drm fixes for 4.8-rc5

-----BEGIN PGP SIGNATURE-----
 Version: GnuPG v1
 
 iQIcBAABAgAGBQJXyRRiAAoJEAx081l5xIa+1T0QAI5Al5CF6pEaQkiDywJkQV9f
 O20m97uknFOikFk8XlwSX1V+3aY5emNmJ/bcD4iKweptmeoZzLmuU0QPkaZNb5jq
 9zyjatJcjwCFO+fHifrL3jfZjwq5Hn9L9RD+aDfKZeMJrD5ds0BIqxW00ensxOEz
 w7oT85b7OHhWThvDkXbTcAVCcO7Aj0lpqxtj10ejbI+yzfmju2YCKKpWTusMskub
 9xTcCvsKD1GylRyATN/sA/fbEN820lP6ySPU++zge3ObXmPAt95BeIDcLaFd0WA3
 bSU6ms30Fg5NuSpE9rOkcbsLjAvGQAbVrGzeMgTY+7SkmreJH21JmjKeKois+bb6
 pQc1eePeDxr0sitj5ItS9DE3Ar/KH7aY5WfsK3iU6+TOzimceo6GeYzz/zoeX8Lc
 NWEX9K7WmvIH4d4/r7D9j43Q23GOPftT1X07EEy+GOoLWIFZZczwfTp3DCqVDwhF
 wCedbAhidn8NDTd0eWck6qS1XCtUMH6u4gC933+Kub4Q3eS/9YRIlizRe3/qzwGw
 WcNniZ4ZCm6/6BWftP+c/lTaMH33ui7I6ZIhB0BG1K3SSPik/fT4W2+vGElXkFdb
 nYapYej+4Mpyq7dSdp4SRekt2ZHnXDdYVuDmMld74oy5O5G7jHHG8QimaveHITrz
 cKxoQTcjA8rPPKex1EnU
 =QKse
 -----END PGP SIGNATURE-----

Merge tag 'drm-fixes-for-4.8-rc5' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 "Contains fixes for imx, amdgpu, vc4, msm and one nouveau ACPI fix"

* tag 'drm-fixes-for-4.8-rc5' of git://people.freedesktop.org/~airlied/linux:
  drm/amdgpu: record error code when ring test failed
  drm/amd/amdgpu: compute ring test fail during S4 on CI
  drm/amd/amdgpu: sdma resume fail during S4 on CI
  drm/nouveau/acpi: use DSM if bridge does not support D3cold
  drm/imx: fix crtc vblank state regression
  drm/imx: Add active plane reconfiguration support
  drm/msm: protect against faults from copy_from_user() in submit ioctl
  drm/msm: fix use of copy_from_user() while holding spinlock
  drm/vc4: Fix oops when userspace hands in a bad BO.
  drm/vc4: Fix overflow mem unreferencing when the binner runs dry.
  drm/vc4: Free hang state before destroying BO cache.
  drm/vc4: Fix handling of a pm_runtime_get_sync() success case.
  drm/vc4: Use drm_malloc_ab to fix large rendering jobs.
  drm/vc4: Use drm_free_large() on handles to match its allocation.

drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c

@@ -280,7 +280,7 @@ void amdgpu_ib_pool_fini(struct amdgpu_device *adev)
 int amdgpu_ib_ring_tests(struct amdgpu_device *adev)
 {
 	unsigned i;
-	int r;
+	int r, ret = 0;
 
 	for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 		struct amdgpu_ring *ring = adev->rings[i];
@@ -301,10 +301,11 @@ int amdgpu_ib_ring_tests(struct amdgpu_device *adev)
 			} else {
 				/* still not good, but we can live with it */
 				DRM_ERROR("amdgpu: failed testing IB on ring %d (%d).\n", i, r);
+				ret = r;
 			}
 		}
 	}
-	return 0;
+	return ret;
 }
 
 /*

drivers/gpu/drm/amd/amdgpu/cik_sdma.c

@@ -52,6 +52,7 @@ static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev);
 static void cik_sdma_set_irq_funcs(struct amdgpu_device *adev);
 static void cik_sdma_set_buffer_funcs(struct amdgpu_device *adev);
 static void cik_sdma_set_vm_pte_funcs(struct amdgpu_device *adev);
+static int cik_sdma_soft_reset(void *handle);
 
 MODULE_FIRMWARE("radeon/bonaire_sdma.bin");
 MODULE_FIRMWARE("radeon/bonaire_sdma1.bin");
@@ -1037,6 +1038,8 @@ static int cik_sdma_resume(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	cik_sdma_soft_reset(handle);
+
 	return cik_sdma_hw_init(adev);
 }
 

drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c

@@ -2755,8 +2755,7 @@ static int gfx_v7_0_cp_compute_resume(struct amdgpu_device *adev)
 	u64 wb_gpu_addr;
 	u32 *buf;
 	struct bonaire_mqd *mqd;
-
-	gfx_v7_0_cp_compute_enable(adev, true);
+	struct amdgpu_ring *ring;
 
 	/* fix up chicken bits */
 	tmp = RREG32(mmCP_CPF_DEBUG);
@@ -2791,7 +2790,7 @@ static int gfx_v7_0_cp_compute_resume(struct amdgpu_device *adev)
 
 	/* init the queues.  Just two for now. */
 	for (i = 0; i < adev->gfx.num_compute_rings; i++) {
-		struct amdgpu_ring *ring = &adev->gfx.compute_ring[i];
+		ring = &adev->gfx.compute_ring[i];
 
 		if (ring->mqd_obj == NULL) {
 			r = amdgpu_bo_create(adev,
@@ -2970,6 +2969,13 @@ static int gfx_v7_0_cp_compute_resume(struct amdgpu_device *adev)
 		amdgpu_bo_unreserve(ring->mqd_obj);
 
 		ring->ready = true;
+	}
+
+	gfx_v7_0_cp_compute_enable(adev, true);
+
+	for (i = 0; i < adev->gfx.num_compute_rings; i++) {
+		ring = &adev->gfx.compute_ring[i];
+
 		r = amdgpu_ring_test_ring(ring);
 		if (r)
 			ring->ready = false;

drivers/gpu/drm/imx/imx-drm-core.c

@@ -171,10 +171,34 @@ static void imx_drm_output_poll_changed(struct drm_device *drm)
 	drm_fbdev_cma_hotplug_event(imxdrm->fbhelper);
 }
 
+static int imx_drm_atomic_check(struct drm_device *dev,
+				struct drm_atomic_state *state)
+{
+	int ret;
+
+	ret = drm_atomic_helper_check_modeset(dev, state);
+	if (ret)
+		return ret;
+
+	ret = drm_atomic_helper_check_planes(dev, state);
+	if (ret)
+		return ret;
+
+	/*
+	 * Check modeset again in case crtc_state->mode_changed is
+	 * updated in plane's ->atomic_check callback.
+	 */
+	ret = drm_atomic_helper_check_modeset(dev, state);
+	if (ret)
+		return ret;
+
+	return ret;
+}
+
 static const struct drm_mode_config_funcs imx_drm_mode_config_funcs = {
 	.fb_create = drm_fb_cma_create,
 	.output_poll_changed = imx_drm_output_poll_changed,
-	.atomic_check = drm_atomic_helper_check,
+	.atomic_check = imx_drm_atomic_check,
 	.atomic_commit = drm_atomic_helper_commit,
 };
 

drivers/gpu/drm/imx/ipuv3-crtc.c

@@ -76,6 +76,8 @@ static void ipu_crtc_disable(struct drm_crtc *crtc)
 		crtc->state->event = NULL;
 	}
 	spin_unlock_irq(&crtc->dev->event_lock);
+
+	drm_crtc_vblank_off(crtc);
 }
 
 static void imx_drm_crtc_reset(struct drm_crtc *crtc)
@@ -175,6 +177,8 @@ static int ipu_crtc_atomic_check(struct drm_crtc *crtc,
 static void ipu_crtc_atomic_begin(struct drm_crtc *crtc,
 				  struct drm_crtc_state *old_crtc_state)
 {
+	drm_crtc_vblank_on(crtc);
+
 	spin_lock_irq(&crtc->dev->event_lock);
 	if (crtc->state->event) {
 		WARN_ON(drm_crtc_vblank_get(crtc));

drivers/gpu/drm/imx/ipuv3-plane.c

@@ -319,13 +319,14 @@ static int ipu_plane_atomic_check(struct drm_plane *plane,
 		return -EINVAL;
 
 	/*
-	 * since we cannot touch active IDMAC channels, we do not support
-	 * resizing the enabled plane or changing its format
+	 * We support resizing active plane or changing its format by
+	 * forcing CRTC mode change and disabling-enabling plane in plane's
+	 * ->atomic_update callback.
 	 */
 	if (old_fb && (state->src_w != old_state->src_w ||
 			      state->src_h != old_state->src_h ||
 			      fb->pixel_format != old_fb->pixel_format))
-		return -EINVAL;
+		crtc_state->mode_changed = true;
 
 	eba = drm_plane_state_to_eba(state);
 
@@ -336,7 +337,7 @@ static int ipu_plane_atomic_check(struct drm_plane *plane,
 		return -EINVAL;
 
 	if (old_fb && fb->pitches[0] != old_fb->pitches[0])
-		return -EINVAL;
+		crtc_state->mode_changed = true;
 
 	switch (fb->pixel_format) {
 	case DRM_FORMAT_YUV420:
@@ -372,7 +373,7 @@ static int ipu_plane_atomic_check(struct drm_plane *plane,
 			return -EINVAL;
 
 		if (old_fb && old_fb->pitches[1] != fb->pitches[1])
-			return -EINVAL;
+			crtc_state->mode_changed = true;
 	}
 
 	return 0;
@@ -392,8 +393,14 @@ static void ipu_plane_atomic_update(struct drm_plane *plane,
 	enum ipu_color_space ics;
 
 	if (old_state->fb) {
-		ipu_plane_atomic_set_base(ipu_plane, old_state);
-		return;
+		struct drm_crtc_state *crtc_state = state->crtc->state;
+
+		if (!crtc_state->mode_changed) {
+			ipu_plane_atomic_set_base(ipu_plane, old_state);
+			return;
+		}
+
+		ipu_disable_plane(plane);
 	}
 
 	switch (ipu_plane->dp_flow) {

drivers/gpu/drm/msm/msm_drv.h

@@ -157,6 +157,12 @@ struct msm_drm_private {
 	struct shrinker shrinker;
 
 	struct msm_vblank_ctrl vblank_ctrl;
+
+	/* task holding struct_mutex.. currently only used in submit path
+	 * to detect and reject faults from copy_from_user() for submit
+	 * ioctl.
+	 */
+	struct task_struct *struct_mutex_task;
 };
 
 struct msm_format {

drivers/gpu/drm/msm/msm_gem.c

@@ -196,11 +196,20 @@ int msm_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct drm_gem_object *obj = vma->vm_private_data;
 	struct drm_device *dev = obj->dev;
+	struct msm_drm_private *priv = dev->dev_private;
 	struct page **pages;
 	unsigned long pfn;
 	pgoff_t pgoff;
 	int ret;
 
+	/* This should only happen if userspace tries to pass a mmap'd
+	 * but unfaulted gem bo vaddr into submit ioctl, triggering
+	 * a page fault while struct_mutex is already held.  This is
+	 * not a valid use-case so just bail.
+	 */
+	if (priv->struct_mutex_task == current)
+		return VM_FAULT_SIGBUS;
+
 	/* Make sure we don't parallel update on a fault, nor move or remove
 	 * something from beneath our feet
 	 */

drivers/gpu/drm/msm/msm_gem_submit.c

@@ -64,6 +64,14 @@ void msm_gem_submit_free(struct msm_gem_submit *submit)
 	kfree(submit);
 }
 
+static inline unsigned long __must_check
+copy_from_user_inatomic(void *to, const void __user *from, unsigned long n)
+{
+	if (access_ok(VERIFY_READ, from, n))
+		return __copy_from_user_inatomic(to, from, n);
+	return -EFAULT;
+}
+
 static int submit_lookup_objects(struct msm_gem_submit *submit,
 		struct drm_msm_gem_submit *args, struct drm_file *file)
 {
@@ -71,6 +79,7 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
 	int ret = 0;
 
 	spin_lock(&file->table_lock);
+	pagefault_disable();
 
 	for (i = 0; i < args->nr_bos; i++) {
 		struct drm_msm_gem_submit_bo submit_bo;
@@ -84,10 +93,15 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
 		 */
 		submit->bos[i].flags = 0;
 
-		ret = copy_from_user(&submit_bo, userptr, sizeof(submit_bo));
-		if (ret) {
-			ret = -EFAULT;
-			goto out_unlock;
+		ret = copy_from_user_inatomic(&submit_bo, userptr, sizeof(submit_bo));
+		if (unlikely(ret)) {
+			pagefault_enable();
+			spin_unlock(&file->table_lock);
+			ret = copy_from_user(&submit_bo, userptr, sizeof(submit_bo));
+			if (ret)
+				goto out;
+			spin_lock(&file->table_lock);
+			pagefault_disable();
 		}
 
 		if (submit_bo.flags & ~MSM_SUBMIT_BO_FLAGS) {
@@ -127,9 +141,12 @@ static int submit_lookup_objects(struct msm_gem_submit *submit,
 	}
 
 out_unlock:
-	submit->nr_bos = i;
+	pagefault_enable();
 	spin_unlock(&file->table_lock);
 
+out:
+	submit->nr_bos = i;
+
 	return ret;
 }
 
@@ -377,6 +394,8 @@ int msm_ioctl_gem_submit(struct drm_device *dev, void *data,
 	if (ret)
 		return ret;
 
+	priv->struct_mutex_task = current;
+
 	submit = submit_create(dev, gpu, args->nr_bos, args->nr_cmds);
 	if (!submit) {
 		ret = -ENOMEM;
@@ -468,6 +487,7 @@ out:
 	if (ret)
 		msm_gem_submit_free(submit);
 out_unlock:
+	priv->struct_mutex_task = NULL;
 	mutex_unlock(&dev->struct_mutex);
 	return ret;
 }

drivers/gpu/drm/nouveau/nouveau_acpi.c

@@ -225,6 +225,17 @@ static bool nouveau_pr3_present(struct pci_dev *pdev)
 	if (!parent_pdev)
 		return false;
 
+	if (!parent_pdev->bridge_d3) {
+		/*
+		 * Parent PCI bridge is currently not power managed.
+		 * Since userspace can change these afterwards to be on
+		 * the safe side we stick with _DSM and prevent usage of
+		 * _PR3 from the bridge.
+		 */
+		pci_d3cold_disable(pdev);
+		return false;
+	}
+
 	parent_adev = ACPI_COMPANION(&parent_pdev->dev);
 	if (!parent_adev)
 		return false;

drivers/gpu/drm/vc4/vc4_drv.c

@@ -57,21 +57,21 @@ static int vc4_get_param_ioctl(struct drm_device *dev, void *data,
 	switch (args->param) {
 	case DRM_VC4_PARAM_V3D_IDENT0:
 		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
-		if (ret)
+		if (ret < 0)
 			return ret;
 		args->value = V3D_READ(V3D_IDENT0);
 		pm_runtime_put(&vc4->v3d->pdev->dev);
 		break;
 	case DRM_VC4_PARAM_V3D_IDENT1:
 		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
-		if (ret)
+		if (ret < 0)
 			return ret;
 		args->value = V3D_READ(V3D_IDENT1);
 		pm_runtime_put(&vc4->v3d->pdev->dev);
 		break;
 	case DRM_VC4_PARAM_V3D_IDENT2:
 		ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev);
-		if (ret)
+		if (ret < 0)
 			return ret;
 		args->value = V3D_READ(V3D_IDENT2);
 		pm_runtime_put(&vc4->v3d->pdev->dev);

drivers/gpu/drm/vc4/vc4_drv.h

@@ -321,6 +321,15 @@ vc4_first_render_job(struct vc4_dev *vc4)
 				struct vc4_exec_info, head);
 }
 
+static inline struct vc4_exec_info *
+vc4_last_render_job(struct vc4_dev *vc4)
+{
+	if (list_empty(&vc4->render_job_list))
+		return NULL;
+	return list_last_entry(&vc4->render_job_list,
+			       struct vc4_exec_info, head);
+}
+
 /**
  * struct vc4_texture_sample_info - saves the offsets into the UBO for texture
  * setup parameters.

drivers/gpu/drm/vc4/vc4_gem.c

@@ -534,8 +534,8 @@ vc4_cl_lookup_bos(struct drm_device *dev,
 		return -EINVAL;
 	}
 
-	exec->bo = kcalloc(exec->bo_count, sizeof(struct drm_gem_cma_object *),
-			   GFP_KERNEL);
+	exec->bo = drm_calloc_large(exec->bo_count,
+				    sizeof(struct drm_gem_cma_object *));
 	if (!exec->bo) {
 		DRM_ERROR("Failed to allocate validated BO pointers\n");
 		return -ENOMEM;
@@ -572,8 +572,8 @@ vc4_cl_lookup_bos(struct drm_device *dev,
 	spin_unlock(&file_priv->table_lock);
 
 fail:
-	kfree(handles);
-	return 0;
+	drm_free_large(handles);
+	return ret;
 }
 
 static int
@@ -608,7 +608,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec)
 	 * read the contents back for validation, and I think the
 	 * bo->vaddr is uncached access.
 	 */
-	temp = kmalloc(temp_size, GFP_KERNEL);
+	temp = drm_malloc_ab(temp_size, 1);
 	if (!temp) {
 		DRM_ERROR("Failed to allocate storage for copying "
 			  "in bin/render CLs.\n");
@@ -675,7 +675,7 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec)
 	ret = vc4_validate_shader_recs(dev, exec);
 
 fail:
-	kfree(temp);
+	drm_free_large(temp);
 	return ret;
 }
 
@@ -688,7 +688,7 @@ vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec)
 	if (exec->bo) {
 		for (i = 0; i < exec->bo_count; i++)
 			drm_gem_object_unreference_unlocked(&exec->bo[i]->base);
-		kfree(exec->bo);
+		drm_free_large(exec->bo);
 	}
 
 	while (!list_empty(&exec->unref_list)) {
@@ -942,8 +942,8 @@ vc4_gem_destroy(struct drm_device *dev)
 		vc4->overflow_mem = NULL;
 	}
 
-	vc4_bo_cache_destroy(dev);
-
 	if (vc4->hang_state)
 		vc4_free_hang_state(dev, vc4->hang_state);
+
+	vc4_bo_cache_destroy(dev);
 }

drivers/gpu/drm/vc4/vc4_irq.c

@@ -83,8 +83,10 @@ vc4_overflow_mem_work(struct work_struct *work)
 
 		spin_lock_irqsave(&vc4->job_lock, irqflags);
 		current_exec = vc4_first_bin_job(vc4);
+		if (!current_exec)
+			current_exec = vc4_last_render_job(vc4);
 		if (current_exec) {
-			vc4->overflow_mem->seqno = vc4->finished_seqno + 1;
+			vc4->overflow_mem->seqno = current_exec->seqno;
 			list_add_tail(&vc4->overflow_mem->unref_head,
 				      &current_exec->unref_list);
 			vc4->overflow_mem = NULL;