dmesg output: [ 2086.402678] [drm] stuck on render ring [ 2086.403370] [drm] GPU HANG: ecode 0:0x00200000, in kwin_x11 [2320], reason: Ring hung, action: reset [ 2086.403371] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [ 2086.403372] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [ 2086.403372] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [ 2086.403372] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [ 2086.403373] [drm] GPU crash dump saved to /sys/class/drm/card0/error [ 2086.403425] [drm:intel_pipe_set_base] *ERROR* pin & fence failed [ 2105.418015] [drm] stuck on render ring [ 2105.418751] [drm] GPU HANG: ecode 0:0x00200000, in kwin_x11 [2320], reason: Ring hung, action: reset the X server was nearly completly frozen after this error, but restarting kwin_x11 from tty helped. software I am using: linux-3.17.1 xorg-server-1.16.1 xf86-video-intel-2.99.916 libdrm-2.4.58 mesa-git-1a17098
Please attach /sys/class/drm/card0/error
Created attachment 108724 [details] /sys/class/drm/card0/error I though I've added it :/ here it is. I see now why: the file was just toio big and I guess I didn't got any error because I've created the bug.
Created attachment 108725 [details] sys/class/drm/card0/error.tar.xz wrong type choosen in bugzilla
It seems the batch is just full of 0xFFFFFFFF's Did this happen right after resuming on suspend?
no, I think I played something via bumblebee/primus before that, not sure though.
Created attachment 109083 [details] /sys/class/drm/card0/error got a similar error again: [ 1492.924345] [drm] stuck on render ring [ 1492.925226] [drm] GPU HANG: ecode 0:0x85dffff8, in steam [4115], reason: Ring hung, action: reset [ 1492.925228] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [ 1492.925229] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [ 1492.925229] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [ 1492.925230] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [ 1492.925230] [drm] GPU crash dump saved to /sys/class/drm/card0/error [ 1498.925206] [drm] stuck on render ring [ 1498.926521] [drm] GPU HANG: ecode 0:0x85dffff8, in steam [4115], reason: Ring hung, action: reset now I just wanted to maximize steam.
Created attachment 109288 [details] /sys/class/drm/card0/error and again, this time I wanted to maximize chromium. After the hang chromium wasn't updating its content. [123027.971860] [drm] stuck on render ring [123027.972737] [drm] GPU HANG: ecode 0:0x00208366, in kwin_x11 [2376], reason: Ring hung, action: reset
Created attachment 109455 [details] /sys/class/drm/card0/error and because this bug annoys me, here is another error state, which came up today. [25186.039463] [drm] stuck on render ring [25186.040310] [drm] GPU HANG: ecode 0:0x002883e6, in kwin_x11 [4096], reason: Ring hung, action: reset [25186.040312] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [25186.040312] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [25186.040313] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [25186.040313] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [25186.040314] [drm] GPU crash dump saved to /sys/class/drm/card0/error Could it be, that this is more like a kwin issue?
Created attachment 109733 [details] another one
some timer after the hang I also get stuff like that before I have to reboot: Nov 19 19:25:31 [kernel] ------------[ cut here ]------------ Nov 19 19:25:31 [kernel] WARNING: CPU: 0 PID: 3848 at drivers/gpu/drm/i915/intel_display.c:3361 intel_crtc_wait_for_pending_flips+0x16f/0x180() Nov 19 19:25:31 [kernel] Modules linked in: nvidia(PO) ttm bbswitch(O) iwldvm tuxedo_wmi(O) btusb [last unloaded: nouveau] Nov 19 19:25:31 [kernel] CPU: 0 PID: 3848 Comm: X Tainted: P O 3.17.3-gentoo #2 Nov 19 19:25:31 [kernel] Hardware name: Notebook P15SM /P15SM , BIOS 1.03.04PM v2 03/12/2014 Nov 19 19:25:31 [kernel] 0000000000000009 ffffffff816dd405 0000000000000000 ffffffff810a3e5d Nov 19 19:25:31 [kernel] 0000000000000000 ffff88041d1b9000 ffff88041cfa8120 ffff88041cf39000 Nov 19 19:25:31 [kernel] 0000000000000000 ffffffff813e7bcf ffff880000000000 ffff8800cb2557c0 Nov 19 19:25:31 [kernel] Call Trace: Nov 19 19:25:31 [kernel] [<ffffffff816dd405>] ? dump_stack+0x49/0x6a Nov 19 19:25:31 [kernel] [<ffffffff810a3e5d>] ? warn_slowpath_common+0x6d/0x90 Nov 19 19:25:31 [kernel] [<ffffffff813e7bcf>] ? intel_crtc_wait_for_pending_flips+0x16f/0x180 Nov 19 19:25:31 [kernel] [<ffffffff810ce5c0>] ? __wake_up_sync+0x10/0x10 Nov 19 19:25:31 [kernel] [<ffffffff813e7c80>] ? intel_primary_plane_disable+0xa0/0xd0 Nov 19 19:25:31 [kernel] [<ffffffff8138b60b>] ? setplane_internal+0x1ab/0x310 Nov 19 19:25:31 [kernel] [<ffffffff8138e96d>] ? drm_mode_setplane+0x10d/0x1d0 Nov 19 19:25:31 [kernel] [<ffffffff81381d6f>] ? drm_ioctl+0x1bf/0x590 Nov 19 19:25:31 [kernel] [<ffffffff8115d51f>] ? new_sync_write+0x6f/0xa0 Nov 19 19:25:31 [kernel] [<ffffffff8119283c>] ? fsnotify+0x23c/0x300 Nov 19 19:25:31 [kernel] [<ffffffff8116df37>] ? do_vfs_ioctl+0x2d7/0x4b0 Nov 19 19:25:31 [kernel] [<ffffffff8115fc28>] ? __sb_end_write+0x28/0x60 Nov 19 19:25:31 [kernel] [<ffffffff8115dbc3>] ? vfs_write+0x123/0x1c0 Nov 19 19:25:31 [kernel] [<ffffffff81177363>] ? __fget+0x63/0xa0 Nov 19 19:25:31 [kernel] [<ffffffff8116e146>] ? SyS_ioctl+0x36/0x80 Nov 19 19:25:31 [kernel] [<ffffffff816e43d2>] ? system_call_fastpath+0x16/0x1b Nov 19 19:25:31 [kernel] ---[ end trace 7165c092cda9b3a9 ]--- Nov 19 19:25:31 [kernel] [drm:intel_pipe_set_base] *ERROR* pipe is still busy with an old pageflip Nov 19 19:26:34 [kernel] ------------[ cut here ]------------ Nov 19 19:26:34 [kernel] WARNING: CPU: 3 PID: 3848 at drivers/gpu/drm/i915/intel_display.c:3361 intel_crtc_wait_for_pending_flips+0x16f/0x180() Nov 19 19:26:34 [kernel] Modules linked in: nvidia(PO) ttm bbswitch(O) iwldvm tuxedo_wmi(O) btusb [last unloaded: nouveau] Nov 19 19:26:34 [kernel] CPU: 3 PID: 3848 Comm: X Tainted: P W O 3.17.3-gentoo #2 Nov 19 19:26:34 [kernel] Hardware name: Notebook P15SM /P15SM , BIOS 1.03.04PM v2 03/12/2014 Nov 19 19:26:34 [kernel] 0000000000000009 ffffffff816dd405 0000000000000000 ffffffff810a3e5d Nov 19 19:26:34 [kernel] 0000000000000000 ffff88041d1b9000 ffff88041cfa8120 ffff88041cf39000 Nov 19 19:26:34 [kernel] ffff88041cf39338 ffffffff813e7bcf 0000000000000000 ffff8800cb2557c0 Nov 19 19:26:34 [kernel] Call Trace: Nov 19 19:26:34 [kernel] [<ffffffff816dd405>] ? dump_stack+0x49/0x6a Nov 19 19:26:34 [kernel] [<ffffffff810a3e5d>] ? warn_slowpath_common+0x6d/0x90 Nov 19 19:26:34 [kernel] [<ffffffff813e7bcf>] ? intel_crtc_wait_for_pending_flips+0x16f/0x180 Nov 19 19:26:34 [kernel] [<ffffffff810ce5c0>] ? __wake_up_sync+0x10/0x10 Nov 19 19:26:34 [kernel] [<ffffffff813f2d77>] ? intel_crtc_set_config+0x9c7/0xe80 Nov 19 19:26:34 [kernel] [<ffffffff8138afac>] ? drm_mode_set_config_internal+0x5c/0xe0 Nov 19 19:26:34 [kernel] [<ffffffff8138eb00>] ? drm_mode_setcrtc+0xd0/0x570 Nov 19 19:26:34 [kernel] [<ffffffff81381d6f>] ? drm_ioctl+0x1bf/0x590 Nov 19 19:26:34 [kernel] [<ffffffff8116df37>] ? do_vfs_ioctl+0x2d7/0x4b0 Nov 19 19:26:34 [kernel] [<ffffffff8115fc28>] ? __sb_end_write+0x28/0x60 Nov 19 19:26:34 [kernel] [<ffffffff8115dbc3>] ? vfs_write+0x123/0x1c0 Nov 19 19:26:34 [kernel] [<ffffffff81177363>] ? __fget+0x63/0xa0 Nov 19 19:26:34 [kernel] [<ffffffff8116e146>] ? SyS_ioctl+0x36/0x80 Nov 19 19:26:34 [kernel] [<ffffffff816e43d2>] ? system_call_fastpath+0x16/0x1b Nov 19 19:26:34 [kernel] ---[ end trace 7165c092cda9b3aa ]--- Nov 19 19:26:34 [kernel] [drm:intel_pipe_set_base] *ERROR* pipe is still busy with an old pageflip
Created attachment 109750 [details] another strange one today this was strange, just wanted to port some wallets to kwallets5, so I pressed "Next", so this happend: [20727.430513] [drm] stuck on render ring [20727.432079] [drm] GPU HANG: ecode 0:0x87d7bffa, in kwin_x11 [1867], reason: Ring hung, action: reset [20727.432086] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [20727.432087] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [20727.432088] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [20727.432090] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [20727.432091] [drm] GPU crash dump saved to /sys/class/drm/card0/error [20733.447306] [drm] stuck on render ring [20733.448224] [drm] GPU HANG: ecode 0:0x85dffffd, in X [1724], reason: Ring hung, action: reset [20739.448176] [drm] stuck on render ring [20739.449445] [drm] GPU HANG: ecode 0:0x85dffffd, in X [1724], reason: Ring hung, action: reset [20739.449530] [drm:i915_context_is_banned] *ERROR* gpu hanging too fast, banning! [20745.444996] [drm] stuck on render ring [20745.445909] [drm] GPU HANG: ecode 0:0x87d7bffa, in kwin_x11 [1867], reason: Ring hung, action: reset [20751.449836] [drm] stuck on render ring [20751.450797] [drm] GPU HANG: ecode 0:0x87d7bffa, in kwin_x11 [1867], reason: Ring hung, action: reset
I have no reason to believe this is anything bug a mesa bug, but for sanity's sake could you please first check with drm-intel-nightly?
the issue stopped to occur for me now. But I am now on linux-3.18.1 and recent mesa-master.
Ok, let's hope it is fixed...
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.