Created attachment 123977 [details] output of /sys/class/drm/card0/error, and list of installed packages. playing with online CAD service onshape, making viewport rotation, scales, moving, got "[drm] stuck on render ring". problem repeteable. cad document is public: https://cad.onshape.com/documents/573ea6d2e4b05d0d35a8fef8/w/de5278f320951921495e09cb/e/dbe8e6ecd88538a1387e34f0 dmesg: [ 1938.185262] [drm] stuck on render ring [ 1938.185271] [drm] GPU crash dump saved to /sys/class/drm/card0/error [ 1938.185272] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [ 1938.185273] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [ 1938.185274] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [ 1938.185275] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [ 1938.188286] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo (0x2e200000 ctx 8) at 0x2e200d50 [ 1945.195027] [drm] stuck on render ring [ 1945.195163] [drm:i915_set_reset_status] *ERROR* render ring hung inside bo (0x2d120000 ctx 1) at 0x2d1202c0 attached: output of drm dump, and starting at line 115408 list of installed packages.
Created attachment 123978 [details] dmesg
Created attachment 123979 [details] drmdump
Created attachment 123980 [details] output of dpkg -l
Created attachment 123981 [details] outptu of xrandr --verbose
Created attachment 123982 [details] output ot intel_reg_dumper
Created attachment 123983 [details] vbios.dump
Created attachment 123984 [details] Xorg.0.log
arch: x86_64 kernel: 3.13.0-61-generic Ubuntu 14.04.4 LTS mb: gigabyte h61m-s1 VGA connection
Repeated with drm.debug=0xe, not X restarted. folowed attachments for this repeat.
Created attachment 123985 [details] intel reg dump 2
Created attachment 123986 [details] Xorg.0.log 2
Created attachment 123987 [details] drm crash dump 2
Created attachment 123988 [details] dmesg 2
(In reply to ski from comment #9) > Repeated with drm.debug=0xe, not X restarted. > folowed attachments for this repeat. NOW x restarted.
Assigning to Mesa product (please let me know if I am mistaken with this GPU Hang). From this error dump, hung is happening in render ring batch with active head at 0x1bc5bd50, with 0x7a000004 (PIPE_CONTROL) as IPEHR (same case in both gpu dump) Batch extract (around 0x1bc5bd50): 0x1bc5bd2c: 0x78021302: 3DSTATE_SAMPLER_STATE_POINTERS: VS mod 1, GS mod 1, PS mod 1 0x1bc5bd30: 0x000076c0: VS sampler state 0x1bc5bd34: 0x00000000: GS sampler state 0x1bc5bd38: 0x00007540: WM sampler state 0x1bc5bd3c: 0x7a000003: PIPE_CONTROL 0x1bc5bd40: 0x00100002: no write, cs stall, stall at scoreboard, 0x1bc5bd44: 0x00000000: destination address 0x1bc5bd48: 0x00000000: immediate dword low 0x1bc5bd4c: 0x00000000: immediate dword high 0x1bc5bd50: 0x7a000003: PIPE_CONTROL 0x1bc5bd54: 0x00004000: qword write, 0x1bc5bd58: 0x03112004: destination address 0x1bc5bd5c: 0x00000000: immediate dword low 0x1bc5bd60: 0x00000000: immediate dword high 0x1bc5bd64: 0x78150003: 3DSTATE_CONSTANT_VS_STATE 0x1bc5bd68: 0x00000000: dword 1 0x1bc5bd6c: 0x00000000: dword 2 0x1bc5bd70: 0x00000000: dword 3 0x1bc5bd74: 0x00000000: dword 4
Please test a new version of Mesa (12 or 13) and mark as REOPENED if you can reproduce and RESOLVED/* if you cannot reproduce.
mesa 12 (ubuntu 16.10). So far so good.
Created attachment 127813 [details] gpu crash dump again, after long idle time.
Sorry no one ever investigated your bug. The good news is that everything seems to be working on kernel 4.9.5 and mesa-13.0.3. Please reopen if you can still reproduce with Mesa 17.0.
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.