System Environment: -------------------------- Mesa: (master)eaf9affa5ec9c5fd919e4207ab80b4677650ac67 Xserver:(master)xorg-server-1.15.99.901 Xf86_video_intel: Xf86_video_intel:(master)2.99.911 Kernel: (drm-intel-nightly) git-ec45c7 Bug detailed description: ---------------------------- It’s not a regression, firstly test ogl-samples. Pre-conditions: - build: https://github.com/g-truc/ogl-samples Test-case: gl-320-fbo-layered Expected outcome: ----------------- https://github.com/g-truc/ogl-samples/blob/master/data/templates/reference/gl-320-fbo-layered.png Actual outcome: ------------------------ visible grid is white instead of black. Please see error.png attached.
Created attachment 96097 [details] render error png for gl-320-fbo-layered
Looking at the code in the test I think the result obtained in Mesa is correct and the reference expected output is actually wrong. When rendering to the main framebuffer, the test first clears it in *white* like this: glBindFramebuffer(GL_FRAMEBUFFER, 0); glClearBufferfv(GL_COLOR, 0, &glm::vec4(1.0f, 1.0f, 1.0f, 1.0f)[0]); Then renders to it with 4 different viewports with these configurations: Viewport 0: x=2, y=2, w=316, h=236 Viewport 1: x=322, y=1, w=316, h=236 Viewport 2: x=322, y=242, w=316, h=236 Viewport 3: x=2, y=242, w=316, h=236 So obviously, there are gaps that are not being rendered and will show in the color in which the color buffer was cleared (i.e. white). I'll ping the developers of ogl-samples about this.
From the og-samples developers: "True, it seems that I forgot to update the reference image." I filed an issue in github so they can track and fix this at some point in ogl-samples. The bug in Mesa should be closed.
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.