Re: [matplotlib-devel] refcount bug
An an extra data point, the attached script runs for over 1000 iterations on Linux. By no means am I suggesting that as a fix ;) ...just a data point for someone on Windows that this is probably Windows-specific. Cheers, Mike Jörgen Stenarson wrote: > Hi, > > I have a problem with matplotlib crashing with a ref count assertion > error. I see this problem intermittently both when using ipython using > %run to execute plot scripts many times and when embedding a plot in a > Tk application. > In both cases I use TkAgg as a backend I have compiled matplotlib r3933 > using mingw32 on windows using the win32_static library. > > The smallest self contained example I have been able to come up with is > this: > > import random,time > from numpy import pi,arange,sin > import pylab > > pylab.ion() > x=arange(0,6*pi,0.1) > for i in range(1000): > print i > pylab.cla() > for i in range(1): > pylab.plot(x,sin(random.random()*pi+x*random.random())) > > which when executed generates the following output. Running this script > several times I get the crash after around 60 iterations. On crash I > also a messagebox, see attached png file. > > Does anyone else see this? What can I do to help narrow down this bug. > > /Jörgen > > > C:\python>python bugtest-matplotlib.py > 0 > 1 > 2 > 3 > 4 > 5 > 6 > 7 > 8 > 9 > 10 > 11 > 12 > 13 > 14 > 15 > 16 > 17 > 18 > 19 > 20 > 21 > 22 > 23 > 24 > 25 > 26 > 27 > 28 > 29 > 30 > 31 > 32 > 33 > 34 > 35 > 36 > 37 > 38 > 39 > 40 > 41 > 42 > 43 > 44 > 45 > 46 > 47 > 48 > 49 > 50 > 51 > 52 > 53 > 54 > 55 > 56 > 57 > 58 > 59 > 60 > 61 > 62 > 63 > Fatal Python error: PyEval_RestoreThread: NULL tstate > > This application has requested the Runtime to terminate it in an unusual > way. > Please contact the application's support team for more information. > > > > > > > > - > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > > > > > ___ > Matplotlib-devel mailing list > Matplotlib-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel -- Michael Droettboom Science Software Branch Operations and Engineering Division Space Telescope Science Institute Operated by AURA for NASA - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Matplotlib-devel mailing list Matplotlib-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
Re: [matplotlib-devel] refcount bug
On 10/11/07, Michael Droettboom <[EMAIL PROTECTED]> wrote: > An an extra data point, the attached script runs for over 1000 > iterations on Linux. By no means am I suggesting that as a fix ;) > ...just a data point for someone on Windows that this is probably > Windows-specific. I also ran to 1000 on solaris -- make sure your build is clean, eg by removing both the "build" subdirectory and the mpl install directory when compiling matplotlib. Also, it would be interesting to know if you get this problem in the Agg or PS backend, eg when just saving figures and not displaying them in a GUI. JDH - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Matplotlib-devel mailing list Matplotlib-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
Re: [matplotlib-devel] refcount bug
You're right I'm stuck on windows. I just realised I forgot to say that I'm using python 2.4. /Jörgen Michael Droettboom skrev: > An an extra data point, the attached script runs for over 1000 > iterations on Linux. By no means am I suggesting that as a fix ;) > ...just a data point for someone on Windows that this is probably > Windows-specific. > > Cheers, > Mike > > Jörgen Stenarson wrote: >> Hi, >> >> I have a problem with matplotlib crashing with a ref count assertion >> error. I see this problem intermittently both when using ipython using >> %run to execute plot scripts many times and when embedding a plot in a >> Tk application. >> In both cases I use TkAgg as a backend I have compiled matplotlib >> r3933 using mingw32 on windows using the win32_static library. >> >> The smallest self contained example I have been able to come up with >> is this: >> >> import random,time >> from numpy import pi,arange,sin >> import pylab >> >> pylab.ion() >> x=arange(0,6*pi,0.1) >> for i in range(1000): >> print i >> pylab.cla() >> for i in range(1): >> pylab.plot(x,sin(random.random()*pi+x*random.random())) >> >> which when executed generates the following output. Running this >> script several times I get the crash after around 60 iterations. On >> crash I also a messagebox, see attached png file. >> >> Does anyone else see this? What can I do to help narrow down this bug. >> >> /Jörgen >> >> >> C:\python>python bugtest-matplotlib.py >> 0 >> 1 >> 2 >> 3 >> 4 >> 5 >> 6 >> 7 >> 8 >> 9 >> 10 >> 11 >> 12 >> 13 >> 14 >> 15 >> 16 >> 17 >> 18 >> 19 >> 20 >> 21 >> 22 >> 23 >> 24 >> 25 >> 26 >> 27 >> 28 >> 29 >> 30 >> 31 >> 32 >> 33 >> 34 >> 35 >> 36 >> 37 >> 38 >> 39 >> 40 >> 41 >> 42 >> 43 >> 44 >> 45 >> 46 >> 47 >> 48 >> 49 >> 50 >> 51 >> 52 >> 53 >> 54 >> 55 >> 56 >> 57 >> 58 >> 59 >> 60 >> 61 >> 62 >> 63 >> Fatal Python error: PyEval_RestoreThread: NULL tstate >> >> This application has requested the Runtime to terminate it in an >> unusual way. >> Please contact the application's support team for more information. >> >> >> >> >> >> >> >> - >> This SF.net email is sponsored by: Splunk Inc. >> Still grepping through log files to find problems? Stop. >> Now Search log events and configuration files using AJAX and a browser. >> Download your FREE copy of Splunk now >> http://get.splunk.com/ >> >> >> >> >> ___ >> Matplotlib-devel mailing list >> Matplotlib-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel > - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Matplotlib-devel mailing list Matplotlib-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
Re: [matplotlib-devel] refcount bug
John Hunter skrev: > On 10/11/07, Michael Droettboom <[EMAIL PROTECTED]> wrote: >> An an extra data point, the attached script runs for over 1000 >> iterations on Linux. By no means am I suggesting that as a fix ;) >> ...just a data point for someone on Windows that this is probably >> Windows-specific. > > I also ran to 1000 on solaris -- make sure your build is clean, eg by > removing both the "build" subdirectory and the mpl install directory > when compiling matplotlib. Also, it would be interesting to know if > you get this problem in the Agg or PS backend, eg when just saving > figures and not displaying them in a GUI. > My build was done on after deleting build and dist. I usually have build/lib-win32-2.4 on my python path, i.e. I do not run the install step. However I just discovered another thing that may narrow the problem down. If I delete the fontManager.cache file I get the crash immidiately on startup at the import of pylab, the only information is the same messagebox as before. Keeping the fontManager.cache file I reran the script for different backends both with and without the pylab.ion() line in the script. ion ioff TkAgg 63 (1) 999 WX 390 (2) 999 WXAgg 57 (3) 999 Agg 999 999 (1) Fatal Python error: PyEval_RestoreThread: NULL tstate This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information. (2) Traceback (most recent call last): File "bugtest-matplotlib.py", line 13, in ? pylab.plot(x,sin(random.random()*pi+x*random.random())) File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\pyplot.py", line 1798, in plot draw_if_interactive() File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\backends\backend_wx.py", line 1 217, in draw_if_interactive figManager.canvas.draw() File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\backends\backend_wx.py", line 9 48, in draw self.figure.draw(self.renderer) File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\figure.py", line 612, in draw for a in self.axes: a.draw(renderer) File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\axes.py", line 1338, in draw a.draw(renderer) File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\axis.py", line 593, in draw tick.draw(renderer) File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\axis.py", line 167, in draw if self.tick1On: self.tick1line.draw(renderer) File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\lines.py", line 526, in draw gc = renderer.new_gc() File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\backends\backend_wx.py", line 3 97, in new_gc self.gc.select() File "c:\python\external\matplotlib\build\lib.win32-2.4\matplotlib\backends\backend_wx.py", line 5 16, in select self.SelectObject(self.bitmap) File "C:\Python24\Lib\site-packages\wx-2.8-msw-unicode\wx\_gdi.py", line 4768, in SelectObject return _gdi_.MemoryDC_SelectObject(*args, **kwargs) wx._core.PyAssertionError: C++ assertion "m_refData && m_refData->GetRefCount() == 1" failed at ..\. .\src\common\object.cpp(347) in wxObject::AllocExclusive(): wxObject::AllocExclusive() failed. (3) No output at all - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ ___ Matplotlib-devel mailing list Matplotlib-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
