[Bug 103584] [bdw] GPU hang in Xorg

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Tue Nov 14 22:24:36 UTC 2017


https://bugs.freedesktop.org/show_bug.cgi?id=103584

--- Comment #13 from Cyril Moreau <cyril.moreauu at gmail.com> ---
I installed glmark2, as it seems to give more information :

#glmark2 -d
Debug: Found 5 matching FB configs.
Debug: GLX chosen config ID: 0x95 Native Visual ID: 0x20
Debug:   Buffer: 32 bits
Debug:      Red: 8 bits
Debug:    Green: 8 bits
Debug:     Blue: 8 bits
Debug:    Alpha: 8 bits
Debug:    Depth: 24 bits
Debug:  Stencil: 8 bits
Debug: Creating XWindow W: 800 H: 600 VisualID: 0x20
Info: =======================================================
Info:     glmark2 2014.03+git20150611.fa71af2d
Info: =======================================================
Info:     OpenGL Information
Info:     GL_VENDOR:     Intel Open Source Technology Center
Info:     GL_RENDERER:   Mesa DRI Intel(R) HD Graphics 5500 (Broadwell GT2) 
Info:     GL_VERSION:    3.0 Mesa 17.0.7
Info: =======================================================
Debug: Loading vertex shader from file None:
Debug: const vec4 MaterialDiffuse = vec4(1.000000, 1.000000, 1.000000,
1.000000);
Debug: const vec4 LightSourcePosition = vec4(20.000000, 20.000000, 10.000000,
1.000000);
Debug: attribute vec3 position;
Debug: attribute vec3 normal;
Debug: attribute vec2 texcoord;
Debug: 
Debug: uniform mat4 ModelViewProjectionMatrix;
Debug: uniform mat4 NormalMatrix;
Debug: 
Debug: varying vec4 Color;
Debug: varying vec2 TextureCoord;
Debug: 
Debug: void main(void)
Debug: {
Debug:     // Transform the normal to eye coordinates
Debug:     vec3 N = normalize(vec3(NormalMatrix * vec4(normal, 1.0)));
Debug: 
Debug:     // The LightSourcePosition is actually its direction for directional
light
Debug:     vec3 L = normalize(LightSourcePosition.xyz);
Debug: 
Debug:     // Multiply the diffuse value by the vertex color (which is fixed in
this case)
Debug:     // to get the actual color that we will use to draw this vertex with
Debug:     float diffuse = max(dot(N, L), 0.0);
Debug:     Color = vec4(diffuse * MaterialDiffuse.rgb, MaterialDiffuse.a);
Debug: 
Debug:     // Set the texture coordinates as a varying
Debug:     TextureCoord = texcoord;
Debug: 
Debug:     // Transform the position to clip coordinates
Debug:     gl_Position = ModelViewProjectionMatrix * vec4(position, 1.0);
Debug: }
Debug: Loading fragment shader from file None:
Debug: #ifdef GL_ES
Debug: precision mediump float;
Debug: #endif
Debug: varying vec4 Color;
Debug: varying vec2 TextureCoord;
Debug: 
Debug: void main(void)
Debug: {
Debug:     gl_FragColor = Color;
Debug: }
Debug: Loading model from 3ds file '/usr/share/glmark2/models/horse.3ds'
Debug:     Object name: HORSE_L_Mate Vertex count: 3582 Face count: 7172
Info: [build] use-vbo=false:intel_do_flush_locked failed: Input/output error



Let me know how can i give you more information.
I am still looking for information or solutions.

BTW, When I am using Mozilla Firefox 56.0 it works
when I am using Chromium 62.0.3202.75 Built on Ubuntu , running on Ubuntu 16.04
or google-chrome-stable, it does not work, getting the same error (Nov  5
14:23:31 PITA kernel: [  335.832424] [drm] stuck on render ring
Nov  5 14:23:31 PITA kernel: [  335.833016] [drm] GPU HANG: ecode
8:0:0x85dfbeff, in chromium [1745], reason: Ring hung, action: reset

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20171114/3c1da190/attachment.html>


More information about the intel-3d-bugs mailing list