > IMO anything like that that doesn't affect the compilation of an individual module should just be forbidden carter_cloud > sounds reasonable to me }}} If there are no objections, I'd like to close this ticket as wontfix, and open a new ticket to show a warning when -main-is/-rtsopts/-with- rtopts/-debug etc. are used in an OPTIONS pragma. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/2710#comment:15> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #4931: hsc2hs emits invalid OPTIONS_GHC pragmas
by GHC 02 Apr '15

02 Apr '15

#4931: hsc2hs emits invalid OPTIONS_GHC pragmas -------------------------------------+------------------------------------- Reporter: awson | Owner: Type: bug | Status: new Priority: low | Milestone: 7.12.1 Component: hsc2hs | Version: 7.0.1 Resolution: | Keywords: newcomer Operating System: Unknown/Multiple | Architecture: Type of failure: None/Unknown | Unknown/Multiple Blocked By: | Test Case: Related Tickets: | Blocking: | Differential Revisions: -------------------------------------+------------------------------------- Comment (by peddie): Thank you for clarifying -- I couldn't tell from the comments that `toArgs` was expected to do something differently. I'll try that approach. I appreciate the guidance, as this is the first time I've dived into the GHC code. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/4931#comment:12> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0