segmentation fault with the viewer
Dear all, I've started playing around with skimage's viewer, and I get a segmentation fault when I try to close the viewer window, or when I click on the window after having the focus on another window. Also, the command viewer.show() is blocking (I'm running ipython --pylab - corresponding to the TkAgg backend). Running a test script from ipython in gdb yields the following message: Program received signal SIGSEGV, Segmentation fault. 0x00007fffd4aa511d in Shiboken::Conversions::isPythonToCppConvertible(SbkConverter*, _object*) () from /usr/lib/x86_64-linux-gnu/libshiboken-python2.7.so.1.1 I'm running Ubuntu 12.04 with Nvidia's proprietary drivers. I am the only one having this problem? Any idea where the seg fault comes from? I reproduce below the test script. Cheers, Emmanuelle ***************** from skimage import data from skimage.viewer import ImageViewer image = data.coins() from skimage.filter import tv_denoise from skimage.viewer.plugins.base import Plugin denoise_plugin = Plugin(image_filter=tv_denoise) from skimage.viewer.widgets import Slider from skimage.viewer.widgets.history import SaveButtons denoise_plugin += Slider('weight', 0.01, 0.5, update_on='release') denoise_plugin += SaveButtons() viewer = ImageViewer(image) viewer += denoise_plugin viewer.show()
I got it to segfault once using the tk backend for pylab, but it wasn't robustly repeatable (randomly happened once out of a lot of focus losses/regains, and never on close). It looked similar to routine segfaults I get for the tk backend, which I've never been able to track down. That's why I usually avoid the tk backend like the plage... Is this reproducible on any other backends? On Saturday, June 29, 2013 4:20:54 PM UTC-5, Emmanuelle Gouillart wrote:
Dear all,
I've started playing around with skimage's viewer, and I get a segmentation fault when I try to close the viewer window, or when I click on the window after having the focus on another window. Also, the command viewer.show() is blocking (I'm running ipython --pylab - corresponding to the TkAgg backend).
Running a test script from ipython in gdb yields the following message:
Program received signal SIGSEGV, Segmentation fault. 0x00007fffd4aa511d in Shiboken::Conversions::isPythonToCppConvertible(SbkConverter*, _object*) () from /usr/lib/x86_64-linux-gnu/libshiboken-python2.7.so.1.1
I'm running Ubuntu 12.04 with Nvidia's proprietary drivers.
I am the only one having this problem? Any idea where the seg fault comes from? I reproduce below the test script.
Cheers, Emmanuelle
*****************
from skimage import data from skimage.viewer import ImageViewer
image = data.coins() from skimage.filter import tv_denoise from skimage.viewer.plugins.base import Plugin
denoise_plugin = Plugin(image_filter=tv_denoise) from skimage.viewer.widgets import Slider from skimage.viewer.widgets.history import SaveButtons
denoise_plugin += Slider('weight', 0.01, 0.5, update_on='release') denoise_plugin += SaveButtons()
viewer = ImageViewer(image) viewer += denoise_plugin viewer.show()
On Sat, Jun 29, 2013 at 02:58:22PM -0700, Josh Warner wrote:
I got it to segfault once using the tk backend for pylab, but it wasn't robustly repeatable (randomly happened once out of a lot of focus losses/ regains, and never on close).
It looked similar to routine segfaults I get for the tk backend, which I've never been able to track down. That's why I usually avoid the tk backend like the plage...
Is this reproducible on any other backends?
Yes, tk, wx, gtk, I get a segfault with all of them (using qt raises an error). Emmanuelle
On Sat, Jun 29, 2013 at 5:03 PM, Emmanuelle Gouillart < emmanuelle.gouillart@nsup.org> wrote:
On Sat, Jun 29, 2013 at 02:58:22PM -0700, Josh Warner wrote:
I got it to segfault once using the tk backend for pylab, but it wasn't robustly repeatable (randomly happened once out of a lot of focus losses/ regains, and never on close).
It looked similar to routine segfaults I get for the tk backend, which I've never been able to track down. That's why I usually avoid the tk backend like the plage...
Is this reproducible on any other backends?
Yes, tk, wx, gtk, I get a segfault with all of them (using qt raises an error).
Emmanuelle
Are you on master? Specifically, a version later than yesterday morning? The viewer-linking PR that was merged yesterday had a number of fixes for PySide. I don't see any issue on PyQt at the moment. -Tony
participants (3)
-
Emmanuelle Gouillart
-
Josh Warner
-
Tony Yu