On Wed, Nov 03, 2010 at 09:29:50AM +0000, Robert Osfield wrote: > Hi David, > > This sounds like a bug in the driver, but given all your fix is has > been to provide a valid initialize variable for each of the variable I > don't see a problem with merging the change. > > Have you seen this problem with other driver doing indirect rendering?
osg::isGLExtensionOrVersionSupported requires version 3.0 or greater or it never calls glGetIntegerv( GL_NUM_EXTENSIONS, &numExt ); and the only non-nVidia drivers I have access to are Mesa with ATI hardware on versions 1.5 and 2.1. On second thought, if I comment out the check to query anyway, 135 GLint numExt = -1443; (gdb) 136 glGetIntegerv( GL_NUM_EXTENSIONS, &numExt ); (gdb) 137 OSG_WARN << "glGetIntegerv GL_NUM_EXTENSIONS is " (gdb) p numExt $4 = -1443 The answer is yes, it isn't nVidia specific. That makes me wonder if it should fall back and query the GL1/2 interface for getting extensions if numExt is 0. > On Fri, Oct 15, 2010 at 6:39 PM, David Fries <[email protected]> wrote: > > Fix remote X11 crash querying GL_NUM_EXTENSIONS -- David Fries <[email protected]> http://fries.net/~david/ (PGP encryption key available) _______________________________________________ osg-submissions mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-submissions-openscenegraph.org
