<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - amdgpu [RX Vega 64] system freeze while gaming" href="https://bugs.freedesktop.org/show_bug.cgi?id=109955#c105">Comment # 105</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - amdgpu [RX Vega 64] system freeze while gaming" href="https://bugs.freedesktop.org/show_bug.cgi?id=109955">bug 109955</a> from <a class="email" href="mailto:rodamorris@gmail.com" title="Rodney A Morris <rodamorris@gmail.com>"> Rodney A Morris</a> <pre>Created <a href="attachment.cgi?id=145462" name="attach_145462" title="dmesg from Stellaris crash 2019-09-20">attachment 145462</a> <a href="attachment.cgi?id=145462&action=edit" title="dmesg from Stellaris crash 2019-09-20">[details]</a> dmesg from Stellaris crash 2019-09-20 I had another lockup on Friday while playing Stellaris again. This time I had the debug kernel running and the mesa debug packages installed. I do not plan to post dmesg and journalctl dumps for future crashes unless the logs indicate a new problem, or I can obtain more information than I previously provided. Like the crash I reported for Hearts of Iron IV, this Stellaris crash seems to be caused by a circular lock dependency. If someone believes my problems are caused by faulty hardware, please let me know. As an FYI, this problem does not seem to manifest under Windows 10, playing the same game. Card: Sapphire Radeon Vega 64 OS Info: /:-------------:\ :-------------------:: -------------------------------- :-----------/shhOHbmp---:\ OS: Fedora release 30 (Thirty) x86_64 /-----------omMMMNNNMMD ---: Kernel: 5.2.15-200.fc30.x86_64 :-----------sMMMMNMNMP. ---: Uptime: 1 day, 22 hours, 37 mins :-----------:MMMdP------- ---\ Packages: 2211 (rpm), 30 (flatpak) ,------------:MMMd-------- ---: Shell: bash 5.0.7 :------------:MMMd------- .---: Resolution: 2560x1440 :---- oNMMMMMMMMMNho .----: DE: GNOME 3.32.2 :-- .+shhhMMMmhhy++ .------/ WM: Mutter :- -------:MMMd--------------: WM Theme: Adwaita :- --------/MMMd-------------; Theme: Adapta-Nokto-Eta [GTK2/3] :- ------/hMMMy------------: Icons: Adwaita [GTK2/3] :-- :dMNdhhdNMMNo------------; Terminal: tilix :---:sdNMMMMNds:------------: CPU: Intel i7-6850K (12) @ 4.000GHz :------:://:-------------:: GPU: AMD ATI Radeon RX Vega 56/64 :---------------------:// Memory: 3097MiB / 32084MiB Mesa info: OpenGL version string: 4.5 (Compatibility Profile) Mesa 19.1.6 Game being played: Stellaris through steam for Linux Native or Wine: Native Crash Type: Screen goes black suddenly while music continues plays for less than a minute; music begins to loop; and computer reboots. Full dmesg attached. Pertinent part of dmesg with debug kernel: [ 2383.732727] perf: interrupt took too long (2502 > 2500), lowering kernel.perf_event_max_sample_rate to 79000 [ 2923.530873] [drm:amdgpu_dm_atomic_commit_tail [amdgpu]] *ERROR* Waiting for fences timed out or interrupted! [ 2928.651952] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring page1 timeout, signaled seq=51954680, emitted seq=51954682 [ 2928.652090] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process pid 0 thread pid 0 [ 2928.652098] amdgpu 0000:06:00.0: GPU reset begin! [ 2928.661852] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring gfx timeout, signaled seq=734676, emitted seq=734677 [ 2928.661898] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process stellaris pid 5395 thread stellaris:cs0 pid 5397 [ 2928.661901] amdgpu 0000:06:00.0: GPU reset begin! [ 2928.661997] ====================================================== [ 2928.661999] WARNING: possible circular locking dependency detected [ 2928.662003] 5.2.15-200.fc30.x86_64+debug #1 Not tainted [ 2928.662005] ------------------------------------------------------ [ 2928.662007] kworker/10:2/974 is trying to acquire lock: [ 2928.662010] 00000000d514cf70 (&(&ring->fence_drv.lock)->rlock){-.-.}, at: dma_fence_remove_callback+0x1a/0x60 [ 2928.662021] but task is already holding lock: [ 2928.662023] 00000000e6ce7c0d (&(&sched->job_list_lock)->rlock){-.-.}, at: drm_sched_stop+0x34/0x130 [gpu_sched] [ 2928.662031] which lock already depends on the new lock. [ 2928.662033] the existing dependency chain (in reverse order) is: [ 2928.662035] -> #1 (&(&sched->job_list_lock)->rlock){-.-.}: [ 2928.662044] _raw_spin_lock_irqsave+0x49/0x83 [ 2928.662049] drm_sched_process_job+0x4d/0x180 [gpu_sched] [ 2928.662052] dma_fence_signal+0x111/0x1a0 [ 2928.662128] amdgpu_fence_process+0xa3/0x100 [amdgpu] [ 2928.662223] sdma_v4_0_process_trap_irq+0x8d/0xa0 [amdgpu] [ 2928.662310] amdgpu_irq_dispatch+0xc0/0x250 [amdgpu] [ 2928.662398] amdgpu_ih_process+0x8d/0x110 [amdgpu] [ 2928.662482] amdgpu_irq_handler+0x1b/0x50 [amdgpu] [ 2928.662487] __handle_irq_event_percpu+0x3f/0x290 [ 2928.662491] handle_irq_event_percpu+0x31/0x80 [ 2928.662495] handle_irq_event+0x34/0x51 [ 2928.662498] handle_edge_irq+0x83/0x1a0 [ 2928.662502] handle_irq+0x1c/0x30 [ 2928.662507] do_IRQ+0x61/0x120 [ 2928.662511] ret_from_intr+0x0/0x22 [ 2928.662517] cpuidle_enter_state+0xc9/0x450 [ 2928.662519] cpuidle_enter+0x29/0x40 [ 2928.662524] do_idle+0x1ec/0x280 [ 2928.662528] cpu_startup_entry+0x19/0x20 [ 2928.662531] start_secondary+0x189/0x1e0 [ 2928.662537] secondary_startup_64+0xa4/0xb0 [ 2928.662539] -> #0 (&(&ring->fence_drv.lock)->rlock){-.-.}: [ 2928.662548] lock_acquire+0xa2/0x1b0 [ 2928.662551] _raw_spin_lock_irqsave+0x49/0x83 [ 2928.662555] dma_fence_remove_callback+0x1a/0x60 [ 2928.662560] drm_sched_stop+0x59/0x130 [gpu_sched] [ 2928.662709] amdgpu_device_pre_asic_reset+0x41/0x20c [amdgpu] [ 2928.662866] amdgpu_device_gpu_recover+0x77/0x788 [amdgpu] [ 2928.663007] amdgpu_job_timedout+0x109/0x130 [amdgpu] [ 2928.663018] drm_sched_job_timedout+0x40/0x70 [gpu_sched] [ 2928.663024] process_one_work+0x272/0x5e0 [ 2928.663029] worker_thread+0x50/0x3b0 [ 2928.663037] kthread+0x108/0x140 [ 2928.663045] ret_from_fork+0x3a/0x50 [ 2928.663048] other info that might help us debug this: [ 2928.663051] Possible unsafe locking scenario: [ 2928.663055] CPU0 CPU1 [ 2928.663059] ---- ---- [ 2928.663062] lock(&(&sched->job_list_lock)->rlock); [ 2928.663068] lock(&(&ring->fence_drv.lock)->rlock); [ 2928.663072] lock(&(&sched->job_list_lock)->rlock); [ 2928.663076] lock(&(&ring->fence_drv.lock)->rlock); [ 2928.663080] *** DEADLOCK *** [ 2928.663085] 5 locks held by kworker/10:2/974: [ 2928.663090] #0: 0000000057c9a435 ((wq_completion)events){+.+.}, at: process_one_work+0x1e9/0x5e0 [ 2928.663100] #1: 00000000aadd5dda ((work_completion)(&(&sched->work_tdr)->work)){+.+.}, at: process_one_work+0x1e9/0x5e0 [ 2928.663108] #2: 0000000007db378b (&adev->lock_reset){+.+.}, at: amdgpu_device_lock_adev+0x17/0x39 [amdgpu] [ 2928.663261] #3: 000000001e0a2926 (&dqm->lock_hidden){+.+.}, at: kgd2kfd_pre_reset+0x30/0x60 [amdgpu] [ 2928.663392] #4: 00000000e6ce7c0d (&(&sched->job_list_lock)->rlock){-.-.}, at: drm_sched_stop+0x34/0x130 [gpu_sched] [ 2928.663403] stack backtrace: [ 2928.663409] CPU: 10 PID: 974 Comm: kworker/10:2 Not tainted 5.2.15-200.fc30.x86_64+debug #1 [ 2928.663413] Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./X99 Taichi, BIOS P1.80 04/06/2018 [ 2928.663423] Workqueue: events drm_sched_job_timedout [gpu_sched] [ 2928.663428] Call Trace: [ 2928.663442] dump_stack+0x85/0xc0 [ 2928.663453] print_circular_bug.cold+0x15c/0x195 [ 2928.663462] __lock_acquire+0x167c/0x1c90 [ 2928.663475] lock_acquire+0xa2/0x1b0 [ 2928.663482] ? dma_fence_remove_callback+0x1a/0x60 [ 2928.663494] _raw_spin_lock_irqsave+0x49/0x83 [ 2928.663499] ? dma_fence_remove_callback+0x1a/0x60 [ 2928.663506] dma_fence_remove_callback+0x1a/0x60 [ 2928.663515] drm_sched_stop+0x59/0x130 [gpu_sched] [ 2928.663663] amdgpu_device_pre_asic_reset+0x41/0x20c [amdgpu] [ 2928.663818] amdgpu_device_gpu_recover+0x77/0x788 [amdgpu] [ 2928.663960] amdgpu_job_timedout+0x109/0x130 [amdgpu] [ 2928.663974] drm_sched_job_timedout+0x40/0x70 [gpu_sched] [ 2928.663981] process_one_work+0x272/0x5e0 [ 2928.663991] worker_thread+0x50/0x3b0 [ 2928.664000] kthread+0x108/0x140 [ 2928.664005] ? process_one_work+0x5e0/0x5e0 [ 2928.664011] ? kthread_park+0x80/0x80 [ 2928.664021] ret_from_fork+0x3a/0x50 [ 2928.681831] pcieport 0000:00:03.0: AER: Uncorrected (Non-Fatal) error received: 0000:00:03.0 [ 2928.681846] pcieport 0000:00:03.0: AER: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, (Requester ID) [ 2928.681851] pcieport 0000:00:03.0: AER: device [8086:6f08] error status/mask=00004000/00000000 [ 2928.681857] pcieport 0000:00:03.0: AER: [14] CmpltTO (First) [ 2928.681963] pcieport 0000:00:03.0: AER: Device recovery failed [ 2933.771664] [drm:drm_atomic_helper_wait_for_flip_done [drm_kms_helper]] *ERROR* [CRTC:47:crtc-0] flip_done timed out [ 2938.890758] [drm:amdgpu_dm_atomic_check [amdgpu]] *ERROR* [CRTC:47:crtc-0] hw_done or flip_done timed out [ 2939.118467] pcieport 0000:00:03.0: AER: Uncorrected (Non-Fatal) error received: 0000:00:03.0 [ 2939.118475] pcieport 0000:00:03.0: AER: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, (Requester ID) [ 2939.118477] pcieport 0000:00:03.0: AER: device [8086:6f08] error status/mask=00004000/00000000 [ 2939.118479] pcieport 0000:00:03.0: AER: [14] CmpltTO (First) [ 2939.118536] pcieport 0000:00:03.0: AER: Device recovery failed [ 2939.141034] pcieport 0000:00:03.0: AER: Multiple Uncorrected (Non-Fatal) error received: 0000:00:03.0 [ 2939.369014] pcieport 0000:00:03.0: AER: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, (Requester ID) [ 2939.369018] pcieport 0000:00:03.0: AER: device [8086:6f08] error status/mask=00004000/00000000 [ 2939.369021] pcieport 0000:00:03.0: AER: [14] CmpltTO (First) [ 2939.369072] pcieport 0000:00:03.0: AER: Device recovery failed [ 2939.369075] pcieport 0000:00:03.0: AER: Multiple Uncorrected (Non-Fatal) error received: 0000:00:03.0 [ 2939.597051] pcieport 0000:00:03.0: AER: PCIe Bus Error: severity=Uncorrected (Non-Fatal), type=Transaction Layer, (Requester ID) [ 2939.597055] pcieport 0000:00:03.0: AER: device [8086:6f08] error status/mask=00004000/00000000 [ 2939.597057] pcieport 0000:00:03.0: AER: [14] CmpltTO (First) [ 2939.597103] pcieport 0000:00:03.0: AER: Device recovery failed [ 2939.597106] pcieport 0000:00:03.0: AER: Multiple Uncorrected (Non-Fatal) error received: 0000:00:03.0 systemd logs: Nothing interesting appears in the logs, not even the information from dmesg. I'm unsure if systemd captured anything from the crash.</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the assignee for the bug.</li> </ul> </body> </html>