«

Hello,
Just downloaded the 2.3.2 bzip2 package and trying to install. When I run
'make', I get the following error:
.....
.....
running build
running build_ext
make: *** [sharedmods] Error 139

$ uname -a
Linux k.b.com 2.4.20-20.8 #1 Mon Aug 18 14:59:07 EDT 2003 i686 i686 i386
GNU/Linux

$ gcc -v
Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/3.2/specs
Configured with:
.../configure --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info
--enable-shared --enable-threads=posix --disable-checking --host=i386-redha
t-linux --with-system-zlib --enable-__cxa_atexit
Thread model: posix
gcc version 3.2 20020903 (Red Hat Linux 8.0 3.2-7)

I regularly build earlier versions on the same machine without problems,
although I keep it(the system) current with up2date. Ideas, anyone (Also, I
am installing to my home directory with --prefix=/home/me passed to
configure, if that matters)?

TIA,
PyUser

>> Just downloaded the 2.3.2 bzip2 package and trying to install. When I
>> run 'make', I get the following error:
>> ....
>> ....
>> running build
>> running build_ext
>> make: *** [sharedmods] Error 139

Do you have any more output than that? I presume the Python executable
itself built okay? If so, does "./python Lib/test/regrtest.py" reveal any
problems besides lots of skipped tests because no extension modules have
been built?

Does building from the .tgz file give you the same problem? There should be
no functional difference between the two, but since you mentioned it, I
thought I would check to be sure you weren't somehow attributing the problem
to the bzip2 package.

Skip

Hi Skip,

> >> running build_ext
> >> make: *** [sharedmods] Error 139
>
> Do you have any more output than that? I presume the Python executable
> itself built okay?

Yes, the Py executable is built. I can import sys, os, string etc. But fails
for
socket, for example (Can't import _socket - as expected).

> If so, does "./python Lib/test/regrtest.py" reveal any
> problems besides lots of skipped tests because no extension modules have
> been built?

It doesn't get far; the first import of random fails, failing to import
math. It's checking
correctly in the new locations once I set the PYTHONHOME variable.

> Does building from the .tgz file give you the same problem? There should
be
> no functional difference between the two, but since you mentioned it, I
> thought I would check to be sure you weren't somehow attributing the
problem
> to the bzip2 package.

Didn't try the .tgz. I just mentioned the package I tried as extra
information. I'll dig around
a bit more.

> Skip
PyUser

>> Hi Skip,
>> >> running build_ext
>> >> make: *** [sharedmods] Error 139
>>
>> Do you have any more output than that? I presume the Python
>> executable itself built okay?

>> Yes, the Py executable is built. I can import sys, os, string
>> etc. But fails for socket, for example (Can't import _socket - as
>> expected).

I am in the midst of building from the 2.3.2 bzip2 file on my Mac OS X
system. It's gone well past the point it apparently crapped out on you (in
the midst of building a large number of Mac-specific modules at the
moment). I doubt there's anything related to the bzip2 file itself.

Had you tried 2.3.1? The changes between 2.3.1 and 2.3.2 were very small.
If the older version built for you, I'd be real surprised to discover the
newer version didn't.

Skip

"PyUser" <> writes:

> Hello,
> Just downloaded the 2.3.2 bzip2 package and trying to install. When I run
> 'make', I get the following error:
> ....
> ....
> running build
> running build_ext
> make: *** [sharedmods] Error 139

That looks like it exited in response to a SIGSEV.

Run "ulimit -c unlimited" and see if you get a core file?

Cheers,
mwh

--
Monte Carlo sampling is no way to understand code.
-- Gordon McMillan, comp.lang.python

> Run "ulimit -c unlimited" and see if you get a core file?

No. Actually, it was choking on compiling the ssl module. Recently, I had
played around with the opensssl package and libcrypto.* was all messed up. I
had a few versions (!) of openssl rpms and probably also built from sources.
Anyway, cleaning up all of that and restoring sanity, everything is fine
now.

Thanks to all who replied.

Cheers!
PyUser


> That looks like it exited in response to a SIGSEV.
>
>
> Cheers,
> mwh
>
> --
> Monte Carlo sampling is no way to understand code.
> -- Gordon McMillan, comp.lang.python