Re: [Gcl-devel] Building GCL on Win32 -- ANSI variant still fails on my rebuilt WinXP machine...

[Donald Winiecki]

After reinstalling build tools, as before, I'm still failing (with segfault) on building the ANSI variant of GCL2.6.8pre on my XP setup.  CLTL1 variant builds without drama.

In looking through the logs I also found indication that the machine flagged possible memory problems.  I've run tests on the memory and all looks okay.

Any ideas what else outside of GCL or the build tools might be producing the issues with building ANSI variant of the current CVS for GCL2.6.8pre?

Best,

_don

On Fri, Jan 28, 2011 at 5:17 PM, Donald Winiecki <address@hidden> wrote:

Yes, it is reproducible -- freakish as it seems. I'll reinstall build
tools and check some stuff in the configuration of the virtual WinXP
drive and report back.

_don

On Thu, Jan 27, 2011 at 12:41 PM, Camm Maguire <address@hidden> wrote:
> Greetings, and thanks for your report!
>
> Donald Winiecki <address@hidden> writes:
>
>> Camm,
>>
>> After my WinXP machine died, you took up the task of verifying that
>> GCL268pre could build on that platform.
>>
>> With the end of the semester I reconfigured the dead machine with
>> Ubuntu 10.10 and installed WinXP under emulation through Sun's (now
>> Oracle's) VirtualBox.
>>
>> I can successfully build the CLtL1 version of GCL, but the ANSI
>> version fails at make.  Did you have success building the ANSI version
>> in your trials?  (Building both variants on my WinVista machines is
>> still unproblematic.)
>>
>
> I have no problems with remote access to an axiom win machine, and
> emulation under wine, building all flavors.
>
>> I'll send the logs for configure and make to your home E-mail since
>> they seem to be too big to fit into the GCL-devel list.  I can send
>> them to anyone else interested also -- just let me know.
>>
>
> Here's the problem:
>
> libtool: link: ranlib .libs/libgmp.a
>      0 [main] sh 3248 open_stackdumpfile: Dumping stack trace to
> sh.exe.stackdump
> ./libtool: line -765:  3248 Segmentation fault      (core dumped)
> ranlib .libs/libgmp.a
> make[4]: *** [libgmp.la] Error 139
> make[4]: Leaving directory `/c/_cvs/gcl268pre_ansi/gmp4'
>
> The ranlib in your build tools segfaulted.  What is surprising is that
> this should have happened with the traditional image too.  Are both
> success and failure reproducible?  In any case, this is outside of
> gcl's control.
>
>> Best wishes for the season and new year!
>>
>
> Likewise!
>
>> _don
>>
>> ~~~~~~~~~~~~~~~~~~~~~~~~~
>> Don Winiecki, Ed.D., Ph.D.
>> Professor
>> Boise State University, College of Engineering
>> Department of Instructional & Performance Technology
>> 1910 University Drive, Boise, Idaho 83725-2070 USA
>> E-mail: address@hidden
>> WWW: http://ipt.boisestate.edu
>> Telephone: (+01) 208 426 1899
>> Fax: (+01) 208 426 1970
>> ~~~~~~~~~~~~~~~~~~~~~~~~d
>>
>> _______________________________________________
>> Gcl-devel mailing list
>> address@hidden
>> http://lists.gnu.org/mailman/listinfo/gcl-devel
>>
>>
>>
>>
>
> --
> Camm Maguire                                        address@hidden
> ==========================================================================
> "The earth is but one country, and mankind its citizens."  --  Baha'u'llah
>