tree: https://gitlab.freedesktop.org/agd5f/linux.git amd-staging-drm-next head: 931794a62f1be00cb2bda1d67798a887c9cd90ba commit: c8a921d49443025e10794342d4433b3f29616409 [2019/2047] drm/amd/amdgpu implement tdr advanced mode config: i386-randconfig-a004-20210315 (attached as .config) compiler: gcc-9 (Debian 9.3.0-22) 9.3.0 reproduce (this is a W=1 build): git remote add agd5f https://gitlab.freedesktop.org/agd5f/linux.git git fetch --no-tags agd5f amd-staging-drm-next git checkout c8a921d49443025e10794342d4433b3f29616409 # save the attached .config to linux build tree make W=1 ARCH=i386 If you fix the issue, kindly add following tag as appropriate Reported-by: kernel test robot <lkp(a)intel.com&gt; All warnings (new ones prefixed by >>): 4677 | void amdgpu_device_recheck_guilty_jobs(struct amdgpu_device *adev, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/amd/amdgpu/amdgpu_device.c: In function 'amdgpu_device_recheck_guilty_jobs': drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:4731:26: error: 'struct drm_gpu_scheduler' has no member named 'num_jobs' 4731 | atomic_dec(&ring->sched.num_jobs); | ^ drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:4738:23: error: 'struct drm_sched_job' has no member named 'node' 4738 | list_del_init(&s_job->node); | ^~ vim +/amdgpu_device_recheck_guilty_jobs +4677 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c 4676 4678 struct amdgpu_hive_info *hive, 4679 struct list_head *device_list_handle, 4680 bool *need_full_reset) 4681 { 4682 int i, r = 0; 4683 4684 for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { 4685 struct amdgpu_ring *ring = adev->rings[i]; 4686 int ret = 0; 4687 struct drm_sched_job *s_job; 4688 4689 if (!ring || !ring->sched.thread) 4690 continue; 4691 4692 s_job = list_first_entry_or_null(&ring->sched.pending_list, 4693 struct drm_sched_job, list); 4694 if (s_job == NULL) 4695 continue; 4696 4697 /* clear job's guilty and depend the folowing step to decide the real one */ 4698 drm_sched_reset_karma(s_job); 4699 drm_sched_resubmit_jobs_ext(&ring->sched, 1); 4700 4701 ret = dma_fence_wait_timeout(s_job->s_fence->parent, false, ring->sched.timeout); 4702 if (ret == 0) { /* timeout */ 4703 DRM_ERROR("Found the real bad job! ring:%s, job_id:%llx\n", 4704 ring->sched.name, s_job->id); 4705 4706 /* set guilty */ 4707 drm_sched_increase_karma(s_job); 4708 retry: 4709 /* do hw reset */ 4710 if (amdgpu_sriov_vf(adev)) { 4711 amdgpu_virt_fini_data_exchange(adev); 4712 r = amdgpu_device_reset_sriov(adev, false); 4713 if (r) 4714 adev->asic_reset_res = r; 4715 } else { 4716 r = amdgpu_do_asic_reset(hive, device_list_handle, 4717 need_full_reset, false); 4718 if (r && r == -EAGAIN) 4719 goto retry; 4720 } 4721 4722 /* 4723 * add reset counter so that the following 4724 * resubmitted job could flush vmid 4725 */ 4726 atomic_inc(&adev->gpu_reset_counter); 4727 continue; 4728 } 4729 4730 /* got the hw fence, signal finished fence */ 4731 atomic_dec(&ring->sched.num_jobs); 4732 dma_fence_get(&s_job->s_fence->finished); 4733 dma_fence_signal(&s_job->s_fence->finished); 4734 dma_fence_put(&s_job->s_fence->finished); 4735 4736 /* remove node from list and free the job */ 4737 spin_lock(&ring->sched.job_list_lock); 4738 list_del_init(&s_job->node); 4739 spin_unlock(&ring->sched.job_list_lock); 4740 ring->sched.ops->free_job(s_job); 4741 } 4742 } 4743 --- 0-DAY CI Kernel Test Service, Intel Corporation https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org