September 2006 - Yhc - Haskell.org

Bug report, bad Data.Char module
by Neil Mitchell 25 Sep '06

25 Sep '06

Hi, A few bugs in one here, if you take a look at Data.Char, protectEsc, you'll find the definition protectEsc p f = f . cont where cont s@(c:_) | p c = "\\&" ++ s cont s = s Note the second equation of cont is not in a group with the first! This function gives an error on the empty list, despite what you might think! So bug 1 is that we have bad libraries. Bug 2 is that the compiler doesn't beep there when it should, I guess this is shared nhc code. Interestingly, Catch was able to spot this bug pretty easily, as soon as it got a program that tickled those bits of the library. Thanks Neil

1 0

Fwd: darcs patch: Add not-yet-working typeOf function for YHC.Dynamic
by Neil Mitchell 25 Sep '06

25 Sep '06

Can Samuel please have this email set as allowed for this list? ---------- Forwarded message ---------- From: Samuel Bronson <naesten(a)gmail.com> Date: Sep 24, 2006 7:33 PM Subject: darcs patch: Add not-yet-working typeOf function for YHC.Dynamic To: ndmitchell(a)gmail.com Sun Sep 24 14:13:22 EDT 2006 Samuel Bronson <naesten(a)gmail.com> * Add not-yet-working typeOf function for YHC.Dynamic For instance, print (typeOf "Hello!") gave me this output: (TyGen "v300") This reminds me of the impossibility of putting type signatures on arbitrary subexpressions without scoped type variables. (Which wouldn't help here anyway, since data isn't scoped...)

2 1

Yhc and plugins architecture
by Cédric Paternotte 20 Sep '06

20 Sep '06

Hi, I was wondering whether Yhc was able to provide dynamic loading of modules. Let's say I'd wanted to do something like : myFiboFunction <- loadModule "myPlugin.hbc" "myFiboFunction" putStrlLn $ myFiboFunction 10 Is it possible to achieve this in the current state of Yhc ? I know there's an API out there to load bytecodes at runtime but I'm not sure it's meant to do that... I'm aware of the hs-plugins library for GHC that allows this kind of hot plugging. Since Yhc is built around the idea of bytecode I figured it should come with built-in dynamic loading facilities. Regards, Cedric

4 4

Broken build
by Neil Mitchell 19 Sep '06

19 Sep '06

Hi, YHC_build(["src\packages\haskell98-1.0\Ix.hbc"], ["src\packages\haskell98-1.0\Ix .hs"]) inst\bin\yhc.exe -i src/packages/haskell98-1.0/ -d src/packages/haskell98-1.0/ - c src\packages\haskell98-1.0\Ix.hs yhc.exe: Error: File not found, YHC.Internal Reason: imported from Ix Looked in: D:\sources\yhc\11_sep_2006\yhc-devel\src\packages\haskell98-1.0\ D:\sources\yhc\11_sep_2006\yhc-devel/inst/lib\yhc\packages\yhc-base\1.0 Thats a most recent version, scons fullclean, scons (core=1) Thanks Neil

1 0

Strange build error
by Samuel Bronson 19 Sep '06

19 Sep '06

When I try to build Yhc, I get this: gcc -o inst/bin/yhi src/runtime/BCKernel/external.o src/runtime/BCKernel/hashtable.o src/runtime/BCKernel/heap.o src/runtime/BCKernel/hsffi.o src/runtime/BCKernel/info.o src/runtime/BCKernel/integer.o src/runtime/BCKernel/iofuncs.o src/runtime/BCKernel/jonkers.o src/runtime/BCKernel/main.o src/runtime/BCKernel/make.o src/runtime/BCKernel/mark.o src/runtime/BCKernel/module.o src/runtime/BCKernel/mutator.o src/runtime/BCKernel/pretty.o src/runtime/BCKernel/primitive.o src/runtime/BCKernel/sanity.o src/runtime/BCKernel/stopcopy.o src/runtime/BCKernel/profile.o src/runtime/BCKernel/foreign.o src/runtime/BCKernel/process.o src/runtime/BCKernel/thread.o src/runtime/BCKernel/stable.o src/runtime/BCKernel/builtin/Concurrent.o src/runtime/BCKernel/builtin/Array.o src/runtime/BCKernel/builtin/FFI.o src/runtime/BCKernel/builtin/IO.o src/runtime/BCKernel/builtin/Prelude.o src/runtime/BCKernel/builtin/PackedString.o src/runtime/BCKernel/builtin/RuntimeAPI.o src/runtime/BCKernel/builtin/System.o depends/ctypes/libffi/src/prep_cif.o depends/ctypes/libffi/src/cfield.o depends/ctypes/libffi/src/x86/ffi.o depends/ctypes/libffi/src/x86/sysv.o -L/usr/local/lib -lgmp -lm -lpthread -ldl YHC_build(["src/packages/yhc-base-1.0/Data/Ix.hbc"], ["src/packages/yhc-base-1.0/Data/Ix.hs"]) inst/bin/yhc -i src/packages/yhc-base-1.0/ -d src/packages/yhc-base-1.0/ -c src/packages/yhc-base-1.0/Data/Ix.hs Compiling Data.Ix ( src/packages/yhc-base-1.0/Data/Ix.hs ) Copy("inst/lib/yhc/packages/yhc-base/1.0/Data/Ix.hi", "src/packages/yhc-base-1.0/Data/Ix.hi") YHC_build(["src/packages/haskell98-1.0/Ix.hbc"], ["src/packages/haskell98-1.0/Ix.hs"]) inst/bin/yhc -i src/packages/haskell98-1.0/ -d src/packages/haskell98-1.0/ -c src/packages/haskell98-1.0/Ix.hs yhc: Error: File not found, Data.Ratio Reason: imported from Ix Looked in: /home/naesten/hacking/haskell/yhc-devel/src/packages/haskell98-1.0/ /home/naesten/hacking/haskell/yhc-devel/inst/lib/yhc/packages/yhc-base/1.0 I can't figure out why Yhc thinks that Ix imports Data.Ratio, because it sure doesn't seem to! The SConscript files for haskell98-1.0 and yhc-base-1.0 are quite bewildering... why is there so much in them anyway? I might like to set up a buildbot, though I don't have much CPU or RAM...

3 3

Issue change notifications?
by Neil Mitchell 17 Sep '06

17 Sep '06

Hi, Google Mail gives the option to send notifications of all changes to the bug tracking system to a mailing list. As a result, that mailing list might be a little bit busier. The solution to this would be to create a separate google groups mailing list just for bug tracking notifications. So, I have two options: 1) Set google issue tracker to send email messages here when issues change 2) Create google group just for issue tracker messages, and then subscribe to it for those who want it. Thoughts? Anyone who lurks and doesn't want to know about issue tracking stuff? I have noticed we're starting to do more developer to developer communication on the issue tracker now, so perhaps its good that gets duplicated here? Thanks Neil

2 3

Build failure
by Neil Mitchell 15 Sep '06

15 Sep '06

Hi, On my home machine, I get the following errors with a fresh pull: Checking for GHC version 6.4.1 or later... Found version 6.4.2. Checking whether byte ordering is bigendian... conftest_0.c no Checking size of char... conftest_1.c 1 Checking size of short... conftest_2.c 2 Checking size of int... conftest_3.c 4 Checking size of long... conftest_4.c 4 Checking size of long long... failed Checking size of float... conftest_6.c 4 Checking size of double... conftest_7.c 8 Checking size of void*... conftest_8.c 4 Failed to check types sizes/endianness. You do have a working C compiler don't y ou? If so then please contact the Yhc developers for help. It seems to have passed almost all the tests, certainly has a working C compiler, but bottoms out. Is the long long test blowing it up? Thanks Neil

1 0

how stable is yhc?
by Kartik Vaddadi 15 Sep '06

15 Sep '06

Hello, I'm building a debugger for Haskell that works by modifying the runtime, and I need to choose between yhc & nhc. Everything I've seen points towards yhc, but I wonder how stable it is. I know it's not stable for production use as a compiler, but as a platform for a debugger? For my purposes, it does not have to work with every program. I find that yhc works fine with almost all of the half-dozen small programs (<1 page) I threw at it, but chokes on many of the hugs demos. So would you suggest I work with yhc rather than nhc? Any comments you can give me are greatly appreciated. Thanks. -- Kartik Vaddadi. Home: www.cse.iitb.ac.in/~kart Blogs: kartik.infogami.com, kartik-log.blogspot.com, kartik-rlog.blogspot.com Alternate mail ID: kartik.vad(a)gmail.com "50% Reservation, 100% Politics" - Protest the Indian government's decision to increase reservation in private educational institutions (yfemumbai.blogspot.com)

3 7

build on x86-64 fails with undefined references
by Kartik Vaddadi 14 Sep '06

14 Sep '06

Hello, I have a 32-bit distro (Kubuntu 6.10) running on a 64-bit kernel. The yhc build fails with 'undefined reference' errors. > scons: Reading SConscript files ... > Checking for architecture... x86_64 > Checking for operating system... linux2 > Bootstrapping... > Done. > scons: done reading SConscript files. > scons: Building targets ... > gcc -o inst/bin/yhi src/runtime/BCKernel/sanity.o > src/runtime/BCKernel/info.o src/runtime/BCKernel/iofuncs.o > src/runtime/BCKernel/external.o src/runtime/BCKernel/profile.o > src/runtime/BCKernel/integer.o src/runtime/BCKernel/hsffi.o > src/runtime/BCKernel/foreign.o src/runtime/BCKernel/jonkers.o > src/runtime/BCKernel/mutator.o src/runtime/BCKernel/process.o > src/runtime/BCKernel/stable.o src/runtime/BCKernel/module.o > src/runtime/BCKernel/primitive.o src/runtime/BCKernel/main.o > src/runtime/BCKernel/hashtable.o src/runtime/BCKernel/pretty.o > src/runtime/BCKernel/thread.o src/runtime/BCKernel/stopcopy.o > src/runtime/BCKernel/make.o src/runtime/BCKernel/mark.o > src/runtime/BCKernel/heap.o src/runtime/BCKernel/builtin/Array.o > src/runtime/BCKernel/builtin/IO.o src/runtime/BCKernel/builtin/Prelude.o > src/runtime/BCKernel/builtin/FFI.o > src/runtime/BCKernel/builtin/RuntimeAPI.o > src/runtime/BCKernel/builtin/Concurrent.o > src/runtime/BCKernel/builtin/System.o > src/runtime/BCKernel/builtin/PackedString.o > depends/ctypes/libffi/src/prep_cif.o depends/ctypes/libffi/src/cfield.o > depends/ctypes/libffi/src/x86/ffi64.o > depends/ctypes/libffi/src/x86/unix64.o depends/ctypes/libffi/src/x86/sysv.o > -L/usr/local/lib -lgmp -lm -lpthread -ldl src/runtime/BCKernel/hsffi.o: In > function `hsffi_call':hsffi.c:(.text+0x3a1): undefined reference to > `ffi_call' src/runtime/BCKernel/hsffi.o: In function > `hsffi_evalContext':hsffi.c:(.text+0x439): undefined reference to > `ffi_call' depends/ctypes/libffi/src/prep_cif.o: In function > `ffi_prep_cif':prep_cif.c:(.text+0x20c): undefined reference to > `ffi_prep_cif_machdep' depends/ctypes/libffi/src/x86/sysv.o: In function > `ffi_closure_SYSV':depends/ctypes/libffi/src/x86/sysv.S:(.text+0x8a): > undefined reference to `ffi_closure_SYSV_inner' collect2: ld returned 1 > exit status > scons: *** [inst/bin/yhi] Error 1 > scons: building terminated because of errors. I replaced uname (in /bin) with a version that prints "i386" for the architecture rather than "x86_64", and the build succeeds, with warnings (some ffi-related). But I'm not sure if my uname hack is a good idea (could it break builds of other software?) Does yhc support 64-bit architectures? And since my distribution is 32-bit, will a 64-bit binary run at all? Do you advise me to use a 32-bit kernel with yhc? Thanks a lot. -- Kartik Vaddadi. Home: www.cse.iitb.ac.in/~kart Blogs: kartik.infogami.com, kartik-log.blogspot.com, kartik-rlog.blogspot.com Alternate mail ID: kartik.vad(a)gmail.com "50% Reservation, 100% Politics" - Protest the Indian government's decision to increase reservation in private educational institutions (yfemumbai.blogspot.com)

3 3

buildbot failure in slack_x86
by andrewjwilkinsonw＠gmail.com 14 Sep '06

14 Sep '06

The Buildbot has detected a failed build of slack_x86. Full details are available at: http://www.indiegigs.co.uk:8010/slack_x86/builds/109 Buildbot URL: http://www.indiegigs.co.uk:8010/ Build Reason: The web-page 'force build' button was pressed by '': Build Source Stamp: HEAD Blamelist: BUILD FAILED: failed darcs sincerely, -The Buildbot

1 0