I have been trying to build GHC, but the ghc.exe that gets built for ghc-inplace always exits with "Error 1". It *can* print its version number and the help, but building anything fails. The build fails making Adjustor.o from Adjustor.c.
I'm trying to build a cvs checkout fpconfig ghc hslibs on Win XP Pro, bootstrapping with the provided GHC 5.04.1 binary and Happy 1.13 binary.
I mentioned this to a colleague and he said he had exactly the same problem (ghc exits with error 1), on Windows and Linux, when building from source tarballs and CVS. We don't normally have any problems running the haskell.org GHC binaries.
Can any of you suggest what we might be doing wrong? We are in the southern hemisphere, so our electrons may be orbiting backwards.
:-) I've never seen this myself, I'm afraid. If you have a Linux build which behaves in the same way, could you try stracing it (i.e. prepend 'strace' to the command which fails) and send us the output? Cheers, Simon
Hi I am also facing the same problem. I tried to build a virgin tree, but with no success as Simon PJ has pointed out. I have attached my strace file. Saswat On Mon, 23 Sep 2002, Simon Marlow wrote:
I have been trying to build GHC, but the ghc.exe that gets built for ghc-inplace always exits with "Error 1". It *can* print its version number and the help, but building anything fails. The build fails making Adjustor.o from Adjustor.c.
I'm trying to build a cvs checkout fpconfig ghc hslibs on Win XP Pro, bootstrapping with the provided GHC 5.04.1 binary and Happy 1.13 binary.
I mentioned this to a colleague and he said he had exactly the same problem (ghc exits with error 1), on Windows and Linux, when building from source tarballs and CVS. We don't normally have any problems running the haskell.org GHC binaries.
Can any of you suggest what we might be doing wrong? We are in the southern hemisphere, so our electrons may be orbiting backwards.
:-) I've never seen this myself, I'm afraid. If you have a Linux build which behaves in the same way, could you try stracing it (i.e. prepend 'strace' to the command which fails) and send us the output?
Cheers, Simon _______________________________________________ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
Use strace with the -ff option. This will produce separate traces for each child, which will be easier to interpret. It looks like whatever is going wrong is happening in a child process (but someone who knows more about the innards of ghc may correct me on this). On Monday 23 September 2002 08:04 pm, Saswat Anand wrote:
Hi
I am also facing the same problem. I tried to build a virgin tree, but with no success as Simon PJ has pointed out.
I have attached my strace file.
Saswat
On Mon, 23 Sep 2002, Simon Marlow wrote:
I have been trying to build GHC, but the ghc.exe that gets built for ghc-inplace always exits with "Error 1". It *can* print its version number and the help, but building anything fails. The build fails making Adjustor.o from Adjustor.c.
I'm trying to build a cvs checkout fpconfig ghc hslibs on Win XP Pro, bootstrapping with the provided GHC 5.04.1 binary and Happy 1.13 binary.
I mentioned this to a colleague and he said he had exactly the same problem (ghc exits with error 1), on Windows and Linux, when building from source tarballs and CVS. We don't normally have any problems running the haskell.org GHC binaries.
Can any of you suggest what we might be doing wrong? We are in the southern hemisphere, so our electrons may be orbiting backwards.
:-) I've never seen this myself, I'm afraid. If you have a Linux build
which behaves in the same way, could you try stracing it (i.e. prepend 'strace' to the command which fails) and send us the output?
Cheers, Simon _______________________________________________ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
-- ----------------------------------- Seth Kurtzberg M. I. S. Corp. 1-480-661-1849
Hi Saswat and Dominic. To build yesterday's CVS HEAD (previously autoconfed) using Windows 2000, GHC 5.04.1, and Mingw32 gcc 2.95.3-7 in the southern hemisphere 10 hours ahead of GMT, I did the following (by memory - the details of the configure options may be shaky): - Start the standard Cygwin Bash prompt - Run the following following commands: export PATH=/cygdrive/c/lang/mingw32/bin::${PATH} make clean; ./configure --build=mingw >& configure.log make >& make.log I had "build.mk" in the "mk" directory set for a performance build and SPLIT_OBJS=NO. I built Happy from scratch. This built Happy, the compiler and RTS, libraries, and some of hslibs stopping at the Win32 library due to a missing file, the name of which escapes me. Thinking that I had to run Green-card to generate that file, I tried building Green-card but GHC failed to build the StdDIS module due to an out of range index - reported separately in ghc-bugs. The build has been a bit shaky for me since the Template Haskell changes. Luckily I preserved a complete HEAD build from just before those changes which runs beautifully including heap profiling and the ObjectIO library. Maybe you should try building the STABLE branch (if you were using HEAD previously)? Cheers Mike Thomas. ----- Original Message ----- From: "Saswat Anand" <saswatan@comp.nus.edu.sg> To: "Simon Marlow" <simonmar@microsoft.com> Cc: "Dominic Cooney" <dominic@dcooney.com>; <glasgow-haskell-users@haskell.org> Sent: Tuesday, September 24, 2002 1:04 PM Subject: RE: Newbie building GHC
Hi
I am also facing the same problem. I tried to build a virgin tree, but with no success as Simon PJ has pointed out.
I have attached my strace file.
Saswat
On Mon, 23 Sep 2002, Simon Marlow wrote:
I have been trying to build GHC, but the ghc.exe that gets built for ghc-inplace always exits with "Error 1". It *can* print its version number and the help, but building anything fails. The build fails making Adjustor.o from Adjustor.c.
I'm trying to build a cvs checkout fpconfig ghc hslibs on Win XP Pro, bootstrapping with the provided GHC 5.04.1 binary and Happy 1.13 binary.
I mentioned this to a colleague and he said he had exactly the same problem (ghc exits with error 1), on Windows and Linux, when building from source tarballs and CVS. We don't normally have any problems running the haskell.org GHC binaries.
Can any of you suggest what we might be doing wrong? We are in the southern hemisphere, so our electrons may be orbiting backwards.
:-) I've never seen this myself, I'm afraid. If you have a Linux build which behaves in the same way, could you try stracing it (i.e. prepend 'strace' to the command which fails) and send us the output?
Cheers, Simon _______________________________________________ Glasgow-haskell-users mailing list Glasgow-haskell-users@haskell.org http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
participants (4)
-
Mike Thomas -
Saswat Anand -
Seth Kurtzberg -
Simon Marlow