New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Segfault running python rendering test #2056

Closed
strk opened this Issue Oct 29, 2013 · 9 comments

Comments

Projects
None yet
3 participants
@strk
Contributor

strk commented Oct 29, 2013

[strk@cdb:/usr/src/mapnik/mapnik-2.2/tests(2.2.x)] ./run_tests.py 
...
python_tests.python_plugin_test.test_python_point_rendering ... Segmentation fault (core dumped)
@springmeyer

This comment has been minimized.

Show comment
Hide comment
@springmeyer

springmeyer Oct 29, 2013

Member

likely a boost bug, see #1968

Member

springmeyer commented Oct 29, 2013

likely a boost bug, see #1968

@strk

This comment has been minimized.

Show comment
Hide comment
@strk

strk Oct 30, 2013

Contributor

Uhm, I'm using boost 1.49.0, the boost ticket ( https://svn.boost.org/trac/boost/ticket/8290 ) refers to 1.52 as the affected version. Do you know of an older non-affected version ?

Contributor

strk commented Oct 30, 2013

Uhm, I'm using boost 1.49.0, the boost ticket ( https://svn.boost.org/trac/boost/ticket/8290 ) refers to 1.52 as the affected version. Do you know of an older non-affected version ?

@springmeyer

This comment has been minimized.

Show comment
Hide comment
@springmeyer

springmeyer Oct 30, 2013

Member

No, I've not looked into this crash in any detail. I generally disable compilation of the python plugin in all my builds. If the python plugin is an important use case for you then please post a gdb backtrace if possible. Otherwise I would just recommend not turning on compilation of the plugin.

Member

springmeyer commented Oct 30, 2013

No, I've not looked into this crash in any detail. I generally disable compilation of the python plugin in all my builds. If the python plugin is an important use case for you then please post a gdb backtrace if possible. Otherwise I would just recommend not turning on compilation of the plugin.

@strk

This comment has been minimized.

Show comment
Hide comment
@strk

strk Oct 30, 2013

Contributor

Pythin is important for debugging, to use nik2img.py and mapnik-speed-check.
Do they use the python plugin ? Or I'll be happy to disable.

Contributor

strk commented Oct 30, 2013

Pythin is important for debugging, to use nik2img.py and mapnik-speed-check.
Do they use the python plugin ? Or I'll be happy to disable.

@springmeyer

This comment has been minimized.

Show comment
Hide comment
@springmeyer

springmeyer Oct 30, 2013

Member

No, the python plugin is unrelated to the python bindings that are used by nik2img and Mapnik-speed-check

Member

springmeyer commented Oct 30, 2013

No, the python plugin is unrelated to the python bindings that are used by nik2img and Mapnik-speed-check

@strk

This comment has been minimized.

Show comment
Hide comment
@strk

strk Oct 30, 2013

Contributor

How do I disable the python plugin ?
Thank you. I confirm dropping python input plugin gives a successful run of the testsuite.

By the way, before re-configuring with INPUT_PLUGINS, ./configure --help said:

INPUT_PLUGINS: Input drivers to include
    (all|none|comma-separated list of names)
    allowed names: sqlite python ogr shape osm postgis occi rasterlite gdal raster geojson csv
    default: sqlite,ogr,shape,postgis,gdal,raster,geojson,csv
    actual: csv gdal geojson ogr osm postgis python raster shape sqlite

Why was actual != default, when I didn't specify it explicitly ? Does this need another ticket ?

Contributor

strk commented Oct 30, 2013

How do I disable the python plugin ?
Thank you. I confirm dropping python input plugin gives a successful run of the testsuite.

By the way, before re-configuring with INPUT_PLUGINS, ./configure --help said:

INPUT_PLUGINS: Input drivers to include
    (all|none|comma-separated list of names)
    allowed names: sqlite python ogr shape osm postgis occi rasterlite gdal raster geojson csv
    default: sqlite,ogr,shape,postgis,gdal,raster,geojson,csv
    actual: csv gdal geojson ogr osm postgis python raster shape sqlite

Why was actual != default, when I didn't specify it explicitly ? Does this need another ticket ?

@springmeyer

This comment has been minimized.

Show comment
Hide comment
@springmeyer

springmeyer Nov 21, 2013

Member

actual is different than default because it looks like you requested the osm plugin or you did INPUT_PLUGINS=all

Member

springmeyer commented Nov 21, 2013

actual is different than default because it looks like you requested the osm plugin or you did INPUT_PLUGINS=all

@marcu

This comment has been minimized.

Show comment
Hide comment
@marcu

marcu Jun 23, 2014

I have the same problem (on Ubuntu 14.04 - with libboost 1.54). What should I do ?

marcu commented Jun 23, 2014

I have the same problem (on Ubuntu 14.04 - with libboost 1.54). What should I do ?

@springmeyer

This comment has been minimized.

Show comment
Hide comment
@springmeyer

springmeyer Jun 24, 2014

Member

@marcu - If you need the python plugin then build boost from source after patching it: https://github.com/mapnik/mapnik-packaging/blob/master/osx/patches/boost_python_shared_ptr_gil.diff. Then build mapnik against that boost. If you don't need the python plugin then just delete it:

rm `mapnik-config --input-plugins`/python.input
Member

springmeyer commented Jun 24, 2014

@marcu - If you need the python plugin then build boost from source after patching it: https://github.com/mapnik/mapnik-packaging/blob/master/osx/patches/boost_python_shared_ptr_gil.diff. Then build mapnik against that boost. If you don't need the python plugin then just delete it:

rm `mapnik-config --input-plugins`/python.input
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment