Merge tag 'v4.8-rc8' into drm-next

Linux 4.8-rc8

There was a lot of fallout in the imx/amdgpu/i915 drivers, so backmerge
it now to avoid troubles.

* tag 'v4.8-rc8': (1442 commits)
  Linux 4.8-rc8
  fault_in_multipages_readable() throws set-but-unused error
  mm: check VMA flags to avoid invalid PROT_NONE NUMA balancing
  radix tree: fix sibling entry handling in radix_tree_descend()
  radix tree test suite: Test radix_tree_replace_slot() for multiorder entries
  fix memory leaks in tracing_buffers_splice_read()
  tracing: Move mutex to protect against resetting of seq data
  MIPS: Fix delay slot emulation count in debugfs
  MIPS: SMP: Fix possibility of deadlock when bringing CPUs online
  mm: delete unnecessary and unsafe init_tlb_ubc()
  huge tmpfs: fix Committed_AS leak
  shmem: fix tmpfs to handle the huge= option properly
  blk-mq: skip unmapped queues in blk_mq_alloc_request_hctx
  MIPS: Fix pre-r6 emulation FPU initialisation
  arm64: kgdb: handle read-only text / modules
  arm64: Call numa_store_cpu_info() earlier.
  locking/hung_task: Fix typo in CONFIG_DETECT_HUNG_TASK help text
  nvme-rdma: only clear queue flags after successful connect
  i2c: qup: skip qup_i2c_suspend if the device is already runtime suspended
  perf/core: Limit matching exclusive events to one PMU
  ...

drivers/gpu/drm/vc4/vc4_bo.c

@@ -144,7 +144,7 @@ static struct list_head *vc4_get_cache_list_for_size(struct drm_device *dev,
 	return &vc4->bo_cache.size_list[page_index];
 }
 
-void vc4_bo_cache_purge(struct drm_device *dev)
+static void vc4_bo_cache_purge(struct drm_device *dev)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 

drivers/gpu/drm/vc4/vc4_drv.c

@@ -58,21 +58,21 @@ 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

@@ -530,8 +530,8 @@ 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;
@@ -568,8 +568,8 @@ 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
@@ -604,7 +604,7 @@ 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");
@@ -671,7 +671,7 @@ 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;
 }
 
@@ -684,7 +684,7 @@ 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)) {
@@ -938,8 +938,8 @@ 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;

drivers/gpu/drm/vc4/vc4_validate_shaders.c

@@ -309,8 +309,14 @@ validate_uniform_address_write(struct vc4_validated_shader_info *validated_shade
 	 * of uniforms on each side.  However, this scheme is easy to
 	 * validate so it's all we allow for now.
 	 */
-
-	if (QPU_GET_FIELD(inst, QPU_SIG) != QPU_SIG_NONE) {
+	switch (QPU_GET_FIELD(inst, QPU_SIG)) {
+	case QPU_SIG_NONE:
+	case QPU_SIG_SCOREBOARD_UNLOCK:
+	case QPU_SIG_COLOR_LOAD:
+	case QPU_SIG_LOAD_TMU0:
+	case QPU_SIG_LOAD_TMU1:
+		break;
+	default:
 		DRM_ERROR("uniforms address change must be "
 			  "normal math\n");
 		return false;