From e78b3197dbf73fc0695dd019e388576d0a551830 Mon Sep 17 00:00:00 2001 From: YuBiao Wang Date: Thu, 5 Aug 2021 10:32:40 +0800 Subject: drm/amd/amdgpu: skip locking delayed work if not initialized. When init failed in early init stage, amdgpu_object has not been initialized, so hasn't the ttm delayed queue functions. Signed-off-by: YuBiao Wang Reviewed-by: Emily.Deng Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_device.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index d7cc45e68dbd..65ccd1144863 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -3826,7 +3826,8 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) { dev_info(adev->dev, "amdgpu: finishing device.\n"); flush_delayed_work(&adev->delayed_init_work); - ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); + if (adev->mman.initialized) + ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); adev->shutdown = true; /* make sure IB test finished before entering exclusive mode -- cgit From c530b02f39850a639b72d01ebbf7e5d745c60831 Mon Sep 17 00:00:00 2001 From: Jack Zhang Date: Wed, 12 May 2021 15:06:35 +0800 Subject: drm/amd/amdgpu embed hw_fence into amdgpu_job Why: Previously hw fence is alloced separately with job. It caused historical lifetime issues and corner cases. The ideal situation is to take fence to manage both job and fence's lifetime, and simplify the design of gpu-scheduler. How: We propose to embed hw_fence into amdgpu_job. 1. We cover the normal job submission by this method. 2. For ib_test, and submit without a parent job keep the legacy way to create a hw fence separately. v2: use AMDGPU_FENCE_FLAG_EMBED_IN_JOB_BIT to show that the fence is embedded in a job. v3: remove redundant variable ring in amdgpu_job v4: add tdr sequence support for this feature. Add a job_run_counter to indicate whether this job is a resubmit job. v5 add missing handling in amdgpu_fence_enable_signaling Signed-off-by: Jingwen Chen Signed-off-by: Jack Zhang Reviewed-by: Andrey Grodzovsky Reviewed by: Monk Liu Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_device.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 65ccd1144863..395779ee44d0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -4449,7 +4449,7 @@ int amdgpu_device_mode1_reset(struct amdgpu_device *adev) int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, struct amdgpu_reset_context *reset_context) { - int i, r = 0; + int i, j, r = 0; struct amdgpu_job *job = NULL; bool need_full_reset = test_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags); @@ -4473,6 +4473,17 @@ int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, if (!ring || !ring->sched.thread) continue; + /*clear job fence from fence drv to avoid force_completion + *leave NULL and vm flush fence in fence drv */ + for (j = 0; j <= ring->fence_drv.num_fences_mask; j++) { + struct dma_fence *old, **ptr; + + ptr = &ring->fence_drv.fences[j]; + old = rcu_dereference_protected(*ptr, 1); + if (old && test_bit(AMDGPU_FENCE_FLAG_EMBED_IN_JOB_BIT, &old->flags)) { + RCU_INIT_POINTER(*ptr, NULL); + } + } /* after all hw jobs are reset, hw fence is meaningless, so force_completion */ amdgpu_fence_driver_force_completion(ring); } -- cgit From 691191a2f458e0176414cb5b3993b0c018cdc58c Mon Sep 17 00:00:00 2001 From: YuBiao Wang Date: Tue, 17 Aug 2021 17:36:33 +0800 Subject: drm/amd/amdgpu:flush ttm delayed work before cancel_sync [Why] In some cases when we unload driver, warning call trace will show up in vram_mgr_fini which claims that LRU is not empty, caused by the ttm bo inside delay deleted queue. [How] We should flush delayed work to make sure the delay deleting is done. Signed-off-by: YuBiao Wang Reviewed-by: Andrey Grodzovsky Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_device.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 395779ee44d0..41cc00e489ac 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -3826,8 +3826,10 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) { dev_info(adev->dev, "amdgpu: finishing device.\n"); flush_delayed_work(&adev->delayed_init_work); - if (adev->mman.initialized) + if (adev->mman.initialized) { + flush_delayed_work(&adev->mman.bdev.wq); ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); + } adev->shutdown = true; /* make sure IB test finished before entering exclusive mode -- cgit From 90a9266269eb9f71af1f323c33e1dca53527bd22 Mon Sep 17 00:00:00 2001 From: Michel Dänzer Date: Tue, 17 Aug 2021 10:23:25 +0200 Subject: drm/amdgpu: Cancel delayed work when GFXOFF is disabled MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit schedule_delayed_work does not push back the work if it was already scheduled before, so amdgpu_device_delay_enable_gfx_off ran ~100 ms after the first time GFXOFF was disabled and re-enabled, even if GFXOFF was disabled and re-enabled again during those 100 ms. This resulted in frame drops / stutter with the upcoming mutter 41 release on Navi 14, due to constantly enabling GFXOFF in the HW and disabling it again (for getting the GPU clock counter). To fix this, call cancel_delayed_work_sync when the disable count transitions from 0 to 1, and only schedule the delayed work on the reverse transition, not if the disable count was already 0. This makes sure the delayed work doesn't run at unexpected times, and allows it to be lock-free. v2: * Use cancel_delayed_work_sync & mutex_trylock instead of mod_delayed_work. v3: * Make amdgpu_device_delay_enable_gfx_off lock-free (Christian König) v4: * Fix race condition between amdgpu_gfx_off_ctrl incrementing adev->gfx.gfx_off_req_count and amdgpu_device_delay_enable_gfx_off checking for it to be 0 (Evan Quan) Cc: stable@vger.kernel.org Reviewed-by: Evan Quan Reviewed-by: Lijo Lazar # v3 Acked-by: Christian König # v3 Signed-off-by: Michel Dänzer Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_device.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 41cc00e489ac..41c6b3aacd37 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -2829,12 +2829,11 @@ static void amdgpu_device_delay_enable_gfx_off(struct work_struct *work) struct amdgpu_device *adev = container_of(work, struct amdgpu_device, gfx.gfx_off_delay_work.work); - mutex_lock(&adev->gfx.gfx_off_mutex); - if (!adev->gfx.gfx_off_state && !adev->gfx.gfx_off_req_count) { - if (!amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, true)) - adev->gfx.gfx_off_state = true; - } - mutex_unlock(&adev->gfx.gfx_off_mutex); + WARN_ON_ONCE(adev->gfx.gfx_off_state); + WARN_ON_ONCE(adev->gfx.gfx_off_req_count); + + if (!amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, true)) + adev->gfx.gfx_off_state = true; } /** -- cgit