On Fri, 2011-05-27 at 10:41 -0400, Eric Sokolowsky wrote: > Robert Osfield wrote: > > Hi All, > > > > I'm just doing a code review from before and after the r12294 rev and > > it looks like the move of the code from a method being called in the > > graphics thread GraphicsWindowX11::swapBufferImplentation() to on > > being called from the main thread GraphcisWindowX11::checkEvents() > > without the display being changed from the one used for the graphics > > thread to the one used for events. If I am correct then it should be > > a simple fix of just replacing the _display variable usage to the > > local display variable. > > > > I will test in the svn/trunk configuration and then with my suggested > > change. Will update you all soon on how I get on. > > > >
Just to chime in, I'm also experiencing this problem. Fedora 14-64, NVidia 270 driver. > I think this is getting close to the problem. I have further narrowed > the problem to the last section of differences in GraphicsWindowX11.cpp > between r12293 and r12294 (I applied all the other parts of the > difference and there is no problem). If I comment out the second while( > XPending(_display) ) loop (see around line 1467) there is no crash. Is > the _display variable changed in the first loop? > > I also tried putting a sleep(1) call just above line 1467 the first time > the function is run and now I can not get it to crash, further > suggesting a race condition that the loop is triggering. > > Eric > > _______________________________________________ > osg-users mailing list > [email protected] > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org > _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
