Merge tag 'topic/drm-misc-2016-01-17' of git://anongit.freedesktop.org/drm-intel into drm-next

Since your main drm-next pull isn't out of the door yet I figured I might
as well flush out drm-misc instead of delaying for 4.6. It's really just
random stuff all over, biggest thing probably connector_mask tracking from
Maarten.

* tag 'topic/drm-misc-2016-01-17' of git://anongit.freedesktop.org/drm-intel: (24 commits)
drm/fb_cma_helper: Remove implicit call to disable_unused_functions
drm/sysfs: use kobj_to_dev()
drm/i915: Init power domains early in driver load
drm: Do not set connector->encoder in drivers
apple-gmux: Add initial documentation
drm: move MODULE_PARM_DESC to other file
drm/edid: index CEA/HDMI mode tables using the VIC
drm/atomic: Remove drm_atomic_connectors_for_crtc.
drm/i915: Update connector_mask during readout, v2.
drm: Remove opencoded drm_gem_object_release_handle()
drm: Do not set outparam on error during GEM handle allocation
drm/docs: more leftovers from the big vtable documentation pile
drm/atomic-helper: Reject legacy flips on a disabled pipe
drm/atomic: add connector mask to drm_crtc_state.
drm/tegra: Use __drm_atomic_helper_reset_connector for subclassing connector state, v2.
drm/atomic: Add __drm_atomic_helper_connector_reset, v2.
drm/i915: Set connector_state->connector using the helper.
drm: Use a normal idr allocation for the obj->name
drm: Only bump object-reference count when adding first handle
drm: Balance error path for GEM handle allocation
...

Documentation/DocBook/gpu.tmpl

···
       entities.
     </para>
     <sect2>
-      <title>Legacy CRTC Helper Operations</title>
-      <itemizedlist>
-        <listitem id="drm-helper-crtc-mode-fixup">
-          <synopsis>bool (*mode_fixup)(struct drm_crtc *crtc,
-                       const struct drm_display_mode *mode,
-                       struct drm_display_mode *adjusted_mode);</synopsis>
-          <para>
-            Let CRTCs adjust the requested mode or reject it completely. This
-            operation returns true if the mode is accepted (possibly after being
-            adjusted) or false if it is rejected.
-          </para>
-          <para>
-            The <methodname>mode_fixup</methodname> operation should reject the
-            mode if it can't reasonably use it. The definition of "reasonable"
-            is currently fuzzy in this context. One possible behaviour would be
-            to set the adjusted mode to the panel timings when a fixed-mode
-            panel is used with hardware capable of scaling. Another behaviour
-            would be to accept any input mode and adjust it to the closest mode
-            supported by the hardware (FIXME: This needs to be clarified).
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>int (*mode_set_base)(struct drm_crtc *crtc, int x, int y,
-                     struct drm_framebuffer *old_fb)</synopsis>
-          <para>
-            Move the CRTC on the current frame buffer (stored in
-            <literal>crtc-&gt;fb</literal>) to position (x,y). Any of the frame
-            buffer, x position or y position may have been modified.
-          </para>
-          <para>
-            This helper operation is optional. If not provided, the
-            <function>drm_crtc_helper_set_config</function> function will fall
-            back to the <methodname>mode_set</methodname> helper operation.
-          </para>
-          <note><para>
-            FIXME: Why are x and y passed as arguments, as they can be accessed
-            through <literal>crtc-&gt;x</literal> and
-            <literal>crtc-&gt;y</literal>?
-          </para></note>
-        </listitem>
-        <listitem>
-          <synopsis>void (*prepare)(struct drm_crtc *crtc);</synopsis>
-          <para>
-            Prepare the CRTC for mode setting. This operation is called after
-            validating the requested mode. Drivers use it to perform
-            device-specific operations required before setting the new mode.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>int (*mode_set)(struct drm_crtc *crtc, struct drm_display_mode *mode,
-                struct drm_display_mode *adjusted_mode, int x, int y,
-                struct drm_framebuffer *old_fb);</synopsis>
-          <para>
-            Set a new mode, position and frame buffer. Depending on the device
-            requirements, the mode can be stored internally by the driver and
-            applied in the <methodname>commit</methodname> operation, or
-            programmed to the hardware immediately.
-          </para>
-          <para>
-            The <methodname>mode_set</methodname> operation returns 0 on success
-	    or a negative error code if an error occurs.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>void (*commit)(struct drm_crtc *crtc);</synopsis>
-          <para>
-            Commit a mode. This operation is called after setting the new mode.
-            Upon return the device must use the new mode and be fully
-            operational.
-          </para>
-        </listitem>
-      </itemizedlist>
-    </sect2>
-    <sect2>
-      <title>Encoder Helper Operations</title>
-      <itemizedlist>
-        <listitem>
-          <synopsis>bool (*mode_fixup)(struct drm_encoder *encoder,
-                       const struct drm_display_mode *mode,
-                       struct drm_display_mode *adjusted_mode);</synopsis>
-          <para>
-            Let encoders adjust the requested mode or reject it completely. This
-            operation returns true if the mode is accepted (possibly after being
-            adjusted) or false if it is rejected. See the
-            <link linkend="drm-helper-crtc-mode-fixup">mode_fixup CRTC helper
-            operation</link> for an explanation of the allowed adjustments.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>void (*prepare)(struct drm_encoder *encoder);</synopsis>
-          <para>
-            Prepare the encoder for mode setting. This operation is called after
-            validating the requested mode. Drivers use it to perform
-            device-specific operations required before setting the new mode.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>void (*mode_set)(struct drm_encoder *encoder,
-                 struct drm_display_mode *mode,
-                 struct drm_display_mode *adjusted_mode);</synopsis>
-          <para>
-            Set a new mode. Depending on the device requirements, the mode can
-            be stored internally by the driver and applied in the
-            <methodname>commit</methodname> operation, or programmed to the
-            hardware immediately.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>void (*commit)(struct drm_encoder *encoder);</synopsis>
-          <para>
-            Commit a mode. This operation is called after setting the new mode.
-            Upon return the device must use the new mode and be fully
-            operational.
-          </para>
-        </listitem>
-      </itemizedlist>
-    </sect2>
-    <sect2>
-      <title>Connector Helper Operations</title>
-      <itemizedlist>
-        <listitem>
-          <synopsis>struct drm_encoder *(*best_encoder)(struct drm_connector *connector);</synopsis>
-          <para>
-            Return a pointer to the best encoder for the connecter. Device that
-            map connectors to encoders 1:1 simply return the pointer to the
-            associated encoder. This operation is mandatory.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>int (*get_modes)(struct drm_connector *connector);</synopsis>
-          <para>
-            Fill the connector's <structfield>probed_modes</structfield> list
-            by parsing EDID data with <function>drm_add_edid_modes</function>,
-            adding standard VESA DMT modes with <function>drm_add_modes_noedid</function>,
-            or calling <function>drm_mode_probed_add</function> directly for every
-            supported mode and return the number of modes it has detected. This
-            operation is mandatory.
-          </para>
-          <para>
-            Note that the caller function will automatically add standard VESA
-            DMT modes up to 1024x768 if the <methodname>get_modes</methodname>
-            helper operation returns no mode and if the connector status is
-            connector_status_connected. There is no need to call
-            <function>drm_add_edid_modes</function> manually in that case.
-          </para>
-          <para>
-            The <structfield>vrefresh</structfield> value is computed by
-            <function>drm_helper_probe_single_connector_modes</function>.
-          </para>
-          <para>
-            When parsing EDID data, <function>drm_add_edid_modes</function> fills the
-            connector <structfield>display_info</structfield>
-            <structfield>width_mm</structfield> and
-            <structfield>height_mm</structfield> fields. When creating modes
-            manually the <methodname>get_modes</methodname> helper operation must
-            set the <structfield>display_info</structfield>
-            <structfield>width_mm</structfield> and
-            <structfield>height_mm</structfield> fields if they haven't been set
-            already (for instance at initialization time when a fixed-size panel is
-            attached to the connector). The mode <structfield>width_mm</structfield>
-            and <structfield>height_mm</structfield> fields are only used internally
-            during EDID parsing and should not be set when creating modes manually.
-          </para>
-        </listitem>
-        <listitem>
-          <synopsis>int (*mode_valid)(struct drm_connector *connector,
-		  struct drm_display_mode *mode);</synopsis>
-          <para>
-            Verify whether a mode is valid for the connector. Return MODE_OK for
-            supported modes and one of the enum drm_mode_status values (MODE_*)
-            for unsupported modes. This operation is optional.
-          </para>
-          <para>
-            As the mode rejection reason is currently not used beside for
-            immediately removing the unsupported mode, an implementation can
-            return MODE_BAD regardless of the exact reason why the mode is not
-            valid.
-          </para>
-          <note><para>
-            Note that the <methodname>mode_valid</methodname> helper operation is
-            only called for modes detected by the device, and
-            <emphasis>not</emphasis> for modes set by the user through the CRTC
-            <methodname>set_config</methodname> operation.
-          </para></note>
-        </listitem>
-      </itemizedlist>
-    </sect2>
-    <sect2>
       <title>Atomic Modeset Helper Functions Reference</title>
       <sect3>
 	<title>Overview</title>
···
 
   <chapter id="modes_of_use">
     <title>Modes of Use</title>
-  <sect1>
-    <title>Manual switching and manual power control</title>
+    <sect1>
+      <title>Manual switching and manual power control</title>
 !Pdrivers/gpu/vga/vga_switcheroo.c Manual switching and manual power control
-  </sect1>
-  <sect1>
-    <title>Driver power control</title>
+    </sect1>
+    <sect1>
+      <title>Driver power control</title>
 !Pdrivers/gpu/vga/vga_switcheroo.c Driver power control
-  </sect1>
+    </sect1>
   </chapter>
 
-  <chapter id="pubfunctions">
-    <title>Public functions</title>
+  <chapter id="api">
+    <title>API</title>
+    <sect1>
+      <title>Public functions</title>
 !Edrivers/gpu/vga/vga_switcheroo.c
-  </chapter>
-
-  <chapter id="pubstructures">
-    <title>Public structures</title>
+    </sect1>
+    <sect1>
+      <title>Public structures</title>
 !Finclude/linux/vga_switcheroo.h vga_switcheroo_handler
 !Finclude/linux/vga_switcheroo.h vga_switcheroo_client_ops
-  </chapter>
-
-  <chapter id="pubconstants">
-    <title>Public constants</title>
+    </sect1>
+    <sect1>
+      <title>Public constants</title>
 !Finclude/linux/vga_switcheroo.h vga_switcheroo_client_id
 !Finclude/linux/vga_switcheroo.h vga_switcheroo_state
-  </chapter>
-
-  <chapter id="privstructures">
-    <title>Private structures</title>
+    </sect1>
+    <sect1>
+      <title>Private structures</title>
 !Fdrivers/gpu/vga/vga_switcheroo.c vgasr_priv
 !Fdrivers/gpu/vga/vga_switcheroo.c vga_switcheroo_client
+    </sect1>
+  </chapter>
+
+  <chapter id="handlers">
+    <title>Handlers</title>
+    <sect1>
+      <title>apple-gmux Handler</title>
+!Pdrivers/platform/x86/apple-gmux.c Overview
+!Pdrivers/platform/x86/apple-gmux.c Interrupt
+      <sect2>
+        <title>Graphics mux</title>
+!Pdrivers/platform/x86/apple-gmux.c Graphics mux
+      </sect2>
+      <sect2>
+        <title>Power control</title>
+!Pdrivers/platform/x86/apple-gmux.c Power control
+      </sect2>
+      <sect2>
+        <title>Backlight control</title>
+!Pdrivers/platform/x86/apple-gmux.c Backlight control
+      </sect2>
+    </sect1>
   </chapter>
 
 !Cdrivers/gpu/vga/vga_switcheroo.c
 !Cinclude/linux/vga_switcheroo.h
+!Cdrivers/platform/x86/apple-gmux.c
 </part>
 
 </book>

drivers/gpu/drm/bridge/dw-hdmi.c

···
 				   &dw_hdmi_connector_funcs,
 				   DRM_MODE_CONNECTOR_HDMIA);
 
-	hdmi->connector.encoder = encoder;
-
 	drm_mode_connector_attach_encoder(&hdmi->connector, encoder);
 
 	return 0;

drivers/gpu/drm/drm_atomic.c

···
 		return -EINVAL;
 	}
 
+	/*
+	 * Reject event generation for when a CRTC is off and stays off.
+	 * It wouldn't be hard to implement this, but userspace has a track
+	 * record of happily burning through 100% cpu (or worse, crash) when the
+	 * display pipe is suspended. To avoid all that fun just reject updates
+	 * that ask for events since likely that indicates a bug in the
+	 * compositor's drawing loop. This is consistent with the vblank IOCTL
+	 * and legacy page_flip IOCTL which also reject service on a disabled
+	 * pipe.
+	 */
+	if (state->event && !state->active && !crtc->state->active) {
+		DRM_DEBUG_ATOMIC("[CRTC:%d] requesting event but off\n",
+				 crtc->base.id);
+		return -EINVAL;
+	}
+
 	return 0;
 }
 
···
 {
 	struct drm_crtc_state *crtc_state;
 
+	if (conn_state->crtc && conn_state->crtc != crtc) {
+		crtc_state = drm_atomic_get_existing_crtc_state(conn_state->state,
+								conn_state->crtc);
+
+		crtc_state->connector_mask &=
+			~(1 << drm_connector_index(conn_state->connector));
+	}
+
 	if (crtc) {
 		crtc_state = drm_atomic_get_crtc_state(conn_state->state, crtc);
 		if (IS_ERR(crtc_state))
 			return PTR_ERR(crtc_state);
+
+		crtc_state->connector_mask |=
+			1 << drm_connector_index(conn_state->connector);
 	}
 
 	conn_state->crtc = crtc;
···
 	return 0;
 }
 EXPORT_SYMBOL(drm_atomic_add_affected_planes);
-
-/**
- * drm_atomic_connectors_for_crtc - count number of connected outputs
- * @state: atomic state
- * @crtc: DRM crtc
- *
- * This function counts all connectors which will be connected to @crtc
- * according to @state. Useful to recompute the enable state for @crtc.
- */
-int
-drm_atomic_connectors_for_crtc(struct drm_atomic_state *state,
-			       struct drm_crtc *crtc)
-{
-	struct drm_connector *connector;
-	struct drm_connector_state *conn_state;
-
-	int i, num_connected_connectors = 0;
-
-	for_each_connector_in_state(state, connector, conn_state, i) {
-		if (conn_state->crtc == crtc)
-			num_connected_connectors++;
-	}
-
-	DRM_DEBUG_ATOMIC("State %p has %i connectors for [CRTC:%d:%s]\n",
-			 state, num_connected_connectors,
-			 crtc->base.id, crtc->name);
-
-	return num_connected_connectors;
-}
-EXPORT_SYMBOL(drm_atomic_connectors_for_crtc);
 
 /**
  * drm_atomic_legacy_backoff - locking backoff for legacy ioctls

drivers/gpu/drm/drm_atomic_helper.c

···
 	 * crtc only changed its mode but has the same set of connectors.
 	 */
 	for_each_crtc_in_state(state, crtc, crtc_state, i) {
-		int num_connectors;
+		bool has_connectors =
+			!!crtc_state->connector_mask;
 
 		/*
 		 * We must set ->active_changed after walking connectors for
···
 		if (ret != 0)
 			return ret;
 
-		num_connectors = drm_atomic_connectors_for_crtc(state,
-								crtc);
-
-		if (crtc_state->enable != !!num_connectors) {
+		if (crtc_state->enable != has_connectors) {
 			DRM_DEBUG_ATOMIC("[CRTC:%d:%s] enabled/connectors mismatch\n",
 					 crtc->base.id, crtc->name);
 
···
 		if (crtc == set->crtc)
 			continue;
 
-		if (!drm_atomic_connectors_for_crtc(state, crtc)) {
+		if (!crtc_state->connector_mask) {
 			ret = drm_atomic_set_mode_prop_for_crtc(crtc_state,
 								NULL);
 			if (ret < 0)
···
 		goto fail;
 	drm_atomic_set_fb_for_plane(plane_state, fb);
 
+	/* Make sure we don't accidentally do a full modeset. */
+	state->allow_modeset = false;
+	if (!crtc_state->active) {
+		DRM_DEBUG_ATOMIC("[CRTC:%d] disabled, rejecting legacy flip\n",
+				 crtc->base.id);
+		ret = -EINVAL;
+		goto fail;
+	}
+
 	ret = drm_atomic_async_commit(state);
 	if (ret != 0)
 		goto fail;
···
 EXPORT_SYMBOL(drm_atomic_helper_plane_destroy_state);
 
 /**
+ * __drm_atomic_helper_connector_reset - reset state on connector
+ * @connector: drm connector
+ * @conn_state: connector state to assign
+ *
+ * Initializes the newly allocated @conn_state and assigns it to
+ * #connector ->state, usually required when initializing the drivers
+ * or when called from the ->reset hook.
+ *
+ * This is useful for drivers that subclass the connector state.
+ */
+void
+__drm_atomic_helper_connector_reset(struct drm_connector *connector,
+				    struct drm_connector_state *conn_state)
+{
+	if (conn_state)
+		conn_state->connector = connector;
+
+	connector->state = conn_state;
+}
+EXPORT_SYMBOL(__drm_atomic_helper_connector_reset);
+
+/**
  * drm_atomic_helper_connector_reset - default ->reset hook for connectors
  * @connector: drm connector
  *
···
  */
 void drm_atomic_helper_connector_reset(struct drm_connector *connector)
 {
-	kfree(connector->state);
-	connector->state = kzalloc(sizeof(*connector->state), GFP_KERNEL);
+	struct drm_connector_state *conn_state =
+		kzalloc(sizeof(*conn_state), GFP_KERNEL);
 
-	if (connector->state)
-		connector->state->connector = connector;
+	kfree(connector->state);
+	__drm_atomic_helper_connector_reset(connector, conn_state);
 }
 EXPORT_SYMBOL(drm_atomic_helper_connector_reset);
 

drivers/gpu/drm/drm_crtc.c

···
 {
 	int i;
 
+	/*
+	 * In the past, drivers have attempted to model the static association
+	 * of connector to encoder in simple connector/encoder devices using a
+	 * direct assignment of connector->encoder = encoder. This connection
+	 * is a logical one and the responsibility of the core, so drivers are
+	 * expected not to mess with this.
+	 *
+	 * Note that the error return should've been enough here, but a large
+	 * majority of drivers ignores the return value, so add in a big WARN
+	 * to get people's attention.
+	 */
+	if (WARN_ON(connector->encoder))
+		return -EINVAL;
+
 	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
 		if (connector->encoder_ids[i] == 0) {
 			connector->encoder_ids[i] = encoder->base.id;

drivers/gpu/drm/drm_drv.c

···
 MODULE_DESCRIPTION(CORE_DESC);
 MODULE_LICENSE("GPL and additional rights");
 MODULE_PARM_DESC(debug, "Enable debug output");
-MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs] (0: never disable, <0: disable immediately)");
-MODULE_PARM_DESC(timestamp_precision_usec, "Max. error on timestamps [usecs]");
-MODULE_PARM_DESC(timestamp_monotonic, "Use monotonic timestamps");
-
 module_param_named(debug, drm_debug, int, 0600);
 
 static DEFINE_SPINLOCK(drm_minor_lock);

drivers/gpu/drm/drm_edid.c

···
 /*
  * Probably taken from CEA-861 spec.
  * This table is converted from xorg's hw/xfree86/modes/xf86EdidModes.c.
+ *
+ * Index using the VIC.
  */
 static const struct drm_display_mode edid_cea_modes[] = {
+	/* 0 - dummy, VICs start at 1 */
+	{ },
 	/* 1 - 640x480@60Hz */
 	{ DRM_MODE("640x480", DRM_MODE_TYPE_DRIVER, 25175, 640, 656,
 		   752, 800, 0, 480, 490, 492, 525, 0,
···
 };
 
 /*
- * HDMI 1.4 4k modes.
+ * HDMI 1.4 4k modes. Index using the VIC.
  */
 static const struct drm_display_mode edid_4k_modes[] = {
+	/* 0 - dummy, VICs start at 1 */
+	{ },
 	/* 1 - 3840x2160@30Hz */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000,
 		   3840, 4016, 4104, 4400, 0,
···
 static u8 drm_match_cea_mode_clock_tolerance(const struct drm_display_mode *to_match,
 					     unsigned int clock_tolerance)
 {
-	u8 mode;
+	u8 vic;
 
 	if (!to_match->clock)
 		return 0;
 
-	for (mode = 0; mode < ARRAY_SIZE(edid_cea_modes); mode++) {
-		const struct drm_display_mode *cea_mode = &edid_cea_modes[mode];
+	for (vic = 1; vic < ARRAY_SIZE(edid_cea_modes); vic++) {
+		const struct drm_display_mode *cea_mode = &edid_cea_modes[vic];
 		unsigned int clock1, clock2;
 
 		/* Check both 60Hz and 59.94Hz */
···
 			continue;
 
 		if (drm_mode_equal_no_clocks(to_match, cea_mode))
-			return mode + 1;
+			return vic;
 	}
 
 	return 0;
···
  */
 u8 drm_match_cea_mode(const struct drm_display_mode *to_match)
 {
-	u8 mode;
+	u8 vic;
 
 	if (!to_match->clock)
 		return 0;
 
-	for (mode = 0; mode < ARRAY_SIZE(edid_cea_modes); mode++) {
-		const struct drm_display_mode *cea_mode = &edid_cea_modes[mode];
+	for (vic = 1; vic < ARRAY_SIZE(edid_cea_modes); vic++) {
+		const struct drm_display_mode *cea_mode = &edid_cea_modes[vic];
 		unsigned int clock1, clock2;
 
 		/* Check both 60Hz and 59.94Hz */
···
 		if ((KHZ2PICOS(to_match->clock) == KHZ2PICOS(clock1) ||
 		     KHZ2PICOS(to_match->clock) == KHZ2PICOS(clock2)) &&
 		    drm_mode_equal_no_clocks_no_stereo(to_match, cea_mode))
-			return mode + 1;
+			return vic;
 	}
 	return 0;
 }
 EXPORT_SYMBOL(drm_match_cea_mode);
+
+static bool drm_valid_cea_vic(u8 vic)
+{
+	return vic > 0 && vic < ARRAY_SIZE(edid_cea_modes);
+}
 
 /**
  * drm_get_cea_aspect_ratio - get the picture aspect ratio corresponding to
···
  */
 enum hdmi_picture_aspect drm_get_cea_aspect_ratio(const u8 video_code)
 {
-	/* return picture aspect ratio for video_code - 1 to access the
-	 * right array element
-	*/
-	return edid_cea_modes[video_code-1].picture_aspect_ratio;
+	return edid_cea_modes[video_code].picture_aspect_ratio;
 }
 EXPORT_SYMBOL(drm_get_cea_aspect_ratio);
 
···
 static u8 drm_match_hdmi_mode_clock_tolerance(const struct drm_display_mode *to_match,
 					      unsigned int clock_tolerance)
 {
-	u8 mode;
+	u8 vic;
 
 	if (!to_match->clock)
 		return 0;
 
-	for (mode = 0; mode < ARRAY_SIZE(edid_4k_modes); mode++) {
-		const struct drm_display_mode *hdmi_mode = &edid_4k_modes[mode];
+	for (vic = 1; vic < ARRAY_SIZE(edid_4k_modes); vic++) {
+		const struct drm_display_mode *hdmi_mode = &edid_4k_modes[vic];
 		unsigned int clock1, clock2;
 
 		/* Make sure to also match alternate clocks */
···
 			continue;
 
 		if (drm_mode_equal_no_clocks(to_match, hdmi_mode))
-			return mode + 1;
+			return vic;
 	}
 
 	return 0;
···
  */
 static u8 drm_match_hdmi_mode(const struct drm_display_mode *to_match)
 {
-	u8 mode;
+	u8 vic;
 
 	if (!to_match->clock)
 		return 0;
 
-	for (mode = 0; mode < ARRAY_SIZE(edid_4k_modes); mode++) {
-		const struct drm_display_mode *hdmi_mode = &edid_4k_modes[mode];
+	for (vic = 1; vic < ARRAY_SIZE(edid_4k_modes); vic++) {
+		const struct drm_display_mode *hdmi_mode = &edid_4k_modes[vic];
 		unsigned int clock1, clock2;
 
 		/* Make sure to also match alternate clocks */
···
 		if ((KHZ2PICOS(to_match->clock) == KHZ2PICOS(clock1) ||
 		     KHZ2PICOS(to_match->clock) == KHZ2PICOS(clock2)) &&
 		    drm_mode_equal_no_clocks_no_stereo(to_match, hdmi_mode))
-			return mode + 1;
+			return vic;
 	}
 	return 0;
+}
+
+static bool drm_valid_hdmi_vic(u8 vic)
+{
+	return vic > 0 && vic < ARRAY_SIZE(edid_4k_modes);
 }
 
 static int
···
 	list_for_each_entry(mode, &connector->probed_modes, head) {
 		const struct drm_display_mode *cea_mode = NULL;
 		struct drm_display_mode *newmode;
-		u8 mode_idx = drm_match_cea_mode(mode) - 1;
+		u8 vic = drm_match_cea_mode(mode);
 		unsigned int clock1, clock2;
 
-		if (mode_idx < ARRAY_SIZE(edid_cea_modes)) {
-			cea_mode = &edid_cea_modes[mode_idx];
+		if (drm_valid_cea_vic(vic)) {
+			cea_mode = &edid_cea_modes[vic];
 			clock2 = cea_mode_alternate_clock(cea_mode);
 		} else {
-			mode_idx = drm_match_hdmi_mode(mode) - 1;
-			if (mode_idx < ARRAY_SIZE(edid_4k_modes)) {
-				cea_mode = &edid_4k_modes[mode_idx];
+			vic = drm_match_hdmi_mode(mode);
+			if (drm_valid_hdmi_vic(vic)) {
+				cea_mode = &edid_4k_modes[vic];
 				clock2 = hdmi_mode_alternate_clock(cea_mode);
 			}
 		}
···
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_display_mode *newmode;
-	u8 cea_mode;
+	u8 vic;
 
 	if (video_db == NULL || video_index >= video_len)
 		return NULL;
 
 	/* CEA modes are numbered 1..127 */
-	cea_mode = (video_db[video_index] & 127) - 1;
-	if (cea_mode >= ARRAY_SIZE(edid_cea_modes))
+	vic = (video_db[video_index] & 127);
+	if (!drm_valid_cea_vic(vic))
 		return NULL;
 
-	newmode = drm_mode_duplicate(dev, &edid_cea_modes[cea_mode]);
+	newmode = drm_mode_duplicate(dev, &edid_cea_modes[vic]);
 	if (!newmode)
 		return NULL;
 
···
 	struct drm_device *dev = connector->dev;
 	struct drm_display_mode *newmode;
 
-	vic--; /* VICs start at 1 */
-	if (vic >= ARRAY_SIZE(edid_4k_modes)) {
+	if (!drm_valid_hdmi_vic(vic)) {
 		DRM_ERROR("Unknown HDMI VIC: %d\n", vic);
 		return 0;
 	}
···
 {
 	const struct drm_display_mode *cea_mode;
 	int clock1, clock2, clock;
-	u8 mode_idx;
+	u8 vic;
 	const char *type;
 
 	/*
 	 * allow 5kHz clock difference either way to account for
 	 * the 10kHz clock resolution limit of detailed timings.
 	 */
-	mode_idx = drm_match_cea_mode_clock_tolerance(mode, 5) - 1;
-	if (mode_idx < ARRAY_SIZE(edid_cea_modes)) {
+	vic = drm_match_cea_mode_clock_tolerance(mode, 5);
+	if (drm_valid_cea_vic(vic)) {
 		type = "CEA";
-		cea_mode = &edid_cea_modes[mode_idx];
+		cea_mode = &edid_cea_modes[vic];
 		clock1 = cea_mode->clock;
 		clock2 = cea_mode_alternate_clock(cea_mode);
 	} else {
-		mode_idx = drm_match_hdmi_mode_clock_tolerance(mode, 5) - 1;
-		if (mode_idx < ARRAY_SIZE(edid_4k_modes)) {
+		vic = drm_match_hdmi_mode_clock_tolerance(mode, 5);
+		if (drm_valid_hdmi_vic(vic)) {
 			type = "HDMI";
-			cea_mode = &edid_4k_modes[mode_idx];
+			cea_mode = &edid_4k_modes[vic];
 			clock1 = cea_mode->clock;
 			clock2 = hdmi_mode_alternate_clock(cea_mode);
 		} else {
···
 		return;
 
 	DRM_DEBUG("detailed mode matches %s VIC %d, adjusting clock %d -> %d\n",
-		  type, mode_idx + 1, mode->clock, clock);
+		  type, vic, mode->clock, clock);
 	mode->clock = clock;
 }
 

drivers/gpu/drm/drm_fb_cma_helper.c

···
 
 	}
 
-	/* disable all the possible outputs/crtcs before entering KMS mode */
-	drm_helper_disable_unused_functions(dev);
-
 	ret = drm_fb_helper_initial_config(helper, preferred_bpp);
 	if (ret < 0) {
 		dev_err(dev->dev, "Failed to set initial hw configuration.\n");

drivers/gpu/drm/drm_fb_helper.c

···
 			goto fail;
 
 		plane = mode_set->crtc->primary;
-		plane_mask |= drm_plane_index(plane);
+		plane_mask |= (1 << drm_plane_index(plane));
 		plane->old_fb = plane->fb;
 	}
 

drivers/gpu/drm/drm_gem.c

···
 static void
 drm_gem_object_handle_unreference_unlocked(struct drm_gem_object *obj)
 {
+	struct drm_device *dev = obj->dev;
+	bool final = false;
+
 	if (WARN_ON(obj->handle_count == 0))
 		return;
 
···
 	* checked for a name
 	*/
 
-	mutex_lock(&obj->dev->object_name_lock);
+	mutex_lock(&dev->object_name_lock);
 	if (--obj->handle_count == 0) {
 		drm_gem_object_handle_free(obj);
 		drm_gem_object_exported_dma_buf_free(obj);
+		final = true;
 	}
-	mutex_unlock(&obj->dev->object_name_lock);
+	mutex_unlock(&dev->object_name_lock);
 
-	drm_gem_object_unreference_unlocked(obj);
+	if (final)
+		drm_gem_object_unreference_unlocked(obj);
+}
+
+/*
+ * Called at device or object close to release the file's
+ * handle references on objects.
+ */
+static int
+drm_gem_object_release_handle(int id, void *ptr, void *data)
+{
+	struct drm_file *file_priv = data;
+	struct drm_gem_object *obj = ptr;
+	struct drm_device *dev = obj->dev;
+
+	if (drm_core_check_feature(dev, DRIVER_PRIME))
+		drm_gem_remove_prime_handles(obj, file_priv);
+	drm_vma_node_revoke(&obj->vma_node, file_priv->filp);
+
+	if (dev->driver->gem_close_object)
+		dev->driver->gem_close_object(obj, file_priv);
+
+	drm_gem_object_handle_unreference_unlocked(obj);
+
+	return 0;
 }
 
 /**
···
 	idr_remove(&filp->object_idr, handle);
 	spin_unlock(&filp->table_lock);
 
-	if (drm_core_check_feature(dev, DRIVER_PRIME))
-		drm_gem_remove_prime_handles(obj, filp);
-	drm_vma_node_revoke(&obj->vma_node, filp->filp);
-
-	if (dev->driver->gem_close_object)
-		dev->driver->gem_close_object(obj, filp);
-	drm_gem_object_handle_unreference_unlocked(obj);
-
+	drm_gem_object_release_handle(handle, obj, filp);
 	return 0;
 }
 EXPORT_SYMBOL(drm_gem_handle_delete);
···
 			   u32 *handlep)
 {
 	struct drm_device *dev = obj->dev;
+	u32 handle;
 	int ret;
 
 	WARN_ON(!mutex_is_locked(&dev->object_name_lock));
+	if (obj->handle_count++ == 0)
+		drm_gem_object_reference(obj);
 
 	/*
 	 * Get the user-visible handle using idr.  Preload and perform
···
 	spin_lock(&file_priv->table_lock);
 
 	ret = idr_alloc(&file_priv->object_idr, obj, 1, 0, GFP_NOWAIT);
-	drm_gem_object_reference(obj);
-	obj->handle_count++;
+
 	spin_unlock(&file_priv->table_lock);
 	idr_preload_end();
+
 	mutex_unlock(&dev->object_name_lock);
-	if (ret < 0) {
-		drm_gem_object_handle_unreference_unlocked(obj);
-		return ret;
-	}
-	*handlep = ret;
+	if (ret < 0)
+		goto err_unref;
+
+	handle = ret;
 
 	ret = drm_vma_node_allow(&obj->vma_node, file_priv->filp);
-	if (ret) {
-		drm_gem_handle_delete(file_priv, *handlep);
-		return ret;
-	}
+	if (ret)
+		goto err_remove;
 
 	if (dev->driver->gem_open_object) {
 		ret = dev->driver->gem_open_object(obj, file_priv);
-		if (ret) {
-			drm_gem_handle_delete(file_priv, *handlep);
-			return ret;
-		}
+		if (ret)
+			goto err_revoke;
 	}
 
+	*handlep = handle;
 	return 0;
+
+err_revoke:
+	drm_vma_node_revoke(&obj->vma_node, file_priv->filp);
+err_remove:
+	spin_lock(&file_priv->table_lock);
+	idr_remove(&file_priv->object_idr, handle);
+	spin_unlock(&file_priv->table_lock);
+err_unref:
+	drm_gem_object_handle_unreference_unlocked(obj);
+	return ret;
 }
 
 /**
···
 		return -ENOENT;
 
 	mutex_lock(&dev->object_name_lock);
-	idr_preload(GFP_KERNEL);
 	/* prevent races with concurrent gem_close. */
 	if (obj->handle_count == 0) {
 		ret = -ENOENT;
···
 	}
 
 	if (!obj->name) {
-		ret = idr_alloc(&dev->object_name_idr, obj, 1, 0, GFP_NOWAIT);
+		ret = idr_alloc(&dev->object_name_idr, obj, 1, 0, GFP_KERNEL);
 		if (ret < 0)
 			goto err;
 
···
 	ret = 0;
 
 err:
-	idr_preload_end();
 	mutex_unlock(&dev->object_name_lock);
 	drm_gem_object_unreference_unlocked(obj);
 	return ret;
···
 {
 	idr_init(&file_private->object_idr);
 	spin_lock_init(&file_private->table_lock);
-}
-
-/*
- * Called at device close to release the file's
- * handle references on objects.
- */
-static int
-drm_gem_object_release_handle(int id, void *ptr, void *data)
-{
-	struct drm_file *file_priv = data;
-	struct drm_gem_object *obj = ptr;
-	struct drm_device *dev = obj->dev;
-
-	if (drm_core_check_feature(dev, DRIVER_PRIME))
-		drm_gem_remove_prime_handles(obj, file_priv);
-	drm_vma_node_revoke(&obj->vma_node, file_priv->filp);
-
-	if (dev->driver->gem_close_object)
-		dev->driver->gem_close_object(obj, file_priv);
-
-	drm_gem_object_handle_unreference_unlocked(obj);
-
-	return 0;
 }
 
 /**

drivers/gpu/drm/drm_irq.c

···
 module_param_named(vblankoffdelay, drm_vblank_offdelay, int, 0600);
 module_param_named(timestamp_precision_usec, drm_timestamp_precision, int, 0600);
 module_param_named(timestamp_monotonic, drm_timestamp_monotonic, int, 0600);
+MODULE_PARM_DESC(vblankoffdelay, "Delay until vblank irq auto-disable [msecs] (0: never disable, <0: disable immediately)");
+MODULE_PARM_DESC(timestamp_precision_usec, "Max. error on timestamps [usecs]");
+MODULE_PARM_DESC(timestamp_monotonic, "Use monotonic timestamps");
 
 static void store_vblank(struct drm_device *dev, unsigned int pipe,
 			 u32 vblank_count_inc,

drivers/gpu/drm/drm_sysfs.c

···
 			 struct bin_attribute *attr, char *buf, loff_t off,
 			 size_t count)
 {
-	struct device *connector_dev = container_of(kobj, struct device, kobj);
+	struct device *connector_dev = kobj_to_dev(kobj);
 	struct drm_connector *connector = to_drm_connector(connector_dev);
 	unsigned char *edid;
 	size_t size;

drivers/gpu/drm/gma500/power.c

···
  */
 int gma_power_suspend(struct device *_dev)
 {
-	struct pci_dev *pdev = container_of(_dev, struct pci_dev, dev);
+	struct pci_dev *pdev = to_pci_dev(_dev);
 	struct drm_device *dev = pci_get_drvdata(pdev);
 	struct drm_psb_private *dev_priv = dev->dev_private;
 
···
  */
 int gma_power_resume(struct device *_dev)
 {
-	struct pci_dev *pdev = container_of(_dev, struct pci_dev, dev);
+	struct pci_dev *pdev = to_pci_dev(_dev);
 	struct drm_device *dev = pci_get_drvdata(pdev);
 
 	mutex_lock(&power_mutex);

drivers/gpu/drm/i2c/tda998x_drv.c

···
 	if (ret)
 		goto err_sysfs;
 
-	priv->connector.encoder = &priv->encoder;
 	drm_mode_connector_attach_encoder(&priv->connector, &priv->encoder);
 
 	return 0;

drivers/gpu/drm/i915/i915_dma.c

···
 	if (ret)
 		goto cleanup_gem_stolen;
 
+	intel_setup_gmbus(dev);
+
 	/* Important: The output setup functions called by modeset_init need
 	 * working irqs for e.g. gmbus and dp aux transfers. */
 	intel_modeset_init(dev);
···
 cleanup_irq:
 	intel_guc_ucode_fini(dev);
 	drm_irq_uninstall(dev);
+	intel_teardown_gmbus(dev);
 cleanup_gem_stolen:
 	i915_gem_cleanup_stolen(dev);
 cleanup_vga_switcheroo:
···
 
 	/* Try to make sure MCHBAR is enabled before poking at it */
 	intel_setup_mchbar(dev);
-	intel_setup_gmbus(dev);
 	intel_opregion_setup(dev);
 
 	i915_gem_load(dev);
···
 	if (dev->pdev->msi_enabled)
 		pci_disable_msi(dev->pdev);
 
-	intel_teardown_gmbus(dev);
 	intel_teardown_mchbar(dev);
 	pm_qos_remove_request(&dev_priv->pm_qos);
 	destroy_workqueue(dev_priv->gpu_error.hangcheck_wq);
···
 
 	intel_csr_ucode_fini(dev_priv);
 
-	intel_teardown_gmbus(dev);
 	intel_teardown_mchbar(dev);
 
 	destroy_workqueue(dev_priv->hotplug.dp_wq);

drivers/gpu/drm/i915/intel_display.c

···
 
 int intel_connector_init(struct intel_connector *connector)
 {
-	struct drm_connector_state *connector_state;
+	drm_atomic_helper_connector_reset(&connector->base);
 
-	connector_state = kzalloc(sizeof *connector_state, GFP_KERNEL);
-	if (!connector_state)
+	if (!connector->base.state)
 		return -ENOMEM;
 
-	connector->base.state = connector_state;
 	return 0;
 }
 
···
 		WARN_ON(drm_atomic_set_mode_for_crtc(crtc->base.state, NULL) < 0);
 		crtc->base.state->active = crtc->active;
 		crtc->base.enabled = crtc->active;
+		crtc->base.state->connector_mask = 0;
 
 		/* Because we only establish the connector -> encoder ->
 		 * crtc links if something is active, this means the
···
 	for_each_intel_connector(dev, connector) {
 		if (connector->get_hw_state(connector)) {
 			connector->base.dpms = DRM_MODE_DPMS_ON;
-			connector->base.encoder = &connector->encoder->base;
+
+			encoder = connector->encoder;
+			connector->base.encoder = &encoder->base;
+
+			if (encoder->base.crtc &&
+			    encoder->base.crtc->state->active) {
+				/*
+				 * This has to be done during hardware readout
+				 * because anything calling .crtc_disable may
+				 * rely on the connector_mask being accurate.
+				 */
+				encoder->base.crtc->state->connector_mask |=
+					1 << drm_connector_index(&connector->base);
+			}
+
 		} else {
 			connector->base.dpms = DRM_MODE_DPMS_OFF;
 			connector->base.encoder = NULL;
···
 	mutex_lock(&dev->struct_mutex);
 	intel_cleanup_gt_powersave(dev);
 	mutex_unlock(&dev->struct_mutex);
+
+	intel_teardown_gmbus(dev);
 }
 
 /*

drivers/gpu/drm/i915/intel_dp_mst.c

···
 	drm_kms_helper_hotplug_event(dev);
 }
 
-static struct drm_dp_mst_topology_cbs mst_cbs = {
+static const struct drm_dp_mst_topology_cbs mst_cbs = {
 	.add_connector = intel_dp_add_mst_connector,
 	.register_connector = intel_dp_register_mst_connector,
 	.destroy_connector = intel_dp_destroy_mst_connector,

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

···
 		dev_warn(drm->dev, "Invalid legacyfb_depth.  Defaulting to 16bpp\n");
 		legacyfb_depth = 16;
 	}
+	drm_helper_disable_unused_functions(drm);
 	imxdrm->fbhelper = drm_fbdev_cma_init(drm, legacyfb_depth,
 				drm->mode_config.num_crtc, MAX_CRTC);
 	if (IS_ERR(imxdrm->fbhelper)) {

drivers/gpu/drm/imx/parallel-display.c

···
 
 	drm_mode_connector_attach_encoder(&imxpd->connector, &imxpd->encoder);
 
-	imxpd->connector.encoder = &imxpd->encoder;
-
 	return 0;
 }
 

drivers/gpu/drm/radeon/radeon_dp_mst.c

···
 	drm_kms_helper_hotplug_event(dev);
 }
 
-struct drm_dp_mst_topology_cbs mst_cbs = {
+const struct drm_dp_mst_topology_cbs mst_cbs = {
 	.add_connector = radeon_dp_add_mst_connector,
 	.register_connector = radeon_dp_register_mst_connector,
 	.destroy_connector = radeon_dp_destroy_mst_connector,

drivers/gpu/drm/shmobile/shmob_drm_crtc.c

···
 	if (ret < 0)
 		goto err_backlight;
 
-	connector->encoder = encoder;
-
 	drm_helper_connector_dpms(connector, DRM_MODE_DPMS_OFF);
 	drm_object_property_set_value(&connector->base,
 		sdev->ddev->mode_config.dpms_property, DRM_MODE_DPMS_OFF);

drivers/gpu/drm/sti/sti_drv.c

···
 
 	drm_mode_config_reset(dev);
 
+	drm_helper_disable_unused_functions(dev);
 	drm_fbdev_cma_init(dev, 32,
 			   dev->mode_config.num_crtc,
 			   dev->mode_config.num_connector);

drivers/gpu/drm/tegra/dsi.c

···
 
 static void tegra_dsi_connector_reset(struct drm_connector *connector)
 {
-	struct tegra_dsi_state *state;
+	struct tegra_dsi_state *state =
+		kzalloc(sizeof(*state), GFP_KERNEL);
 
-	kfree(connector->state);
-	connector->state = NULL;
-
-	state = kzalloc(sizeof(*state), GFP_KERNEL);
-	if (state)
-		connector->state = &state->base;
+	if (state) {
+		kfree(connector->state);
+		__drm_atomic_helper_connector_reset(connector, &state->base);
+	}
 }
 
 static struct drm_connector_state *

drivers/gpu/drm/tilcdc/tilcdc_drv.c

···
 			break;
 	}
 
+	drm_helper_disable_unused_functions(dev);
 	priv->fbdev = drm_fbdev_cma_init(dev, bpp,
 			dev->mode_config.num_crtc,
 			dev->mode_config.num_connector);

drivers/gpu/drm/vc4/vc4_crtc.c

···
 	/* The pixelvalve can only feed one encoder (and encoders are
 	 * 1:1 with connectors.)
 	 */
-	if (drm_atomic_connectors_for_crtc(state->state, crtc) > 1)
+	if (hweight32(state->connector_mask) > 1)
 		return -EINVAL;
 
 	drm_atomic_crtc_state_for_each_plane(plane, state) {

drivers/gpu/vga/vga_switcheroo.c

···
  * for the inactive GPU.) Also, muxes are often used to cut power to the
  * discrete GPU while it is not used.
  *
- * DRM drivers register GPUs with vga_switcheroo, these are heretoforth called
+ * DRM drivers register GPUs with vga_switcheroo, these are henceforth called
  * clients. The mux is called the handler. Muxless machines also register a
  * handler to control the power state of the discrete GPU, its ->switchto
  * callback is a no-op for obvious reasons. The discrete GPU is often equipped

drivers/platform/x86/apple-gmux.c

···
  *
  *  Copyright (C) Canonical Ltd. <[e-mail]>
  *  Copyright (C) 2010-2012 Andreas Heider <[e-mail]>
+ *  Copyright (C) 2015 Lukas Wunner <[e-mail]>
  *
  *  This program is free software; you can redistribute it and/or modify
  *  it under the terms of the GNU General Public License version 2 as
···
 #include <linux/vgaarb.h>
 #include <acpi/video.h>
 #include <asm/io.h>
+
+/**
+ * DOC: Overview
+ *
+ * :1:  http://www.latticesemi.com/en/Products/FPGAandCPLD/LatticeXP2.aspx
+ * :2:  http://www.renesas.com/products/mpumcu/h8s/h8s2100/h8s2113/index.jsp
+ *
+ * gmux is a microcontroller built into the MacBook Pro to support dual GPUs:
+ * A {1}[Lattice XP2] on pre-retinas, a {2}[Renesas R4F2113] on retinas.
+ *
+ * (The MacPro6,1 2013 also has a gmux, however it is unclear why since it has
+ * dual GPUs but no built-in display.)
+ *
+ * gmux is connected to the LPC bus of the southbridge. Its I/O ports are
+ * accessed differently depending on the microcontroller: Driver functions
+ * to access a pre-retina gmux are infixed `_pio_`, those for a retina gmux
+ * are infixed `_index_`.
+ */
 
 struct apple_gmux_data {
 	unsigned long iostart;
···
 	return false;
 }
 
+/**
+ * DOC: Backlight control
+ *
+ * :3:  http://www.ti.com/lit/ds/symlink/lp8543.pdf
+ * :4:  http://www.ti.com/lit/ds/symlink/lp8545.pdf
+ *
+ * On single GPU MacBooks, the PWM signal for the backlight is generated by
+ * the GPU. On dual GPU MacBook Pros by contrast, either GPU may be suspended
+ * to conserve energy. Hence the PWM signal needs to be generated by a separate
+ * backlight driver which is controlled by gmux. The earliest generation
+ * MBP5 2008/09 uses a {3}[TI LP8543] backlight driver. All newer models
+ * use a {4}[TI LP8545].
+ */
+
 static int gmux_get_brightness(struct backlight_device *bd)
 {
 	struct apple_gmux_data *gmux_data = bl_get_data(bd);
···
 	.update_status = gmux_update_status,
 };
 
+/**
+ * DOC: Graphics mux
+ *
+ * :5:  http://pimg-fpiw.uspto.gov/fdd/07/870/086/0.pdf
+ * :6:  http://www.nxp.com/documents/data_sheet/CBTL06141.pdf
+ * :7:  http://www.ti.com/lit/ds/symlink/hd3ss212.pdf
+ * :8:  https://www.pericom.com/assets/Datasheets/PI3VDP12412.pdf
+ * :9:  http://www.ti.com/lit/ds/symlink/sn74lv4066a.pdf
+ * :10: http://pdf.datasheetarchive.com/indexerfiles/Datasheets-SW16/DSASW00308511.pdf
+ * :11: http://www.ti.com/lit/ds/symlink/ts3ds10224.pdf
+ *
+ * On pre-retinas, the LVDS outputs of both GPUs feed into gmux which muxes
+ * either of them to the panel. One of the tricks gmux has up its sleeve is
+ * to lengthen the blanking interval of its output during a switch to
+ * synchronize it with the GPU switched to. This allows for a flicker-free
+ * switch that is imperceptible by the user ({5}[US 8,687,007 B2]).
+ *
+ * On retinas, muxing is no longer done by gmux itself, but by a separate
+ * chip which is controlled by gmux. The chip is triple sourced, it is
+ * either an {6}[NXP CBTL06142], {7}[TI HD3SS212] or {8}[Pericom PI3VDP12412].
+ * The panel is driven with eDP instead of LVDS since the pixel clock
+ * required for retina resolution exceeds LVDS' limits.
+ *
+ * Pre-retinas are able to switch the panel's DDC pins separately.
+ * This is handled by a {9}[TI SN74LV4066A] which is controlled by gmux.
+ * The inactive GPU can thus probe the panel's EDID without switching over
+ * the entire panel. Retinas lack this functionality as the chips used for
+ * eDP muxing are incapable of switching the AUX channel separately (see
+ * the linked data sheets, Pericom would be capable but this is unused).
+ * However the retina panel has the NO_AUX_HANDSHAKE_LINK_TRAINING bit set
+ * in its DPCD, allowing the inactive GPU to skip the AUX handshake and
+ * set up the output with link parameters pre-calibrated by the active GPU.
+ *
+ * The external DP port is only fully switchable on the first two unibody
+ * MacBook Pro generations, MBP5 2008/09 and MBP6 2010. This is done by an
+ * {6}[NXP CBTL06141] which is controlled by gmux. It's the predecessor of the
+ * eDP mux on retinas, the difference being support for 2.7 versus 5.4 Gbit/s.
+ *
+ * The following MacBook Pro generations replaced the external DP port with a
+ * combined DP/Thunderbolt port and lost the ability to switch it between GPUs,
+ * connecting it either to the discrete GPU or the Thunderbolt controller.
+ * Oddly enough, while the full port is no longer switchable, AUX and HPD
+ * are still switchable by way of an {10}[NXP CBTL03062] (on pre-retinas
+ * MBP8 2011 and MBP9 2012) or two {11}[TI TS3DS10224] (on retinas) under the
+ * control of gmux. Since the integrated GPU is missing the main link,
+ * external displays appear to it as phantoms which fail to link-train.
+ *
+ * gmux receives the HPD signal of all display connectors and sends an
+ * interrupt on hotplug. On generations which cannot switch external ports,
+ * the discrete GPU can then be woken to drive the newly connected display.
+ * The ability to switch AUX on these generations could be used to improve
+ * reliability of hotplug detection by having the integrated GPU poll the
+ * ports while the discrete GPU is asleep, but currently we do not make use
+ * of this feature.
+ *
+ * gmux' initial switch state on bootup is user configurable via the EFI
+ * variable `gpu-power-prefs-fa4ce28d-b62f-4c99-9cc3-6815686e30f9` (5th byte,
+ * 1 = IGD, 0 = DIS). Based on this setting, the EFI firmware tells gmux to
+ * switch the panel and the external DP connector and allocates a framebuffer
+ * for the selected GPU.
+ */
+
 static int gmux_switchto(enum vga_switcheroo_client_id id)
 {
 	if (id == VGA_SWITCHEROO_IGD) {
···
 
 	return 0;
 }
+
+/**
+ * DOC: Power control
+ *
+ * gmux is able to cut power to the discrete GPU. It automatically takes care
+ * of the correct sequence to tear down and bring up the power rails for
+ * core voltage, VRAM and PCIe.
+ */
 
 static int gmux_set_discrete_state(struct apple_gmux_data *gmux_data,
 				   enum vga_switcheroo_state state)
···
 	.power_state = gmux_set_power_state,
 	.get_client_id = gmux_get_client_id,
 };
+
+/**
+ * DOC: Interrupt
+ *
+ * gmux is also connected to a GPIO pin of the southbridge and thereby is able
+ * to trigger an ACPI GPE. On the MBP5 2008/09 it's GPIO pin 22 of the Nvidia
+ * MCP79, on all following generations it's GPIO pin 6 of the Intel PCH.
+ * The GPE merely signals that an interrupt occurred, the actual type of event
+ * is identified by reading a gmux register.
+ */
 
 static inline void gmux_disable_interrupts(struct apple_gmux_data *gmux_data)
 {

include/drm/drm_atomic.h

···
 drm_atomic_add_affected_planes(struct drm_atomic_state *state,
 			       struct drm_crtc *crtc);
 
-int
-drm_atomic_connectors_for_crtc(struct drm_atomic_state *state,
-			       struct drm_crtc *crtc);
-
 void drm_atomic_legacy_backoff(struct drm_atomic_state *state);
 
 void

include/drm/drm_atomic_helper.h

···
 void drm_atomic_helper_plane_destroy_state(struct drm_plane *plane,
 					  struct drm_plane_state *state);
 
+void __drm_atomic_helper_connector_reset(struct drm_connector *connector,
+					 struct drm_connector_state *conn_state);
 void drm_atomic_helper_connector_reset(struct drm_connector *connector);
 void
 __drm_atomic_helper_connector_duplicate_state(struct drm_connector *connector,

include/drm/drm_crtc.h

···
  * @active_changed: crtc_state->active has been toggled.
  * @connectors_changed: connectors to this crtc have been updated
  * @plane_mask: bitmask of (1 << drm_plane_index(plane)) of attached planes
+ * @connector_mask: bitmask of (1 << drm_connector_index(connector)) of attached connectors
  * @last_vblank_count: for helpers and drivers to capture the vblank of the
  * 	update to ensure framebuffer cleanup isn't done too early
  * @adjusted_mode: for use by helpers and drivers to compute adjusted mode timings
···
 	 * on plane_mask being accurate!
 	 */
 	u32 plane_mask;
+
+	u32 connector_mask;
 
 	/* last_vblank_count: for vblank waits before cleanup */
 	u32 last_vblank_count;
···
 	 * ->page_flip() operation is already pending the callback should return
 	 * -EBUSY. Pageflips on a disabled CRTC (either by setting a NULL mode
 	 * or just runtime disabled through DPMS respectively the new atomic
-	 * "ACTIVE" state) should result in an -EINVAL error code.
+	 * "ACTIVE" state) should result in an -EINVAL error code. Note that
+	 * drm_atomic_helper_page_flip() checks this already for atomic drivers.
 	 */
 	int (*page_flip)(struct drm_crtc *crtc,
 			 struct drm_framebuffer *fb,

include/drm/drm_dp_mst_helper.h

···
 struct drm_dp_mst_topology_mgr {
 
 	struct device *dev;
-	struct drm_dp_mst_topology_cbs *cbs;
+	const struct drm_dp_mst_topology_cbs *cbs;
 	int max_dpcd_transaction_bytes;
 	struct drm_dp_aux *aux; /* auxch for this topology mgr to use */
 	int max_payloads;

include/drm/drm_modeset_helper_vtables.h

···
 	 * Atomic drivers which need to inspect and adjust more state should
 	 * instead use the @atomic_check callback.
 	 *
+	 * Also beware that neither core nor helpers filter modes before
+	 * passing them to the driver: While the list of modes that is
+	 * advertised to userspace is filtered using the connector's
+	 * ->mode_valid() callback, neither the core nor the helpers do any
+	 * filtering on modes passed in from userspace when setting a mode. It
+	 * is therefore possible for userspace to pass in a mode that was
+	 * previously filtered out using ->mode_valid() or add a custom mode
+	 * that wasn't probed from EDID or similar to begin with.  Even though
+	 * this is an advanced feature and rarely used nowadays, some users rely
+	 * on being able to specify modes manually so drivers must be prepared
+	 * to deal with it. Specifically this means that all drivers need not
+	 * only validate modes in ->mode_valid() but also in ->mode_fixup() to
+	 * make sure invalid modes passed in from userspace are rejected.
+	 *
 	 * RETURNS:
 	 *
 	 * True if an acceptable configuration is possible, false if the modeset
···
 	 * This callback is used by the legacy CRTC helpers to set a new
 	 * framebuffer and scanout position. It is optional and used as an
 	 * optimized fast-path instead of a full mode set operation with all the
-	 * resulting flickering. Since it can't update other planes it's
+	 * resulting flickering. If it is not present
+	 * drm_crtc_helper_set_config() will fall back to a full modeset, using
+	 * the ->mode_set() callback. Since it can't update other planes it's
 	 * incompatible with atomic modeset support.
 	 *
 	 * This callback is only used by the CRTC helpers and deprecated.
···
 	 * Atomic drivers which need to inspect and adjust more state should
 	 * instead use the @atomic_check callback.
 	 *
+	 * Also beware that neither core nor helpers filter modes before
+	 * passing them to the driver: While the list of modes that is
+	 * advertised to userspace is filtered using the connector's
+	 * ->mode_valid() callback, neither the core nor the helpers do any
+	 * filtering on modes passed in from userspace when setting a mode. It
+	 * is therefore possible for userspace to pass in a mode that was
+	 * previously filtered out using ->mode_valid() or add a custom mode
+	 * that wasn't probed from EDID or similar to begin with.  Even though
+	 * this is an advanced feature and rarely used nowadays, some users rely
+	 * on being able to specify modes manually so drivers must be prepared
+	 * to deal with it. Specifically this means that all drivers need not
+	 * only validate modes in ->mode_valid() but also in ->mode_fixup() to
+	 * make sure invalid modes passed in from userspace are rejected.
+	 *
 	 * RETURNS:
 	 *
 	 * True if an acceptable configuration is possible, false if the modeset
···
 	 * In this function drivers then parse the modes in the EDID and add
 	 * them by calling drm_add_edid_modes(). But connectors that driver a
 	 * fixed panel can also manually add specific modes using
-	 * drm_mode_probed_add(). Finally drivers that support audio probably
-	 * want to update the ELD data, too, using drm_edid_to_eld().
+	 * drm_mode_probed_add(). Drivers which manually add modes should also
+	 * make sure that the @display_info, @width_mm and @height_mm fields of the
+	 * struct #drm_connector are filled in.
+	 *
+	 * Virtual drivers that just want some standard VESA mode with a given
+	 * resolution can call drm_add_modes_noedid(), and mark the preferred
+	 * one using drm_set_preferred_mode().
+	 *
+	 * Finally drivers that support audio probably want to update the ELD
+	 * data, too, using drm_edid_to_eld().
 	 *
 	 * This function is only called after the ->detect() hook has indicated
 	 * that a sink is connected and when the EDID isn't overridden through