Looking for Solaris 2.6/Python 1.5.2 make test fix

Thomas Wouters thomas at xs4all.net
Fri Jan 21 17:50:34 EST 2000


On Fri, Jan 21, 2000 at 04:30:20PM -0600, Rod Haper wrote:

> I'm also getting a segmentation fault on Red Hat Linux 6.1 at the same
> point in 'make test' but no core file is generated:

> rm -f ./Lib/test/*.py[co]
> PYTHONPATH= ./python ./Lib/test/regrtest.py 
> make: *** [test] Segmentation fault
> PYTHONPATH= ./python ./Lib/test/regrtest.py 
> make: *** [test] Segmentation fault

Something probably went wrong with building your Python binary. Can you
start the python executable ? do you get the prompt ? Can you execute code ?
Python should run fine on rh6.1, unless you have a buggy compiler or a buggy
computer, or you built a buggy module statically. How were you building it ?
Which version of python was it, how did you run 'configure' and what did you
change to Modules/Setup ? (You *do* need to change a few things in there.)

> lvirden at cas.org wrote:
> > 
> > When I configure then do a make all and then make test, Python core dumps:
> > 
> > rm -f /ldatae/Misc/python/Python-1.5.2/Lib/test/*.py[co]
> > PYTHONPATH= ./python /ldatae/Misc/python/Python-1.5.2/Lib/test/regrtest.py
> > Bus Error - core dumped
> > make: [test] Error 138 (ignored)
> > PYTHONPATH= ./python /ldatae/Misc/python/Python-1.5.2/Lib/test/regrtest.py
> > Bus Error - core dumped
> > make: *** [test] Error 138

The same goes for solaris, though i cannot be 100% certain that it works
there -- haven't installed Python on the one solaris box we have (but
that'll change if we start using it) but it would strike me as preposterous
if it should fail to work ;) The same goes here, however. Did you build any
things statically ? What happens if you build them dynamically ? Can you
possibly then start python, but not import that specific module ? 

Other helpful hints towards the problem: Did you get any warnings during
compilation ? (Especially the optional modules -- you may have given the
wrong library directory for linking, for instance, or linking with the wrong
library version, compared to the include files.) If either of you can't
figure it out and need a hand, feel free to mail me with the additional info
(or the list, if you think the wonderful spam is suitable for this bunch o'
vikings ;)

-- 
Thomas Wouters <thomas at xs4all.net>

Hi! I'm a .signature virus! copy me into your .signature file to help me spread!




More information about the Python-list mailing list