I see Grisha has posted another patch which I will test next. It's
taking longer to test these things than I thought because of the
slowness of qemu combined with my stumbling around in FreeBSD. Please be
patient. :)
In the mean time I've tested with the LD_PRELOAD and
PYTHON_WITH_THREAD_SUPPORT hacks. It compiles but the unit tests fail.
Even though the unit tests fail, apache seems to load mod_python.so but
has a problem initializing it. This is an improvement.
When I manually start apache with
apachectl -k start -f /home/jim/tmp/mod_python/test/conf/test.conf
I get the following in the test/logs/error_log:
[Mon Sep 12 19:49:33 2005] [notice] mod_python: Creating 8 session
mutexes based on 256 max processes and 0 max threads.
[Mon Sep 12 19:49:33 2005] [emerg] (2)No such file or directory:
Couldn't create accept lock
Apache then exits.
I thought I'd also try installing and running apache and test a simple
script. Results are better, but still negative (and I'm not sure why
it's different than using test.conf).
I installed mod_python as root and started apache with apachectl -k
start. Apache loads mod_python.so and seems to initialize mod_python
without a problem.
Using mod_python.publisher and the following test script:
mp/index.py
-----------
def index(req):
req.content_type = 'text/plain'
req.write('hello world')
The correct content is returned (hello world) but with the following
http-error.log output:
[Mon Sep 12 21:38:57 2005] [notice] mod_python: Creating 8 session
mutexes based on 150 max processes and 0 max threads.
[Mon Sep 12 21:38:58 2005] [notice] Apache/2.0.53 (FreeBSD)
mod_python/3.2.0-dev-20050912 Python/2.4.1 configured -- resuming normal
operations
[Mon Sep 12 21:39:45 2005] [notice] mod_python: (Re)importing module
'mod_python.publisher'
[Mon Sep 12 21:39:49 2005] [notice] [client ::1] Publisher loading page
/usr/local/www/data/mp/index.py
Fatal Python error: PyThreadState_Delete: tstate is still current
[Mon Sep 12 21:39:49 2005] [notice] child pid 8449 exit signal Abort
trap (6)
Jim
