Merge tag 'drm-fixes-for-v4.15-rc5' of git://people.freedesktop.org/~airlied/linux

Pull drm fixes from Dave Airlie:
 "I've got most of two weeks worth of fixes here due to being on
  holidays last week.

  The main things are:

  - Core:
     * Syncobj fd reference count fix
     * Leasing ioctl misuse fix

   - nouveau regression fixes

   - further amdgpu DC fixes

   - sun4i regression fixes

  I'm not sure I'll see many fixes over next couple of weeks, we'll see
  how we go"

* tag 'drm-fixes-for-v4.15-rc5' of git://people.freedesktop.org/~airlied/linux: (27 commits)
  drm/syncobj: Stop reusing the same struct file for all syncobj -> fd
  drm: move lease init after validation in drm_lease_create
  drm/plane: Make framebuffer refcounting the responsibility of setplane_internal callers
  drm/sun4i: hdmi: Move the mode_valid callback to the encoder
  drm/nouveau: fix obvious memory leak
  drm/i915: Protect DDI port to DPLL map from theoretical race.
  drm/i915/lpe: Remove double-encapsulation of info string
  drm/sun4i: Fix error path handling
  drm/nouveau: use alternate memory type for system-memory buffers with kind != 0
  drm/nouveau: avoid GPU page sizes > PAGE_SIZE for buffer objects in host memory
  drm/nouveau/mmu/gp10b: use correct implementation
  drm/nouveau/pci: do a msi rearm on init
  drm/nouveau/imem/nv50: fix refcount_t warning
  drm/nouveau/bios/dp: support DP Info Table 2.0
  drm/nouveau/fbcon: fix NULL pointer access in nouveau_fbcon_destroy
  drm/amd/display: Fix rehook MST display not light back on
  drm/amd/display: fix missing pixel clock adjustment for dongle
  drm/amd/display: set chroma taps to 1 when not scaling
  drm/amd/display: add pipe locking before front end programing
  drm/sun4i: validate modes for HDMI
  ...

drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c

@@ -2467,7 +2467,7 @@ static int gfx_v9_0_kiq_kcq_enable(struct amdgpu_device *adev)
 				  PACKET3_MAP_QUEUES_PIPE(ring->pipe) |
 				  PACKET3_MAP_QUEUES_ME((ring->me == 1 ? 0 : 1)) |
 				  PACKET3_MAP_QUEUES_QUEUE_TYPE(0) | /*queue_type: normal compute queue */
-				  PACKET3_MAP_QUEUES_ALLOC_FORMAT(1) | /* alloc format: all_on_one_pipe */
+				  PACKET3_MAP_QUEUES_ALLOC_FORMAT(0) | /* alloc format: all_on_one_pipe */
 				  PACKET3_MAP_QUEUES_ENGINE_SEL(0) | /* engine_sel: compute */
 				  PACKET3_MAP_QUEUES_NUM_QUEUES(1)); /* num_queues: must be 1 */
 		amdgpu_ring_write(kiq_ring, PACKET3_MAP_QUEUES_DOORBELL_OFFSET(ring->doorbell_index));

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c

@@ -2336,7 +2336,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 		       const struct dm_connector_state *dm_state)
 {
 	struct drm_display_mode *preferred_mode = NULL;
-	const struct drm_connector *drm_connector;
+	struct drm_connector *drm_connector;
 	struct dc_stream_state *stream = NULL;
 	struct drm_display_mode mode = *drm_mode;
 	bool native_mode_found = false;
@@ -2355,11 +2355,13 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 
 	if (!aconnector->dc_sink) {
 		/*
-		 * Exclude MST from creating fake_sink
-		 * TODO: need to enable MST into fake_sink feature
+		 * Create dc_sink when necessary to MST
+		 * Don't apply fake_sink to MST
 		 */
-		if (aconnector->mst_port)
-			goto stream_create_fail;
+		if (aconnector->mst_port) {
+			dm_dp_mst_dc_sink_create(drm_connector);
+			goto mst_dc_sink_create_done;
+		}
 
 		if (create_fake_sink(aconnector))
 			goto stream_create_fail;
@@ -2410,6 +2412,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 stream_create_fail:
 dm_state_null:
 drm_connector_null:
+mst_dc_sink_create_done:
 	return stream;
 }
 

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h

@@ -189,6 +189,8 @@ struct amdgpu_dm_connector {
 	struct mutex hpd_lock;
 
 	bool fake_enable;
+
+	bool mst_connected;
 };
 
 #define to_amdgpu_dm_connector(x) container_of(x, struct amdgpu_dm_connector, base)

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c

@@ -185,6 +185,42 @@ static int dm_connector_update_modes(struct drm_connector *connector,
 	return ret;
 }
 
+void dm_dp_mst_dc_sink_create(struct drm_connector *connector)
+{
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+	struct edid *edid;
+	struct dc_sink *dc_sink;
+	struct dc_sink_init_data init_params = {
+			.link = aconnector->dc_link,
+			.sink_signal = SIGNAL_TYPE_DISPLAY_PORT_MST };
+
+	edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port);
+
+	if (!edid) {
+		drm_mode_connector_update_edid_property(
+			&aconnector->base,
+			NULL);
+		return;
+	}
+
+	aconnector->edid = edid;
+
+	dc_sink = dc_link_add_remote_sink(
+		aconnector->dc_link,
+		(uint8_t *)aconnector->edid,
+		(aconnector->edid->extensions + 1) * EDID_LENGTH,
+		&init_params);
+
+	dc_sink->priv = aconnector;
+	aconnector->dc_sink = dc_sink;
+
+	amdgpu_dm_add_sink_to_freesync_module(
+			connector, aconnector->edid);
+
+	drm_mode_connector_update_edid_property(
+					&aconnector->base, aconnector->edid);
+}
+
 static int dm_dp_mst_get_modes(struct drm_connector *connector)
 {
 	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
@@ -311,6 +347,7 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 			drm_mode_connector_set_path_property(connector, pathprop);
 
 			drm_connector_list_iter_end(&conn_iter);
+			aconnector->mst_connected = true;
 			return &aconnector->base;
 		}
 	}
@@ -363,6 +400,8 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 	 */
 	amdgpu_dm_connector_funcs_reset(connector);
 
+	aconnector->mst_connected = true;
+
 	DRM_INFO("DM_MST: added connector: %p [id: %d] [master: %p]\n",
 			aconnector, connector->base.id, aconnector->mst_port);
 
@@ -394,6 +433,8 @@ static void dm_dp_destroy_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 	drm_mode_connector_update_edid_property(
 			&aconnector->base,
 			NULL);
+
+	aconnector->mst_connected = false;
 }
 
 static void dm_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr)
@@ -404,10 +445,18 @@ static void dm_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr)
 	drm_kms_helper_hotplug_event(dev);
 }
 
+static void dm_dp_mst_link_status_reset(struct drm_connector *connector)
+{
+	mutex_lock(&connector->dev->mode_config.mutex);
+	drm_mode_connector_set_link_status_property(connector, DRM_MODE_LINK_STATUS_BAD);
+	mutex_unlock(&connector->dev->mode_config.mutex);
+}
+
 static void dm_dp_mst_register_connector(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	struct amdgpu_device *adev = dev->dev_private;
+	struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
 
 	if (adev->mode_info.rfbdev)
 		drm_fb_helper_add_one_connector(&adev->mode_info.rfbdev->helper, connector);
@@ -416,6 +465,8 @@ static void dm_dp_mst_register_connector(struct drm_connector *connector)
 
 	drm_connector_register(connector);
 
+	if (aconnector->mst_connected)
+		dm_dp_mst_link_status_reset(connector);
 }
 
 static const struct drm_dp_mst_topology_cbs dm_mst_cbs = {

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.h

@@ -31,5 +31,6 @@ struct amdgpu_dm_connector;
 
 void amdgpu_dm_initialize_dp_connector(struct amdgpu_display_manager *dm,
 				       struct amdgpu_dm_connector *aconnector);
+void dm_dp_mst_dc_sink_create(struct drm_connector *connector);
 
 #endif

drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c

@@ -900,6 +900,15 @@ bool dcn_validate_bandwidth(
 			v->override_vta_ps[input_idx] = pipe->plane_res.scl_data.taps.v_taps;
 			v->override_hta_pschroma[input_idx] = pipe->plane_res.scl_data.taps.h_taps_c;
 			v->override_vta_pschroma[input_idx] = pipe->plane_res.scl_data.taps.v_taps_c;
+			/*
+			 * Spreadsheet doesn't handle taps_c is one properly,
+			 * need to force Chroma to always be scaled to pass
+			 * bandwidth validation.
+			 */
+			if (v->override_hta_pschroma[input_idx] == 1)
+				v->override_hta_pschroma[input_idx] = 2;
+			if (v->override_vta_pschroma[input_idx] == 1)
+				v->override_vta_pschroma[input_idx] = 2;
 			v->source_scan[input_idx] = (pipe->plane_state->rotation % 2) ? dcn_bw_vert : dcn_bw_hor;
 		}
 		if (v->is_line_buffer_bpp_fixed == dcn_bw_yes)

drivers/gpu/drm/amd/display/dc/core/dc_link.c

@@ -1801,7 +1801,7 @@ static void disable_link(struct dc_link *link, enum signal_type signal)
 		link->link_enc->funcs->disable_output(link->link_enc, signal, link);
 }
 
-bool dp_active_dongle_validate_timing(
+static bool dp_active_dongle_validate_timing(
 		const struct dc_crtc_timing *timing,
 		const struct dc_dongle_caps *dongle_caps)
 {
@@ -1833,6 +1833,8 @@ bool dp_active_dongle_validate_timing(
 	/* Check Color Depth and Pixel Clock */
 	if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
 		required_pix_clk /= 2;
+	else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+		required_pix_clk = required_pix_clk * 2 / 3;
 
 	switch (timing->display_color_depth) {
 	case COLOR_DEPTH_666:

drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c

@@ -2866,16 +2866,19 @@ static void dce110_apply_ctx_for_surface(
 		int num_planes,
 		struct dc_state *context)
 {
-	int i, be_idx;
+	int i;
 
 	if (num_planes == 0)
 		return;
 
-	be_idx = -1;
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		if (stream == context->res_ctx.pipe_ctx[i].stream) {
-			be_idx = context->res_ctx.pipe_ctx[i].stream_res.tg->inst;
-			break;
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+		struct pipe_ctx *old_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+
+		if (stream == pipe_ctx->stream) {
+			if (!pipe_ctx->top_pipe &&
+				(pipe_ctx->plane_state || old_pipe_ctx->plane_state))
+				dc->hwss.pipe_control_lock(dc, pipe_ctx, true);
 		}
 	}
 
@@ -2895,9 +2898,22 @@ static void dce110_apply_ctx_for_surface(
 					context->stream_count);
 
 		dce110_program_front_end_for_pipe(dc, pipe_ctx);
+
+		dc->hwss.update_plane_addr(dc, pipe_ctx);
+
 		program_surface_visibility(dc, pipe_ctx);
 
 	}
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+		struct pipe_ctx *old_pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+
+		if ((stream == pipe_ctx->stream) &&
+			(!pipe_ctx->top_pipe) &&
+			(pipe_ctx->plane_state || old_pipe_ctx->plane_state))
+			dc->hwss.pipe_control_lock(dc, pipe_ctx, false);
+	}
 }
 
 static void dce110_power_down_fe(struct dc *dc, int fe_idx)

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c

@@ -159,11 +159,10 @@ bool dpp_get_optimal_number_of_taps(
 			scl_data->taps.h_taps = 1;
 		if (IDENTITY_RATIO(scl_data->ratios.vert))
 			scl_data->taps.v_taps = 1;
-		/*
-		 * Spreadsheet doesn't handle taps_c is one properly,
-		 * need to force Chroma to always be scaled to pass
-		 * bandwidth validation.
-		 */
+		if (IDENTITY_RATIO(scl_data->ratios.horz_c))
+			scl_data->taps.h_taps_c = 1;
+		if (IDENTITY_RATIO(scl_data->ratios.vert_c))
+			scl_data->taps.v_taps_c = 1;
 	}
 
 	return true;

drivers/gpu/drm/drm_lease.c

@@ -220,17 +220,6 @@ static struct drm_master *drm_lease_create(struct drm_master *lessor, struct idr
 
 	mutex_lock(&dev->mode_config.idr_mutex);
 
-	/* Insert the new lessee into the tree */
-	id = idr_alloc(&(drm_lease_owner(lessor)->lessee_idr), lessee, 1, 0, GFP_KERNEL);
-	if (id < 0) {
-		error = id;
-		goto out_lessee;
-	}
-
-	lessee->lessee_id = id;
-	lessee->lessor = drm_master_get(lessor);
-	list_add_tail(&lessee->lessee_list, &lessor->lessees);
-
 	idr_for_each_entry(leases, entry, object) {
 		error = 0;
 		if (!idr_find(&dev->mode_config.crtc_idr, object))
@@ -246,6 +235,17 @@ static struct drm_master *drm_lease_create(struct drm_master *lessor, struct idr
 		}
 	}
 
+	/* Insert the new lessee into the tree */
+	id = idr_alloc(&(drm_lease_owner(lessor)->lessee_idr), lessee, 1, 0, GFP_KERNEL);
+	if (id < 0) {
+		error = id;
+		goto out_lessee;
+	}
+
+	lessee->lessee_id = id;
+	lessee->lessor = drm_master_get(lessor);
+	list_add_tail(&lessee->lessee_list, &lessor->lessees);
+
 	/* Move the leases over */
 	lessee->leases = *leases;
 	DRM_DEBUG_LEASE("new lessee %d %p, lessor %d %p\n", lessee->lessee_id, lessee, lessor->lessee_id, lessor);

drivers/gpu/drm/drm_plane.c

@@ -558,11 +558,10 @@ int drm_plane_check_pixel_format(const struct drm_plane *plane, u32 format)
 }
 
 /*
- * setplane_internal - setplane handler for internal callers
+ * __setplane_internal - setplane handler for internal callers
  *
- * Note that we assume an extra reference has already been taken on fb.  If the
- * update fails, this reference will be dropped before return; if it succeeds,
- * the previous framebuffer (if any) will be unreferenced instead.
+ * This function will take a reference on the new fb for the plane
+ * on success.
  *
  * src_{x,y,w,h} are provided in 16.16 fixed point format
  */
@@ -630,14 +629,12 @@ static int __setplane_internal(struct drm_plane *plane,
 	if (!ret) {
 		plane->crtc = crtc;
 		plane->fb = fb;
-		fb = NULL;
+		drm_framebuffer_get(plane->fb);
 	} else {
 		plane->old_fb = NULL;
 	}
 
 out:
-	if (fb)
-		drm_framebuffer_put(fb);
 	if (plane->old_fb)
 		drm_framebuffer_put(plane->old_fb);
 	plane->old_fb = NULL;
@@ -685,6 +682,7 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 	struct drm_plane *plane;
 	struct drm_crtc *crtc = NULL;
 	struct drm_framebuffer *fb = NULL;
+	int ret;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
@@ -717,15 +715,16 @@ int drm_mode_setplane(struct drm_device *dev, void *data,
 		}
 	}
 
-	/*
-	 * setplane_internal will take care of deref'ing either the old or new
-	 * framebuffer depending on success.
-	 */
-	return setplane_internal(plane, crtc, fb,
-				 plane_req->crtc_x, plane_req->crtc_y,
-				 plane_req->crtc_w, plane_req->crtc_h,
-				 plane_req->src_x, plane_req->src_y,
-				 plane_req->src_w, plane_req->src_h);
+	ret = setplane_internal(plane, crtc, fb,
+				plane_req->crtc_x, plane_req->crtc_y,
+				plane_req->crtc_w, plane_req->crtc_h,
+				plane_req->src_x, plane_req->src_y,
+				plane_req->src_w, plane_req->src_h);
+
+	if (fb)
+		drm_framebuffer_put(fb);
+
+	return ret;
 }
 
 static int drm_mode_cursor_universal(struct drm_crtc *crtc,
@@ -788,13 +787,12 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 		src_h = fb->height << 16;
 	}
 
-	/*
-	 * setplane_internal will take care of deref'ing either the old or new
-	 * framebuffer depending on success.
-	 */
 	ret = __setplane_internal(crtc->cursor, crtc, fb,
-				crtc_x, crtc_y, crtc_w, crtc_h,
-				0, 0, src_w, src_h, ctx);
+				  crtc_x, crtc_y, crtc_w, crtc_h,
+				  0, 0, src_w, src_h, ctx);
+
+	if (fb)
+		drm_framebuffer_put(fb);
 
 	/* Update successful; save new cursor position, if necessary */
 	if (ret == 0 && req->flags & DRM_MODE_CURSOR_MOVE) {

drivers/gpu/drm/drm_syncobj.c

@@ -369,40 +369,26 @@ static const struct file_operations drm_syncobj_file_fops = {
 	.release = drm_syncobj_file_release,
 };
 
-static int drm_syncobj_alloc_file(struct drm_syncobj *syncobj)
-{
-	struct file *file = anon_inode_getfile("syncobj_file",
-					       &drm_syncobj_file_fops,
-					       syncobj, 0);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	drm_syncobj_get(syncobj);
-	if (cmpxchg(&syncobj->file, NULL, file)) {
-		/* lost the race */
-		fput(file);
-	}
-
-	return 0;
-}
-
 int drm_syncobj_get_fd(struct drm_syncobj *syncobj, int *p_fd)
 {
-	int ret;
+	struct file *file;
 	int fd;
 
 	fd = get_unused_fd_flags(O_CLOEXEC);
 	if (fd < 0)
 		return fd;
 
-	if (!syncobj->file) {
-		ret = drm_syncobj_alloc_file(syncobj);
-		if (ret) {
-			put_unused_fd(fd);
-			return ret;
-		}
+	file = anon_inode_getfile("syncobj_file",
+				  &drm_syncobj_file_fops,
+				  syncobj, 0);
+	if (IS_ERR(file)) {
+		put_unused_fd(fd);
+		return PTR_ERR(file);
 	}
-	fd_install(fd, syncobj->file);
+
+	drm_syncobj_get(syncobj);
+	fd_install(fd, file);
+
 	*p_fd = fd;
 	return 0;
 }
@@ -422,31 +408,24 @@ static int drm_syncobj_handle_to_fd(struct drm_file *file_private,
 	return ret;
 }
 
-static struct drm_syncobj *drm_syncobj_fdget(int fd)
-{
-	struct file *file = fget(fd);
-
-	if (!file)
-		return NULL;
-	if (file->f_op != &drm_syncobj_file_fops)
-		goto err;
-
-	return file->private_data;
-err:
-	fput(file);
-	return NULL;
-};
-
 static int drm_syncobj_fd_to_handle(struct drm_file *file_private,
 				    int fd, u32 *handle)
 {
-	struct drm_syncobj *syncobj = drm_syncobj_fdget(fd);
+	struct drm_syncobj *syncobj;
+	struct file *file;
 	int ret;
 
-	if (!syncobj)
+	file = fget(fd);
+	if (!file)
 		return -EINVAL;
 
+	if (file->f_op != &drm_syncobj_file_fops) {
+		fput(file);
+		return -EINVAL;
+	}
+
 	/* take a reference to put in the idr */
+	syncobj = file->private_data;
 	drm_syncobj_get(syncobj);
 
 	idr_preload(GFP_KERNEL);
@@ -455,12 +434,14 @@ static int drm_syncobj_fd_to_handle(struct drm_file *file_private,
 	spin_unlock(&file_private->syncobj_table_lock);
 	idr_preload_end();
 
-	if (ret < 0) {
-		fput(syncobj->file);
-		return ret;
-	}
-	*handle = ret;
-	return 0;
+	if (ret > 0) {
+		*handle = ret;
+		ret = 0;
+	} else
+		drm_syncobj_put(syncobj);
+
+	fput(file);
+	return ret;
 }
 
 static int drm_syncobj_import_sync_file_fence(struct drm_file *file_private,

drivers/gpu/drm/i915/i915_gem.c

@@ -330,17 +330,10 @@ int i915_gem_object_unbind(struct drm_i915_gem_object *obj)
 	 * must wait for all rendering to complete to the object (as unbinding
 	 * must anyway), and retire the requests.
 	 */
-	ret = i915_gem_object_wait(obj,
-				   I915_WAIT_INTERRUPTIBLE |
-				   I915_WAIT_LOCKED |
-				   I915_WAIT_ALL,
-				   MAX_SCHEDULE_TIMEOUT,
-				   NULL);
+	ret = i915_gem_object_set_to_cpu_domain(obj, false);
 	if (ret)
 		return ret;
 
-	i915_gem_retire_requests(to_i915(obj->base.dev));
-
 	while ((vma = list_first_entry_or_null(&obj->vma_list,
 					       struct i915_vma,
 					       obj_link))) {

drivers/gpu/drm/i915/i915_sw_fence.c

@@ -367,6 +367,7 @@ struct i915_sw_dma_fence_cb {
 	struct dma_fence *dma;
 	struct timer_list timer;
 	struct irq_work work;
+	struct rcu_head rcu;
 };
 
 static void timer_i915_sw_fence_wake(struct timer_list *t)
@@ -406,7 +407,7 @@ static void irq_i915_sw_fence_work(struct irq_work *wrk)
 	del_timer_sync(&cb->timer);
 	dma_fence_put(cb->dma);
 
-	kfree(cb);
+	kfree_rcu(cb, rcu);
 }
 
 int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,

drivers/gpu/drm/i915/intel_breadcrumbs.c

@@ -186,7 +186,7 @@ void intel_engine_disarm_breadcrumbs(struct intel_engine_cs *engine)
 	struct intel_wait *wait, *n, *first;
 
 	if (!b->irq_armed)
-		return;
+		goto wakeup_signaler;
 
 	/* We only disarm the irq when we are idle (all requests completed),
 	 * so if the bottom-half remains asleep, it missed the request
@@ -208,6 +208,14 @@ void intel_engine_disarm_breadcrumbs(struct intel_engine_cs *engine)
 	b->waiters = RB_ROOT;
 
 	spin_unlock_irq(&b->rb_lock);
+
+	/*
+	 * The signaling thread may be asleep holding a reference to a request,
+	 * that had its signaling cancelled prior to being preempted. We need
+	 * to kick the signaler, just in case, to release any such reference.
+	 */
+wakeup_signaler:
+	wake_up_process(b->signaler);
 }
 
 static bool use_fake_irq(const struct intel_breadcrumbs *b)
@@ -651,23 +659,15 @@ static int intel_breadcrumbs_signaler(void *arg)
 		}
 
 		if (unlikely(do_schedule)) {
-			DEFINE_WAIT(exec);
-
 			if (kthread_should_park())
 				kthread_parkme();
 
-			if (kthread_should_stop()) {
-				GEM_BUG_ON(request);
+			if (unlikely(kthread_should_stop())) {
+				i915_gem_request_put(request);
 				break;
 			}
 
-			if (request)
-				add_wait_queue(&request->execute, &exec);
-
 			schedule();
-
-			if (request)
-				remove_wait_queue(&request->execute, &exec);
 		}
 		i915_gem_request_put(request);
 	} while (1);

drivers/gpu/drm/i915/intel_ddi.c

@@ -2128,6 +2128,8 @@ static void intel_ddi_clk_select(struct intel_encoder *encoder,
 	if (WARN_ON(!pll))
 		return;
 
+	 mutex_lock(&dev_priv->dpll_lock);
+
 	if (IS_CANNONLAKE(dev_priv)) {
 		/* Configure DPCLKA_CFGCR0 to map the DPLL to the DDI. */
 		val = I915_READ(DPCLKA_CFGCR0);
@@ -2157,6 +2159,8 @@ static void intel_ddi_clk_select(struct intel_encoder *encoder,
 	} else if (INTEL_INFO(dev_priv)->gen < 9) {
 		I915_WRITE(PORT_CLK_SEL(port), hsw_pll_to_ddi_pll_sel(pll));
 	}
+
+	mutex_unlock(&dev_priv->dpll_lock);
 }
 
 static void intel_ddi_clk_disable(struct intel_encoder *encoder)

drivers/gpu/drm/i915/intel_display.c

@@ -9944,11 +9944,10 @@ found:
 	}
 
 	ret = intel_modeset_setup_plane_state(state, crtc, mode, fb, 0, 0);
+	drm_framebuffer_put(fb);
 	if (ret)
 		goto fail;
 
-	drm_framebuffer_put(fb);
-
 	ret = drm_atomic_set_mode_for_crtc(&crtc_state->base, mode);
 	if (ret)
 		goto fail;

drivers/gpu/drm/i915/intel_lpe_audio.c

@@ -193,7 +193,7 @@ static bool lpe_audio_detect(struct drm_i915_private *dev_priv)
 		};
 
 		if (!pci_dev_present(atom_hdaudio_ids)) {
-			DRM_INFO("%s\n", "HDaudio controller not detected, using LPE audio instead\n");
+			DRM_INFO("HDaudio controller not detected, using LPE audio instead\n");
 			lpe_present = true;
 		}
 	}

drivers/gpu/drm/nouveau/nouveau_bo.c

@@ -224,7 +224,7 @@ nouveau_bo_new(struct nouveau_cli *cli, u64 size, int align,
 		/* Determine if we can get a cache-coherent map, forcing
 		 * uncached mapping if we can't.
 		 */
-		if (mmu->type[drm->ttm.type_host].type & NVIF_MEM_UNCACHED)
+		if (!nouveau_drm_use_coherent_gpu_mapping(drm))
 			nvbo->force_coherent = true;
 	}
 
@@ -262,7 +262,8 @@ nouveau_bo_new(struct nouveau_cli *cli, u64 size, int align,
 		if (cli->device.info.family > NV_DEVICE_INFO_V0_CURIE &&
 		    (flags & TTM_PL_FLAG_VRAM) && !vmm->page[i].vram)
 			continue;
-		if ((flags & TTM_PL_FLAG_TT  ) && !vmm->page[i].host)
+		if ((flags & TTM_PL_FLAG_TT) &&
+		    (!vmm->page[i].host || vmm->page[i].shift > PAGE_SHIFT))
 			continue;
 
 		/* Select this page size if it's the first that supports

drivers/gpu/drm/nouveau/nouveau_drv.h

@@ -157,8 +157,8 @@ struct nouveau_drm {
 		struct nvif_object copy;
 		int mtrr;
 		int type_vram;
-		int type_host;
-		int type_ncoh;
+		int type_host[2];
+		int type_ncoh[2];
 	} ttm;
 
 	/* GEM interface support */
@@ -217,6 +217,13 @@ nouveau_drm(struct drm_device *dev)
 	return dev->dev_private;
 }
 
+static inline bool
+nouveau_drm_use_coherent_gpu_mapping(struct nouveau_drm *drm)
+{
+	struct nvif_mmu *mmu = &drm->client.mmu;
+	return !(mmu->type[drm->ttm.type_host[0]].type & NVIF_MEM_UNCACHED);
+}
+
 int nouveau_pmops_suspend(struct device *);
 int nouveau_pmops_resume(struct device *);
 bool nouveau_pmops_runtime(void);

drivers/gpu/drm/nouveau/nouveau_fbcon.c

@@ -429,7 +429,7 @@ nouveau_fbcon_destroy(struct drm_device *dev, struct nouveau_fbdev *fbcon)
 	drm_fb_helper_unregister_fbi(&fbcon->helper);
 	drm_fb_helper_fini(&fbcon->helper);
 
-	if (nouveau_fb->nvbo) {
+	if (nouveau_fb && nouveau_fb->nvbo) {
 		nouveau_vma_del(&nouveau_fb->vma);
 		nouveau_bo_unmap(nouveau_fb->nvbo);
 		nouveau_bo_unpin(nouveau_fb->nvbo);

drivers/gpu/drm/nouveau/nouveau_mem.c

@@ -103,10 +103,10 @@ nouveau_mem_host(struct ttm_mem_reg *reg, struct ttm_dma_tt *tt)
 	u8 type;
 	int ret;
 
-	if (mmu->type[drm->ttm.type_host].type & NVIF_MEM_UNCACHED)
-		type = drm->ttm.type_ncoh;
+	if (!nouveau_drm_use_coherent_gpu_mapping(drm))
+		type = drm->ttm.type_ncoh[!!mem->kind];
 	else
-		type = drm->ttm.type_host;
+		type = drm->ttm.type_host[0];
 
 	if (mem->kind && !(mmu->type[type].type & NVIF_MEM_KIND))
 		mem->comp = mem->kind = 0;

drivers/gpu/drm/nouveau/nouveau_ttm.c

@@ -235,27 +235,46 @@ nouveau_ttm_global_release(struct nouveau_drm *drm)
 	drm->ttm.mem_global_ref.release = NULL;
 }
 
-int
-nouveau_ttm_init(struct nouveau_drm *drm)
+static int
+nouveau_ttm_init_host(struct nouveau_drm *drm, u8 kind)
 {
-	struct nvkm_device *device = nvxx_device(&drm->client.device);
-	struct nvkm_pci *pci = device->pci;
 	struct nvif_mmu *mmu = &drm->client.mmu;
-	struct drm_device *dev = drm->dev;
-	int typei, ret;
+	int typei;
 
 	typei = nvif_mmu_type(mmu, NVIF_MEM_HOST | NVIF_MEM_MAPPABLE |
-						   NVIF_MEM_COHERENT);
+					    kind | NVIF_MEM_COHERENT);
 	if (typei < 0)
 		return -ENOSYS;
 
-	drm->ttm.type_host = typei;
+	drm->ttm.type_host[!!kind] = typei;
 
-	typei = nvif_mmu_type(mmu, NVIF_MEM_HOST | NVIF_MEM_MAPPABLE);
+	typei = nvif_mmu_type(mmu, NVIF_MEM_HOST | NVIF_MEM_MAPPABLE | kind);
 	if (typei < 0)
 		return -ENOSYS;
 
-	drm->ttm.type_ncoh = typei;
+	drm->ttm.type_ncoh[!!kind] = typei;
+	return 0;
+}
+
+int
+nouveau_ttm_init(struct nouveau_drm *drm)
+{
+	struct nvkm_device *device = nvxx_device(&drm->client.device);
+	struct nvkm_pci *pci = device->pci;
+	struct nvif_mmu *mmu = &drm->client.mmu;
+	struct drm_device *dev = drm->dev;
+	int typei, ret;
+
+	ret = nouveau_ttm_init_host(drm, 0);
+	if (ret)
+		return ret;
+
+	if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA &&
+	    drm->client.device.info.chipset != 0x50) {
+		ret = nouveau_ttm_init_host(drm, NVIF_MEM_KIND);
+		if (ret)
+			return ret;
+	}
 
 	if (drm->client.device.info.platform != NV_DEVICE_INFO_V0_SOC &&
 	    drm->client.device.info.family >= NV_DEVICE_INFO_V0_TESLA) {

drivers/gpu/drm/nouveau/nouveau_vmm.c

@@ -67,8 +67,8 @@ nouveau_vma_del(struct nouveau_vma **pvma)
 			nvif_vmm_put(&vma->vmm->vmm, &tmp);
 		}
 		list_del(&vma->head);
-		*pvma = NULL;
 		kfree(*pvma);
+		*pvma = NULL;
 	}
 }
 

drivers/gpu/drm/nouveau/nvkm/engine/device/base.c

@@ -2369,7 +2369,7 @@ nv13b_chipset = {
 	.imem = gk20a_instmem_new,
 	.ltc = gp100_ltc_new,
 	.mc = gp10b_mc_new,
-	.mmu = gf100_mmu_new,
+	.mmu = gp10b_mmu_new,
 	.secboot = gp10b_secboot_new,
 	.pmu = gm20b_pmu_new,
 	.timer = gk20a_timer_new,

drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c

@@ -36,6 +36,7 @@ nvbios_dp_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 			if (data) {
 				*ver = nvbios_rd08(bios, data + 0x00);
 				switch (*ver) {
+				case 0x20:
 				case 0x21:
 				case 0x30:
 				case 0x40:
@@ -63,6 +64,7 @@ nvbios_dpout_entry(struct nvkm_bios *bios, u8 idx,
 	if (data && idx < *cnt) {
 		u16 outp = nvbios_rd16(bios, data + *hdr + idx * *len);
 		switch (*ver * !!outp) {
+		case 0x20:
 		case 0x21:
 		case 0x30:
 			*hdr = nvbios_rd08(bios, data + 0x04);
@@ -96,12 +98,16 @@ nvbios_dpout_parse(struct nvkm_bios *bios, u8 idx,
 		info->type = nvbios_rd16(bios, data + 0x00);
 		info->mask = nvbios_rd16(bios, data + 0x02);
 		switch (*ver) {
+		case 0x20:
+			info->mask |= 0x00c0; /* match any link */
+			/* fall-through */
 		case 0x21:
 		case 0x30:
 			info->flags     = nvbios_rd08(bios, data + 0x05);
 			info->script[0] = nvbios_rd16(bios, data + 0x06);
 			info->script[1] = nvbios_rd16(bios, data + 0x08);
-			info->lnkcmp    = nvbios_rd16(bios, data + 0x0a);
+			if (*len >= 0x0c)
+				info->lnkcmp    = nvbios_rd16(bios, data + 0x0a);
 			if (*len >= 0x0f) {
 				info->script[2] = nvbios_rd16(bios, data + 0x0c);
 				info->script[3] = nvbios_rd16(bios, data + 0x0e);
@@ -170,6 +176,7 @@ nvbios_dpcfg_parse(struct nvkm_bios *bios, u16 outp, u8 idx,
 	memset(info, 0x00, sizeof(*info));
 	if (data) {
 		switch (*ver) {
+		case 0x20:
 		case 0x21:
 			info->dc    = nvbios_rd08(bios, data + 0x02);
 			info->pe    = nvbios_rd08(bios, data + 0x03);

drivers/gpu/drm/nouveau/nvkm/subdev/instmem/nv50.c

@@ -249,7 +249,7 @@ nv50_instobj_acquire(struct nvkm_memory *memory)
 			iobj->base.memory.ptrs = &nv50_instobj_fast;
 		else
 			iobj->base.memory.ptrs = &nv50_instobj_slow;
-		refcount_inc(&iobj->maps);
+		refcount_set(&iobj->maps, 1);
 	}
 
 	mutex_unlock(&imem->subdev.mutex);

drivers/gpu/drm/nouveau/nvkm/subdev/pci/base.c

@@ -136,6 +136,13 @@ nvkm_pci_init(struct nvkm_subdev *subdev)
 		return ret;
 
 	pci->irq = pdev->irq;
+
+	/* Ensure MSI interrupts are armed, for the case where there are
+	 * already interrupts pending (for whatever reason) at load time.
+	 */
+	if (pci->msi)
+		pci->func->msi_rearm(pci);
+
 	return ret;
 }
 

drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c

@@ -175,11 +175,31 @@ static void sun4i_hdmi_mode_set(struct drm_encoder *encoder,
 	writel(val, hdmi->base + SUN4I_HDMI_VID_TIMING_POL_REG);
 }
 
+static enum drm_mode_status sun4i_hdmi_mode_valid(struct drm_encoder *encoder,
+					const struct drm_display_mode *mode)
+{
+	struct sun4i_hdmi *hdmi = drm_encoder_to_sun4i_hdmi(encoder);
+	unsigned long rate = mode->clock * 1000;
+	unsigned long diff = rate / 200; /* +-0.5% allowed by HDMI spec */
+	long rounded_rate;
+
+	/* 165 MHz is the typical max pixelclock frequency for HDMI <= 1.2 */
+	if (rate > 165000000)
+		return MODE_CLOCK_HIGH;
+	rounded_rate = clk_round_rate(hdmi->tmds_clk, rate);
+	if (rounded_rate > 0 &&
+	    max_t(unsigned long, rounded_rate, rate) -
+	    min_t(unsigned long, rounded_rate, rate) < diff)
+		return MODE_OK;
+	return MODE_NOCLOCK;
+}
+
 static const struct drm_encoder_helper_funcs sun4i_hdmi_helper_funcs = {
 	.atomic_check	= sun4i_hdmi_atomic_check,
 	.disable	= sun4i_hdmi_disable,
 	.enable		= sun4i_hdmi_enable,
 	.mode_set	= sun4i_hdmi_mode_set,
+	.mode_valid	= sun4i_hdmi_mode_valid,
 };
 
 static const struct drm_encoder_funcs sun4i_hdmi_funcs = {

drivers/gpu/drm/sun4i/sun4i_tcon.c

@@ -724,12 +724,12 @@ static int sun4i_tcon_bind(struct device *dev, struct device *master,
 	if (IS_ERR(tcon->crtc)) {
 		dev_err(dev, "Couldn't create our CRTC\n");
 		ret = PTR_ERR(tcon->crtc);
-		goto err_free_clocks;
+		goto err_free_dotclock;
 	}
 
 	ret = sun4i_rgb_init(drm, tcon);
 	if (ret < 0)
-		goto err_free_clocks;
+		goto err_free_dotclock;
 
 	if (tcon->quirks->needs_de_be_mux) {
 		/*

drivers/gpu/drm/ttm/ttm_page_alloc.c

@@ -455,6 +455,7 @@ ttm_pool_shrink_scan(struct shrinker *shrink, struct shrink_control *sc)
 		freed += (nr_free_pool - shrink_pages) << pool->order;
 		if (freed >= sc->nr_to_scan)
 			break;
+		shrink_pages <<= pool->order;
 	}
 	mutex_unlock(&lock);
 	return freed;
@@ -543,7 +544,7 @@ static int ttm_alloc_new_pages(struct list_head *pages, gfp_t gfp_flags,
 	int r = 0;
 	unsigned i, j, cpages;
 	unsigned npages = 1 << order;
-	unsigned max_cpages = min(count, (unsigned)NUM_PAGES_TO_ALLOC);
+	unsigned max_cpages = min(count << order, (unsigned)NUM_PAGES_TO_ALLOC);
 
 	/* allocate array for page caching change */
 	caching_array = kmalloc(max_cpages*sizeof(struct page *), GFP_KERNEL);