Created attachment 53020 [details] Code that show the bug. Hi, when I execute the code in attachment (buildable with scons, freeglut, and glu), I get the image (in attachment as well), using gallium nvc0 or gallium swrast. Teapot should not be transparent...
Created attachment 53021 [details] Here is the image showing display artifact
I've narrowed the bug : the faulty command is glutInitDisplayMode(GLUT_RGBA | GLUT_DOUBLE); Adding an additionnal GLUT_DEPTH param make the code works. AMD and Nvidia proprietary drivers seem to force GLUT_DEPTH in display mode when depth_test are enabled. I have no idea if there is a spec that tells if such a behavior is legit or not, I'm marking this as NEEDINFO.
(In reply to comment #2) > I've narrowed the bug : > the faulty command is > > glutInitDisplayMode(GLUT_RGBA | GLUT_DOUBLE); > > Adding an additionnal GLUT_DEPTH param make the code works. > > AMD and Nvidia proprietary drivers seem to force GLUT_DEPTH in display mode > when depth_test are enabled. I have no idea if there is a spec that tells if > such a behavior is legit or not, I'm marking this as NEEDINFO. I think what's really happening is you're getting a GLX visual with a depth buffer just by luck with AMD/NVIDIA. glxinfo will show you which visuals have a depth buffer. With the NVIDIA driver, most visuals have a depth buffer. Probably the same thing with AMD's driver. If your application uses depth testing you _must_ call glutInitDisplayMode() with GLUT_DEPTH. Similarly with stencil and accum buffers. I think we can close this as not-a-bug.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.