Created attachment 117295 [details] drm/card0/error log [50942.976162] [drm] stuck on render ring [50942.980758] [drm] GPU HANG: ecode 3:0:0x0263ffc2, in Xorg [1620], reason: Ring hung, action: reset [50942.980771] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [50942.980777] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [50942.980783] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [50942.980789] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [50942.980796] [drm] GPU crash dump saved to /sys/class/drm/card0/error [50942.980931] ------------[ cut here ]------------ [50942.981023] WARNING: CPU: 2 PID: 5314 at /home/kernel/COD/linux/drivers/gpu/drm/i915/intel_display.c:3291 intel_crtc_wait_for_pending_flips+0x138/0x1f0 [i915]() [50942.981027] WARN_ON(ret) [50942.981032] Modules linked in: drbg ansi_cprng ctr ccm zram lz4_compress arc4 acer_wmi sparse_keymap iwlmvm uvcvideo mac80211 videobuf2_vmalloc videobuf2_memops dm_multipath videobuf2_core snd_hda_codec_realtek v4l2_common videodev snd_hda_codec_generic scsi_dh coretemp bnep media snd_hda_intel snd_hda_codec iwlwifi snd_hda_core rfcomm snd_hwdep bluetooth snd_pcm kvm snd_seq_midi snd_seq_midi_event snd_rawmidi cfg80211 snd_seq snd_seq_device joydev input_leds rtsx_pci_ms snd_timer memstick serio_raw snd soundcore shpchp lpc_ich mac_hid binfmt_misc dm_mirror dm_region_hash dm_log i915 rtsx_pci_sdmmc i2c_algo_bit psmouse drm_kms_helper ahci drm libahci r8169 rtsx_pci mii video wmi [50942.981167] CPU: 2 PID: 5314 Comm: kworker/u8:2 Not tainted 4.2.0-999-generic #201507192201 [50942.981174] Hardware name: Packard Bell dot s/SJE06_PT, BIOS V1.15 10/21/2011 [50942.981247] Workqueue: i915-hangcheck i915_hangcheck_elapsed [i915] [50942.981255] 00000000 00000000 c01c1d5c c16f69a4 c01c1d9c c01c1d8c c10632d7 f8be48dd [50942.981270] c01c1db8 000014c2 f8bd6c20 00000cdb f8b7c6b8 f8b7c6b8 f5e38000 00000001 [50942.981285] f382fc34 c01c1da4 c1063343 00000009 c01c1d9c f8be48dd c01c1db8 c01c1ddc [50942.981300] Call Trace: [50942.981316] [<c16f69a4>] dump_stack+0x41/0x52 [50942.981327] [<c10632d7>] warn_slowpath_common+0x87/0xc0 [50942.981404] [<f8b7c6b8>] ? intel_crtc_wait_for_pending_flips+0x138/0x1f0 [i915] [50942.981479] [<f8b7c6b8>] ? intel_crtc_wait_for_pending_flips+0x138/0x1f0 [i915] [50942.981488] [<c1063343>] warn_slowpath_fmt+0x33/0x40 [50942.981563] [<f8b7c6b8>] intel_crtc_wait_for_pending_flips+0x138/0x1f0 [i915] [50942.981615] [<f88a490c>] ? drm_modeset_lock_all_crtcs+0x8c/0xa0 [drm] [50942.981690] [<f8b7df37>] intel_crtc_disable_planes+0x37/0xf0 [i915] [50942.981764] [<f8b7e09a>] intel_prepare_reset+0x6a/0x80 [i915] [50942.981831] [<f8b52709>] i915_handle_error+0x259/0x6e0 [i915] [50942.981844] [<c10b2f0e>] ? vprintk_emit+0x2de/0x540 [50942.981853] [<c10b32e1>] ? vprintk_default+0x41/0x60 [50942.981920] [<f8b52e33>] i915_hangcheck_elapsed+0x253/0x420 [i915] [50942.981931] [<c1078d4a>] process_one_work+0x11a/0x360 [50942.981939] [<c107932d>] worker_thread+0xed/0x430 [50942.981947] [<c1079240>] ? rescuer_thread+0x2b0/0x2b0 [50942.981955] [<c107d9bb>] kthread+0x9b/0xb0 [50942.981966] [<c16fce81>] ret_from_kernel_thread+0x21/0x30 [50942.981974] [<c107d920>] ? kthread_create_on_node+0x110/0x110 [50942.981982] ---[ end trace 14191389931e7b38 ]--- [50943.327312] drm/i915: Resetting chip after gpu hang
Hmm, doesn't exhibit obvious incoherency - but that doesn't mean that there wasn't! I have a theory that we need: http://cgit.freedesktop.org/~ickle/linux-2.6/commit/?h=nightly&id=198d70d03c8cce69d01233c5e68db1e9e6e5e676
I'm on 4.2-rc - apply patch by hands?
Ok, 4.2.0-rc3-00020-g5a5ca73 with manually applied patch work longer, gpu hang after some hours of work. [72592.964165] [drm] stuck on render ring [72592.968841] [drm] GPU HANG: ecode 3:0:0x0263ff38, in Xorg [1581], reason: Ring hung, action: reset [72592.968849] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace. [72592.968853] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel [72592.968857] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue. [72592.968861] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it. [72592.968865] [drm] GPU crash dump saved to /sys/class/drm/card0/error [72592.969016] ------------[ cut here ]------------ [72592.969108] WARNING: CPU: 2 PID: 19943 at drivers/gpu/drm/i915/intel_display.c:3291 intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915]() [72592.969115] WARN_ON(ret) [72592.969120] Modules linked in: drbg ansi_cprng ctr ccm zram lz4_compress bnep rfcomm bluetooth arc4 uvcvideo iwlmvm videobuf2_vmalloc videobuf2_memops videobuf2_core v4l2_common mac80211 acer_wmi videodev sparse_keymap media snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel dm_multipath snd_hda_codec scsi_dh snd_hda_core snd_hwdep coretemp snd_pcm snd_seq_midi snd_seq_midi_event snd_rawmidi iwlwifi kvm joydev snd_seq serio_raw rtsx_pci_ms cfg80211 snd_seq_device snd_timer memstick snd lpc_ich soundcore shpchp mac_hid binfmt_misc dm_mirror dm_region_hash dm_log i915 rtsx_pci_sdmmc i2c_algo_bit drm_kms_helper ahci psmouse drm libahci r8169 rtsx_pci mii wmi video [72592.969266] CPU: 2 PID: 19943 Comm: kworker/u8:1 Not tainted 4.2.0-rc3-00020-g5a5ca73-dirty #2 [72592.969273] Hardware name: Packard Bell dot s/SJE06_PT, BIOS V1.15 10/21/2011 [72592.969350] Workqueue: i915-hangcheck i915_hangcheck_elapsed [i915] [72592.969360] 00000002 00000000 d7703d5c c1746f99 d7703da0 d7703d90 c10597d1 f887d173 [72592.969376] d7703dbc 00004de7 f886f5c0 00000cdb f8811ffe 00000cdb f8811ffe f1ff1c34 [72592.969393] f5fd8000 f1f1b800 d7703da8 c1059843 00000009 d7703da0 f887d173 d7703dbc [72592.969409] Call Trace: [72592.969428] [<c1746f99>] dump_stack+0x41/0x52 [72592.969441] [<c10597d1>] warn_slowpath_common+0x91/0xd0 [72592.969523] [<f8811ffe>] ? intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915] [72592.969602] [<f8811ffe>] ? intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915] [72592.969613] [<c1059843>] warn_slowpath_fmt+0x33/0x40 [72592.969698] [<f8811ffe>] intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915] [72592.969766] [<f82a0bc4>] ? drm_modeset_lock_all_crtcs+0x94/0xa0 [drm] [72592.969849] [<f8813a47>] intel_crtc_disable_planes+0x37/0xf0 [i915] [72592.969935] [<f8813bca>] intel_prepare_reset+0x7a/0x90 [i915] [72592.970009] [<f87e63f7>] i915_handle_error+0x147/0x6e0 [i915] [72592.970032] [<c10aee4e>] ? vprintk_default+0x3e/0x50 [72592.970104] [<f87e6be5>] i915_hangcheck_elapsed+0x205/0x3c0 [i915] [72592.970116] [<c1001af2>] ? __switch_to+0x192/0x500 [72592.970129] [<c10703d2>] process_one_work+0x122/0x3a0 [72592.970137] [<c1070d09>] worker_thread+0x39/0x440 [72592.970144] [<c1070cd0>] ? rescuer_thread+0x2e0/0x2e0 [72592.970152] [<c1075698>] kthread+0xa8/0xc0 [72592.970161] [<c174cac1>] ret_from_kernel_thread+0x21/0x30 [72592.970168] [<c10755f0>] ? kthread_create_on_node+0x120/0x120 [72592.970174] ---[ end trace 2293beecb2f21023 ]--- [72593.318823] drm/i915: Resetting chip after gpu hang [72648.988231] [drm] stuck on render ring [72648.993593] [drm] GPU HANG: ecode 3:0:0x0263ff38, in Xorg [1581], reason: Ring hung, action: reset [72648.993767] ------------[ cut here ]------------ [72648.993841] WARNING: CPU: 1 PID: 19943 at drivers/gpu/drm/i915/intel_display.c:3291 intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915]() [72648.993845] WARN_ON(ret) [cut same trace] [72648.994601] ---[ end trace 2293beecb2f21024 ]--- [72649.033249] drm/i915: Resetting chip after gpu hang [72770.000179] [drm] stuck on render ring [72770.004814] [drm] GPU HANG: ecode 3:0:0x0263ff38, in Xorg [1581], reason: Ring hung, action: reset [72770.004982] ------------[ cut here ]------------ [72770.005076] WARNING: CPU: 1 PID: 19943 at drivers/gpu/drm/i915/intel_display.c:3291 intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915]() [72770.005083] WARN_ON(ret) [cut same trace] [72889.994144] ---[ end trace 2293beecb2f21026 ]--- [72890.031126] drm/i915: Resetting chip after gpu hang [73129.976187] [drm] stuck on render ring [73129.980704] [drm] GPU HANG: ecode 3:0:0x0263ff38, in Xorg [1581], reason: Ring hung, action: reset [73129.980852] ------------[ cut here ]------------ [73129.980927] WARNING: CPU: 3 PID: 20087 at drivers/gpu/drm/i915/intel_display.c:3291 intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915]() [73129.980932] WARN_ON(ret) [cut same trace] [73129.981690] ---[ end trace 2293beecb2f21027 ]--- [73130.018470] drm/i915: Resetting chip after gpu hang [73189.976199] [drm] stuck on render ring [73189.981577] [drm] GPU HANG: ecode 3:0:0x0263ff38, in Xorg [1581], reason: Ring hung, action: reset [73189.981797] ------------[ cut here ]------------ [73189.981943] WARNING: CPU: 1 PID: 20137 at drivers/gpu/drm/i915/intel_display.c:3291 intel_crtc_wait_for_pending_flips+0x15e/0x200 [i915]() [73189.981955] WARN_ON(ret) [cut same trace] [73189.983551] ---[ end trace 2293beecb2f21028 ]--- [73190.021730] drm/i915: Resetting chip after gpu hang
Created attachment 117323 [details] second hung drm/card0/error log
6 days on 4.3-rc1 don't show any GPU hangs. So close this ticket.
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.