Merge tag 'drm-intel-next-2019-12-23' of git://anongit.freedesktop.org/drm/drm-intel into drm-next

i915 features for v5.6:

- Separate hardware and uapi state (Maarten)

- Expose a number of sprite and plane formats (Ville)

- DDC symlink in HDMI connector sysfs directory (Andrzej Pietrasiewicz)

- Improve obj->mm.lock nesting lock annotation (Daniel)
  (Includes lockdep changes)

- Selftest improvements across the board (Chris)

- ICL/TGL VDSC support on DSI (Jani, Vandita)

- TGL DSB fixes (Animesh, Lucas, Tvrtko)

- VBT parsing improvements and fixes (Lucas, Matt, José, Jani, Dan Carpenter)

- Fix LPSS vs. PMIC PWM backlight use on BYT/CHT (Hans)
  (Includes ACPI+MFD changes)

- Display state, crtc, plane code refactoring (Ville)

- Set opregion chpd value to indicate the driver handles hotplug (Hans de Goede)

- DSI updates and fixes, TGL pipe D support, port mapping (José, Jani, Vandita)

- Make HDCP 2.2 support cover CFL (Juston Li)

- Fix CML PCI IDs and ULT (Shawn Lee)

- CMP-V PCH fix (Imre)

- TGL: Add another TGL PCH ID (James)

- EHL/JSL: Add new PCI IDs (James)

- Rename pipe update tracepoints (Ville)

- Fix FBC on GLK+ (Ville)

- GuC fixes and improvements (Daniele, Don Hiatt, Stuart Summers, Matthew Brost)

- Display debugfs improvements (Ville)

- Hotplug/irq fixes (Matt)

- PSR fixes and improvements (José)

- DRM_I915_GEM_MMAP_OFFSET ioctl (Abdiel)

- Static analysis fixes (Colin Ian King)

- Register sysctl path globally (Venkata Sandeep Dhanalakota)

- Introduce new macros for tracing (Venkata Sandeep Dhanalakota)

- Migrate gt towards intel_uncore_read/write (Andi)

- Add rps frequency translation helpers (Andi)

- Fix TGL transcoder clock off sequence (José)

- Fix TGL port A audio (Kai Vehmanen)

- TGL render decompression (DK)

- GEM/GT improvements and fixes across the board (Chris)

- Couple of backmerges (Jani)

Signed-off-by: Dave Airlie <airlied@redhat.com>

# gpg: Signature made Tue 24 Dec 2019 03:20:48 AM AEST
# gpg:                using RSA key D398079D26ABEE6F
# gpg: Good signature from "Jani Nikula <jani.nikula@intel.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 1565 A65B 77B0 632E 1124  E59C D398 079D 26AB EE6F

# Conflicts:
#	drivers/gpu/drm/i915/display/intel_fbc.c
#	drivers/gpu/drm/i915/gt/intel_lrc.c
#	drivers/gpu/drm/i915/i915_gem.c
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/87lfr3rkry.fsf@intel.com

232 files changed, 12356 insertions, 7834 deletions

diff --git a/drivers/acpi/acpi_lpss.c b/drivers/acpi/acpi_lpss.c
index 70f740b09684..db18df6cb330 100644
--- a/drivers/acpi/acpi_lpss.c
+++ b/drivers/acpi/acpi_lpss.c
@@ -69,10 +69,6 @@ ACPI_MODULE_NAME("acpi_lpss");
 #define LPSS_SAVE_CTX			BIT(4)
 #define LPSS_NO_D3_DELAY		BIT(5)
 
-/* Crystal Cove PMIC shares same ACPI ID between different platforms */
-#define BYT_CRC_HRV			2
-#define CHT_CRC_HRV			3
-
 struct lpss_private_data;
 
 struct lpss_device_desc {
@@ -158,7 +154,7 @@ static void lpss_deassert_reset(struct lpss_private_data *pdata)
  */
 static struct pwm_lookup byt_pwm_lookup[] = {
 	PWM_LOOKUP_WITH_MODULE("80860F09:00", 0, "0000:00:02.0",
-			       "pwm_backlight", 0, PWM_POLARITY_NORMAL,
+			       "pwm_soc_backlight", 0, PWM_POLARITY_NORMAL,
 			       "pwm-lpss-platform"),
 };
 
@@ -170,8 +166,7 @@ static void byt_pwm_setup(struct lpss_private_data *pdata)
 	if (!adev->pnp.unique_id || strcmp(adev->pnp.unique_id, "1"))
 		return;
 
-	if (!acpi_dev_present("INT33FD", NULL, BYT_CRC_HRV))
-		pwm_add_table(byt_pwm_lookup, ARRAY_SIZE(byt_pwm_lookup));
+	pwm_add_table(byt_pwm_lookup, ARRAY_SIZE(byt_pwm_lookup));
 }
 
 #define LPSS_I2C_ENABLE			0x6c
@@ -204,7 +199,7 @@ static void byt_i2c_setup(struct lpss_private_data *pdata)
 /* BSW PWM used for backlight control by the i915 driver */
 static struct pwm_lookup bsw_pwm_lookup[] = {
 	PWM_LOOKUP_WITH_MODULE("80862288:00", 0, "0000:00:02.0",
-			       "pwm_backlight", 0, PWM_POLARITY_NORMAL,
+			       "pwm_soc_backlight", 0, PWM_POLARITY_NORMAL,
 			       "pwm-lpss-platform"),
 };
 
diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index d3e3ea3b5843..7041323a7bff 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -54,6 +54,9 @@ config DRM_DEBUG_MM
 
 	  If in doubt, say "N".
 
+config DRM_EXPORT_FOR_TESTS
+	bool
+
 config DRM_DEBUG_SELFTEST
 	tristate "kselftests for DRM"
 	depends on DRM
@@ -61,6 +64,7 @@ config DRM_DEBUG_SELFTEST
 	select PRIME_NUMBERS
 	select DRM_LIB_RANDOM
 	select DRM_KMS_HELPER
+	select DRM_EXPORT_FOR_TESTS if m
 	default n
 	help
 	  This option provides kernel modules that can be used to run
diff --git a/drivers/gpu/drm/drm_atomic_state_helper.c b/drivers/gpu/drm/drm_atomic_state_helper.c
index d0a937fb0c56..7cf3cf936547 100644
--- a/drivers/gpu/drm/drm_atomic_state_helper.c
+++ b/drivers/gpu/drm/drm_atomic_state_helper.c
@@ -58,6 +58,22 @@
  */
 
 /**
+ * __drm_atomic_helper_crtc_state_reset - reset the CRTC state
+ * @crtc_state: atomic CRTC state, must not be NULL
+ * @crtc: CRTC object, must not be NULL
+ *
+ * Initializes the newly allocated @crtc_state with default
+ * values. This is useful for drivers that subclass the CRTC state.
+ */
+void
+__drm_atomic_helper_crtc_state_reset(struct drm_crtc_state *crtc_state,
+				     struct drm_crtc *crtc)
+{
+	crtc_state->crtc = crtc;
+}
+EXPORT_SYMBOL(__drm_atomic_helper_crtc_state_reset);
+
+/**
  * __drm_atomic_helper_crtc_reset - reset state on CRTC
  * @crtc: drm CRTC
  * @crtc_state: CRTC state to assign
@@ -74,7 +90,7 @@ __drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
 			       struct drm_crtc_state *crtc_state)
 {
 	if (crtc_state)
-		crtc_state->crtc = crtc;
+		__drm_atomic_helper_crtc_state_reset(crtc_state, crtc);
 
 	crtc->state = crtc_state;
 }
@@ -212,23 +228,43 @@ void drm_atomic_helper_crtc_destroy_state(struct drm_crtc *crtc,
 EXPORT_SYMBOL(drm_atomic_helper_crtc_destroy_state);
 
 /**
- * __drm_atomic_helper_plane_reset - resets planes state to default values
+ * __drm_atomic_helper_plane_state_reset - resets plane state to default values
+ * @plane_state: atomic plane state, must not be NULL
  * @plane: plane object, must not be NULL
- * @state: atomic plane state, must not be NULL
  *
- * Initializes plane state to default. This is useful for drivers that subclass
- * the plane state.
+ * Initializes the newly allocated @plane_state with default
+ * values. This is useful for drivers that subclass the CRTC state.
  */
-void __drm_atomic_helper_plane_reset(struct drm_plane *plane,
-				     struct drm_plane_state *state)
+void __drm_atomic_helper_plane_state_reset(struct drm_plane_state *plane_state,
+					   struct drm_plane *plane)
 {
-	state->plane = plane;
-	state->rotation = DRM_MODE_ROTATE_0;
+	plane_state->plane = plane;
+	plane_state->rotation = DRM_MODE_ROTATE_0;
 
-	state->alpha = DRM_BLEND_ALPHA_OPAQUE;
-	state->pixel_blend_mode = DRM_MODE_BLEND_PREMULTI;
+	plane_state->alpha = DRM_BLEND_ALPHA_OPAQUE;
+	plane_state->pixel_blend_mode = DRM_MODE_BLEND_PREMULTI;
+}
+EXPORT_SYMBOL(__drm_atomic_helper_plane_state_reset);
 
-	plane->state = state;
+/**
+ * __drm_atomic_helper_plane_reset - reset state on plane
+ * @plane: drm plane
+ * @plane_state: plane state to assign
+ *
+ * Initializes the newly allocated @plane_state and assigns it to
+ * the &drm_crtc->state pointer of @plane, usually required when
+ * initializing the drivers or when called from the &drm_plane_funcs.reset
+ * hook.
+ *
+ * This is useful for drivers that subclass the plane state.
+ */
+void __drm_atomic_helper_plane_reset(struct drm_plane *plane,
+				     struct drm_plane_state *plane_state)
+{
+	if (plane_state)
+		__drm_atomic_helper_plane_state_reset(plane_state, plane);
+
+	plane->state = plane_state;
 }
 EXPORT_SYMBOL(__drm_atomic_helper_plane_reset);
 
@@ -336,6 +372,22 @@ void drm_atomic_helper_plane_destroy_state(struct drm_plane *plane,
 EXPORT_SYMBOL(drm_atomic_helper_plane_destroy_state);
 
 /**
+ * __drm_atomic_helper_connector_state_reset - reset the connector state
+ * @conn_state: atomic connector state, must not be NULL
+ * @connector: connectotr object, must not be NULL
+ *
+ * Initializes the newly allocated @conn_state with default
+ * values. This is useful for drivers that subclass the connector state.
+ */
+void
+__drm_atomic_helper_connector_state_reset(struct drm_connector_state *conn_state,
+					  struct drm_connector *connector)
+{
+	conn_state->connector = connector;
+}
+EXPORT_SYMBOL(__drm_atomic_helper_connector_state_reset);
+
+/**
  * __drm_atomic_helper_connector_reset - reset state on connector
  * @connector: drm connector
  * @conn_state: connector state to assign
@@ -352,7 +404,7 @@ __drm_atomic_helper_connector_reset(struct drm_connector *connector,
 				    struct drm_connector_state *conn_state)
 {
 	if (conn_state)
-		conn_state->connector = connector;
+		__drm_atomic_helper_connector_state_reset(conn_state, connector);
 
 	connector->state = conn_state;
 }
diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c
index 8b5a06013ccc..92d16724f949 100644
--- a/drivers/gpu/drm/drm_file.c
+++ b/drivers/gpu/drm/drm_file.c
@@ -31,7 +31,9 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
+#include <linux/anon_inodes.h>
 #include <linux/dma-fence.h>
+#include <linux/file.h>
 #include <linux/module.h>
 #include <linux/pci.h>
 #include <linux/poll.h>
@@ -754,3 +756,43 @@ void drm_send_event(struct drm_device *dev, struct drm_pending_event *e)
 	spin_unlock_irqrestore(&dev->event_lock, irqflags);
 }
 EXPORT_SYMBOL(drm_send_event);
+
+/**
+ * mock_drm_getfile - Create a new struct file for the drm device
+ * @minor: drm minor to wrap (e.g. #drm_device.primary)
+ * @flags: file creation mode (O_RDWR etc)
+ *
+ * This create a new struct file that wraps a DRM file context around a
+ * DRM minor. This mimicks userspace opening e.g. /dev/dri/card0, but without
+ * invoking userspace. The struct file may be operated on using its f_op
+ * (the drm_device.driver.fops) to mimick userspace operations, or be supplied
+ * to userspace facing functions as an internal/anonymous client.
+ *
+ * RETURNS:
+ * Pointer to newly created struct file, ERR_PTR on failure.
+ */
+struct file *mock_drm_getfile(struct drm_minor *minor, unsigned int flags)
+{
+	struct drm_device *dev = minor->dev;
+	struct drm_file *priv;
+	struct file *file;
+
+	priv = drm_file_alloc(minor);
+	if (IS_ERR(priv))
+		return ERR_CAST(priv);
+
+	file = anon_inode_getfile("drm", dev->driver->fops, priv, flags);
+	if (IS_ERR(file)) {
+		drm_file_free(priv);
+		return file;
+	}
+
+	/* Everyone shares a single global address space */
+	file->f_mapping = dev->anon_inode->i_mapping;
+
+	drm_dev_get(dev);
+	priv->filp = file;
+
+	return file;
+}
+EXPORT_SYMBOL_FOR_TESTS_ONLY(mock_drm_getfile);
diff --git a/drivers/gpu/drm/i915/Kconfig.debug b/drivers/gpu/drm/i915/Kconfig.debug
index 438040ff0179..1cb28c20807c 100644
--- a/drivers/gpu/drm/i915/Kconfig.debug
+++ b/drivers/gpu/drm/i915/Kconfig.debug
@@ -27,6 +27,7 @@ config DRM_I915_DEBUG
 	select X86_MSR # used by igt/pm_rpm
 	select DRM_VGEM # used by igt/prime_vgem (dmabuf interop checks)
 	select DRM_DEBUG_MM if DRM=y
+	select DRM_EXPORT_FOR_TESTS if m
 	select DRM_DEBUG_SELFTEST
 	select DMABUF_SELFTESTS
 	select SW_SYNC # signaling validation framework (igt/syncobj*)
@@ -149,6 +150,7 @@ config DRM_I915_SELFTEST
 	bool "Enable selftests upon driver load"
 	depends on DRM_I915
 	default n
+	select DRM_EXPORT_FOR_TESTS if m
 	select FAULT_INJECTION
 	select PRIME_NUMBERS
 	help
diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile
index 90dcf09f52cc..b0c53661f62b 100644
--- a/drivers/gpu/drm/i915/Makefile
+++ b/drivers/gpu/drm/i915/Makefile
@@ -75,6 +75,9 @@ i915-$(CONFIG_PERF_EVENTS) += i915_pmu.o
 # "Graphics Technology" (aka we talk to the gpu)
 obj-y += gt/
 gt-y += \
+	gt/debugfs_engines.o \
+	gt/debugfs_gt.o \
+	gt/debugfs_gt_pm.o \
 	gt/intel_breadcrumbs.o \
 	gt/intel_context.o \
 	gt/intel_engine_cs.o \
@@ -259,6 +262,7 @@ i915-$(CONFIG_DRM_I915_SELFTEST) += \
 	selftests/i915_selftest.o \
 	selftests/igt_flush_test.o \
 	selftests/igt_live_test.o \
+	selftests/igt_mmap.o \
 	selftests/igt_reset.o \
 	selftests/igt_spinner.o
 
diff --git a/drivers/gpu/drm/i915/display/icl_dsi.c b/drivers/gpu/drm/i915/display/icl_dsi.c
index 325df29b0447..006b1a297e6f 100644
--- a/drivers/gpu/drm/i915/display/icl_dsi.c
+++ b/drivers/gpu/drm/i915/display/icl_dsi.c
@@ -34,6 +34,7 @@
 #include "intel_ddi.h"
 #include "intel_dsi.h"
 #include "intel_panel.h"
+#include "intel_vdsc.h"
 
 static inline int header_credits_available(struct drm_i915_private *dev_priv,
 					   enum transcoder dsi_trans)
@@ -276,7 +277,7 @@ static void configure_dual_link_mode(struct intel_encoder *encoder,
 
 	if (intel_dsi->dual_link == DSI_DUAL_LINK_FRONT_BACK) {
 		const struct drm_display_mode *adjusted_mode =
-					&pipe_config->base.adjusted_mode;
+					&pipe_config->hw.adjusted_mode;
 		u32 dss_ctl2;
 		u16 hactive = adjusted_mode->crtc_hdisplay;
 		u16 dl_buffer_depth;
@@ -301,18 +302,31 @@ static void configure_dual_link_mode(struct intel_encoder *encoder,
 	I915_WRITE(DSS_CTL1, dss_ctl1);
 }
 
-static void gen11_dsi_program_esc_clk_div(struct intel_encoder *encoder)
+/* aka DSI 8X clock */
+static int afe_clk(struct intel_encoder *encoder,
+		   const struct intel_crtc_state *crtc_state)
+{
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+	int bpp;
+
+	if (crtc_state->dsc.compression_enable)
+		bpp = crtc_state->dsc.compressed_bpp;
+	else
+		bpp = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format);
+
+	return DIV_ROUND_CLOSEST(intel_dsi->pclk * bpp, intel_dsi->lane_count);
+}
+
+static void gen11_dsi_program_esc_clk_div(struct intel_encoder *encoder,
+					  const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	enum port port;
-	u32 bpp = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format);
-	u32 afe_clk_khz; /* 8X Clock */
+	int afe_clk_khz;
 	u32 esc_clk_div_m;
 
-	afe_clk_khz = DIV_ROUND_CLOSEST(intel_dsi->pclk * bpp,
-					intel_dsi->lane_count);
-
+	afe_clk_khz = afe_clk(encoder, crtc_state);
 	esc_clk_div_m = DIV_ROUND_UP(afe_clk_khz, DSI_MAX_ESC_CLK);
 
 	for_each_dsi_port(port, intel_dsi->ports) {
@@ -490,7 +504,9 @@ static void gen11_dsi_enable_ddi_buffer(struct intel_encoder *encoder)
 	}
 }
 
-static void gen11_dsi_setup_dphy_timings(struct intel_encoder *encoder)
+static void
+gen11_dsi_setup_dphy_timings(struct intel_encoder *encoder,
+			     const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
@@ -531,7 +547,7 @@ static void gen11_dsi_setup_dphy_timings(struct intel_encoder *encoder)
 	 * leave all fields at HW default values.
 	 */
 	if (IS_GEN(dev_priv, 11)) {
-		if (intel_dsi_bitrate(intel_dsi) <= 800000) {
+		if (afe_clk(encoder, crtc_state) <= 800000) {
 			for_each_dsi_port(port, intel_dsi->ports) {
 				tmp = I915_READ(DPHY_TA_TIMING_PARAM(port));
 				tmp &= ~TA_SURE_MASK;
@@ -625,7 +641,7 @@ gen11_dsi_configure_transcoder(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	enum pipe pipe = intel_crtc->pipe;
 	u32 tmp;
 	enum port port;
@@ -641,7 +657,7 @@ gen11_dsi_configure_transcoder(struct intel_encoder *encoder,
 			tmp |= EOTP_DISABLED;
 
 		/* enable link calibration if freq > 1.5Gbps */
-		if (intel_dsi_bitrate(intel_dsi) >= 1500 * 1000) {
+		if (afe_clk(encoder, pipe_config) >= 1500 * 1000) {
 			tmp &= ~LINK_CALIBRATION_MASK;
 			tmp |= CALIBRATION_ENABLED_INITIAL_ONLY;
 		}
@@ -667,22 +683,26 @@ gen11_dsi_configure_transcoder(struct intel_encoder *encoder,
 
 		/* select pixel format */
 		tmp &= ~PIX_FMT_MASK;
-		switch (intel_dsi->pixel_format) {
-		default:
-			MISSING_CASE(intel_dsi->pixel_format);
-			/* fallthrough */
-		case MIPI_DSI_FMT_RGB565:
-			tmp |= PIX_FMT_RGB565;
-			break;
-		case MIPI_DSI_FMT_RGB666_PACKED:
-			tmp |= PIX_FMT_RGB666_PACKED;
-			break;
-		case MIPI_DSI_FMT_RGB666:
-			tmp |= PIX_FMT_RGB666_LOOSE;
-			break;
-		case MIPI_DSI_FMT_RGB888:
-			tmp |= PIX_FMT_RGB888;
-			break;
+		if (pipe_config->dsc.compression_enable) {
+			tmp |= PIX_FMT_COMPRESSED;
+		} else {
+			switch (intel_dsi->pixel_format) {
+			default:
+				MISSING_CASE(intel_dsi->pixel_format);
+				/* fallthrough */
+			case MIPI_DSI_FMT_RGB565:
+				tmp |= PIX_FMT_RGB565;
+				break;
+			case MIPI_DSI_FMT_RGB666_PACKED:
+				tmp |= PIX_FMT_RGB666_PACKED;
+				break;
+			case MIPI_DSI_FMT_RGB666:
+				tmp |= PIX_FMT_RGB666_LOOSE;
+				break;
+			case MIPI_DSI_FMT_RGB888:
+				tmp |= PIX_FMT_RGB888;
+				break;
+			}
 		}
 
 		if (INTEL_GEN(dev_priv) >= 12) {
@@ -745,6 +765,9 @@ gen11_dsi_configure_transcoder(struct intel_encoder *encoder,
 		case PIPE_C:
 			tmp |= TRANS_DDI_EDP_INPUT_C_ONOFF;
 			break;
+		case PIPE_D:
+			tmp |= TRANS_DDI_EDP_INPUT_D_ONOFF;
+			break;
 		}
 
 		/* enable DDI buffer */
@@ -763,12 +786,12 @@ gen11_dsi_configure_transcoder(struct intel_encoder *encoder,
 
 static void
 gen11_dsi_set_transcoder_timings(struct intel_encoder *encoder,
-				 const struct intel_crtc_state *pipe_config)
+				 const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	const struct drm_display_mode *adjusted_mode =
-					&pipe_config->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	enum port port;
 	enum transcoder dsi_trans;
 	/* horizontal timings */
@@ -776,11 +799,25 @@ gen11_dsi_set_transcoder_timings(struct intel_encoder *encoder,
 	u16 hback_porch;
 	/* vertical timings */
 	u16 vtotal, vactive, vsync_start, vsync_end, vsync_shift;
+	int mul = 1, div = 1;
+
+	/*
+	 * Adjust horizontal timings (htotal, hsync_start, hsync_end) to account
+	 * for slower link speed if DSC is enabled.
+	 *
+	 * The compression frequency ratio is the ratio between compressed and
+	 * non-compressed link speeds, and simplifies down to the ratio between
+	 * compressed and non-compressed bpp.
+	 */
+	if (crtc_state->dsc.compression_enable) {
+		mul = crtc_state->dsc.compressed_bpp;
+		div = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format);
+	}
 
 	hactive = adjusted_mode->crtc_hdisplay;
-	htotal = adjusted_mode->crtc_htotal;
-	hsync_start = adjusted_mode->crtc_hsync_start;
-	hsync_end = adjusted_mode->crtc_hsync_end;
+	htotal = DIV_ROUND_UP(adjusted_mode->crtc_htotal * mul, div);
+	hsync_start = DIV_ROUND_UP(adjusted_mode->crtc_hsync_start * mul, div);
+	hsync_end = DIV_ROUND_UP(adjusted_mode->crtc_hsync_end * mul, div);
 	hsync_size  = hsync_end - hsync_start;
 	hback_porch = (adjusted_mode->crtc_htotal -
 		       adjusted_mode->crtc_hsync_end);
@@ -904,7 +941,8 @@ static void gen11_dsi_enable_transcoder(struct intel_encoder *encoder)
 	}
 }
 
-static void gen11_dsi_setup_timeouts(struct intel_encoder *encoder)
+static void gen11_dsi_setup_timeouts(struct intel_encoder *encoder,
+				     const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
@@ -919,7 +957,7 @@ static void gen11_dsi_setup_timeouts(struct intel_encoder *encoder)
 	 * TIME_NS = (BYTE_CLK_COUNT * 8 * 10^6)/ Bitrate
 	 * ESCAPE_CLK_COUNT  = TIME_NS/ESC_CLK_NS
 	 */
-	divisor = intel_dsi_tlpx_ns(intel_dsi) * intel_dsi_bitrate(intel_dsi) * 1000;
+	divisor = intel_dsi_tlpx_ns(intel_dsi) * afe_clk(encoder, crtc_state) * 1000;
 	mul = 8 * 1000000;
 	hs_tx_timeout = DIV_ROUND_UP(intel_dsi->hs_tx_timeout * mul,
 				     divisor);
@@ -955,7 +993,7 @@ static void gen11_dsi_setup_timeouts(struct intel_encoder *encoder)
 
 static void
 gen11_dsi_enable_port_and_phy(struct intel_encoder *encoder,
-			      const struct intel_crtc_state *pipe_config)
+			      const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 
@@ -972,13 +1010,13 @@ gen11_dsi_enable_port_and_phy(struct intel_encoder *encoder,
 	gen11_dsi_enable_ddi_buffer(encoder);
 
 	/* setup D-PHY timings */
-	gen11_dsi_setup_dphy_timings(encoder);
+	gen11_dsi_setup_dphy_timings(encoder, crtc_state);
 
 	/* step 4h: setup DSI protocol timeouts */
-	gen11_dsi_setup_timeouts(encoder);
+	gen11_dsi_setup_timeouts(encoder, crtc_state);
 
 	/* Step (4h, 4i, 4j, 4k): Configure transcoder */
-	gen11_dsi_configure_transcoder(encoder, pipe_config);
+	gen11_dsi_configure_transcoder(encoder, crtc_state);
 
 	/* Step 4l: Gate DDI clocks */
 	if (IS_GEN(dev_priv, 11))
@@ -1025,14 +1063,14 @@ static void gen11_dsi_powerup_panel(struct intel_encoder *encoder)
 }
 
 static void gen11_dsi_pre_pll_enable(struct intel_encoder *encoder,
-				     const struct intel_crtc_state *pipe_config,
+				     const struct intel_crtc_state *crtc_state,
 				     const struct drm_connector_state *conn_state)
 {
 	/* step2: enable IO power */
 	gen11_dsi_enable_io_power(encoder);
 
 	/* step3: enable DSI PLL */
-	gen11_dsi_program_esc_clk_div(encoder);
+	gen11_dsi_program_esc_clk_div(encoder, crtc_state);
 }
 
 static void gen11_dsi_pre_enable(struct intel_encoder *encoder,
@@ -1050,6 +1088,8 @@ static void gen11_dsi_pre_enable(struct intel_encoder *encoder,
 	/* step5: program and powerup panel */
 	gen11_dsi_powerup_panel(encoder);
 
+	intel_dsc_enable(encoder, pipe_config);
+
 	/* step6c: configure transcoder timings */
 	gen11_dsi_set_transcoder_timings(encoder, pipe_config);
 
@@ -1211,12 +1251,42 @@ static void gen11_dsi_disable(struct intel_encoder *encoder,
 	gen11_dsi_disable_io_power(encoder);
 }
 
+static void gen11_dsi_post_disable(struct intel_encoder *encoder,
+				   const struct intel_crtc_state *old_crtc_state,
+				   const struct drm_connector_state *old_conn_state)
+{
+	intel_crtc_vblank_off(old_crtc_state);
+
+	intel_dsc_disable(old_crtc_state);
+
+	skylake_scaler_disable(old_crtc_state);
+}
+
+static enum drm_mode_status gen11_dsi_mode_valid(struct drm_connector *connector,
+						 struct drm_display_mode *mode)
+{
+	/* FIXME: DSC? */
+	return intel_dsi_mode_valid(connector, mode);
+}
+
 static void gen11_dsi_get_timings(struct intel_encoder *encoder,
 				  struct intel_crtc_state *pipe_config)
 {
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	struct drm_display_mode *adjusted_mode =
-					&pipe_config->base.adjusted_mode;
+					&pipe_config->hw.adjusted_mode;
+
+	if (pipe_config->dsc.compressed_bpp) {
+		int div = pipe_config->dsc.compressed_bpp;
+		int mul = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format);
+
+		adjusted_mode->crtc_htotal =
+			DIV_ROUND_UP(adjusted_mode->crtc_htotal * mul, div);
+		adjusted_mode->crtc_hsync_start =
+			DIV_ROUND_UP(adjusted_mode->crtc_hsync_start * mul, div);
+		adjusted_mode->crtc_hsync_end =
+			DIV_ROUND_UP(adjusted_mode->crtc_hsync_end * mul, div);
+	}
 
 	if (intel_dsi->dual_link) {
 		adjusted_mode->crtc_hdisplay *= 2;
@@ -1242,22 +1312,66 @@ static void gen11_dsi_get_config(struct intel_encoder *encoder,
 				 struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 
+	intel_dsc_get_config(encoder, pipe_config);
+
 	/* FIXME: adapt icl_ddi_clock_get() for DSI and use that? */
 	pipe_config->port_clock =
 		cnl_calc_wrpll_link(dev_priv, &pipe_config->dpll_hw_state);
 
-	pipe_config->base.adjusted_mode.crtc_clock = intel_dsi->pclk;
+	pipe_config->hw.adjusted_mode.crtc_clock = intel_dsi->pclk;
 	if (intel_dsi->dual_link)
-		pipe_config->base.adjusted_mode.crtc_clock *= 2;
+		pipe_config->hw.adjusted_mode.crtc_clock *= 2;
 
 	gen11_dsi_get_timings(encoder, pipe_config);
 	pipe_config->output_types |= BIT(INTEL_OUTPUT_DSI);
 	pipe_config->pipe_bpp = bdw_get_pipemisc_bpp(crtc);
 }
 
+static int gen11_dsi_dsc_compute_config(struct intel_encoder *encoder,
+					struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
+	int dsc_max_bpc = INTEL_GEN(dev_priv) >= 12 ? 12 : 10;
+	bool use_dsc;
+	int ret;
+
+	use_dsc = intel_bios_get_dsc_params(encoder, crtc_state, dsc_max_bpc);
+	if (!use_dsc)
+		return 0;
+
+	if (crtc_state->pipe_bpp < 8 * 3)
+		return -EINVAL;
+
+	/* FIXME: split only when necessary */
+	if (crtc_state->dsc.slice_count > 1)
+		crtc_state->dsc.dsc_split = true;
+
+	vdsc_cfg->convert_rgb = true;
+
+	ret = intel_dsc_compute_params(encoder, crtc_state);
+	if (ret)
+		return ret;
+
+	/* DSI specific sanity checks on the common code */
+	WARN_ON(vdsc_cfg->vbr_enable);
+	WARN_ON(vdsc_cfg->simple_422);
+	WARN_ON(vdsc_cfg->pic_width % vdsc_cfg->slice_width);
+	WARN_ON(vdsc_cfg->slice_height < 8);
+	WARN_ON(vdsc_cfg->pic_height % vdsc_cfg->slice_height);
+
+	ret = drm_dsc_compute_rc_parameters(vdsc_cfg);
+	if (ret)
+		return ret;
+
+	crtc_state->dsc.compression_enable = true;
+
+	return 0;
+}
+
 static int gen11_dsi_compute_config(struct intel_encoder *encoder,
 				    struct intel_crtc_state *pipe_config,
 				    struct drm_connector_state *conn_state)
@@ -1265,11 +1379,11 @@ static int gen11_dsi_compute_config(struct intel_encoder *encoder,
 	struct intel_dsi *intel_dsi = container_of(encoder, struct intel_dsi,
 						   base);
 	struct intel_connector *intel_connector = intel_dsi->attached_connector;
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	const struct drm_display_mode *fixed_mode =
 					intel_connector->panel.fixed_mode;
 	struct drm_display_mode *adjusted_mode =
-					&pipe_config->base.adjusted_mode;
+					&pipe_config->hw.adjusted_mode;
 
 	pipe_config->output_format = INTEL_OUTPUT_FORMAT_RGB;
 	intel_fixed_panel_mode(fixed_mode, adjusted_mode);
@@ -1283,8 +1397,17 @@ static int gen11_dsi_compute_config(struct intel_encoder *encoder,
 	else
 		pipe_config->cpu_transcoder = TRANSCODER_DSI_0;
 
+	if (intel_dsi->pixel_format == MIPI_DSI_FMT_RGB888)
+		pipe_config->pipe_bpp = 24;
+	else
+		pipe_config->pipe_bpp = 18;
+
 	pipe_config->clock_set = true;
-	pipe_config->port_clock = intel_dsi_bitrate(intel_dsi) / 5;
+
+	if (gen11_dsi_dsc_compute_config(encoder, pipe_config))
+		DRM_DEBUG_KMS("Attempting to use DSC failed\n");
+
+	pipe_config->port_clock = afe_clk(encoder, pipe_config) / 5;
 
 	return 0;
 }
@@ -1292,8 +1415,13 @@ static int gen11_dsi_compute_config(struct intel_encoder *encoder,
 static void gen11_dsi_get_power_domains(struct intel_encoder *encoder,
 					struct intel_crtc_state *crtc_state)
 {
-	get_dsi_io_power_domains(to_i915(encoder->base.dev),
-				 enc_to_intel_dsi(&encoder->base));
+	struct drm_i915_private *i915 = to_i915(encoder->base.dev);
+
+	get_dsi_io_power_domains(i915, enc_to_intel_dsi(&encoder->base));
+
+	if (crtc_state->dsc.compression_enable)
+		intel_display_power_get(i915,
+					intel_dsc_power_domain(crtc_state));
 }
 
 static bool gen11_dsi_get_hw_state(struct intel_encoder *encoder,
@@ -1325,6 +1453,9 @@ static bool gen11_dsi_get_hw_state(struct intel_encoder *encoder,
 		case TRANS_DDI_EDP_INPUT_C_ONOFF:
 			*pipe = PIPE_C;
 			break;
+		case TRANS_DDI_EDP_INPUT_D_ONOFF:
+			*pipe = PIPE_D;
+			break;
 		default:
 			DRM_ERROR("Invalid PIPE input\n");
 			goto out;
@@ -1360,7 +1491,7 @@ static const struct drm_connector_funcs gen11_dsi_connector_funcs = {
 
 static const struct drm_connector_helper_funcs gen11_dsi_connector_helper_funcs = {
 	.get_modes = intel_dsi_get_modes,
-	.mode_valid = intel_dsi_mode_valid,
+	.mode_valid = gen11_dsi_mode_valid,
 	.atomic_check = intel_digital_connector_atomic_check,
 };
 
@@ -1577,6 +1708,7 @@ void icl_dsi_init(struct drm_i915_private *dev_priv)
 	encoder->pre_pll_enable = gen11_dsi_pre_pll_enable;
 	encoder->pre_enable = gen11_dsi_pre_enable;
 	encoder->disable = gen11_dsi_disable;
+	encoder->post_disable = gen11_dsi_post_disable;
 	encoder->port = port;
 	encoder->get_config = gen11_dsi_get_config;
 	encoder->update_pipe = intel_panel_update_backlight;
diff --git a/drivers/gpu/drm/i915/display/intel_atomic.c b/drivers/gpu/drm/i915/display/intel_atomic.c
index c2875b10adf9..fd0026fc3618 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic.c
+++ b/drivers/gpu/drm/i915/display/intel_atomic.c
@@ -186,13 +186,22 @@ intel_digital_connector_duplicate_state(struct drm_connector *connector)
 struct drm_crtc_state *
 intel_crtc_duplicate_state(struct drm_crtc *crtc)
 {
+	const struct intel_crtc_state *old_crtc_state = to_intel_crtc_state(crtc->state);
 	struct intel_crtc_state *crtc_state;
 
-	crtc_state = kmemdup(crtc->state, sizeof(*crtc_state), GFP_KERNEL);
+	crtc_state = kmemdup(old_crtc_state, sizeof(*crtc_state), GFP_KERNEL);
 	if (!crtc_state)
 		return NULL;
 
-	__drm_atomic_helper_crtc_duplicate_state(crtc, &crtc_state->base);
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &crtc_state->uapi);
+
+	/* copy color blobs */
+	if (crtc_state->hw.degamma_lut)
+		drm_property_blob_get(crtc_state->hw.degamma_lut);
+	if (crtc_state->hw.ctm)
+		drm_property_blob_get(crtc_state->hw.ctm);
+	if (crtc_state->hw.gamma_lut)
+		drm_property_blob_get(crtc_state->hw.gamma_lut);
 
 	crtc_state->update_pipe = false;
 	crtc_state->disable_lp_wm = false;
@@ -205,7 +214,29 @@ intel_crtc_duplicate_state(struct drm_crtc *crtc)
 	crtc_state->fb_bits = 0;
 	crtc_state->update_planes = 0;
 
-	return &crtc_state->base;
+	return &crtc_state->uapi;
+}
+
+static void intel_crtc_put_color_blobs(struct intel_crtc_state *crtc_state)
+{
+	drm_property_blob_put(crtc_state->hw.degamma_lut);
+	drm_property_blob_put(crtc_state->hw.gamma_lut);
+	drm_property_blob_put(crtc_state->hw.ctm);
+}
+
+void intel_crtc_free_hw_state(struct intel_crtc_state *crtc_state)
+{
+	intel_crtc_put_color_blobs(crtc_state);
+}
+
+void intel_crtc_copy_color_blobs(struct intel_crtc_state *crtc_state)
+{
+	drm_property_replace_blob(&crtc_state->hw.degamma_lut,
+				  crtc_state->uapi.degamma_lut);
+	drm_property_replace_blob(&crtc_state->hw.gamma_lut,
+				  crtc_state->uapi.gamma_lut);
+	drm_property_replace_blob(&crtc_state->hw.ctm,
+				  crtc_state->uapi.ctm);
 }
 
 /**
@@ -220,7 +251,11 @@ void
 intel_crtc_destroy_state(struct drm_crtc *crtc,
 			 struct drm_crtc_state *state)
 {
-	drm_atomic_helper_crtc_destroy_state(crtc, state);
+	struct intel_crtc_state *crtc_state = to_intel_crtc_state(state);
+
+	__drm_atomic_helper_crtc_destroy_state(&crtc_state->uapi);
+	intel_crtc_free_hw_state(crtc_state);
+	kfree(crtc_state);
 }
 
 static void intel_atomic_setup_scaler(struct intel_crtc_scaler_state *scaler_state,
@@ -249,10 +284,10 @@ static void intel_atomic_setup_scaler(struct intel_crtc_scaler_state *scaler_sta
 		return;
 
 	/* set scaler mode */
-	if (plane_state && plane_state->base.fb &&
-	    plane_state->base.fb->format->is_yuv &&
-	    plane_state->base.fb->format->num_planes > 1) {
-		struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	if (plane_state && plane_state->hw.fb &&
+	    plane_state->hw.fb->format->is_yuv &&
+	    plane_state->hw.fb->format->num_planes > 1) {
+		struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 		if (IS_GEN(dev_priv, 9) &&
 		    !IS_GEMINILAKE(dev_priv)) {
 			mode = SKL_PS_SCALER_MODE_NV12;
@@ -319,7 +354,7 @@ int intel_atomic_setup_scalers(struct drm_i915_private *dev_priv,
 	struct intel_plane_state *plane_state = NULL;
 	struct intel_crtc_scaler_state *scaler_state =
 		&crtc_state->scaler_state;
-	struct drm_atomic_state *drm_state = crtc_state->base.state;
+	struct drm_atomic_state *drm_state = crtc_state->uapi.state;
 	struct intel_atomic_state *intel_state = to_intel_atomic_state(drm_state);
 	int num_scalers_need;
 	int i;
diff --git a/drivers/gpu/drm/i915/display/intel_atomic.h b/drivers/gpu/drm/i915/display/intel_atomic.h
index 49d5cb1b9e0a..7b49623419ba 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic.h
+++ b/drivers/gpu/drm/i915/display/intel_atomic.h
@@ -36,6 +36,8 @@ intel_digital_connector_duplicate_state(struct drm_connector *connector);
 struct drm_crtc_state *intel_crtc_duplicate_state(struct drm_crtc *crtc);
 void intel_crtc_destroy_state(struct drm_crtc *crtc,
 			       struct drm_crtc_state *state);
+void intel_crtc_free_hw_state(struct intel_crtc_state *crtc_state);
+void intel_crtc_copy_color_blobs(struct intel_crtc_state *crtc_state);
 struct drm_atomic_state *intel_atomic_state_alloc(struct drm_device *dev);
 void intel_atomic_state_clear(struct drm_atomic_state *state);
 
diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.c b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
index 98f557a9f8ee..3e97af682b1b 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic_plane.c
+++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.c
@@ -41,6 +41,16 @@
 #include "intel_pm.h"
 #include "intel_sprite.h"
 
+static void intel_plane_state_reset(struct intel_plane_state *plane_state,
+				    struct intel_plane *plane)
+{
+	memset(plane_state, 0, sizeof(*plane_state));
+
+	__drm_atomic_helper_plane_state_reset(&plane_state->uapi, &plane->base);
+
+	plane_state->scaler_id = -1;
+}
+
 struct intel_plane *intel_plane_alloc(void)
 {
 	struct intel_plane_state *plane_state;
@@ -56,8 +66,9 @@ struct intel_plane *intel_plane_alloc(void)
 		return ERR_PTR(-ENOMEM);
 	}
 
-	__drm_atomic_helper_plane_reset(&plane->base, &plane_state->base);
-	plane_state->scaler_id = -1;
+	intel_plane_state_reset(plane_state, plane);
+
+	plane->base.state = &plane_state->uapi;
 
 	return plane;
 }
@@ -80,22 +91,24 @@ void intel_plane_free(struct intel_plane *plane)
 struct drm_plane_state *
 intel_plane_duplicate_state(struct drm_plane *plane)
 {
-	struct drm_plane_state *state;
 	struct intel_plane_state *intel_state;
 
-	intel_state = kmemdup(plane->state, sizeof(*intel_state), GFP_KERNEL);
+	intel_state = to_intel_plane_state(plane->state);
+	intel_state = kmemdup(intel_state, sizeof(*intel_state), GFP_KERNEL);
 
 	if (!intel_state)
 		return NULL;
 
-	state = &intel_state->base;
-
-	__drm_atomic_helper_plane_duplicate_state(plane, state);
+	__drm_atomic_helper_plane_duplicate_state(plane, &intel_state->uapi);
 
 	intel_state->vma = NULL;
 	intel_state->flags = 0;
 
-	return state;
+	/* add reference to fb */
+	if (intel_state->hw.fb)
+		drm_framebuffer_get(intel_state->hw.fb);
+
+	return &intel_state->uapi;
 }
 
 /**
@@ -110,18 +123,22 @@ void
 intel_plane_destroy_state(struct drm_plane *plane,
 			  struct drm_plane_state *state)
 {
-	WARN_ON(to_intel_plane_state(state)->vma);
+	struct intel_plane_state *plane_state = to_intel_plane_state(state);
+	WARN_ON(plane_state->vma);
 
-	drm_atomic_helper_plane_destroy_state(plane, state);
+	__drm_atomic_helper_plane_destroy_state(&plane_state->uapi);
+	if (plane_state->hw.fb)
+		drm_framebuffer_put(plane_state->hw.fb);
+	kfree(plane_state);
 }
 
 unsigned int intel_plane_data_rate(const struct intel_crtc_state *crtc_state,
 				   const struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int cpp;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	cpp = fb->format->cpp[0];
@@ -144,10 +161,10 @@ bool intel_plane_calc_min_cdclk(struct intel_atomic_state *state,
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	const struct intel_plane_state *plane_state =
 		intel_atomic_get_new_plane_state(state, plane);
-	struct intel_crtc *crtc = to_intel_crtc(plane_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(plane_state->hw.crtc);
 	struct intel_crtc_state *crtc_state;
 
-	if (!plane_state->base.visible || !plane->min_cdclk)
+	if (!plane_state->uapi.visible || !plane->min_cdclk)
 		return false;
 
 	crtc_state = intel_atomic_get_new_crtc_state(state, crtc);
@@ -176,23 +193,52 @@ bool intel_plane_calc_min_cdclk(struct intel_atomic_state *state,
 	return false;
 }
 
+static void intel_plane_clear_hw_state(struct intel_plane_state *plane_state)
+{
+	if (plane_state->hw.fb)
+		drm_framebuffer_put(plane_state->hw.fb);
+
+	memset(&plane_state->hw, 0, sizeof(plane_state->hw));
+}
+
+void intel_plane_copy_uapi_to_hw_state(struct intel_plane_state *plane_state,
+				       const struct intel_plane_state *from_plane_state)
+{
+	intel_plane_clear_hw_state(plane_state);
+
+	plane_state->hw.crtc = from_plane_state->uapi.crtc;
+	plane_state->hw.fb = from_plane_state->uapi.fb;
+	if (plane_state->hw.fb)
+		drm_framebuffer_get(plane_state->hw.fb);
+
+	plane_state->hw.alpha = from_plane_state->uapi.alpha;
+	plane_state->hw.pixel_blend_mode =
+		from_plane_state->uapi.pixel_blend_mode;
+	plane_state->hw.rotation = from_plane_state->uapi.rotation;
+	plane_state->hw.color_encoding = from_plane_state->uapi.color_encoding;
+	plane_state->hw.color_range = from_plane_state->uapi.color_range;
+}
+
 int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_state,
 					struct intel_crtc_state *new_crtc_state,
 					const struct intel_plane_state *old_plane_state,
 					struct intel_plane_state *new_plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(new_plane_state->base.plane);
-	const struct drm_framebuffer *fb = new_plane_state->base.fb;
+	struct intel_plane *plane = to_intel_plane(new_plane_state->uapi.plane);
+	const struct drm_framebuffer *fb;
 	int ret;
 
+	intel_plane_copy_uapi_to_hw_state(new_plane_state, new_plane_state);
+	fb = new_plane_state->hw.fb;
+
 	new_crtc_state->active_planes &= ~BIT(plane->id);
 	new_crtc_state->nv12_planes &= ~BIT(plane->id);
 	new_crtc_state->c8_planes &= ~BIT(plane->id);
 	new_crtc_state->data_rate[plane->id] = 0;
 	new_crtc_state->min_cdclk[plane->id] = 0;
-	new_plane_state->base.visible = false;
+	new_plane_state->uapi.visible = false;
 
-	if (!new_plane_state->base.crtc && !old_plane_state->base.crtc)
+	if (!new_plane_state->hw.crtc && !old_plane_state->hw.crtc)
 		return 0;
 
 	ret = plane->check_plane(new_crtc_state, new_plane_state);
@@ -200,18 +246,18 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_
 		return ret;
 
 	/* FIXME pre-g4x don't work like this */
-	if (new_plane_state->base.visible)
+	if (new_plane_state->uapi.visible)
 		new_crtc_state->active_planes |= BIT(plane->id);
 
-	if (new_plane_state->base.visible &&
-	    drm_format_info_is_yuv_semiplanar(fb->format))
+	if (new_plane_state->uapi.visible &&
+	    intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier))
 		new_crtc_state->nv12_planes |= BIT(plane->id);
 
-	if (new_plane_state->base.visible &&
+	if (new_plane_state->uapi.visible &&
 	    fb->format->format == DRM_FORMAT_C8)
 		new_crtc_state->c8_planes |= BIT(plane->id);
 
-	if (new_plane_state->base.visible || old_plane_state->base.visible)
+	if (new_plane_state->uapi.visible || old_plane_state->uapi.visible)
 		new_crtc_state->update_planes |= BIT(plane->id);
 
 	new_crtc_state->data_rate[plane->id] =
@@ -225,11 +271,11 @@ static struct intel_crtc *
 get_crtc_from_states(const struct intel_plane_state *old_plane_state,
 		     const struct intel_plane_state *new_plane_state)
 {
-	if (new_plane_state->base.crtc)
-		return to_intel_crtc(new_plane_state->base.crtc);
+	if (new_plane_state->uapi.crtc)
+		return to_intel_crtc(new_plane_state->uapi.crtc);
 
-	if (old_plane_state->base.crtc)
-		return to_intel_crtc(old_plane_state->base.crtc);
+	if (old_plane_state->uapi.crtc)
+		return to_intel_crtc(old_plane_state->uapi.crtc);
 
 	return NULL;
 }
@@ -246,7 +292,7 @@ int intel_plane_atomic_check(struct intel_atomic_state *state,
 	const struct intel_crtc_state *old_crtc_state;
 	struct intel_crtc_state *new_crtc_state;
 
-	new_plane_state->base.visible = false;
+	new_plane_state->uapi.visible = false;
 	if (!crtc)
 		return 0;
 
@@ -307,26 +353,16 @@ void intel_update_plane(struct intel_plane *plane,
 			const struct intel_crtc_state *crtc_state,
 			const struct intel_plane_state *plane_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 
 	trace_intel_update_plane(&plane->base, crtc);
 	plane->update_plane(plane, crtc_state, plane_state);
 }
 
-void intel_update_slave(struct intel_plane *plane,
-			const struct intel_crtc_state *crtc_state,
-			const struct intel_plane_state *plane_state)
-{
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-
-	trace_intel_update_plane(&plane->base, crtc);
-	plane->update_slave(plane, crtc_state, plane_state);
-}
-
 void intel_disable_plane(struct intel_plane *plane,
 			 const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 
 	trace_intel_disable_plane(&plane->base, crtc);
 	plane->disable_plane(plane, crtc_state);
@@ -355,25 +391,9 @@ void skl_update_planes_on_crtc(struct intel_atomic_state *state,
 		struct intel_plane_state *new_plane_state =
 			intel_atomic_get_new_plane_state(state, plane);
 
-		if (new_plane_state->base.visible) {
+		if (new_plane_state->uapi.visible ||
+		    new_plane_state->planar_slave) {
 			intel_update_plane(plane, new_crtc_state, new_plane_state);
-		} else if (new_plane_state->planar_slave) {
-			struct intel_plane *master =
-				new_plane_state->planar_linked_plane;
-
-			/*
-			 * We update the slave plane from this function because
-			 * programming it from the master plane's update_plane
-			 * callback runs into issues when the Y plane is
-			 * reassigned, disabled or used by a different plane.
-			 *
-			 * The slave plane is updated with the master plane's
-			 * plane_state.
-			 */
-			new_plane_state =
-				intel_atomic_get_new_plane_state(state, master);
-
-			intel_update_slave(plane, new_crtc_state, new_plane_state);
 		} else {
 			intel_disable_plane(plane, new_crtc_state);
 		}
@@ -395,7 +415,7 @@ void i9xx_update_planes_on_crtc(struct intel_atomic_state *state,
 		    !(update_mask & BIT(plane->id)))
 			continue;
 
-		if (new_plane_state->base.visible)
+		if (new_plane_state->uapi.visible)
 			intel_update_plane(plane, new_crtc_state, new_plane_state);
 		else
 			intel_disable_plane(plane, new_crtc_state);
diff --git a/drivers/gpu/drm/i915/display/intel_atomic_plane.h b/drivers/gpu/drm/i915/display/intel_atomic_plane.h
index e61e9a82aadf..5cedafdddb55 100644
--- a/drivers/gpu/drm/i915/display/intel_atomic_plane.h
+++ b/drivers/gpu/drm/i915/display/intel_atomic_plane.h
@@ -20,12 +20,11 @@ extern const struct drm_plane_helper_funcs intel_plane_helper_funcs;
 
 unsigned int intel_plane_data_rate(const struct intel_crtc_state *crtc_state,
 				   const struct intel_plane_state *plane_state);
+void intel_plane_copy_uapi_to_hw_state(struct intel_plane_state *plane_state,
+				       const struct intel_plane_state *from_plane_state);
 void intel_update_plane(struct intel_plane *plane,
 			const struct intel_crtc_state *crtc_state,
 			const struct intel_plane_state *plane_state);
-void intel_update_slave(struct intel_plane *plane,
-			const struct intel_crtc_state *crtc_state,
-			const struct intel_plane_state *plane_state);
 void intel_disable_plane(struct intel_plane *plane,
 			 const struct intel_crtc_state *crtc_state);
 struct intel_plane *intel_plane_alloc(void);
diff --git a/drivers/gpu/drm/i915/display/intel_audio.c b/drivers/gpu/drm/i915/display/intel_audio.c
index 85e6b2bbb34f..27710098d056 100644
--- a/drivers/gpu/drm/i915/display/intel_audio.c
+++ b/drivers/gpu/drm/i915/display/intel_audio.c
@@ -234,7 +234,7 @@ static const struct hdmi_aud_ncts hdmi_aud_ncts_36bpp[] = {
 static u32 audio_config_hdmi_pixel_clock(const struct intel_crtc_state *crtc_state)
 {
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	int i;
 
 	for (i = 0; i < ARRAY_SIZE(hdmi_audio_clock); i++) {
@@ -555,7 +555,7 @@ static void ilk_audio_codec_disable(struct intel_encoder *encoder,
 				    const struct drm_connector_state *old_conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	enum pipe pipe = crtc->pipe;
 	enum port port = encoder->port;
 	u32 tmp, eldv;
@@ -602,7 +602,7 @@ static void ilk_audio_codec_enable(struct intel_encoder *encoder,
 				   const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_connector *connector = conn_state->connector;
 	enum pipe pipe = crtc->pipe;
 	enum port port = encoder->port;
@@ -692,10 +692,10 @@ void intel_audio_codec_enable(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct i915_audio_component *acomp = dev_priv->audio_component;
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_connector *connector = conn_state->connector;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	enum port port = encoder->port;
 	enum pipe pipe = crtc->pipe;
 
@@ -753,7 +753,7 @@ void intel_audio_codec_disable(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct i915_audio_component *acomp = dev_priv->audio_component;
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	enum port port = encoder->port;
 	enum pipe pipe = crtc->pipe;
 
diff --git a/drivers/gpu/drm/i915/display/intel_bios.c b/drivers/gpu/drm/i915/display/intel_bios.c
index 63c1bd4c2954..8beac06e3f10 100644
--- a/drivers/gpu/drm/i915/display/intel_bios.c
+++ b/drivers/gpu/drm/i915/display/intel_bios.c
@@ -29,6 +29,7 @@
 #include <drm/i915_drm.h>
 
 #include "display/intel_display.h"
+#include "display/intel_display_types.h"
 #include "display/intel_gmbus.h"
 
 #include "i915_drv.h"
@@ -58,6 +59,13 @@
  * that.
  */
 
+/* Wrapper for VBT child device config */
+struct display_device_data {
+	struct child_device_config child;
+	struct dsc_compression_parameters_entry *dsc;
+	struct list_head node;
+};
+
 #define	SLAVE_ADDR1	0x70
 #define	SLAVE_ADDR2	0x72
 
@@ -202,17 +210,12 @@ get_lvds_fp_timing(const struct bdb_header *bdb,
 	return (const struct lvds_fp_timing *)((const u8 *)bdb + ofs);
 }
 
-/* Try to find integrated panel data */
+/* Parse general panel options */
 static void
-parse_lfp_panel_data(struct drm_i915_private *dev_priv,
-		     const struct bdb_header *bdb)
+parse_panel_options(struct drm_i915_private *dev_priv,
+		    const struct bdb_header *bdb)
 {
 	const struct bdb_lvds_options *lvds_options;
-	const struct bdb_lvds_lfp_data *lvds_lfp_data;
-	const struct bdb_lvds_lfp_data_ptrs *lvds_lfp_data_ptrs;
-	const struct lvds_dvo_timing *panel_dvo_timing;
-	const struct lvds_fp_timing *fp_timing;
-	struct drm_display_mode *panel_fixed_mode;
 	int panel_type;
 	int drrs_mode;
 	int ret;
@@ -261,6 +264,19 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 		DRM_DEBUG_KMS("DRRS not supported (VBT input)\n");
 		break;
 	}
+}
+
+/* Try to find integrated panel timing data */
+static void
+parse_lfp_panel_dtd(struct drm_i915_private *dev_priv,
+		    const struct bdb_header *bdb)
+{
+	const struct bdb_lvds_lfp_data *lvds_lfp_data;
+	const struct bdb_lvds_lfp_data_ptrs *lvds_lfp_data_ptrs;
+	const struct lvds_dvo_timing *panel_dvo_timing;
+	const struct lvds_fp_timing *fp_timing;
+	struct drm_display_mode *panel_fixed_mode;
+	int panel_type = dev_priv->vbt.panel_type;
 
 	lvds_lfp_data = find_section(bdb, BDB_LVDS_LFP_DATA);
 	if (!lvds_lfp_data)
@@ -282,7 +298,7 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 
 	dev_priv->vbt.lfp_lvds_vbt_mode = panel_fixed_mode;
 
-	DRM_DEBUG_KMS("Found panel mode in BIOS VBT tables:\n");
+	DRM_DEBUG_KMS("Found panel mode in BIOS VBT legacy lfp table:\n");
 	drm_mode_debug_printmodeline(panel_fixed_mode);
 
 	fp_timing = get_lvds_fp_timing(bdb, lvds_lfp_data,
@@ -300,6 +316,98 @@ parse_lfp_panel_data(struct drm_i915_private *dev_priv,
 }
 
 static void
+parse_generic_dtd(struct drm_i915_private *dev_priv,
+		  const struct bdb_header *bdb)
+{
+	const struct bdb_generic_dtd *generic_dtd;
+	const struct generic_dtd_entry *dtd;
+	struct drm_display_mode *panel_fixed_mode;
+	int num_dtd;
+
+	generic_dtd = find_section(bdb, BDB_GENERIC_DTD);
+	if (!generic_dtd)
+		return;
+
+	if (generic_dtd->gdtd_size < sizeof(struct generic_dtd_entry)) {
+		DRM_ERROR("GDTD size %u is too small.\n",
+			  generic_dtd->gdtd_size);
+		return;
+	} else if (generic_dtd->gdtd_size !=
+		   sizeof(struct generic_dtd_entry)) {
+		DRM_ERROR("Unexpected GDTD size %u\n", generic_dtd->gdtd_size);
+		/* DTD has unknown fields, but keep going */
+	}
+
+	num_dtd = (get_blocksize(generic_dtd) -
+		   sizeof(struct bdb_generic_dtd)) / generic_dtd->gdtd_size;
+	if (dev_priv->vbt.panel_type >= num_dtd) {
+		DRM_ERROR("Panel type %d not found in table of %d DTD's\n",
+			  dev_priv->vbt.panel_type, num_dtd);
+		return;
+	}
+
+	dtd = &generic_dtd->dtd[dev_priv->vbt.panel_type];
+
+	panel_fixed_mode = kzalloc(sizeof(*panel_fixed_mode), GFP_KERNEL);
+	if (!panel_fixed_mode)
+		return;
+
+	panel_fixed_mode->hdisplay = dtd->hactive;
+	panel_fixed_mode->hsync_start =
+		panel_fixed_mode->hdisplay + dtd->hfront_porch;
+	panel_fixed_mode->hsync_end =
+		panel_fixed_mode->hsync_start + dtd->hsync;
+	panel_fixed_mode->htotal = panel_fixed_mode->hsync_end;
+
+	panel_fixed_mode->vdisplay = dtd->vactive;
+	panel_fixed_mode->vsync_start =
+		panel_fixed_mode->vdisplay + dtd->vfront_porch;
+	panel_fixed_mode->vsync_end =
+		panel_fixed_mode->vsync_start + dtd->vsync;
+	panel_fixed_mode->vtotal = panel_fixed_mode->vsync_end;
+
+	panel_fixed_mode->clock = dtd->pixel_clock;
+	panel_fixed_mode->width_mm = dtd->width_mm;
+	panel_fixed_mode->height_mm = dtd->height_mm;
+
+	panel_fixed_mode->type = DRM_MODE_TYPE_PREFERRED;
+	drm_mode_set_name(panel_fixed_mode);
+
+	if (dtd->hsync_positive_polarity)
+		panel_fixed_mode->flags |= DRM_MODE_FLAG_PHSYNC;
+	else
+		panel_fixed_mode->flags |= DRM_MODE_FLAG_NHSYNC;
+
+	if (dtd->vsync_positive_polarity)
+		panel_fixed_mode->flags |= DRM_MODE_FLAG_PVSYNC;
+	else
+		panel_fixed_mode->flags |= DRM_MODE_FLAG_NVSYNC;
+
+	DRM_DEBUG_KMS("Found panel mode in BIOS VBT generic dtd table:\n");
+	drm_mode_debug_printmodeline(panel_fixed_mode);
+
+	dev_priv->vbt.lfp_lvds_vbt_mode = panel_fixed_mode;
+}
+
+static void
+parse_panel_dtd(struct drm_i915_private *dev_priv,
+		const struct bdb_header *bdb)
+{
+	/*
+	 * Older VBTs provided provided DTD information for internal displays
+	 * through the "LFP panel DTD" block (42).  As of VBT revision 229,
+	 * that block is now deprecated and DTD information should be provided
+	 * via a newer "generic DTD" block (58).  Just to be safe, we'll
+	 * try the new generic DTD block first on VBT >= 229, but still fall
+	 * back to trying the old LFP block if that fails.
+	 */
+	if (bdb->version >= 229)
+		parse_generic_dtd(dev_priv, bdb);
+	if (!dev_priv->vbt.lfp_lvds_vbt_mode)
+		parse_lfp_panel_dtd(dev_priv, bdb);
+}
+
+static void
 parse_lfp_backlight(struct drm_i915_private *dev_priv,
 		    const struct bdb_header *bdb)
 {
@@ -449,8 +557,9 @@ static void
 parse_sdvo_device_mapping(struct drm_i915_private *dev_priv, u8 bdb_version)
 {
 	struct sdvo_device_mapping *mapping;
+	const struct display_device_data *devdata;
 	const struct child_device_config *child;
-	int i, count = 0;
+	int count = 0;
 
 	/*
 	 * Only parse SDVO mappings on gens that could have SDVO. This isn't
@@ -461,8 +570,8 @@ parse_sdvo_device_mapping(struct drm_i915_private *dev_priv, u8 bdb_version)
 		return;
 	}
 
-	for (i = 0, count = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		child = &devdata->child;
 
 		if (child->slave_addr != SLAVE_ADDR1 &&
 		    child->slave_addr != SLAVE_ADDR2) {
@@ -552,16 +661,45 @@ parse_driver_features(struct drm_i915_private *dev_priv,
 			dev_priv->vbt.int_lvds_support = 0;
 	}
 
-	DRM_DEBUG_KMS("DRRS State Enabled:%d\n", driver->drrs_enabled);
+	if (bdb->version < 228) {
+		DRM_DEBUG_KMS("DRRS State Enabled:%d\n", driver->drrs_enabled);
+		/*
+		 * If DRRS is not supported, drrs_type has to be set to 0.
+		 * This is because, VBT is configured in such a way that
+		 * static DRRS is 0 and DRRS not supported is represented by
+		 * driver->drrs_enabled=false
+		 */
+		if (!driver->drrs_enabled)
+			dev_priv->vbt.drrs_type = DRRS_NOT_SUPPORTED;
+
+		dev_priv->vbt.psr.enable = driver->psr_enabled;
+	}
+}
+
+static void
+parse_power_conservation_features(struct drm_i915_private *dev_priv,
+				  const struct bdb_header *bdb)
+{
+	const struct bdb_lfp_power *power;
+	u8 panel_type = dev_priv->vbt.panel_type;
+
+	if (bdb->version < 228)
+		return;
+
+	power = find_section(bdb, BDB_LVDS_POWER);
+	if (!power)
+		return;
+
+	dev_priv->vbt.psr.enable = power->psr & BIT(panel_type);
+
 	/*
 	 * If DRRS is not supported, drrs_type has to be set to 0.
 	 * This is because, VBT is configured in such a way that
 	 * static DRRS is 0 and DRRS not supported is represented by
-	 * driver->drrs_enabled=false
+	 * power->drrs & BIT(panel_type)=false
 	 */
-	if (!driver->drrs_enabled)
+	if (!(power->drrs & BIT(panel_type)))
 		dev_priv->vbt.drrs_type = DRRS_NOT_SUPPORTED;
-	dev_priv->vbt.psr.enable = driver->psr_enabled;
 }
 
 static void
@@ -1230,6 +1368,57 @@ err:
 	memset(dev_priv->vbt.dsi.sequence, 0, sizeof(dev_priv->vbt.dsi.sequence));
 }
 
+static void
+parse_compression_parameters(struct drm_i915_private *i915,
+			     const struct bdb_header *bdb)
+{
+	const struct bdb_compression_parameters *params;
+	struct display_device_data *devdata;
+	const struct child_device_config *child;
+	u16 block_size;
+	int index;
+
+	if (bdb->version < 198)
+		return;
+
+	params = find_section(bdb, BDB_COMPRESSION_PARAMETERS);
+	if (params) {
+		/* Sanity checks */
+		if (params->entry_size != sizeof(params->data[0])) {
+			DRM_DEBUG_KMS("VBT: unsupported compression param entry size\n");
+			return;
+		}
+
+		block_size = get_blocksize(params);
+		if (block_size < sizeof(*params)) {
+			DRM_DEBUG_KMS("VBT: expected 16 compression param entries\n");
+			return;
+		}
+	}
+
+	list_for_each_entry(devdata, &i915->vbt.display_devices, node) {
+		child = &devdata->child;
+
+		if (!child->compression_enable)
+			continue;
+
+		if (!params) {
+			DRM_DEBUG_KMS("VBT: compression params not available\n");
+			continue;
+		}
+
+		if (child->compression_method_cps) {
+			DRM_DEBUG_KMS("VBT: CPS compression not supported\n");
+			continue;
+		}
+
+		index = child->compression_structure_index;
+
+		devdata->dsc = kmemdup(&params->data[index],
+				       sizeof(*devdata->dsc), GFP_KERNEL);
+	}
+}
+
 static u8 translate_iboost(u8 val)
 {
 	static const u8 mapping[] = { 1, 3, 7 }; /* See VBT spec */
@@ -1246,7 +1435,7 @@ static enum port get_port_by_ddc_pin(struct drm_i915_private *i915, u8 ddc_pin)
 	const struct ddi_vbt_port_info *info;
 	enum port port;
 
-	for (port = PORT_A; port < I915_MAX_PORTS; port++) {
+	for_each_port(port) {
 		info = &i915->vbt.ddi_port_info[port];
 
 		if (info->child && ddc_pin == info->alternate_ddc_pin)
@@ -1297,7 +1486,7 @@ static enum port get_port_by_aux_ch(struct drm_i915_private *i915, u8 aux_ch)
 	const struct ddi_vbt_port_info *info;
 	enum port port;
 
-	for (port = PORT_A; port < I915_MAX_PORTS; port++) {
+	for_each_port(port) {
 		info = &i915->vbt.ddi_port_info[port];
 
 		if (info->child && aux_ch == info->alternate_aux_channel)
@@ -1418,9 +1607,10 @@ static enum port dvo_port_to_port(u8 dvo_port)
 }
 
 static void parse_ddi_port(struct drm_i915_private *dev_priv,
-			   const struct child_device_config *child,
+			   struct display_device_data *devdata,
 			   u8 bdb_version)
 {
+	const struct child_device_config *child = &devdata->child;
 	struct ddi_vbt_port_info *info;
 	bool is_dvi, is_hdmi, is_dp, is_edp, is_crt;
 	enum port port;
@@ -1443,7 +1633,7 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv,
 	is_hdmi = is_dvi && (child->device_type & DEVICE_TYPE_NOT_HDMI_OUTPUT) == 0;
 	is_edp = is_dp && (child->device_type & DEVICE_TYPE_INTERNAL_CONNECTOR);
 
-	if (port == PORT_A && is_dvi) {
+	if (port == PORT_A && is_dvi && INTEL_GEN(dev_priv) < 12) {
 		DRM_DEBUG_KMS("VBT claims port A supports DVI%s, ignoring\n",
 			      is_hdmi ? "/HDMI" : "");
 		is_dvi = false;
@@ -1461,26 +1651,11 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv,
 	if (bdb_version >= 209)
 		info->supports_tbt = child->tbt;
 
-	DRM_DEBUG_KMS("Port %c VBT info: CRT:%d DVI:%d HDMI:%d DP:%d eDP:%d LSPCON:%d USB-Type-C:%d TBT:%d\n",
+	DRM_DEBUG_KMS("Port %c VBT info: CRT:%d DVI:%d HDMI:%d DP:%d eDP:%d LSPCON:%d USB-Type-C:%d TBT:%d DSC:%d\n",
 		      port_name(port), is_crt, is_dvi, is_hdmi, is_dp, is_edp,
 		      HAS_LSPCON(dev_priv) && child->lspcon,
-		      info->supports_typec_usb, info->supports_tbt);
-
-	if (is_edp && is_dvi)
-		DRM_DEBUG_KMS("Internal DP port %c is TMDS compatible\n",
-			      port_name(port));
-	if (is_crt && port != PORT_E)
-		DRM_DEBUG_KMS("Port %c is analog\n", port_name(port));
-	if (is_crt && (is_dvi || is_dp))
-		DRM_DEBUG_KMS("Analog port %c is also DP or TMDS compatible\n",
-			      port_name(port));
-	if (is_dvi && (port == PORT_A || port == PORT_E))
-		DRM_DEBUG_KMS("Port %c is TMDS compatible\n", port_name(port));
-	if (!is_dvi && !is_dp && !is_crt)
-		DRM_DEBUG_KMS("Port %c is not DP/TMDS/CRT compatible\n",
-			      port_name(port));
-	if (is_edp && (port == PORT_B || port == PORT_C || port == PORT_E))
-		DRM_DEBUG_KMS("Port %c is internal DP\n", port_name(port));
+		      info->supports_typec_usb, info->supports_tbt,
+		      devdata->dsc != NULL);
 
 	if (is_dvi) {
 		u8 ddc_pin;
@@ -1509,6 +1684,7 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv,
 			      port_name(port),
 			      hdmi_level_shift);
 		info->hdmi_level_shift = hdmi_level_shift;
+		info->hdmi_level_shift_set = true;
 	}
 
 	if (bdb_version >= 204) {
@@ -1571,8 +1747,7 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv,
 
 static void parse_ddi_ports(struct drm_i915_private *dev_priv, u8 bdb_version)
 {
-	const struct child_device_config *child;
-	int i;
+	struct display_device_data *devdata;
 
 	if (!HAS_DDI(dev_priv) && !IS_CHERRYVIEW(dev_priv))
 		return;
@@ -1580,11 +1755,8 @@ static void parse_ddi_ports(struct drm_i915_private *dev_priv, u8 bdb_version)
 	if (bdb_version < 155)
 		return;
 
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
-
-		parse_ddi_port(dev_priv, child, bdb_version);
-	}
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node)
+		parse_ddi_port(dev_priv, devdata, bdb_version);
 }
 
 static void
@@ -1592,8 +1764,9 @@ parse_general_definitions(struct drm_i915_private *dev_priv,
 			  const struct bdb_header *bdb)
 {
 	const struct bdb_general_definitions *defs;
+	struct display_device_data *devdata;
 	const struct child_device_config *child;
-	int i, child_device_num, count;
+	int i, child_device_num;
 	u8 expected_size;
 	u16 block_size;
 	int bus_pin;
@@ -1649,26 +1822,7 @@ parse_general_definitions(struct drm_i915_private *dev_priv,
 
 	/* get the number of child device */
 	child_device_num = (block_size - sizeof(*defs)) / defs->child_dev_size;
-	count = 0;
-	/* get the number of child device that is present */
-	for (i = 0; i < child_device_num; i++) {
-		child = child_device_ptr(defs, i);
-		if (!child->device_type)
-			continue;
-		count++;
-	}
-	if (!count) {
-		DRM_DEBUG_KMS("no child dev is parsed from VBT\n");
-		return;
-	}
-	dev_priv->vbt.child_dev = kcalloc(count, sizeof(*child), GFP_KERNEL);
-	if (!dev_priv->vbt.child_dev) {
-		DRM_DEBUG_KMS("No memory space for child device\n");
-		return;
-	}
 
-	dev_priv->vbt.child_dev_num = count;
-	count = 0;
 	for (i = 0; i < child_device_num; i++) {
 		child = child_device_ptr(defs, i);
 		if (!child->device_type)
@@ -1677,23 +1831,29 @@ parse_general_definitions(struct drm_i915_private *dev_priv,
 		DRM_DEBUG_KMS("Found VBT child device with type 0x%x\n",
 			      child->device_type);
 
+		devdata = kzalloc(sizeof(*devdata), GFP_KERNEL);
+		if (!devdata)
+			break;
+
 		/*
 		 * Copy as much as we know (sizeof) and is available
-		 * (child_dev_size) of the child device. Accessing the data must
-		 * depend on VBT version.
+		 * (child_dev_size) of the child device config. Accessing the
+		 * data must depend on VBT version.
 		 */
-		memcpy(dev_priv->vbt.child_dev + count, child,
+		memcpy(&devdata->child, child,
 		       min_t(size_t, defs->child_dev_size, sizeof(*child)));
-		count++;
+
+		list_add_tail(&devdata->node, &dev_priv->vbt.display_devices);
 	}
+
+	if (list_empty(&dev_priv->vbt.display_devices))
+		DRM_DEBUG_KMS("no child dev is parsed from VBT\n");
 }
 
 /* Common defaults which may be overridden by VBT. */
 static void
 init_vbt_defaults(struct drm_i915_private *dev_priv)
 {
-	enum port port;
-
 	dev_priv->vbt.crt_ddc_pin = GMBUS_PIN_VGADDC;
 
 	/* Default to having backlight */
@@ -1721,13 +1881,6 @@ init_vbt_defaults(struct drm_i915_private *dev_priv)
 	dev_priv->vbt.lvds_ssc_freq = intel_bios_ssc_frequency(dev_priv,
 			!HAS_PCH_SPLIT(dev_priv));
 	DRM_DEBUG_KMS("Set default to SSC at %d kHz\n", dev_priv->vbt.lvds_ssc_freq);
-
-	for (port = PORT_A; port < I915_MAX_PORTS; port++) {
-		struct ddi_vbt_port_info *info =
-			&dev_priv->vbt.ddi_port_info[port];
-
-		info->hdmi_level_shift = HDMI_LEVEL_SHIFT_UNKNOWN;
-	}
 }
 
 /* Defaults to initialize only if there is no VBT. */
@@ -1736,7 +1889,7 @@ init_vbt_missing_defaults(struct drm_i915_private *dev_priv)
 {
 	enum port port;
 
-	for (port = PORT_A; port < I915_MAX_PORTS; port++) {
+	for_each_port(port) {
 		struct ddi_vbt_port_info *info =
 			&dev_priv->vbt.ddi_port_info[port];
 		enum phy phy = intel_port_to_phy(dev_priv, port);
@@ -1787,6 +1940,13 @@ bool intel_bios_is_valid_vbt(const void *buf, size_t size)
 		return false;
 	}
 
+	if (vbt->vbt_size > size) {
+		DRM_DEBUG_DRIVER("VBT incomplete (vbt_size overflows)\n");
+		return false;
+	}
+
+	size = vbt->vbt_size;
+
 	if (range_overflows_t(size_t,
 			      vbt->bdb_offset,
 			      sizeof(struct bdb_header),
@@ -1804,28 +1964,61 @@ bool intel_bios_is_valid_vbt(const void *buf, size_t size)
 	return vbt;
 }
 
-static const struct vbt_header *find_vbt(void __iomem *bios, size_t size)
+static struct vbt_header *oprom_get_vbt(struct drm_i915_private *dev_priv)
 {
-	size_t i;
+	struct pci_dev *pdev = dev_priv->drm.pdev;
+	void __iomem *p = NULL, *oprom;
+	struct vbt_header *vbt;
+	u16 vbt_size;
+	size_t i, size;
 
-	/* Scour memory looking for the VBT signature. */
-	for (i = 0; i + 4 < size; i++) {
-		void *vbt;
+	oprom = pci_map_rom(pdev, &size);
+	if (!oprom)
+		return NULL;
 
-		if (ioread32(bios + i) != *((const u32 *) "$VBT"))
+	/* Scour memory looking for the VBT signature. */
+	for (i = 0; i + 4 < size; i += 4) {
+		if (ioread32(oprom + i) != *((const u32 *)"$VBT"))
 			continue;
 
-		/*
-		 * This is the one place where we explicitly discard the address
-		 * space (__iomem) of the BIOS/VBT.
-		 */
-		vbt = (void __force *) bios + i;
-		if (intel_bios_is_valid_vbt(vbt, size - i))
-			return vbt;
-
+		p = oprom + i;
+		size -= i;
 		break;
 	}
 
+	if (!p)
+		goto err_unmap_oprom;
+
+	if (sizeof(struct vbt_header) > size) {
+		DRM_DEBUG_DRIVER("VBT header incomplete\n");
+		goto err_unmap_oprom;
+	}
+
+	vbt_size = ioread16(p + offsetof(struct vbt_header, vbt_size));
+	if (vbt_size > size) {
+		DRM_DEBUG_DRIVER("VBT incomplete (vbt_size overflows)\n");
+		goto err_unmap_oprom;
+	}
+
+	/* The rest will be validated by intel_bios_is_valid_vbt() */
+	vbt = kmalloc(vbt_size, GFP_KERNEL);
+	if (!vbt)
+		goto err_unmap_oprom;
+
+	memcpy_fromio(vbt, p, vbt_size);
+
+	if (!intel_bios_is_valid_vbt(vbt, vbt_size))
+		goto err_free_vbt;
+
+	pci_unmap_rom(pdev, oprom);
+
+	return vbt;
+
+err_free_vbt:
+	kfree(vbt);
+err_unmap_oprom:
+	pci_unmap_rom(pdev, oprom);
+
 	return NULL;
 }
 
@@ -1839,10 +2032,11 @@ static const struct vbt_header *find_vbt(void __iomem *bios, size_t size)
  */
 void intel_bios_init(struct drm_i915_private *dev_priv)
 {
-	struct pci_dev *pdev = dev_priv->drm.pdev;
 	const struct vbt_header *vbt = dev_priv->opregion.vbt;
+	struct vbt_header *oprom_vbt = NULL;
 	const struct bdb_header *bdb;
-	u8 __iomem *bios = NULL;
+
+	INIT_LIST_HEAD(&dev_priv->vbt.display_devices);
 
 	if (!HAS_DISPLAY(dev_priv) || !INTEL_DISPLAY_ENABLED(dev_priv)) {
 		DRM_DEBUG_KMS("Skipping VBT init due to disabled display.\n");
@@ -1853,15 +2047,11 @@ void intel_bios_init(struct drm_i915_private *dev_priv)
 
 	/* If the OpRegion does not have VBT, look in PCI ROM. */
 	if (!vbt) {
-		size_t size;
-
-		bios = pci_map_rom(pdev, &size);
-		if (!bios)
+		oprom_vbt = oprom_get_vbt(dev_priv);
+		if (!oprom_vbt)
 			goto out;
 
-		vbt = find_vbt(bios, size);
-		if (!vbt)
-			goto out;
+		vbt = oprom_vbt;
 
 		DRM_DEBUG_KMS("Found valid VBT in PCI ROM\n");
 	}
@@ -1874,15 +2064,20 @@ void intel_bios_init(struct drm_i915_private *dev_priv)
 	/* Grab useful general definitions */
 	parse_general_features(dev_priv, bdb);
 	parse_general_definitions(dev_priv, bdb);
-	parse_lfp_panel_data(dev_priv, bdb);
+	parse_panel_options(dev_priv, bdb);
+	parse_panel_dtd(dev_priv, bdb);
 	parse_lfp_backlight(dev_priv, bdb);
 	parse_sdvo_panel_data(dev_priv, bdb);
 	parse_driver_features(dev_priv, bdb);
+	parse_power_conservation_features(dev_priv, bdb);
 	parse_edp(dev_priv, bdb);
 	parse_psr(dev_priv, bdb);
 	parse_mipi_config(dev_priv, bdb);
 	parse_mipi_sequence(dev_priv, bdb);
 
+	/* Depends on child device list */
+	parse_compression_parameters(dev_priv, bdb);
+
 	/* Further processing on pre-parsed data */
 	parse_sdvo_device_mapping(dev_priv, bdb->version);
 	parse_ddi_ports(dev_priv, bdb->version);
@@ -1893,8 +2088,7 @@ out:
 		init_vbt_missing_defaults(dev_priv);
 	}
 
-	if (bios)
-		pci_unmap_rom(pdev, bios);
+	kfree(oprom_vbt);
 }
 
 /**
@@ -1903,9 +2097,14 @@ out:
  */
 void intel_bios_driver_remove(struct drm_i915_private *dev_priv)
 {
-	kfree(dev_priv->vbt.child_dev);
-	dev_priv->vbt.child_dev = NULL;
-	dev_priv->vbt.child_dev_num = 0;
+	struct display_device_data *devdata, *n;
+
+	list_for_each_entry_safe(devdata, n, &dev_priv->vbt.display_devices, node) {
+		list_del(&devdata->node);
+		kfree(devdata->dsc);
+		kfree(devdata);
+	}
+
 	kfree(dev_priv->vbt.sdvo_lvds_vbt_mode);
 	dev_priv->vbt.sdvo_lvds_vbt_mode = NULL;
 	kfree(dev_priv->vbt.lfp_lvds_vbt_mode);
@@ -1929,17 +2128,18 @@ void intel_bios_driver_remove(struct drm_i915_private *dev_priv)
  */
 bool intel_bios_is_tv_present(struct drm_i915_private *dev_priv)
 {
+	const struct display_device_data *devdata;
 	const struct child_device_config *child;
-	int i;
 
 	if (!dev_priv->vbt.int_tv_support)
 		return false;
 
-	if (!dev_priv->vbt.child_dev_num)
+	if (list_empty(&dev_priv->vbt.display_devices))
 		return true;
 
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		child = &devdata->child;
+
 		/*
 		 * If the device type is not TV, continue.
 		 */
@@ -1971,14 +2171,14 @@ bool intel_bios_is_tv_present(struct drm_i915_private *dev_priv)
  */
 bool intel_bios_is_lvds_present(struct drm_i915_private *dev_priv, u8 *i2c_pin)
 {
+	const struct display_device_data *devdata;
 	const struct child_device_config *child;
-	int i;
 
-	if (!dev_priv->vbt.child_dev_num)
+	if (list_empty(&dev_priv->vbt.display_devices))
 		return true;
 
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		child = &devdata->child;
 
 		/* If the device type is not LFP, continue.
 		 * We have to check both the new identifiers as well as the
@@ -2020,6 +2220,7 @@ bool intel_bios_is_lvds_present(struct drm_i915_private *dev_priv, u8 *i2c_pin)
  */
 bool intel_bios_is_port_present(struct drm_i915_private *dev_priv, enum port port)
 {
+	const struct display_device_data *devdata;
 	const struct child_device_config *child;
 	static const struct {
 		u16 dp, hdmi;
@@ -2030,7 +2231,6 @@ bool intel_bios_is_port_present(struct drm_i915_private *dev_priv, enum port por
 		[PORT_E] = { DVO_PORT_DPE, DVO_PORT_HDMIE, },
 		[PORT_F] = { DVO_PORT_DPF, DVO_PORT_HDMIF, },
 	};
-	int i;
 
 	if (HAS_DDI(dev_priv)) {
 		const struct ddi_vbt_port_info *port_info =
@@ -2045,11 +2245,8 @@ bool intel_bios_is_port_present(struct drm_i915_private *dev_priv, enum port por
 	if (WARN_ON(port == PORT_A) || port >= ARRAY_SIZE(port_mapping))
 		return false;
 
-	if (!dev_priv->vbt.child_dev_num)
-		return false;
-
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		child = &devdata->child;
 
 		if ((child->dvo_port == port_mapping[port].dp ||
 		     child->dvo_port == port_mapping[port].hdmi) &&
@@ -2070,6 +2267,7 @@ bool intel_bios_is_port_present(struct drm_i915_private *dev_priv, enum port por
  */
 bool intel_bios_is_port_edp(struct drm_i915_private *dev_priv, enum port port)
 {
+	const struct display_device_data *devdata;
 	const struct child_device_config *child;
 	static const short port_mapping[] = {
 		[PORT_B] = DVO_PORT_DPB,
@@ -2078,16 +2276,12 @@ bool intel_bios_is_port_edp(struct drm_i915_private *dev_priv, enum port port)
 		[PORT_E] = DVO_PORT_DPE,
 		[PORT_F] = DVO_PORT_DPF,
 	};
-	int i;
 
 	if (HAS_DDI(dev_priv))
 		return dev_priv->vbt.ddi_port_info[port].supports_edp;
 
-	if (!dev_priv->vbt.child_dev_num)
-		return false;
-
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		child = &devdata->child;
 
 		if (child->dvo_port == port_mapping[port] &&
 		    (child->device_type & DEVICE_TYPE_eDP_BITS) ==
@@ -2136,13 +2330,10 @@ static bool child_dev_is_dp_dual_mode(const struct child_device_config *child,
 bool intel_bios_is_port_dp_dual_mode(struct drm_i915_private *dev_priv,
 				     enum port port)
 {
-	const struct child_device_config *child;
-	int i;
-
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	const struct display_device_data *devdata;
 
-		if (child_dev_is_dp_dual_mode(child, port))
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		if (child_dev_is_dp_dual_mode(&devdata->child, port))
 			return true;
 	}
 
@@ -2159,12 +2350,12 @@ bool intel_bios_is_port_dp_dual_mode(struct drm_i915_private *dev_priv,
 bool intel_bios_is_dsi_present(struct drm_i915_private *dev_priv,
 			       enum port *port)
 {
+	const struct display_device_data *devdata;
 	const struct child_device_config *child;
 	u8 dvo_port;
-	int i;
 
-	for (i = 0; i < dev_priv->vbt.child_dev_num; i++) {
-		child = dev_priv->vbt.child_dev + i;
+	list_for_each_entry(devdata, &dev_priv->vbt.display_devices, node) {
+		child = &devdata->child;
 
 		if (!(child->device_type & DEVICE_TYPE_MIPI_OUTPUT))
 			continue;
@@ -2188,6 +2379,104 @@ bool intel_bios_is_dsi_present(struct drm_i915_private *dev_priv,
 	return false;
 }
 
+static void fill_dsc(struct intel_crtc_state *crtc_state,
+		     struct dsc_compression_parameters_entry *dsc,
+		     int dsc_max_bpc)
+{
+	struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
+	int bpc = 8;
+
+	vdsc_cfg->dsc_version_major = dsc->version_major;
+	vdsc_cfg->dsc_version_minor = dsc->version_minor;
+
+	if (dsc->support_12bpc && dsc_max_bpc >= 12)
+		bpc = 12;
+	else if (dsc->support_10bpc && dsc_max_bpc >= 10)
+		bpc = 10;
+	else if (dsc->support_8bpc && dsc_max_bpc >= 8)
+		bpc = 8;
+	else
+		DRM_DEBUG_KMS("VBT: Unsupported BPC %d for DCS\n",
+			      dsc_max_bpc);
+
+	crtc_state->pipe_bpp = bpc * 3;
+
+	crtc_state->dsc.compressed_bpp = min(crtc_state->pipe_bpp,
+					     VBT_DSC_MAX_BPP(dsc->max_bpp));
+
+	/*
+	 * FIXME: This is ugly, and slice count should take DSC engine
+	 * throughput etc. into account.
+	 *
+	 * Also, per spec DSI supports 1, 2, 3 or 4 horizontal slices.
+	 */
+	if (dsc->slices_per_line & BIT(2)) {
+		crtc_state->dsc.slice_count = 4;
+	} else if (dsc->slices_per_line & BIT(1)) {
+		crtc_state->dsc.slice_count = 2;
+	} else {
+		/* FIXME */
+		if (!(dsc->slices_per_line & BIT(0)))
+			DRM_DEBUG_KMS("VBT: Unsupported DSC slice count for DSI\n");
+
+		crtc_state->dsc.slice_count = 1;
+	}
+
+	if (crtc_state->hw.adjusted_mode.crtc_hdisplay %
+	    crtc_state->dsc.slice_count != 0)
+		DRM_DEBUG_KMS("VBT: DSC hdisplay %d not divisible by slice count %d\n",
+			      crtc_state->hw.adjusted_mode.crtc_hdisplay,
+			      crtc_state->dsc.slice_count);
+
+	/*
+	 * FIXME: Use VBT rc_buffer_block_size and rc_buffer_size for the
+	 * implementation specific physical rate buffer size. Currently we use
+	 * the required rate buffer model size calculated in
+	 * drm_dsc_compute_rc_parameters() according to VESA DSC Annex E.
+	 *
+	 * The VBT rc_buffer_block_size and rc_buffer_size definitions
+	 * correspond to DP 1.4 DPCD offsets 0x62 and 0x63. The DP DSC
+	 * implementation should also use the DPCD (or perhaps VBT for eDP)
+	 * provided value for the buffer size.
+	 */
+
+	/* FIXME: DSI spec says bpc + 1 for this one */
+	vdsc_cfg->line_buf_depth = VBT_DSC_LINE_BUFFER_DEPTH(dsc->line_buffer_depth);
+
+	vdsc_cfg->block_pred_enable = dsc->block_prediction_enable;
+
+	vdsc_cfg->slice_height = dsc->slice_height;
+}
+
+/* FIXME: initially DSI specific */
+bool intel_bios_get_dsc_params(struct intel_encoder *encoder,
+			       struct intel_crtc_state *crtc_state,
+			       int dsc_max_bpc)
+{
+	struct drm_i915_private *i915 = to_i915(encoder->base.dev);
+	const struct display_device_data *devdata;
+	const struct child_device_config *child;
+
+	list_for_each_entry(devdata, &i915->vbt.display_devices, node) {
+		child = &devdata->child;
+
+		if (!(child->device_type & DEVICE_TYPE_MIPI_OUTPUT))
+			continue;
+
+		if (child->dvo_port - DVO_PORT_MIPIA == encoder->port) {
+			if (!devdata->dsc)
+				return false;
+
+			if (crtc_state)
+				fill_dsc(crtc_state, devdata->dsc, dsc_max_bpc);
+
+			return true;
+		}
+	}
+
+	return false;
+}
+
 /**
  * intel_bios_is_port_hpd_inverted - is HPD inverted for %port
  * @i915:	i915 device instance
diff --git a/drivers/gpu/drm/i915/display/intel_bios.h b/drivers/gpu/drm/i915/display/intel_bios.h
index 98f064828a57..d6a0c29d37ac 100644
--- a/drivers/gpu/drm/i915/display/intel_bios.h
+++ b/drivers/gpu/drm/i915/display/intel_bios.h
@@ -35,6 +35,8 @@
 #include <drm/i915_drm.h>
 
 struct drm_i915_private;
+struct intel_crtc_state;
+struct intel_encoder;
 enum port;
 
 enum intel_backlight_type {
@@ -242,5 +244,8 @@ bool intel_bios_is_port_hpd_inverted(const struct drm_i915_private *i915,
 bool intel_bios_is_lspcon_present(const struct drm_i915_private *i915,
 				  enum port port);
 enum aux_ch intel_bios_port_aux_ch(struct drm_i915_private *dev_priv, enum port port);
+bool intel_bios_get_dsc_params(struct intel_encoder *encoder,
+			       struct intel_crtc_state *crtc_state,
+			       int dsc_max_bpc);
 
 #endif /* _INTEL_BIOS_H_ */
diff --git a/drivers/gpu/drm/i915/display/intel_bw.c b/drivers/gpu/drm/i915/display/intel_bw.c
index 22e83f857de8..dcb66a33be9b 100644
--- a/drivers/gpu/drm/i915/display/intel_bw.c
+++ b/drivers/gpu/drm/i915/display/intel_bw.c
@@ -15,7 +15,7 @@ struct intel_qgv_point {
 };
 
 struct intel_qgv_info {
-	struct intel_qgv_point points[3];
+	struct intel_qgv_point points[I915_NUM_QGV_POINTS];
 	u8 num_points;
 	u8 num_channels;
 	u8 t_bl;
@@ -264,6 +264,9 @@ static unsigned int icl_max_bw(struct drm_i915_private *dev_priv,
 
 void intel_bw_init_hw(struct drm_i915_private *dev_priv)
 {
+	if (!HAS_DISPLAY(dev_priv))
+		return;
+
 	if (IS_GEN(dev_priv, 12))
 		icl_get_bw_info(dev_priv, &tgl_sa_info);
 	else if (IS_GEN(dev_priv, 11))
@@ -273,17 +276,29 @@ void intel_bw_init_hw(struct drm_i915_private *dev_priv)
 static unsigned int intel_max_data_rate(struct drm_i915_private *dev_priv,
 					int num_planes)
 {
-	if (INTEL_GEN(dev_priv) >= 11)
+	if (INTEL_GEN(dev_priv) >= 11) {
+		/*
+		 * Any bw group has same amount of QGV points
+		 */
+		const struct intel_bw_info *bi =
+			&dev_priv->max_bw[0];
+		unsigned int min_bw = UINT_MAX;
+		int i;
+
 		/*
 		 * FIXME with SAGV disabled maybe we can assume
 		 * point 1 will always be used? Seems to match
 		 * the behaviour observed in the wild.
 		 */
-		return min3(icl_max_bw(dev_priv, num_planes, 0),
-			    icl_max_bw(dev_priv, num_planes, 1),
-			    icl_max_bw(dev_priv, num_planes, 2));
-	else
+		for (i = 0; i < bi->num_qgv_points; i++) {
+			unsigned int bw = icl_max_bw(dev_priv, num_planes, i);
+
+			min_bw = min(bw, min_bw);
+		}
+		return min_bw;
+	} else {
 		return UINT_MAX;
+	}
 }
 
 static unsigned int intel_bw_crtc_num_active_planes(const struct intel_crtc_state *crtc_state)
@@ -297,7 +312,7 @@ static unsigned int intel_bw_crtc_num_active_planes(const struct intel_crtc_stat
 
 static unsigned int intel_bw_crtc_data_rate(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	unsigned int data_rate = 0;
 	enum plane_id plane_id;
 
@@ -318,7 +333,7 @@ static unsigned int intel_bw_crtc_data_rate(const struct intel_crtc_state *crtc_
 void intel_bw_crtc_update(struct intel_bw_state *bw_state,
 			  const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 
 	bw_state->data_rate[crtc->pipe] =
 		intel_bw_crtc_data_rate(crtc_state);
diff --git a/drivers/gpu/drm/i915/display/intel_cdclk.c b/drivers/gpu/drm/i915/display/intel_cdclk.c
index ed8c7ce62119..7d1ab1e5b7c3 100644
--- a/drivers/gpu/drm/i915/display/intel_cdclk.c
+++ b/drivers/gpu/drm/i915/display/intel_cdclk.c
@@ -1904,7 +1904,7 @@ intel_set_cdclk_post_plane_update(struct drm_i915_private *dev_priv,
 
 static int intel_pixel_rate_to_cdclk(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	int pixel_rate = crtc_state->pixel_rate;
 
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
@@ -1922,7 +1922,7 @@ static int intel_pixel_rate_to_cdclk(const struct intel_crtc_state *crtc_state)
 
 static int intel_planes_min_cdclk(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_plane *plane;
 	int min_cdclk = 0;
@@ -1936,10 +1936,10 @@ static int intel_planes_min_cdclk(const struct intel_crtc_state *crtc_state)
 int intel_crtc_compute_min_cdclk(const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(crtc_state->base.crtc->dev);
+		to_i915(crtc_state->uapi.crtc->dev);
 	int min_cdclk;
 
-	if (!crtc_state->base.enable)
+	if (!crtc_state->hw.enable)
 		return 0;
 
 	min_cdclk = intel_pixel_rate_to_cdclk(crtc_state);
@@ -2076,7 +2076,7 @@ static int bxt_compute_min_voltage_level(struct intel_atomic_state *state)
 	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i) {
 		int ret;
 
-		if (crtc_state->base.enable)
+		if (crtc_state->hw.enable)
 			min_voltage_level = crtc_state->min_voltage_level;
 		else
 			min_voltage_level = 0;
@@ -2170,7 +2170,7 @@ static int skl_dpll0_vco(struct intel_atomic_state *state)
 		vco = dev_priv->skl_preferred_vco_freq;
 
 	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i) {
-		if (!crtc_state->base.enable)
+		if (!crtc_state->hw.enable)
 			continue;
 
 		if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_EDP))
@@ -2283,11 +2283,11 @@ static int intel_modeset_all_pipes(struct intel_atomic_state *state)
 		if (IS_ERR(crtc_state))
 			return PTR_ERR(crtc_state);
 
-		if (!crtc_state->base.active ||
-		    drm_atomic_crtc_needs_modeset(&crtc_state->base))
+		if (!crtc_state->hw.active ||
+		    drm_atomic_crtc_needs_modeset(&crtc_state->uapi))
 			continue;
 
-		crtc_state->base.mode_changed = true;
+		crtc_state->uapi.mode_changed = true;
 
 		ret = drm_atomic_add_affected_connectors(&state->base,
 							 &crtc->base);
@@ -2368,7 +2368,7 @@ int intel_modeset_calc_cdclk(struct intel_atomic_state *state)
 		if (IS_ERR(crtc_state))
 			return PTR_ERR(crtc_state);
 
-		if (drm_atomic_crtc_needs_modeset(&crtc_state->base))
+		if (drm_atomic_crtc_needs_modeset(&crtc_state->uapi))
 			pipe = INVALID_PIPE;
 	} else {
 		pipe = INVALID_PIPE;
diff --git a/drivers/gpu/drm/i915/display/intel_color.c b/drivers/gpu/drm/i915/display/intel_color.c
index aa3a063549c3..3980e8b50c28 100644
--- a/drivers/gpu/drm/i915/display/intel_color.c
+++ b/drivers/gpu/drm/i915/display/intel_color.c
@@ -117,10 +117,10 @@ static bool lut_is_legacy(const struct drm_property_blob *lut)
 
 static bool crtc_state_is_legacy_gamma(const struct intel_crtc_state *crtc_state)
 {
-	return !crtc_state->base.degamma_lut &&
-		!crtc_state->base.ctm &&
-		crtc_state->base.gamma_lut &&
-		lut_is_legacy(crtc_state->base.gamma_lut);
+	return !crtc_state->hw.degamma_lut &&
+		!crtc_state->hw.ctm &&
+		crtc_state->hw.gamma_lut &&
+		lut_is_legacy(crtc_state->hw.gamma_lut);
 }
 
 /*
@@ -205,7 +205,7 @@ static void icl_update_output_csc(struct intel_crtc *crtc,
 
 static bool ilk_csc_limited_range(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	/*
 	 * FIXME if there's a gamma LUT after the CSC, we should
@@ -219,7 +219,7 @@ static bool ilk_csc_limited_range(const struct intel_crtc_state *crtc_state)
 static void ilk_csc_convert_ctm(const struct intel_crtc_state *crtc_state,
 				u16 coeffs[9])
 {
-	const struct drm_color_ctm *ctm = crtc_state->base.ctm->data;
+	const struct drm_color_ctm *ctm = crtc_state->hw.ctm->data;
 	const u64 *input;
 	u64 temp[9];
 	int i;
@@ -270,11 +270,11 @@ static void ilk_csc_convert_ctm(const struct intel_crtc_state *crtc_state,
 
 static void ilk_load_csc_matrix(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	bool limited_color_range = ilk_csc_limited_range(crtc_state);
 
-	if (crtc_state->base.ctm) {
+	if (crtc_state->hw.ctm) {
 		u16 coeff[9];
 
 		ilk_csc_convert_ctm(crtc_state, coeff);
@@ -309,10 +309,10 @@ static void ilk_load_csc_matrix(const struct intel_crtc_state *crtc_state)
 
 static void icl_load_csc_matrix(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
-	if (crtc_state->base.ctm) {
+	if (crtc_state->hw.ctm) {
 		u16 coeff[9];
 
 		ilk_csc_convert_ctm(crtc_state, coeff);
@@ -338,12 +338,12 @@ static void icl_load_csc_matrix(const struct intel_crtc_state *crtc_state)
  */
 static void cherryview_load_csc_matrix(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
-	if (crtc_state->base.ctm) {
-		const struct drm_color_ctm *ctm = crtc_state->base.ctm->data;
+	if (crtc_state->hw.ctm) {
+		const struct drm_color_ctm *ctm = crtc_state->hw.ctm->data;
 		u16 coeffs[9] = {};
 		int i;
 
@@ -404,7 +404,7 @@ static u32 ilk_lut_10(const struct drm_color_lut *color)
 static void i9xx_load_luts_internal(const struct intel_crtc_state *crtc_state,
 				    const struct drm_property_blob *blob)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	int i;
@@ -435,12 +435,12 @@ static void i9xx_load_luts_internal(const struct intel_crtc_state *crtc_state,
 
 static void i9xx_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	i9xx_load_luts_internal(crtc_state, crtc_state->base.gamma_lut);
+	i9xx_load_luts_internal(crtc_state, crtc_state->hw.gamma_lut);
 }
 
 static void i9xx_color_commit(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	u32 val;
@@ -453,7 +453,7 @@ static void i9xx_color_commit(const struct intel_crtc_state *crtc_state)
 
 static void ilk_color_commit(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	u32 val;
@@ -468,7 +468,7 @@ static void ilk_color_commit(const struct intel_crtc_state *crtc_state)
 
 static void hsw_color_commit(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	I915_WRITE(GAMMA_MODE(crtc->pipe), crtc_state->gamma_mode);
@@ -478,7 +478,7 @@ static void hsw_color_commit(const struct intel_crtc_state *crtc_state)
 
 static void skl_color_commit(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	u32 val = 0;
@@ -524,8 +524,8 @@ static void i965_load_lut_10p6(struct intel_crtc *crtc,
 
 static void i965_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
 		i9xx_load_luts(crtc_state);
@@ -547,8 +547,8 @@ static void ilk_load_lut_10(struct intel_crtc *crtc,
 
 static void ilk_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
 		i9xx_load_luts(crtc_state);
@@ -654,9 +654,9 @@ static void ivb_load_lut_ext_max(struct intel_crtc *crtc)
 
 static void ivb_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
-	const struct drm_property_blob *degamma_lut = crtc_state->base.degamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	const struct drm_property_blob *degamma_lut = crtc_state->hw.degamma_lut;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT) {
 		i9xx_load_luts(crtc_state);
@@ -677,9 +677,9 @@ static void ivb_load_luts(const struct intel_crtc_state *crtc_state)
 
 static void bdw_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
-	const struct drm_property_blob *degamma_lut = crtc_state->base.degamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	const struct drm_property_blob *degamma_lut = crtc_state->hw.degamma_lut;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT) {
 		i9xx_load_luts(crtc_state);
@@ -700,11 +700,11 @@ static void bdw_load_luts(const struct intel_crtc_state *crtc_state)
 
 static void glk_load_degamma_lut(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	const u32 lut_size = INTEL_INFO(dev_priv)->color.degamma_lut_size;
-	const struct drm_color_lut *lut = crtc_state->base.degamma_lut->data;
+	const struct drm_color_lut *lut = crtc_state->hw.degamma_lut->data;
 	u32 i;
 
 	/*
@@ -739,7 +739,7 @@ static void glk_load_degamma_lut(const struct intel_crtc_state *crtc_state)
 
 static void glk_load_degamma_lut_linear(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	const u32 lut_size = INTEL_INFO(dev_priv)->color.degamma_lut_size;
@@ -766,8 +766,8 @@ static void glk_load_degamma_lut_linear(const struct intel_crtc_state *crtc_stat
 
 static void glk_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 
 	/*
 	 * On GLK+ both pipe CSC and degamma LUT are controlled
@@ -777,7 +777,7 @@ static void glk_load_luts(const struct intel_crtc_state *crtc_state)
 	 * the degama LUT so that we don't have to reload
 	 * it every time the pipe CSC is being enabled.
 	 */
-	if (crtc_state->base.degamma_lut)
+	if (crtc_state->hw.degamma_lut)
 		glk_load_degamma_lut(crtc_state);
 	else
 		glk_load_degamma_lut_linear(crtc_state);
@@ -808,7 +808,7 @@ static void
 icl_load_gcmax(const struct intel_crtc_state *crtc_state,
 	       const struct drm_color_lut *color)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
 
@@ -822,8 +822,8 @@ icl_load_gcmax(const struct intel_crtc_state *crtc_state,
 static void
 icl_program_gamma_superfine_segment(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *blob = crtc_state->base.gamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *blob = crtc_state->hw.gamma_lut;
 	const struct drm_color_lut *lut = blob->data;
 	struct intel_dsb *dsb = intel_dsb_get(crtc);
 	enum pipe pipe = crtc->pipe;
@@ -854,8 +854,8 @@ icl_program_gamma_superfine_segment(const struct intel_crtc_state *crtc_state)
 static void
 icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *blob = crtc_state->base.gamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *blob = crtc_state->hw.gamma_lut;
 	const struct drm_color_lut *lut = blob->data;
 	const struct drm_color_lut *entry;
 	struct intel_dsb *dsb = intel_dsb_get(crtc);
@@ -910,11 +910,11 @@ icl_program_gamma_multi_segment(const struct intel_crtc_state *crtc_state)
 
 static void icl_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_dsb *dsb = intel_dsb_get(crtc);
 
-	if (crtc_state->base.degamma_lut)
+	if (crtc_state->hw.degamma_lut)
 		glk_load_degamma_lut(crtc_state);
 
 	switch (crtc_state->gamma_mode & GAMMA_MODE_MODE_MASK) {
@@ -990,9 +990,9 @@ static void chv_load_cgm_gamma(struct intel_crtc *crtc,
 
 static void chv_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
-	const struct drm_property_blob *degamma_lut = crtc_state->base.degamma_lut;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	const struct drm_property_blob *degamma_lut = crtc_state->hw.degamma_lut;
 
 	cherryview_load_csc_matrix(crtc_state);
 
@@ -1010,35 +1010,35 @@ static void chv_load_luts(const struct intel_crtc_state *crtc_state)
 
 void intel_color_load_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	dev_priv->display.load_luts(crtc_state);
 }
 
 void intel_color_commit(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	dev_priv->display.color_commit(crtc_state);
 }
 
 static bool intel_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(state, crtc);
 
-	return !old_crtc_state->base.gamma_lut &&
-		!old_crtc_state->base.degamma_lut;
+	return !old_crtc_state->hw.gamma_lut &&
+		!old_crtc_state->hw.degamma_lut;
 }
 
 static bool chv_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(state, crtc);
 
@@ -1050,14 +1050,14 @@ static bool chv_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
 	if (old_crtc_state->cgm_mode || new_crtc_state->cgm_mode)
 		return false;
 
-	return !old_crtc_state->base.gamma_lut;
+	return !old_crtc_state->hw.gamma_lut;
 }
 
 static bool glk_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(state, crtc);
 
@@ -1068,19 +1068,19 @@ static bool glk_can_preload_luts(const struct intel_crtc_state *new_crtc_state)
 	 * linear hardware degamma mid scanout.
 	 */
 	return !old_crtc_state->csc_enable &&
-		!old_crtc_state->base.gamma_lut;
+		!old_crtc_state->hw.gamma_lut;
 }
 
 int intel_color_check(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	return dev_priv->display.color_check(crtc_state);
 }
 
 void intel_color_get_config(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	if (dev_priv->display.read_luts)
 		dev_priv->display.read_luts(crtc_state);
@@ -1104,16 +1104,16 @@ static bool need_plane_update(struct intel_plane *plane,
 static int
 intel_color_add_affected_planes(struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(state, crtc);
 	struct intel_plane *plane;
 
-	if (!new_crtc_state->base.active ||
-	    drm_atomic_crtc_needs_modeset(&new_crtc_state->base))
+	if (!new_crtc_state->hw.active ||
+	    drm_atomic_crtc_needs_modeset(&new_crtc_state->uapi))
 		return 0;
 
 	if (new_crtc_state->gamma_enable == old_crtc_state->gamma_enable &&
@@ -1155,9 +1155,9 @@ static int check_lut_size(const struct drm_property_blob *lut, int expected)
 
 static int check_luts(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	const struct drm_property_blob *gamma_lut = crtc_state->base.gamma_lut;
-	const struct drm_property_blob *degamma_lut = crtc_state->base.degamma_lut;
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
+	const struct drm_property_blob *gamma_lut = crtc_state->hw.gamma_lut;
+	const struct drm_property_blob *degamma_lut = crtc_state->hw.degamma_lut;
 	int gamma_length, degamma_length;
 	u32 gamma_tests, degamma_tests;
 
@@ -1205,7 +1205,7 @@ static int i9xx_color_check(struct intel_crtc_state *crtc_state)
 		return ret;
 
 	crtc_state->gamma_enable =
-		crtc_state->base.gamma_lut &&
+		crtc_state->hw.gamma_lut &&
 		!crtc_state->c8_planes;
 
 	crtc_state->gamma_mode = i9xx_gamma_mode(crtc_state);
@@ -1226,11 +1226,11 @@ static u32 chv_cgm_mode(const struct intel_crtc_state *crtc_state)
 	if (crtc_state_is_legacy_gamma(crtc_state))
 		return 0;
 
-	if (crtc_state->base.degamma_lut)
+	if (crtc_state->hw.degamma_lut)
 		cgm_mode |= CGM_PIPE_MODE_DEGAMMA;
-	if (crtc_state->base.ctm)
+	if (crtc_state->hw.ctm)
 		cgm_mode |= CGM_PIPE_MODE_CSC;
-	if (crtc_state->base.gamma_lut)
+	if (crtc_state->hw.gamma_lut)
 		cgm_mode |= CGM_PIPE_MODE_GAMMA;
 
 	return cgm_mode;
@@ -1306,7 +1306,7 @@ static int ilk_color_check(struct intel_crtc_state *crtc_state)
 		return ret;
 
 	crtc_state->gamma_enable =
-		crtc_state->base.gamma_lut &&
+		crtc_state->hw.gamma_lut &&
 		!crtc_state->c8_planes;
 
 	/*
@@ -1334,8 +1334,8 @@ static u32 ivb_gamma_mode(const struct intel_crtc_state *crtc_state)
 	if (!crtc_state->gamma_enable ||
 	    crtc_state_is_legacy_gamma(crtc_state))
 		return GAMMA_MODE_MODE_8BIT;
-	else if (crtc_state->base.gamma_lut &&
-		 crtc_state->base.degamma_lut)
+	else if (crtc_state->hw.gamma_lut &&
+		 crtc_state->hw.degamma_lut)
 		return GAMMA_MODE_MODE_SPLIT;
 	else
 		return GAMMA_MODE_MODE_10BIT;
@@ -1349,7 +1349,7 @@ static u32 ivb_csc_mode(const struct intel_crtc_state *crtc_state)
 	 * CSC comes after the LUT in degamma, RGB->YCbCr,
 	 * and RGB full->limited range mode.
 	 */
-	if (crtc_state->base.degamma_lut ||
+	if (crtc_state->hw.degamma_lut ||
 	    crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB ||
 	    limited_color_range)
 		return 0;
@@ -1367,13 +1367,13 @@ static int ivb_color_check(struct intel_crtc_state *crtc_state)
 		return ret;
 
 	crtc_state->gamma_enable =
-		(crtc_state->base.gamma_lut ||
-		 crtc_state->base.degamma_lut) &&
+		(crtc_state->hw.gamma_lut ||
+		 crtc_state->hw.degamma_lut) &&
 		!crtc_state->c8_planes;
 
 	crtc_state->csc_enable =
 		crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB ||
-		crtc_state->base.ctm || limited_color_range;
+		crtc_state->hw.ctm || limited_color_range;
 
 	crtc_state->gamma_mode = ivb_gamma_mode(crtc_state);
 
@@ -1406,14 +1406,14 @@ static int glk_color_check(struct intel_crtc_state *crtc_state)
 		return ret;
 
 	crtc_state->gamma_enable =
-		crtc_state->base.gamma_lut &&
+		crtc_state->hw.gamma_lut &&
 		!crtc_state->c8_planes;
 
 	/* On GLK+ degamma LUT is controlled by csc_enable */
 	crtc_state->csc_enable =
-		crtc_state->base.degamma_lut ||
+		crtc_state->hw.degamma_lut ||
 		crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB ||
-		crtc_state->base.ctm || crtc_state->limited_color_range;
+		crtc_state->hw.ctm || crtc_state->limited_color_range;
 
 	crtc_state->gamma_mode = glk_gamma_mode(crtc_state);
 
@@ -1432,14 +1432,14 @@ static u32 icl_gamma_mode(const struct intel_crtc_state *crtc_state)
 {
 	u32 gamma_mode = 0;
 
-	if (crtc_state->base.degamma_lut)
+	if (crtc_state->hw.degamma_lut)
 		gamma_mode |= PRE_CSC_GAMMA_ENABLE;
 
-	if (crtc_state->base.gamma_lut &&
+	if (crtc_state->hw.gamma_lut &&
 	    !crtc_state->c8_planes)
 		gamma_mode |= POST_CSC_GAMMA_ENABLE;
 
-	if (!crtc_state->base.gamma_lut ||
+	if (!crtc_state->hw.gamma_lut ||
 	    crtc_state_is_legacy_gamma(crtc_state))
 		gamma_mode |= GAMMA_MODE_MODE_8BIT;
 	else
@@ -1452,7 +1452,7 @@ static u32 icl_csc_mode(const struct intel_crtc_state *crtc_state)
 {
 	u32 csc_mode = 0;
 
-	if (crtc_state->base.ctm)
+	if (crtc_state->hw.ctm)
 		csc_mode |= ICL_CSC_ENABLE;
 
 	if (crtc_state->output_format != INTEL_OUTPUT_FORMAT_RGB ||
@@ -1540,7 +1540,7 @@ static int glk_gamma_precision(const struct intel_crtc_state *crtc_state)
 
 int intel_color_get_gamma_bit_precision(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (HAS_GMCH(dev_priv)) {
@@ -1646,7 +1646,7 @@ static u32 intel_color_lut_pack(u32 val, u32 bit_precision)
 static struct drm_property_blob *
 i9xx_read_lut_8(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	struct drm_property_blob *blob;
@@ -1683,13 +1683,13 @@ static void i9xx_read_luts(struct intel_crtc_state *crtc_state)
 	if (!crtc_state->gamma_enable)
 		return;
 
-	crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+	crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
 }
 
 static struct drm_property_blob *
 i965_read_lut_10p6(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 	enum pipe pipe = crtc->pipe;
@@ -1733,15 +1733,15 @@ static void i965_read_luts(struct intel_crtc_state *crtc_state)
 		return;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
 	else
-		crtc_state->base.gamma_lut = i965_read_lut_10p6(crtc_state);
+		crtc_state->hw.gamma_lut = i965_read_lut_10p6(crtc_state);
 }
 
 static struct drm_property_blob *
 chv_read_cgm_lut(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 	enum pipe pipe = crtc->pipe;
@@ -1775,7 +1775,7 @@ chv_read_cgm_lut(const struct intel_crtc_state *crtc_state)
 static void chv_read_luts(struct intel_crtc_state *crtc_state)
 {
 	if (crtc_state->cgm_mode & CGM_PIPE_MODE_GAMMA)
-		crtc_state->base.gamma_lut = chv_read_cgm_lut(crtc_state);
+		crtc_state->hw.gamma_lut = chv_read_cgm_lut(crtc_state);
 	else
 		i965_read_luts(crtc_state);
 }
@@ -1783,7 +1783,7 @@ static void chv_read_luts(struct intel_crtc_state *crtc_state)
 static struct drm_property_blob *
 ilk_read_lut_10(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 lut_size = INTEL_INFO(dev_priv)->color.gamma_lut_size;
 	enum pipe pipe = crtc->pipe;
@@ -1822,15 +1822,15 @@ static void ilk_read_luts(struct intel_crtc_state *crtc_state)
 		return;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
 	else
-		crtc_state->base.gamma_lut = ilk_read_lut_10(crtc_state);
+		crtc_state->hw.gamma_lut = ilk_read_lut_10(crtc_state);
 }
 
 static struct drm_property_blob *
 glk_read_lut_10(const struct intel_crtc_state *crtc_state, u32 prec_index)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	int hw_lut_size = ivb_lut_10_size(prec_index);
 	enum pipe pipe = crtc->pipe;
@@ -1871,9 +1871,9 @@ static void glk_read_luts(struct intel_crtc_state *crtc_state)
 		return;
 
 	if (crtc_state->gamma_mode == GAMMA_MODE_MODE_8BIT)
-		crtc_state->base.gamma_lut = i9xx_read_lut_8(crtc_state);
+		crtc_state->hw.gamma_lut = i9xx_read_lut_8(crtc_state);
 	else
-		crtc_state->base.gamma_lut = glk_read_lut_10(crtc_state, PAL_PREC_INDEX_VALUE(0));
+		crtc_state->hw.gamma_lut = glk_read_lut_10(crtc_state, PAL_PREC_INDEX_VALUE(0));
 }
 
 void intel_color_init(struct intel_crtc *crtc)
diff --git a/drivers/gpu/drm/i915/display/intel_crt.c b/drivers/gpu/drm/i915/display/intel_crt.c
index 39cc6d79dc85..b2b1336ecdb6 100644
--- a/drivers/gpu/drm/i915/display/intel_crt.c
+++ b/drivers/gpu/drm/i915/display/intel_crt.c
@@ -132,9 +132,9 @@ static void intel_crt_get_config(struct intel_encoder *encoder,
 {
 	pipe_config->output_types |= BIT(INTEL_OUTPUT_ANALOG);
 
-	pipe_config->base.adjusted_mode.flags |= intel_crt_get_flags(encoder);
+	pipe_config->hw.adjusted_mode.flags |= intel_crt_get_flags(encoder);
 
-	pipe_config->base.adjusted_mode.crtc_clock = pipe_config->port_clock;
+	pipe_config->hw.adjusted_mode.crtc_clock = pipe_config->port_clock;
 }
 
 static void hsw_crt_get_config(struct intel_encoder *encoder,
@@ -144,13 +144,13 @@ static void hsw_crt_get_config(struct intel_encoder *encoder,
 
 	intel_ddi_get_config(encoder, pipe_config);
 
-	pipe_config->base.adjusted_mode.flags &= ~(DRM_MODE_FLAG_PHSYNC |
+	pipe_config->hw.adjusted_mode.flags &= ~(DRM_MODE_FLAG_PHSYNC |
 					      DRM_MODE_FLAG_NHSYNC |
 					      DRM_MODE_FLAG_PVSYNC |
 					      DRM_MODE_FLAG_NVSYNC);
-	pipe_config->base.adjusted_mode.flags |= intel_crt_get_flags(encoder);
+	pipe_config->hw.adjusted_mode.flags |= intel_crt_get_flags(encoder);
 
-	pipe_config->base.adjusted_mode.crtc_clock = lpt_get_iclkip(dev_priv);
+	pipe_config->hw.adjusted_mode.crtc_clock = lpt_get_iclkip(dev_priv);
 }
 
 /* Note: The caller is required to filter out dpms modes not supported by the
@@ -161,8 +161,8 @@ static void intel_crt_set_dpms(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_crt *crt = intel_encoder_to_crt(encoder);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_display_mode *adjusted_mode = &crtc_state->hw.adjusted_mode;
 	u32 adpa;
 
 	if (INTEL_GEN(dev_priv) >= 5)
@@ -241,6 +241,14 @@ static void hsw_post_disable_crt(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 
+	intel_crtc_vblank_off(old_crtc_state);
+
+	intel_disable_pipe(old_crtc_state);
+
+	intel_ddi_disable_transcoder_func(old_crtc_state);
+
+	ironlake_pfit_disable(old_crtc_state);
+
 	intel_ddi_disable_pipe_clock(old_crtc_state);
 
 	pch_post_disable_crt(encoder, old_crtc_state, old_conn_state);
@@ -271,14 +279,14 @@ static void hsw_pre_enable_crt(struct intel_encoder *encoder,
 			       const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum pipe pipe = crtc->pipe;
 
 	WARN_ON(!crtc_state->has_pch_encoder);
 
 	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 
-	dev_priv->display.fdi_link_train(crtc, crtc_state);
+	hsw_fdi_link_train(encoder, crtc_state);
 
 	intel_ddi_enable_pipe_clock(crtc_state);
 }
@@ -288,7 +296,7 @@ static void hsw_enable_crt(struct intel_encoder *encoder,
 			   const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum pipe pipe = crtc->pipe;
 
 	WARN_ON(!crtc_state->has_pch_encoder);
@@ -358,7 +366,7 @@ static int intel_crt_compute_config(struct intel_encoder *encoder,
 				    struct drm_connector_state *conn_state)
 {
 	struct drm_display_mode *adjusted_mode =
-		&pipe_config->base.adjusted_mode;
+		&pipe_config->hw.adjusted_mode;
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN)
 		return -EINVAL;
@@ -373,7 +381,7 @@ static int pch_crt_compute_config(struct intel_encoder *encoder,
 				  struct drm_connector_state *conn_state)
 {
 	struct drm_display_mode *adjusted_mode =
-		&pipe_config->base.adjusted_mode;
+		&pipe_config->hw.adjusted_mode;
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN)
 		return -EINVAL;
@@ -390,7 +398,7 @@ static int hsw_crt_compute_config(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct drm_display_mode *adjusted_mode =
-		&pipe_config->base.adjusted_mode;
+		&pipe_config->hw.adjusted_mode;
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLSCAN)
 		return -EINVAL;
diff --git a/drivers/gpu/drm/i915/display/intel_ddi.c b/drivers/gpu/drm/i915/display/intel_ddi.c
index 2a27fb5d7dc6..c9ba7d7f3787 100644
--- a/drivers/gpu/drm/i915/display/intel_ddi.c
+++ b/drivers/gpu/drm/i915/display/intel_ddi.c
@@ -902,11 +902,10 @@ icl_get_combo_buf_trans(struct drm_i915_private *dev_priv, int type, int rate,
 
 static int intel_ddi_hdmi_level(struct drm_i915_private *dev_priv, enum port port)
 {
+	struct ddi_vbt_port_info *port_info = &dev_priv->vbt.ddi_port_info[port];
 	int n_entries, level, default_entry;
 	enum phy phy = intel_port_to_phy(dev_priv, port);
 
-	level = dev_priv->vbt.ddi_port_info[port].hdmi_level_shift;
-
 	if (INTEL_GEN(dev_priv) >= 12) {
 		if (intel_phy_is_combo(dev_priv, phy))
 			icl_get_combo_buf_trans(dev_priv, INTEL_OUTPUT_HDMI,
@@ -941,12 +940,14 @@ static int intel_ddi_hdmi_level(struct drm_i915_private *dev_priv, enum port por
 		return 0;
 	}
 
-	/* Choose a good default if VBT is badly populated */
-	if (level == HDMI_LEVEL_SHIFT_UNKNOWN || level >= n_entries)
-		level = default_entry;
-
 	if (WARN_ON_ONCE(n_entries == 0))
 		return 0;
+
+	if (port_info->hdmi_level_shift_set)
+		level = port_info->hdmi_level_shift;
+	else
+		level = default_entry;
+
 	if (WARN_ON_ONCE(level >= n_entries))
 		level = n_entries - 1;
 
@@ -1106,18 +1107,14 @@ static u32 icl_pll_to_ddi_clk_sel(struct intel_encoder *encoder,
  * DDI A (which is used for eDP)
  */
 
-void hsw_fdi_link_train(struct intel_crtc *crtc,
+void hsw_fdi_link_train(struct intel_encoder *encoder,
 			const struct intel_crtc_state *crtc_state)
 {
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_encoder *encoder;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 temp, i, rx_ctl_val, ddi_pll_sel;
 
-	for_each_encoder_on_crtc(dev, &crtc->base, encoder) {
-		WARN_ON(encoder->type != INTEL_OUTPUT_ANALOG);
-		intel_prepare_dp_ddi_buffers(encoder, crtc_state);
-	}
+	intel_prepare_dp_ddi_buffers(encoder, crtc_state);
 
 	/* Set the FDI_RX_MISC pwrdn lanes and the 2 workarounds listed at the
 	 * mode set "sequence for CRT port" document:
@@ -1542,7 +1539,7 @@ static void ddi_dotclock_get(struct intel_crtc_state *pipe_config)
 	if (pipe_config->pixel_multiplier)
 		dotclock /= pipe_config->pixel_multiplier;
 
-	pipe_config->base.adjusted_mode.crtc_clock = dotclock;
+	pipe_config->hw.adjusted_mode.crtc_clock = dotclock;
 }
 
 static void icl_ddi_clock_get(struct intel_encoder *encoder,
@@ -1758,7 +1755,7 @@ static void intel_ddi_clock_get(struct intel_encoder *encoder,
 void intel_ddi_set_dp_msa(const struct intel_crtc_state *crtc_state,
 			  const struct drm_connector_state *conn_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 temp;
@@ -1815,22 +1812,6 @@ void intel_ddi_set_dp_msa(const struct intel_crtc_state *crtc_state,
 	I915_WRITE(TRANS_MSA_MISC(cpu_transcoder), temp);
 }
 
-void intel_ddi_set_vc_payload_alloc(const struct intel_crtc_state *crtc_state,
-				    bool state)
-{
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
-	u32 temp;
-
-	temp = I915_READ(TRANS_DDI_FUNC_CTL(cpu_transcoder));
-	if (state == true)
-		temp |= TRANS_DDI_DP_VC_PAYLOAD_ALLOC;
-	else
-		temp &= ~TRANS_DDI_DP_VC_PAYLOAD_ALLOC;
-	I915_WRITE(TRANS_DDI_FUNC_CTL(cpu_transcoder), temp);
-}
-
 /*
  * Returns the TRANS_DDI_FUNC_CTL value based on CRTC state.
  *
@@ -1840,7 +1821,7 @@ void intel_ddi_set_vc_payload_alloc(const struct intel_crtc_state *crtc_state,
 static u32
 intel_ddi_transcoder_func_reg_val_get(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_encoder *encoder = intel_ddi_get_crtc_encoder(crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
@@ -1872,9 +1853,9 @@ intel_ddi_transcoder_func_reg_val_get(const struct intel_crtc_state *crtc_state)
 		BUG();
 	}
 
-	if (crtc_state->base.adjusted_mode.flags & DRM_MODE_FLAG_PVSYNC)
+	if (crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_PVSYNC)
 		temp |= TRANS_DDI_PVSYNC;
-	if (crtc_state->base.adjusted_mode.flags & DRM_MODE_FLAG_PHSYNC)
+	if (crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_PHSYNC)
 		temp |= TRANS_DDI_PHSYNC;
 
 	if (cpu_transcoder == TRANSCODER_EDP) {
@@ -1930,12 +1911,14 @@ intel_ddi_transcoder_func_reg_val_get(const struct intel_crtc_state *crtc_state)
 
 void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 temp;
 
 	temp = intel_ddi_transcoder_func_reg_val_get(crtc_state);
+	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DP_MST))
+		temp |= TRANS_DDI_DP_VC_PAYLOAD_ALLOC;
 	I915_WRITE(TRANS_DDI_FUNC_CTL(cpu_transcoder), temp);
 }
 
@@ -1946,7 +1929,7 @@ void intel_ddi_enable_transcoder_func(const struct intel_crtc_state *crtc_state)
 static void
 intel_ddi_config_transcoder_func(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 temp;
@@ -1958,7 +1941,7 @@ intel_ddi_config_transcoder_func(const struct intel_crtc_state *crtc_state)
 
 void intel_ddi_disable_transcoder_func(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	i915_reg_t reg = TRANS_DDI_FUNC_CTL(cpu_transcoder);
@@ -2256,7 +2239,7 @@ static void intel_ddi_get_power_domains(struct intel_encoder *encoder,
 
 void intel_ddi_enable_pipe_clock(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_encoder *encoder = intel_ddi_get_crtc_encoder(crtc);
 	enum port port = encoder->port;
@@ -2274,7 +2257,7 @@ void intel_ddi_enable_pipe_clock(const struct intel_crtc_state *crtc_state)
 
 void intel_ddi_disable_pipe_clock(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 
 	if (cpu_transcoder != TRANSCODER_EDP) {
@@ -3016,11 +2999,38 @@ static void icl_unmap_plls_to_ports(struct intel_encoder *encoder)
 	mutex_unlock(&dev_priv->dpll_lock);
 }
 
+static void icl_sanitize_port_clk_off(struct drm_i915_private *dev_priv,
+				      u32 port_mask, bool ddi_clk_needed)
+{
+	enum port port;
+	u32 val;
+
+	val = I915_READ(ICL_DPCLKA_CFGCR0);
+	for_each_port_masked(port, port_mask) {
+		enum phy phy = intel_port_to_phy(dev_priv, port);
+		bool ddi_clk_off = val & icl_dpclka_cfgcr0_clk_off(dev_priv,
+								   phy);
+
+		if (ddi_clk_needed == !ddi_clk_off)
+			continue;
+
+		/*
+		 * Punt on the case now where clock is gated, but it would
+		 * be needed by the port. Something else is really broken then.
+		 */
+		if (WARN_ON(ddi_clk_needed))
+			continue;
+
+		DRM_NOTE("PHY %c is disabled/in DSI mode with an ungated DDI clock, gate it\n",
+			 phy_name(phy));
+		val |= icl_dpclka_cfgcr0_clk_off(dev_priv, phy);
+		I915_WRITE(ICL_DPCLKA_CFGCR0, val);
+	}
+}
+
 void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	u32 val;
-	enum port port;
 	u32 port_mask;
 	bool ddi_clk_needed;
 
@@ -3069,29 +3079,7 @@ void icl_sanitize_encoder_pll_mapping(struct intel_encoder *encoder)
 		ddi_clk_needed = false;
 	}
 
-	val = I915_READ(ICL_DPCLKA_CFGCR0);
-	for_each_port_masked(port, port_mask) {
-		enum phy phy = intel_port_to_phy(dev_priv, port);
-
-		bool ddi_clk_ungated = !(val &
-					 icl_dpclka_cfgcr0_clk_off(dev_priv,
-								   phy));
-
-		if (ddi_clk_needed == ddi_clk_ungated)
-			continue;
-
-		/*
-		 * Punt on the case now where clock is gated, but it would
-		 * be needed by the port. Something else is really broken then.
-		 */
-		if (WARN_ON(ddi_clk_needed))
-			continue;
-
-		DRM_NOTE("PHY %c is disabled/in DSI mode with an ungated DDI clock, gate it\n",
-			 phy_name(port));
-		val |= icl_dpclka_cfgcr0_clk_off(dev_priv, phy);
-		I915_WRITE(ICL_DPCLKA_CFGCR0, val);
-	}
+	icl_sanitize_port_clk_off(dev_priv, port_mask, ddi_clk_needed);
 }
 
 static void intel_ddi_clk_select(struct intel_encoder *encoder,
@@ -3359,7 +3347,7 @@ static void intel_ddi_disable_fec_state(struct intel_encoder *encoder,
 static void
 tgl_clear_psr2_transcoder_exitline(const struct intel_crtc_state *cstate)
 {
-	struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(cstate->uapi.crtc->dev);
 	u32 val;
 
 	if (!cstate->dc3co_exitline)
@@ -3374,7 +3362,7 @@ static void
 tgl_set_psr2_transcoder_exitline(const struct intel_crtc_state *cstate)
 {
 	u32 val, exit_scanlines;
-	struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(cstate->uapi.crtc->dev);
 
 	if (!cstate->dc3co_exitline)
 		return;
@@ -3392,8 +3380,8 @@ static void tgl_dc3co_exitline_compute_config(struct intel_encoder *encoder,
 					      struct intel_crtc_state *cstate)
 {
 	u32 exit_scanlines;
-	struct drm_i915_private *dev_priv = to_i915(cstate->base.crtc->dev);
-	u32 crtc_vdisplay = cstate->base.adjusted_mode.crtc_vdisplay;
+	struct drm_i915_private *dev_priv = to_i915(cstate->uapi.crtc->dev);
+	u32 crtc_vdisplay = cstate->hw.adjusted_mode.crtc_vdisplay;
 
 	cstate->dc3co_exitline = 0;
 
@@ -3401,11 +3389,11 @@ static void tgl_dc3co_exitline_compute_config(struct intel_encoder *encoder,
 		return;
 
 	/* B.Specs:49196 DC3CO only works with pipeA and DDIA.*/
-	if (to_intel_crtc(cstate->base.crtc)->pipe != PIPE_A ||
+	if (to_intel_crtc(cstate->uapi.crtc)->pipe != PIPE_A ||
 	    encoder->port != PORT_A)
 		return;
 
-	if (!cstate->has_psr2 || !cstate->base.active)
+	if (!cstate->has_psr2 || !cstate->hw.active)
 		return;
 
 	/*
@@ -3413,7 +3401,7 @@ static void tgl_dc3co_exitline_compute_config(struct intel_encoder *encoder,
 	 * PSR2 transcoder Early Exit scanlines = ROUNDUP(200 / line time) + 1
 	 */
 	exit_scanlines =
-		intel_usecs_to_scanlines(&cstate->base.adjusted_mode, 200) + 1;
+		intel_usecs_to_scanlines(&cstate->hw.adjusted_mode, 200) + 1;
 
 	if (WARN_ON(exit_scanlines > crtc_vdisplay))
 		return;
@@ -3425,7 +3413,7 @@ static void tgl_dc3co_exitline_compute_config(struct intel_encoder *encoder,
 static void tgl_dc3co_exitline_get_config(struct intel_crtc_state *crtc_state)
 {
 	u32 val;
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	if (INTEL_GEN(dev_priv) < 12)
 		return;
@@ -3455,47 +3443,86 @@ static void tgl_ddi_pre_enable_dp(struct intel_encoder *encoder,
 	intel_dp->regs.dp_tp_ctl = TGL_DP_TP_CTL(transcoder);
 	intel_dp->regs.dp_tp_status = TGL_DP_TP_STATUS(transcoder);
 
-	/* 1.a got on intel_atomic_commit_tail() */
+	/*
+	 * 1. Enable Power Wells
+	 *
+	 * This was handled at the beginning of intel_atomic_commit_tail(),
+	 * before we called down into this function.
+	 */
 
-	/* 2. */
+	/* 2. Enable Panel Power if PPS is required */
 	intel_edp_panel_on(intel_dp);
 
 	/*
-	 * 1.b, 3. and 4.a is done before tgl_ddi_pre_enable_dp() by:
-	 * haswell_crtc_enable()->intel_encoders_pre_pll_enable() and
-	 * haswell_crtc_enable()->intel_enable_shared_dpll()
+	 * 3. For non-TBT Type-C ports, set FIA lane count
+	 * (DFLEXDPSP.DPX4TXLATC)
+	 *
+	 * This was done before tgl_ddi_pre_enable_dp by
+	 * haswell_crtc_enable()->intel_encoders_pre_pll_enable().
 	 */
 
-	/* 4.b */
+	/*
+	 * 4. Enable the port PLL.
+	 *
+	 * The PLL enabling itself was already done before this function by
+	 * haswell_crtc_enable()->intel_enable_shared_dpll().  We need only
+	 * configure the PLL to port mapping here.
+	 */
 	intel_ddi_clk_select(encoder, crtc_state);
 
-	/* 5. */
+	/* 5. If IO power is controlled through PWR_WELL_CTL, Enable IO Power */
 	if (!intel_phy_is_tc(dev_priv, phy) ||
 	    dig_port->tc_mode != TC_PORT_TBT_ALT)
 		intel_display_power_get(dev_priv,
 					dig_port->ddi_io_power_domain);
 
-	/* 6. */
+	/* 6. Program DP_MODE */
 	icl_program_mg_dp_mode(dig_port, crtc_state);
 
 	/*
-	 * 7.a - Steps in this function should only be executed over MST
-	 * master, what will be taken in care by MST hook
-	 * intel_mst_pre_enable_dp()
+	 * 7. The rest of the below are substeps under the bspec's "Enable and
+	 * Train Display Port" step.  Note that steps that are specific to
+	 * MST will be handled by intel_mst_pre_enable_dp() before/after it
+	 * calls into this function.  Also intel_mst_pre_enable_dp() only calls
+	 * us when active_mst_links==0, so any steps designated for "single
+	 * stream or multi-stream master transcoder" can just be performed
+	 * unconditionally here.
+	 */
+
+	/*
+	 * 7.a Configure Transcoder Clock Select to direct the Port clock to the
+	 * Transcoder.
 	 */
 	intel_ddi_enable_pipe_clock(crtc_state);
 
-	/* 7.b */
+	/*
+	 * 7.b Configure TRANS_DDI_FUNC_CTL DDI Select, DDI Mode Select & MST
+	 * Transport Select
+	 */
 	intel_ddi_config_transcoder_func(crtc_state);
 
-	/* 7.d */
+	/*
+	 * 7.c Configure & enable DP_TP_CTL with link training pattern 1
+	 * selected
+	 *
+	 * This will be handled by the intel_dp_start_link_train() farther
+	 * down this function.
+	 */
+
+	/*
+	 * 7.d Type C with DP alternate or fixed/legacy/static connection -
+	 * Disable PHY clock gating per Type-C DDI Buffer page
+	 */
 	icl_phy_set_clock_gating(dig_port, false);
 
-	/* 7.e */
+	/* 7.e Configure voltage swing and related IO settings */
 	tgl_ddi_vswing_sequence(encoder, crtc_state->port_clock, level,
 				encoder->type);
 
-	/* 7.f */
+	/*
+	 * 7.f Combo PHY: Configure PORT_CL_DW10 Static Power Down to power up
+	 * the used lanes of the DDI.
+	 */
 	if (intel_phy_is_combo(dev_priv, phy)) {
 		bool lane_reversal =
 			dig_port->saved_port_bits & DDI_BUF_PORT_REVERSAL;
@@ -3505,7 +3532,14 @@ static void tgl_ddi_pre_enable_dp(struct intel_encoder *encoder,
 					       lane_reversal);
 	}
 
-	/* 7.g */
+	/*
+	 * 7.g Configure and enable DDI_BUF_CTL
+	 * 7.h Wait for DDI_BUF_CTL DDI Idle Status = 0b (Not Idle), timeout
+	 *     after 500 us.
+	 *
+	 * We only configure what the register value will be here.  Actual
+	 * enabling happens during link training farther down.
+	 */
 	intel_ddi_init_dp_buf_reg(encoder);
 
 	if (!is_mst)
@@ -3518,10 +3552,17 @@ static void tgl_ddi_pre_enable_dp(struct intel_encoder *encoder,
 	 * training
 	 */
 	intel_dp_sink_set_fec_ready(intel_dp, crtc_state);
-	/* 7.c, 7.h, 7.i, 7.j */
+
+	/*
+	 * 7.i Follow DisplayPort specification training sequence (see notes for
+	 *     failure handling)
+	 * 7.j If DisplayPort multi-stream - Set DP_TP_CTL link training to Idle
+	 *     Pattern, wait for 5 idle patterns (DP_TP_STATUS Min_Idles_Sent)
+	 *     (timeout after 800 us)
+	 */
 	intel_dp_start_link_train(intel_dp);
 
-	/* 7.k */
+	/* 7.k Set DP_TP_CTL link training to Normal */
 	if (!is_trans_port_sync_mode(crtc_state))
 		intel_dp_stop_link_train(intel_dp);
 
@@ -3534,7 +3575,7 @@ static void tgl_ddi_pre_enable_dp(struct intel_encoder *encoder,
 	 * so not enabling it for now.
 	 */
 
-	/* 7.l */
+	/* 7.l Configure and enable FEC if needed */
 	intel_ddi_enable_fec(encoder, crtc_state);
 	intel_dsc_enable(encoder, crtc_state);
 }
@@ -3677,7 +3718,7 @@ static void intel_ddi_pre_enable(struct intel_encoder *encoder,
 				 const struct intel_crtc_state *crtc_state,
 				 const struct drm_connector_state *conn_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
@@ -3761,17 +3802,25 @@ static void intel_ddi_post_disable_dp(struct intel_encoder *encoder,
 					  INTEL_OUTPUT_DP_MST);
 	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
 
-	if (!is_mst) {
+	/*
+	 * Power down sink before disabling the port, otherwise we end
+	 * up getting interrupts from the sink on detecting link loss.
+	 */
+	intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
+
+	if (INTEL_GEN(dev_priv) < 12 && !is_mst)
 		intel_ddi_disable_pipe_clock(old_crtc_state);
-		/*
-		 * Power down sink before disabling the port, otherwise we end
-		 * up getting interrupts from the sink on detecting link loss.
-		 */
-		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
-	}
 
 	intel_disable_ddi_buf(encoder, old_crtc_state);
 
+	/*
+	 * From TGL spec: "If single stream or multi-stream master transcoder:
+	 * Configure Transcoder Clock select to direct no clock to the
+	 * transcoder"
+	 */
+	if (INTEL_GEN(dev_priv) >= 12)
+		intel_ddi_disable_pipe_clock(old_crtc_state);
+
 	intel_edp_panel_vdd_on(intel_dp);
 	intel_edp_panel_off(intel_dp);
 
@@ -3807,11 +3856,49 @@ static void intel_ddi_post_disable_hdmi(struct intel_encoder *encoder,
 	intel_dp_dual_mode_set_tmds_output(intel_hdmi, false);
 }
 
+static void icl_disable_transcoder_port_sync(const struct intel_crtc_state *old_crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	i915_reg_t reg;
+	u32 trans_ddi_func_ctl2_val;
+
+	if (old_crtc_state->master_transcoder == INVALID_TRANSCODER)
+		return;
+
+	DRM_DEBUG_KMS("Disabling Transcoder Port Sync on Slave Transcoder %s\n",
+		      transcoder_name(old_crtc_state->cpu_transcoder));
+
+	reg = TRANS_DDI_FUNC_CTL2(old_crtc_state->cpu_transcoder);
+	trans_ddi_func_ctl2_val = ~(PORT_SYNC_MODE_ENABLE |
+				    PORT_SYNC_MODE_MASTER_SELECT_MASK);
+	I915_WRITE(reg, trans_ddi_func_ctl2_val);
+}
+
 static void intel_ddi_post_disable(struct intel_encoder *encoder,
 				   const struct intel_crtc_state *old_crtc_state,
 				   const struct drm_connector_state *old_conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	struct intel_digital_port *dig_port = enc_to_dig_port(&encoder->base);
+	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
+	bool is_tc_port = intel_phy_is_tc(dev_priv, phy);
+
+	intel_crtc_vblank_off(old_crtc_state);
+
+	intel_disable_pipe(old_crtc_state);
+
+	if (INTEL_GEN(dev_priv) >= 11)
+		icl_disable_transcoder_port_sync(old_crtc_state);
+
+	intel_ddi_disable_transcoder_func(old_crtc_state);
+
+	intel_dsc_disable(old_crtc_state);
+
+	if (INTEL_GEN(dev_priv) >= 9)
+		skylake_scaler_disable(old_crtc_state);
+	else
+		ironlake_pfit_disable(old_crtc_state);
 
 	/*
 	 * When called from DP MST code:
@@ -3835,6 +3922,13 @@ static void intel_ddi_post_disable(struct intel_encoder *encoder,
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		icl_unmap_plls_to_ports(encoder);
+
+	if (intel_crtc_has_dp_encoder(old_crtc_state) || is_tc_port)
+		intel_display_power_put_unchecked(dev_priv,
+						  intel_ddi_main_link_aux_domain(dig_port));
+
+	if (is_tc_port)
+		intel_tc_port_put_link(dig_port);
 }
 
 void intel_ddi_fdi_post_disable(struct intel_encoder *encoder,
@@ -4107,7 +4201,7 @@ intel_ddi_update_prepare(struct intel_atomic_state *state,
 	WARN_ON(crtc && crtc->active);
 
 	intel_tc_port_get_link(enc_to_dig_port(&encoder->base), required_lanes);
-	if (crtc_state && crtc_state->base.active)
+	if (crtc_state && crtc_state->hw.active)
 		intel_update_active_dpll(state, crtc, encoder);
 }
 
@@ -4147,61 +4241,44 @@ intel_ddi_pre_pll_enable(struct intel_encoder *encoder,
 						crtc_state->lane_lat_optim_mask);
 }
 
-static void
-intel_ddi_post_pll_disable(struct intel_encoder *encoder,
-			   const struct intel_crtc_state *crtc_state,
-			   const struct drm_connector_state *conn_state)
-{
-	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_digital_port *dig_port = enc_to_dig_port(&encoder->base);
-	enum phy phy = intel_port_to_phy(dev_priv, encoder->port);
-	bool is_tc_port = intel_phy_is_tc(dev_priv, phy);
-
-	if (intel_crtc_has_dp_encoder(crtc_state) || is_tc_port)
-		intel_display_power_put_unchecked(dev_priv,
-						  intel_ddi_main_link_aux_domain(dig_port));
-
-	if (is_tc_port)
-		intel_tc_port_put_link(dig_port);
-}
-
 static void intel_ddi_prepare_link_retrain(struct intel_dp *intel_dp)
 {
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
 	struct drm_i915_private *dev_priv =
 		to_i915(intel_dig_port->base.base.dev);
 	enum port port = intel_dig_port->base.port;
-	u32 val;
+	u32 dp_tp_ctl, ddi_buf_ctl;
 	bool wait = false;
 
-	if (I915_READ(intel_dp->regs.dp_tp_ctl) & DP_TP_CTL_ENABLE) {
-		val = I915_READ(DDI_BUF_CTL(port));
-		if (val & DDI_BUF_CTL_ENABLE) {
-			val &= ~DDI_BUF_CTL_ENABLE;
-			I915_WRITE(DDI_BUF_CTL(port), val);
+	dp_tp_ctl = I915_READ(intel_dp->regs.dp_tp_ctl);
+
+	if (dp_tp_ctl & DP_TP_CTL_ENABLE) {
+		ddi_buf_ctl = I915_READ(DDI_BUF_CTL(port));
+		if (ddi_buf_ctl & DDI_BUF_CTL_ENABLE) {
+			I915_WRITE(DDI_BUF_CTL(port),
+				   ddi_buf_ctl & ~DDI_BUF_CTL_ENABLE);
 			wait = true;
 		}
 
-		val = I915_READ(intel_dp->regs.dp_tp_ctl);
-		val &= ~(DP_TP_CTL_ENABLE | DP_TP_CTL_LINK_TRAIN_MASK);
-		val |= DP_TP_CTL_LINK_TRAIN_PAT1;
-		I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
+		dp_tp_ctl &= ~(DP_TP_CTL_ENABLE | DP_TP_CTL_LINK_TRAIN_MASK);
+		dp_tp_ctl |= DP_TP_CTL_LINK_TRAIN_PAT1;
+		I915_WRITE(intel_dp->regs.dp_tp_ctl, dp_tp_ctl);
 		POSTING_READ(intel_dp->regs.dp_tp_ctl);
 
 		if (wait)
 			intel_wait_ddi_buf_idle(dev_priv, port);
 	}
 
-	val = DP_TP_CTL_ENABLE |
-	      DP_TP_CTL_LINK_TRAIN_PAT1 | DP_TP_CTL_SCRAMBLE_DISABLE;
+	dp_tp_ctl = DP_TP_CTL_ENABLE |
+		    DP_TP_CTL_LINK_TRAIN_PAT1 | DP_TP_CTL_SCRAMBLE_DISABLE;
 	if (intel_dp->link_mst)
-		val |= DP_TP_CTL_MODE_MST;
+		dp_tp_ctl |= DP_TP_CTL_MODE_MST;
 	else {
-		val |= DP_TP_CTL_MODE_SST;
+		dp_tp_ctl |= DP_TP_CTL_MODE_SST;
 		if (drm_dp_enhanced_frame_cap(intel_dp->dpcd))
-			val |= DP_TP_CTL_ENHANCED_FRAME_ENABLE;
+			dp_tp_ctl |= DP_TP_CTL_ENHANCED_FRAME_ENABLE;
 	}
-	I915_WRITE(intel_dp->regs.dp_tp_ctl, val);
+	I915_WRITE(intel_dp->regs.dp_tp_ctl, dp_tp_ctl);
 	POSTING_READ(intel_dp->regs.dp_tp_ctl);
 
 	intel_dp->DP |= DDI_BUF_CTL_ENABLE;
@@ -4237,7 +4314,7 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 			  struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	enum transcoder cpu_transcoder = pipe_config->cpu_transcoder;
 	u32 temp, flags = 0;
 
@@ -4245,6 +4322,8 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 	if (WARN_ON(transcoder_is_dsi(cpu_transcoder)))
 		return;
 
+	intel_dsc_get_config(encoder, pipe_config);
+
 	temp = I915_READ(TRANS_DDI_FUNC_CTL(cpu_transcoder));
 	if (temp & TRANS_DDI_PHSYNC)
 		flags |= DRM_MODE_FLAG_PHSYNC;
@@ -4255,7 +4334,7 @@ void intel_ddi_get_config(struct intel_encoder *encoder,
 	else
 		flags |= DRM_MODE_FLAG_NVSYNC;
 
-	pipe_config->base.adjusted_mode.flags |= flags;
+	pipe_config->hw.adjusted_mode.flags |= flags;
 
 	switch (temp & TRANS_DDI_BPC_MASK) {
 	case TRANS_DDI_BPC_6:
@@ -4404,7 +4483,7 @@ static int intel_ddi_compute_config(struct intel_encoder *encoder,
 				    struct intel_crtc_state *pipe_config,
 				    struct drm_connector_state *conn_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	enum port port = encoder->port;
 	int ret;
@@ -4538,7 +4617,7 @@ static int intel_hdmi_reset_link(struct intel_encoder *encoder,
 
 	WARN_ON(!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_HDMI));
 
-	if (!crtc_state->base.active)
+	if (!crtc_state->hw.active)
 		return 0;
 
 	if (!crtc_state->hdmi_high_tmds_clock_ratio &&
@@ -4709,8 +4788,7 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 	struct ddi_vbt_port_info *port_info =
 		&dev_priv->vbt.ddi_port_info[port];
 	struct intel_digital_port *intel_dig_port;
-	struct intel_encoder *intel_encoder;
-	struct drm_encoder *encoder;
+	struct intel_encoder *encoder;
 	bool init_hdmi, init_dp, init_lspcon = false;
 	enum phy phy = intel_port_to_phy(dev_priv, port);
 
@@ -4739,31 +4817,30 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 	if (!intel_dig_port)
 		return;
 
-	intel_encoder = &intel_dig_port->base;
-	encoder = &intel_encoder->base;
+	encoder = &intel_dig_port->base;
 
-	drm_encoder_init(&dev_priv->drm, encoder, &intel_ddi_funcs,
+	drm_encoder_init(&dev_priv->drm, &encoder->base, &intel_ddi_funcs,
 			 DRM_MODE_ENCODER_TMDS, "DDI %c", port_name(port));
 
-	intel_encoder->hotplug = intel_ddi_hotplug;
-	intel_encoder->compute_output_type = intel_ddi_compute_output_type;
-	intel_encoder->compute_config = intel_ddi_compute_config;
-	intel_encoder->enable = intel_enable_ddi;
-	intel_encoder->pre_pll_enable = intel_ddi_pre_pll_enable;
-	intel_encoder->post_pll_disable = intel_ddi_post_pll_disable;
-	intel_encoder->pre_enable = intel_ddi_pre_enable;
-	intel_encoder->disable = intel_disable_ddi;
-	intel_encoder->post_disable = intel_ddi_post_disable;
-	intel_encoder->update_pipe = intel_ddi_update_pipe;
-	intel_encoder->get_hw_state = intel_ddi_get_hw_state;
-	intel_encoder->get_config = intel_ddi_get_config;
-	intel_encoder->suspend = intel_dp_encoder_suspend;
-	intel_encoder->get_power_domains = intel_ddi_get_power_domains;
-	intel_encoder->type = INTEL_OUTPUT_DDI;
-	intel_encoder->power_domain = intel_port_to_power_domain(port);
-	intel_encoder->port = port;
-	intel_encoder->cloneable = 0;
-	intel_encoder->pipe_mask = ~0;
+	encoder->hotplug = intel_ddi_hotplug;
+	encoder->compute_output_type = intel_ddi_compute_output_type;
+	encoder->compute_config = intel_ddi_compute_config;
+	encoder->enable = intel_enable_ddi;
+	encoder->pre_pll_enable = intel_ddi_pre_pll_enable;
+	encoder->pre_enable = intel_ddi_pre_enable;
+	encoder->disable = intel_disable_ddi;
+	encoder->post_disable = intel_ddi_post_disable;
+	encoder->update_pipe = intel_ddi_update_pipe;
+	encoder->get_hw_state = intel_ddi_get_hw_state;
+	encoder->get_config = intel_ddi_get_config;
+	encoder->suspend = intel_dp_encoder_suspend;
+	encoder->get_power_domains = intel_ddi_get_power_domains;
+
+	encoder->type = INTEL_OUTPUT_DDI;
+	encoder->power_domain = intel_port_to_power_domain(port);
+	encoder->port = port;
+	encoder->cloneable = 0;
+	encoder->pipe_mask = ~0;
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		intel_dig_port->saved_port_bits = I915_READ(DDI_BUF_CTL(port)) &
@@ -4771,6 +4848,7 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 	else
 		intel_dig_port->saved_port_bits = I915_READ(DDI_BUF_CTL(port)) &
 			(DDI_BUF_PORT_REVERSAL | DDI_A_4_LANES);
+
 	intel_dig_port->dp.output_reg = INVALID_MMIO_REG;
 	intel_dig_port->max_lanes = intel_ddi_max_lanes(intel_dig_port);
 	intel_dig_port->aux_ch = intel_bios_port_aux_ch(dev_priv, port);
@@ -4781,8 +4859,8 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 
 		intel_tc_port_init(intel_dig_port, is_legacy);
 
-		intel_encoder->update_prepare = intel_ddi_update_prepare;
-		intel_encoder->update_complete = intel_ddi_update_complete;
+		encoder->update_prepare = intel_ddi_update_prepare;
+		encoder->update_complete = intel_ddi_update_complete;
 	}
 
 	WARN_ON(port > PORT_I);
@@ -4798,7 +4876,7 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 
 	/* In theory we don't need the encoder->type check, but leave it just in
 	 * case we have some really bad VBTs... */
-	if (intel_encoder->type != INTEL_OUTPUT_EDP && init_hdmi) {
+	if (encoder->type != INTEL_OUTPUT_EDP && init_hdmi) {
 		if (!intel_ddi_init_hdmi_connector(intel_dig_port))
 			goto err;
 	}
@@ -4822,6 +4900,6 @@ void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port)
 	return;
 
 err:
-	drm_encoder_cleanup(encoder);
+	drm_encoder_cleanup(&encoder->base);
 	kfree(intel_dig_port);
 }
diff --git a/drivers/gpu/drm/i915/display/intel_ddi.h b/drivers/gpu/drm/i915/display/intel_ddi.h
index 19aeab1246ee..167c6579d972 100644
--- a/drivers/gpu/drm/i915/display/intel_ddi.h
+++ b/drivers/gpu/drm/i915/display/intel_ddi.h
@@ -22,7 +22,7 @@ struct intel_encoder;
 void intel_ddi_fdi_post_disable(struct intel_encoder *intel_encoder,
 				const struct intel_crtc_state *old_crtc_state,
 				const struct drm_connector_state *old_conn_state);
-void hsw_fdi_link_train(struct intel_crtc *crtc,
+void hsw_fdi_link_train(struct intel_encoder *encoder,
 			const struct intel_crtc_state *crtc_state);
 void intel_ddi_init(struct drm_i915_private *dev_priv, enum port port);
 bool intel_ddi_get_hw_state(struct intel_encoder *encoder, enum pipe *pipe);
diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
index 434f563603a6..1860da0a493e 100644
--- a/drivers/gpu/drm/i915/display/intel_display.c
+++ b/drivers/gpu/drm/i915/display/intel_display.c
@@ -86,8 +86,8 @@
 /* Primary plane formats for gen <= 3 */
 static const u32 i8xx_primary_formats[] = {
 	DRM_FORMAT_C8,
-	DRM_FORMAT_RGB565,
 	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_RGB565,
 	DRM_FORMAT_XRGB8888,
 };
 
@@ -112,6 +112,21 @@ static const u32 i965_primary_formats[] = {
 	DRM_FORMAT_XBGR16161616F,
 };
 
+/* Primary plane formats for vlv/chv */
+static const u32 vlv_primary_formats[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ARGB2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR16161616F,
+};
+
 static const u64 i9xx_format_modifiers[] = {
 	I915_FORMAT_MOD_X_TILED,
 	DRM_FORMAT_MOD_LINEAR,
@@ -149,14 +164,11 @@ static void vlv_prepare_pll(struct intel_crtc *crtc,
 			    const struct intel_crtc_state *pipe_config);
 static void chv_prepare_pll(struct intel_crtc *crtc,
 			    const struct intel_crtc_state *pipe_config);
-static void intel_crtc_init_scalers(struct intel_crtc *crtc,
-				    struct intel_crtc_state *crtc_state);
 static void skylake_pfit_enable(const struct intel_crtc_state *crtc_state);
-static void ironlake_pfit_disable(const struct intel_crtc_state *old_crtc_state);
 static void ironlake_pfit_enable(const struct intel_crtc_state *crtc_state);
 static void intel_modeset_setup_hw_state(struct drm_device *dev,
 					 struct drm_modeset_acquire_ctx *ctx);
-static void intel_pre_disable_primary_noatomic(struct drm_crtc *crtc);
+static struct intel_crtc_state *intel_crtc_state_alloc(struct intel_crtc *crtc);
 
 struct intel_limit {
 	struct {
@@ -530,7 +542,7 @@ icl_wa_scalerclkgating(struct drm_i915_private *dev_priv, enum pipe pipe,
 static bool
 needs_modeset(const struct intel_crtc_state *state)
 {
-	return drm_atomic_crtc_needs_modeset(&state->base);
+	return drm_atomic_crtc_needs_modeset(&state->uapi);
 }
 
 bool
@@ -547,6 +559,12 @@ is_trans_port_sync_master(const struct intel_crtc_state *crtc_state)
 		crtc_state->sync_mode_slaves_mask);
 }
 
+static bool
+is_trans_port_sync_slave(const struct intel_crtc_state *crtc_state)
+{
+	return crtc_state->master_transcoder != INVALID_TRANSCODER;
+}
+
 /*
  * Platform specific helpers to calculate the port PLL loopback- (clock.m),
  * and post-divider (clock.p) values, pre- (clock.vco) and post-divided fast
@@ -658,7 +676,7 @@ i9xx_select_p2_div(const struct intel_limit *limit,
 		   const struct intel_crtc_state *crtc_state,
 		   int target)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	if (intel_crtc_has_type(crtc_state, INTEL_OUTPUT_LVDS)) {
 		/*
@@ -694,7 +712,7 @@ i9xx_find_best_dpll(const struct intel_limit *limit,
 		    int target, int refclk, struct dpll *match_clock,
 		    struct dpll *best_clock)
 {
-	struct drm_device *dev = crtc_state->base.crtc->dev;
+	struct drm_device *dev = crtc_state->uapi.crtc->dev;
 	struct dpll clock;
 	int err = target;
 
@@ -752,7 +770,7 @@ pnv_find_best_dpll(const struct intel_limit *limit,
 		   int target, int refclk, struct dpll *match_clock,
 		   struct dpll *best_clock)
 {
-	struct drm_device *dev = crtc_state->base.crtc->dev;
+	struct drm_device *dev = crtc_state->uapi.crtc->dev;
 	struct dpll clock;
 	int err = target;
 
@@ -808,7 +826,7 @@ g4x_find_best_dpll(const struct intel_limit *limit,
 		   int target, int refclk, struct dpll *match_clock,
 		   struct dpll *best_clock)
 {
-	struct drm_device *dev = crtc_state->base.crtc->dev;
+	struct drm_device *dev = crtc_state->uapi.crtc->dev;
 	struct dpll clock;
 	int max_n;
 	bool found = false;
@@ -902,7 +920,7 @@ vlv_find_best_dpll(const struct intel_limit *limit,
 		   int target, int refclk, struct dpll *match_clock,
 		   struct dpll *best_clock)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_device *dev = crtc->base.dev;
 	struct dpll clock;
 	unsigned int bestppm = 1000000;
@@ -962,7 +980,7 @@ chv_find_best_dpll(const struct intel_limit *limit,
 		   int target, int refclk, struct dpll *match_clock,
 		   struct dpll *best_clock)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_device *dev = crtc->base.dev;
 	unsigned int best_error_ppm;
 	struct dpll clock;
@@ -1025,33 +1043,6 @@ bool bxt_find_best_dpll(struct intel_crtc_state *crtc_state,
 				  NULL, best_clock);
 }
 
-bool intel_crtc_active(struct intel_crtc *crtc)
-{
-	/* Be paranoid as we can arrive here with only partial
-	 * state retrieved from the hardware during setup.
-	 *
-	 * We can ditch the adjusted_mode.crtc_clock check as soon
-	 * as Haswell has gained clock readout/fastboot support.
-	 *
-	 * We can ditch the crtc->primary->state->fb check as soon as we can
-	 * properly reconstruct framebuffers.
-	 *
-	 * FIXME: The intel_crtc->active here should be switched to
-	 * crtc->state->active once we have proper CRTC states wired up
-	 * for atomic.
-	 */
-	return crtc->active && crtc->base.primary->state->fb &&
-		crtc->config->base.adjusted_mode.crtc_clock;
-}
-
-enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
-					     enum pipe pipe)
-{
-	struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
-
-	return crtc->config->cpu_transcoder;
-}
-
 static bool pipe_scanline_is_moving(struct drm_i915_private *dev_priv,
 				    enum pipe pipe)
 {
@@ -1095,7 +1086,7 @@ static void intel_wait_for_pipe_scanline_moving(struct intel_crtc *crtc)
 static void
 intel_wait_for_pipe_off(const struct intel_crtc_state *old_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (INTEL_GEN(dev_priv) >= 4) {
@@ -1145,11 +1136,15 @@ static void assert_fdi_tx(struct drm_i915_private *dev_priv,
 			  enum pipe pipe, bool state)
 {
 	bool cur_state;
-	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
-								      pipe);
 
 	if (HAS_DDI(dev_priv)) {
-		/* DDI does not have a specific FDI_TX register */
+		/*
+		 * DDI does not have a specific FDI_TX register.
+		 *
+		 * FDI is never fed from EDP transcoder
+		 * so pipe->transcoder cast is fine here.
+		 */
+		enum transcoder cpu_transcoder = (enum transcoder)pipe;
 		u32 val = I915_READ(TRANS_DDI_FUNC_CTL(cpu_transcoder));
 		cur_state = !!(val & TRANS_DDI_FUNC_ENABLE);
 	} else {
@@ -1266,11 +1261,9 @@ void assert_panel_unlocked(struct drm_i915_private *dev_priv, enum pipe pipe)
 }
 
 void assert_pipe(struct drm_i915_private *dev_priv,
-		 enum pipe pipe, bool state)
+		 enum transcoder cpu_transcoder, bool state)
 {
 	bool cur_state;
-	enum transcoder cpu_transcoder = intel_pipe_to_cpu_transcoder(dev_priv,
-								      pipe);
 	enum intel_display_power_domain power_domain;
 	intel_wakeref_t wakeref;
 
@@ -1290,8 +1283,9 @@ void assert_pipe(struct drm_i915_private *dev_priv,
 	}
 
 	I915_STATE_WARN(cur_state != state,
-	     "pipe %c assertion failure (expected %s, current %s)\n",
-			pipe_name(pipe), onoff(state), onoff(cur_state));
+			"transcoder %s assertion failure (expected %s, current %s)\n",
+			transcoder_name(cpu_transcoder),
+			onoff(state), onoff(cur_state));
 }
 
 static void assert_plane(struct intel_plane *plane, bool state)
@@ -1418,7 +1412,7 @@ static void vlv_enable_pll(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
-	assert_pipe_disabled(dev_priv, pipe);
+	assert_pipe_disabled(dev_priv, pipe_config->cpu_transcoder);
 
 	/* PLL is protected by panel, make sure we can write it */
 	assert_panel_unlocked(dev_priv, pipe);
@@ -1467,7 +1461,7 @@ static void chv_enable_pll(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
-	assert_pipe_disabled(dev_priv, pipe);
+	assert_pipe_disabled(dev_priv, pipe_config->cpu_transcoder);
 
 	/* PLL is protected by panel, make sure we can write it */
 	assert_panel_unlocked(dev_priv, pipe);
@@ -1514,7 +1508,7 @@ static void i9xx_enable_pll(struct intel_crtc *crtc,
 	u32 dpll = crtc_state->dpll_hw_state.dpll;
 	int i;
 
-	assert_pipe_disabled(dev_priv, crtc->pipe);
+	assert_pipe_disabled(dev_priv, crtc_state->cpu_transcoder);
 
 	/* PLL is protected by panel, make sure we can write it */
 	if (i9xx_has_pps(dev_priv))
@@ -1554,7 +1548,7 @@ static void i9xx_enable_pll(struct intel_crtc *crtc,
 
 static void i9xx_disable_pll(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
@@ -1563,7 +1557,7 @@ static void i9xx_disable_pll(const struct intel_crtc_state *crtc_state)
 		return;
 
 	/* Make sure the pipe isn't still relying on us */
-	assert_pipe_disabled(dev_priv, pipe);
+	assert_pipe_disabled(dev_priv, crtc_state->cpu_transcoder);
 
 	I915_WRITE(DPLL(pipe), DPLL_VGA_MODE_DIS);
 	POSTING_READ(DPLL(pipe));
@@ -1574,7 +1568,7 @@ static void vlv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 	u32 val;
 
 	/* Make sure the pipe isn't still relying on us */
-	assert_pipe_disabled(dev_priv, pipe);
+	assert_pipe_disabled(dev_priv, (enum transcoder)pipe);
 
 	val = DPLL_INTEGRATED_REF_CLK_VLV |
 		DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS;
@@ -1591,7 +1585,7 @@ static void chv_disable_pll(struct drm_i915_private *dev_priv, enum pipe pipe)
 	u32 val;
 
 	/* Make sure the pipe isn't still relying on us */
-	assert_pipe_disabled(dev_priv, pipe);
+	assert_pipe_disabled(dev_priv, (enum transcoder)pipe);
 
 	val = DPLL_SSC_REF_CLK_CHV |
 		DPLL_REF_CLK_ENABLE_VLV | DPLL_VGA_MODE_DIS;
@@ -1645,7 +1639,7 @@ void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
 
 static void ironlake_enable_pch_transcoder(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	i915_reg_t reg;
@@ -1659,11 +1653,16 @@ static void ironlake_enable_pch_transcoder(const struct intel_crtc_state *crtc_s
 	assert_fdi_rx_enabled(dev_priv, pipe);
 
 	if (HAS_PCH_CPT(dev_priv)) {
-		/* Workaround: Set the timing override bit before enabling the
-		 * pch transcoder. */
 		reg = TRANS_CHICKEN2(pipe);
 		val = I915_READ(reg);
+		/*
+		 * Workaround: Set the timing override bit
+		 * before enabling the pch transcoder.
+		 */
 		val |= TRANS_CHICKEN2_TIMING_OVERRIDE;
+		/* Configure frame start delay to match the CPU */
+		val &= ~TRANS_CHICKEN2_FRAME_START_DELAY_MASK;
+		val |= TRANS_CHICKEN2_FRAME_START_DELAY(0);
 		I915_WRITE(reg, val);
 	}
 
@@ -1672,6 +1671,10 @@ static void ironlake_enable_pch_transcoder(const struct intel_crtc_state *crtc_s
 	pipeconf_val = I915_READ(PIPECONF(pipe));
 
 	if (HAS_PCH_IBX(dev_priv)) {
+		/* Configure frame start delay to match the CPU */
+		val &= ~TRANS_FRAME_START_DELAY_MASK;
+		val |= TRANS_FRAME_START_DELAY(0);
+
 		/*
 		 * Make the BPC in transcoder be consistent with
 		 * that in pipeconf reg. For HDMI we must use 8bpc
@@ -1709,9 +1712,12 @@ static void lpt_enable_pch_transcoder(struct drm_i915_private *dev_priv,
 	assert_fdi_tx_enabled(dev_priv, (enum pipe) cpu_transcoder);
 	assert_fdi_rx_enabled(dev_priv, PIPE_A);
 
-	/* Workaround: set timing override bit. */
 	val = I915_READ(TRANS_CHICKEN2(PIPE_A));
+	/* Workaround: set timing override bit. */
 	val |= TRANS_CHICKEN2_TIMING_OVERRIDE;
+	/* Configure frame start delay to match the CPU */
+	val &= ~TRANS_CHICKEN2_FRAME_START_DELAY_MASK;
+	val |= TRANS_CHICKEN2_FRAME_START_DELAY(0);
 	I915_WRITE(TRANS_CHICKEN2(PIPE_A), val);
 
 	val = TRANS_ENABLE;
@@ -1789,7 +1795,7 @@ enum pipe intel_crtc_pch_transcoder(struct intel_crtc *crtc)
 
 static u32 intel_crtc_max_vblank_count(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	/*
 	 * On i965gm the hardware frame counter reads
@@ -1809,16 +1815,25 @@ static u32 intel_crtc_max_vblank_count(const struct intel_crtc_state *crtc_state
 
 static void intel_crtc_vblank_on(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 
+	assert_vblank_disabled(&crtc->base);
 	drm_crtc_set_max_vblank_count(&crtc->base,
 				      intel_crtc_max_vblank_count(crtc_state));
 	drm_crtc_vblank_on(&crtc->base);
 }
 
+void intel_crtc_vblank_off(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+
+	drm_crtc_vblank_off(&crtc->base);
+	assert_vblank_disabled(&crtc->base);
+}
+
 static void intel_enable_pipe(const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = new_crtc_state->cpu_transcoder;
 	enum pipe pipe = crtc->pipe;
@@ -1874,9 +1889,9 @@ static void intel_enable_pipe(const struct intel_crtc_state *new_crtc_state)
 		intel_wait_for_pipe_scanline_moving(crtc);
 }
 
-static void intel_disable_pipe(const struct intel_crtc_state *old_crtc_state)
+void intel_disable_pipe(const struct intel_crtc_state *old_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = old_crtc_state->cpu_transcoder;
 	enum pipe pipe = crtc->pipe;
@@ -1919,6 +1934,66 @@ static unsigned int intel_tile_size(const struct drm_i915_private *dev_priv)
 	return IS_GEN(dev_priv, 2) ? 2048 : 4096;
 }
 
+static bool is_ccs_plane(const struct drm_framebuffer *fb, int plane)
+{
+	if (!is_ccs_modifier(fb->modifier))
+		return false;
+
+	return plane >= fb->format->num_planes / 2;
+}
+
+static bool is_gen12_ccs_modifier(u64 modifier)
+{
+	return modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS;
+}
+
+static bool is_gen12_ccs_plane(const struct drm_framebuffer *fb, int plane)
+{
+	return is_gen12_ccs_modifier(fb->modifier) && is_ccs_plane(fb, plane);
+}
+
+static bool is_aux_plane(const struct drm_framebuffer *fb, int plane)
+{
+	if (is_ccs_modifier(fb->modifier))
+		return is_ccs_plane(fb, plane);
+
+	return plane == 1;
+}
+
+static int main_to_ccs_plane(const struct drm_framebuffer *fb, int main_plane)
+{
+	WARN_ON(!is_ccs_modifier(fb->modifier) ||
+		(main_plane && main_plane >= fb->format->num_planes / 2));
+
+	return fb->format->num_planes / 2 + main_plane;
+}
+
+static int ccs_to_main_plane(const struct drm_framebuffer *fb, int ccs_plane)
+{
+	WARN_ON(!is_ccs_modifier(fb->modifier) ||
+		ccs_plane < fb->format->num_planes / 2);
+
+	return ccs_plane - fb->format->num_planes / 2;
+}
+
+/* Return either the main plane's CCS or - if not a CCS FB - UV plane */
+static int
+intel_main_to_aux_plane(const struct drm_framebuffer *fb, int main_plane)
+{
+	if (is_ccs_modifier(fb->modifier))
+		return main_to_ccs_plane(fb, main_plane);
+
+	return 1;
+}
+
+bool
+intel_format_info_is_yuv_semiplanar(const struct drm_format_info *info,
+				    uint64_t modifier)
+{
+	return info->is_yuv &&
+	       info->num_planes == (is_ccs_modifier(modifier) ? 4 : 2);
+}
+
 static unsigned int
 intel_tile_width_bytes(const struct drm_framebuffer *fb, int color_plane)
 {
@@ -1934,16 +2009,20 @@ intel_tile_width_bytes(const struct drm_framebuffer *fb, int color_plane)
 		else
 			return 512;
 	case I915_FORMAT_MOD_Y_TILED_CCS:
-		if (color_plane == 1)
+		if (is_ccs_plane(fb, color_plane))
 			return 128;
 		/* fall through */
+	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
+		if (is_ccs_plane(fb, color_plane))
+			return 64;
+		/* fall through */
 	case I915_FORMAT_MOD_Y_TILED:
 		if (IS_GEN(dev_priv, 2) || HAS_128_BYTE_Y_TILING(dev_priv))
 			return 128;
 		else
 			return 512;
 	case I915_FORMAT_MOD_Yf_TILED_CCS:
-		if (color_plane == 1)
+		if (is_ccs_plane(fb, color_plane))
 			return 128;
 		/* fall through */
 	case I915_FORMAT_MOD_Yf_TILED:
@@ -1970,6 +2049,9 @@ intel_tile_width_bytes(const struct drm_framebuffer *fb, int color_plane)
 static unsigned int
 intel_tile_height(const struct drm_framebuffer *fb, int color_plane)
 {
+	if (is_gen12_ccs_plane(fb, color_plane))
+		return 1;
+
 	return intel_tile_size(to_i915(fb->dev)) /
 		intel_tile_width_bytes(fb, color_plane);
 }
@@ -1983,7 +2065,7 @@ static void intel_tile_dims(const struct drm_framebuffer *fb, int color_plane,
 	unsigned int cpp = fb->format->cpp[color_plane];
 
 	*tile_width = tile_width_bytes / cpp;
-	*tile_height = intel_tile_size(to_i915(fb->dev)) / tile_width_bytes;
+	*tile_height = intel_tile_height(fb, color_plane);
 }
 
 unsigned int
@@ -2060,7 +2142,7 @@ static unsigned int intel_surf_alignment(const struct drm_framebuffer *fb,
 	struct drm_i915_private *dev_priv = to_i915(fb->dev);
 
 	/* AUX_DIST needs only 4K alignment */
-	if (color_plane == 1)
+	if (is_aux_plane(fb, color_plane))
 		return 4096;
 
 	switch (fb->modifier) {
@@ -2070,6 +2152,8 @@ static unsigned int intel_surf_alignment(const struct drm_framebuffer *fb,
 		if (INTEL_GEN(dev_priv) >= 9)
 			return 256 * 1024;
 		return 0;
+	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
+		return 16 * 1024;
 	case I915_FORMAT_MOD_Y_TILED_CCS:
 	case I915_FORMAT_MOD_Yf_TILED_CCS:
 	case I915_FORMAT_MOD_Y_TILED:
@@ -2083,7 +2167,7 @@ static unsigned int intel_surf_alignment(const struct drm_framebuffer *fb,
 
 static bool intel_plane_uses_fence(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 
 	return INTEL_GEN(dev_priv) < 4 ||
@@ -2126,19 +2210,18 @@ intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
 	 * pin/unpin/fence and not more.
 	 */
 	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
-	i915_gem_object_lock(obj);
 
 	atomic_inc(&dev_priv->gpu_error.pending_fb_pin);
 
-	pinctl = 0;
-
-	/* Valleyview is definitely limited to scanning out the first
+	/*
+	 * Valleyview is definitely limited to scanning out the first
 	 * 512MiB. Lets presume this behaviour was inherited from the
 	 * g4x display engine and that all earlier gen are similarly
 	 * limited. Testing suggests that it is a little more
 	 * complicated than this. For example, Cherryview appears quite
 	 * happy to scanout from anywhere within its global aperture.
 	 */
+	pinctl = 0;
 	if (HAS_GMCH(dev_priv))
 		pinctl |= PIN_MAPPABLE;
 
@@ -2150,7 +2233,8 @@ intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
 	if (uses_fence && i915_vma_is_map_and_fenceable(vma)) {
 		int ret;
 
-		/* Install a fence for tiled scan-out. Pre-i965 always needs a
+		/*
+		 * Install a fence for tiled scan-out. Pre-i965 always needs a
 		 * fence, whereas 965+ only requires a fence if using
 		 * framebuffer compression.  For simplicity, we always, when
 		 * possible, install a fence as the cost is not that onerous.
@@ -2180,8 +2264,6 @@ intel_pin_and_fence_fb_obj(struct drm_framebuffer *fb,
 	i915_vma_get(vma);
 err:
 	atomic_dec(&dev_priv->gpu_error.pending_fb_pin);
-
-	i915_gem_object_unlock(obj);
 	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
 	return vma;
 }
@@ -2216,7 +2298,7 @@ u32 intel_fb_xy_to_linear(int x, int y,
 			  const struct intel_plane_state *state,
 			  int color_plane)
 {
-	const struct drm_framebuffer *fb = state->base.fb;
+	const struct drm_framebuffer *fb = state->hw.fb;
 	unsigned int cpp = fb->format->cpp[color_plane];
 	unsigned int pitch = state->color_plane[color_plane].stride;
 
@@ -2264,9 +2346,10 @@ static u32 intel_adjust_tile_offset(int *x, int *y,
 	return new_offset;
 }
 
-static bool is_surface_linear(u64 modifier, int color_plane)
+static bool is_surface_linear(const struct drm_framebuffer *fb, int color_plane)
 {
-	return modifier == DRM_FORMAT_MOD_LINEAR;
+	return fb->modifier == DRM_FORMAT_MOD_LINEAR ||
+	       is_gen12_ccs_plane(fb, color_plane);
 }
 
 static u32 intel_adjust_aligned_offset(int *x, int *y,
@@ -2281,7 +2364,7 @@ static u32 intel_adjust_aligned_offset(int *x, int *y,
 
 	WARN_ON(new_offset > old_offset);
 
-	if (!is_surface_linear(fb->modifier, color_plane)) {
+	if (!is_surface_linear(fb, color_plane)) {
 		unsigned int tile_size, tile_width, tile_height;
 		unsigned int pitch_tiles;
 
@@ -2317,8 +2400,8 @@ static u32 intel_plane_adjust_aligned_offset(int *x, int *y,
 					     int color_plane,
 					     u32 old_offset, u32 new_offset)
 {
-	return intel_adjust_aligned_offset(x, y, state->base.fb, color_plane,
-					   state->base.rotation,
+	return intel_adjust_aligned_offset(x, y, state->hw.fb, color_plane,
+					   state->hw.rotation,
 					   state->color_plane[color_plane].stride,
 					   old_offset, new_offset);
 }
@@ -2351,7 +2434,7 @@ static u32 intel_compute_aligned_offset(struct drm_i915_private *dev_priv,
 	if (alignment)
 		alignment--;
 
-	if (!is_surface_linear(fb->modifier, color_plane)) {
+	if (!is_surface_linear(fb, color_plane)) {
 		unsigned int tile_size, tile_width, tile_height;
 		unsigned int tile_rows, tiles, pitch_tiles;
 
@@ -2392,10 +2475,10 @@ static u32 intel_plane_compute_aligned_offset(int *x, int *y,
 					      const struct intel_plane_state *state,
 					      int color_plane)
 {
-	struct intel_plane *intel_plane = to_intel_plane(state->base.plane);
+	struct intel_plane *intel_plane = to_intel_plane(state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(intel_plane->base.dev);
-	const struct drm_framebuffer *fb = state->base.fb;
-	unsigned int rotation = state->base.rotation;
+	const struct drm_framebuffer *fb = state->hw.fb;
+	unsigned int rotation = state->hw.rotation;
 	int pitch = state->color_plane[color_plane].stride;
 	u32 alignment;
 
@@ -2453,6 +2536,7 @@ static unsigned int intel_fb_modifier_to_tiling(u64 fb_modifier)
 		return I915_TILING_X;
 	case I915_FORMAT_MOD_Y_TILED:
 	case I915_FORMAT_MOD_Y_TILED_CCS:
+	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
 		return I915_TILING_Y;
 	default:
 		return I915_TILING_NONE;
@@ -2473,7 +2557,7 @@ static unsigned int intel_fb_modifier_to_tiling(u64 fb_modifier)
  * us a ratio of one byte in the CCS for each 8x16 pixels in the
  * main surface.
  */
-static const struct drm_format_info ccs_formats[] = {
+static const struct drm_format_info skl_ccs_formats[] = {
 	{ .format = DRM_FORMAT_XRGB8888, .depth = 24, .num_planes = 2,
 	  .cpp = { 4, 1, }, .hsub = 8, .vsub = 16, },
 	{ .format = DRM_FORMAT_XBGR8888, .depth = 24, .num_planes = 2,
@@ -2484,6 +2568,28 @@ static const struct drm_format_info ccs_formats[] = {
 	  .cpp = { 4, 1, }, .hsub = 8, .vsub = 16, .has_alpha = true, },
 };
 
+/*
+ * Gen-12 compression uses 4 bits of CCS data for each cache line pair in the
+ * main surface. And each 64B CCS cache line represents an area of 4x1 Y-tiles
+ * in the main surface. With 4 byte pixels and each Y-tile having dimensions of
+ * 32x32 pixels, the ratio turns out to 1B in the CCS for every 2x32 pixels in
+ * the main surface.
+ */
+static const struct drm_format_info gen12_ccs_formats[] = {
+	{ .format = DRM_FORMAT_XRGB8888, .depth = 24, .num_planes = 2,
+	  .char_per_block = { 4, 1 }, .block_w = { 1, 2 }, .block_h = { 1, 1 },
+	  .hsub = 1, .vsub = 1, },
+	{ .format = DRM_FORMAT_XBGR8888, .depth = 24, .num_planes = 2,
+	  .char_per_block = { 4, 1 }, .block_w = { 1, 2 }, .block_h = { 1, 1 },
+	  .hsub = 1, .vsub = 1, },
+	{ .format = DRM_FORMAT_ARGB8888, .depth = 32, .num_planes = 2,
+	  .char_per_block = { 4, 1 }, .block_w = { 1, 2 }, .block_h = { 1, 1 },
+	  .hsub = 1, .vsub = 1, .has_alpha = true },
+	{ .format = DRM_FORMAT_ABGR8888, .depth = 32, .num_planes = 2,
+	  .char_per_block = { 4, 1 }, .block_w = { 1, 2 }, .block_h = { 1, 1 },
+	  .hsub = 1, .vsub = 1, .has_alpha = true },
+};
+
 static const struct drm_format_info *
 lookup_format_info(const struct drm_format_info formats[],
 		   int num_formats, u32 format)
@@ -2504,8 +2610,12 @@ intel_get_format_info(const struct drm_mode_fb_cmd2 *cmd)
 	switch (cmd->modifier[0]) {
 	case I915_FORMAT_MOD_Y_TILED_CCS:
 	case I915_FORMAT_MOD_Yf_TILED_CCS:
-		return lookup_format_info(ccs_formats,
-					  ARRAY_SIZE(ccs_formats),
+		return lookup_format_info(skl_ccs_formats,
+					  ARRAY_SIZE(skl_ccs_formats),
+					  cmd->pixel_format);
+	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
+		return lookup_format_info(gen12_ccs_formats,
+					  ARRAY_SIZE(gen12_ccs_formats),
 					  cmd->pixel_format);
 	default:
 		return NULL;
@@ -2514,10 +2624,17 @@ intel_get_format_info(const struct drm_mode_fb_cmd2 *cmd)
 
 bool is_ccs_modifier(u64 modifier)
 {
-	return modifier == I915_FORMAT_MOD_Y_TILED_CCS ||
+	return modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS ||
+	       modifier == I915_FORMAT_MOD_Y_TILED_CCS ||
 	       modifier == I915_FORMAT_MOD_Yf_TILED_CCS;
 }
 
+static int gen12_ccs_aux_stride(struct drm_framebuffer *fb, int ccs_plane)
+{
+	return DIV_ROUND_UP(fb->pitches[ccs_to_main_plane(fb, ccs_plane)],
+			    512) * 64;
+}
+
 u32 intel_plane_fb_max_stride(struct drm_i915_private *dev_priv,
 			      u32 pixel_format, u64 modifier)
 {
@@ -2562,8 +2679,9 @@ static u32
 intel_fb_stride_alignment(const struct drm_framebuffer *fb, int color_plane)
 {
 	struct drm_i915_private *dev_priv = to_i915(fb->dev);
+	u32 tile_width;
 
-	if (fb->modifier == DRM_FORMAT_MOD_LINEAR) {
+	if (is_surface_linear(fb, color_plane)) {
 		u32 max_stride = intel_plane_fb_max_stride(dev_priv,
 							   fb->format->format,
 							   fb->modifier);
@@ -2572,20 +2690,41 @@ intel_fb_stride_alignment(const struct drm_framebuffer *fb, int color_plane)
 		 * To make remapping with linear generally feasible
 		 * we need the stride to be page aligned.
 		 */
-		if (fb->pitches[color_plane] > max_stride)
+		if (fb->pitches[color_plane] > max_stride &&
+		    !is_ccs_modifier(fb->modifier))
 			return intel_tile_size(dev_priv);
 		else
 			return 64;
-	} else {
-		return intel_tile_width_bytes(fb, color_plane);
 	}
+
+	tile_width = intel_tile_width_bytes(fb, color_plane);
+	if (is_ccs_modifier(fb->modifier) && color_plane == 0) {
+		/*
+		 * Display WA #0531: skl,bxt,kbl,glk
+		 *
+		 * Render decompression and plane width > 3840
+		 * combined with horizontal panning requires the
+		 * plane stride to be a multiple of 4. We'll just
+		 * require the entire fb to accommodate that to avoid
+		 * potential runtime errors at plane configuration time.
+		 */
+		if (IS_GEN(dev_priv, 9) && fb->width > 3840)
+			tile_width *= 4;
+		/*
+		 * The main surface pitch must be padded to a multiple of four
+		 * tile widths.
+		 */
+		else if (INTEL_GEN(dev_priv) >= 12)
+			tile_width *= 4;
+	}
+	return tile_width;
 }
 
 bool intel_plane_can_remap(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int i;
 
 	/* We don't want to deal with remapping with cursors */
@@ -2623,16 +2762,16 @@ bool intel_plane_can_remap(const struct intel_plane_state *plane_state)
 
 static bool intel_plane_needs_remap(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	u32 stride, max_stride;
 
 	/*
 	 * No remapping for invisible planes since we don't have
 	 * an actual source viewport to remap.
 	 */
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return false;
 
 	if (!intel_plane_can_remap(plane_state))
@@ -2649,12 +2788,167 @@ static bool intel_plane_needs_remap(const struct intel_plane_state *plane_state)
 	return stride > max_stride;
 }
 
+static void
+intel_fb_plane_get_subsampling(int *hsub, int *vsub,
+			       const struct drm_framebuffer *fb,
+			       int color_plane)
+{
+	int main_plane;
+
+	if (color_plane == 0) {
+		*hsub = 1;
+		*vsub = 1;
+
+		return;
+	}
+
+	/*
+	 * TODO: Deduct the subsampling from the char block for all CCS
+	 * formats and planes.
+	 */
+	if (!is_gen12_ccs_plane(fb, color_plane)) {
+		*hsub = fb->format->hsub;
+		*vsub = fb->format->vsub;
+
+		return;
+	}
+
+	main_plane = ccs_to_main_plane(fb, color_plane);
+	*hsub = drm_format_info_block_width(fb->format, color_plane) /
+		drm_format_info_block_width(fb->format, main_plane);
+
+	/*
+	 * The min stride check in the core framebuffer_check() function
+	 * assumes that format->hsub applies to every plane except for the
+	 * first plane. That's incorrect for the CCS AUX plane of the first
+	 * plane, but for the above check to pass we must define the block
+	 * width with that subsampling applied to it. Adjust the width here
+	 * accordingly, so we can calculate the actual subsampling factor.
+	 */
+	if (main_plane == 0)
+		*hsub *= fb->format->hsub;
+
+	*vsub = 32;
+}
+static int
+intel_fb_check_ccs_xy(struct drm_framebuffer *fb, int ccs_plane, int x, int y)
+{
+	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
+	int main_plane;
+	int hsub, vsub;
+	int tile_width, tile_height;
+	int ccs_x, ccs_y;
+	int main_x, main_y;
+
+	if (!is_ccs_plane(fb, ccs_plane))
+		return 0;
+
+	intel_tile_dims(fb, ccs_plane, &tile_width, &tile_height);
+	intel_fb_plane_get_subsampling(&hsub, &vsub, fb, ccs_plane);
+
+	tile_width *= hsub;
+	tile_height *= vsub;
+
+	ccs_x = (x * hsub) % tile_width;
+	ccs_y = (y * vsub) % tile_height;
+
+	main_plane = ccs_to_main_plane(fb, ccs_plane);
+	main_x = intel_fb->normal[main_plane].x % tile_width;
+	main_y = intel_fb->normal[main_plane].y % tile_height;
+
+	/*
+	 * CCS doesn't have its own x/y offset register, so the intra CCS tile
+	 * x/y offsets must match between CCS and the main surface.
+	 */
+	if (main_x != ccs_x || main_y != ccs_y) {
+		DRM_DEBUG_KMS("Bad CCS x/y (main %d,%d ccs %d,%d) full (main %d,%d ccs %d,%d)\n",
+			      main_x, main_y,
+			      ccs_x, ccs_y,
+			      intel_fb->normal[main_plane].x,
+			      intel_fb->normal[main_plane].y,
+			      x, y);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void
+intel_fb_plane_dims(int *w, int *h, struct drm_framebuffer *fb, int color_plane)
+{
+	int hsub, vsub;
+
+	intel_fb_plane_get_subsampling(&hsub, &vsub, fb, color_plane);
+	*w = fb->width / hsub;
+	*h = fb->height / vsub;
+}
+
+/*
+ * Setup the rotated view for an FB plane and return the size the GTT mapping
+ * requires for this view.
+ */
+static u32
+setup_fb_rotation(int plane, const struct intel_remapped_plane_info *plane_info,
+		  u32 gtt_offset_rotated, int x, int y,
+		  unsigned int width, unsigned int height,
+		  unsigned int tile_size,
+		  unsigned int tile_width, unsigned int tile_height,
+		  struct drm_framebuffer *fb)
+{
+	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
+	struct intel_rotation_info *rot_info = &intel_fb->rot_info;
+	unsigned int pitch_tiles;
+	struct drm_rect r;
+
+	/* Y or Yf modifiers required for 90/270 rotation */
+	if (fb->modifier != I915_FORMAT_MOD_Y_TILED &&
+	    fb->modifier != I915_FORMAT_MOD_Yf_TILED)
+		return 0;
+
+	if (WARN_ON(plane >= ARRAY_SIZE(rot_info->plane)))
+		return 0;
+
+	rot_info->plane[plane] = *plane_info;
+
+	intel_fb->rotated[plane].pitch = plane_info->height * tile_height;
+
+	/* rotate the x/y offsets to match the GTT view */
+	drm_rect_init(&r, x, y, width, height);
+	drm_rect_rotate(&r,
+			plane_info->width * tile_width,
+			plane_info->height * tile_height,
+			DRM_MODE_ROTATE_270);
+	x = r.x1;
+	y = r.y1;
+
+	/* rotate the tile dimensions to match the GTT view */
+	pitch_tiles = intel_fb->rotated[plane].pitch / tile_height;
+	swap(tile_width, tile_height);
+
+	/*
+	 * We only keep the x/y offsets, so push all of the
+	 * gtt offset into the x/y offsets.
+	 */
+	intel_adjust_tile_offset(&x, &y,
+				 tile_width, tile_height,
+				 tile_size, pitch_tiles,
+				 gtt_offset_rotated * tile_size, 0);
+
+	/*
+	 * First pixel of the framebuffer from
+	 * the start of the rotated gtt mapping.
+	 */
+	intel_fb->rotated[plane].x = x;
+	intel_fb->rotated[plane].y = y;
+
+	return plane_info->width * plane_info->height;
+}
+
 static int
 intel_fill_fb_info(struct drm_i915_private *dev_priv,
 		   struct drm_framebuffer *fb)
 {
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
-	struct intel_rotation_info *rot_info = &intel_fb->rot_info;
 	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	u32 gtt_offset_rotated = 0;
 	unsigned int max_size = 0;
@@ -2669,8 +2963,7 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 		int ret;
 
 		cpp = fb->format->cpp[i];
-		width = drm_framebuffer_plane_width(fb->width, fb, i);
-		height = drm_framebuffer_plane_height(fb->height, fb, i);
+		intel_fb_plane_dims(&width, &height, fb, i);
 
 		ret = intel_fb_offset_to_xy(&x, &y, fb, i);
 		if (ret) {
@@ -2679,36 +2972,9 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 			return ret;
 		}
 
-		if (is_ccs_modifier(fb->modifier) && i == 1) {
-			int hsub = fb->format->hsub;
-			int vsub = fb->format->vsub;
-			int tile_width, tile_height;
-			int main_x, main_y;
-			int ccs_x, ccs_y;
-
-			intel_tile_dims(fb, i, &tile_width, &tile_height);
-			tile_width *= hsub;
-			tile_height *= vsub;
-
-			ccs_x = (x * hsub) % tile_width;
-			ccs_y = (y * vsub) % tile_height;
-			main_x = intel_fb->normal[0].x % tile_width;
-			main_y = intel_fb->normal[0].y % tile_height;
-
-			/*
-			 * CCS doesn't have its own x/y offset register, so the intra CCS tile
-			 * x/y offsets must match between CCS and the main surface.
-			 */
-			if (main_x != ccs_x || main_y != ccs_y) {
-				DRM_DEBUG_KMS("Bad CCS x/y (main %d,%d ccs %d,%d) full (main %d,%d ccs %d,%d)\n",
-					      main_x, main_y,
-					      ccs_x, ccs_y,
-					      intel_fb->normal[0].x,
-					      intel_fb->normal[0].y,
-					      x, y);
-				return -EINVAL;
-			}
-		}
+		ret = intel_fb_check_ccs_xy(fb, i, x, y);
+		if (ret)
+			return ret;
 
 		/*
 		 * The fence (if used) is aligned to the start of the object
@@ -2739,23 +3005,21 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 						      tile_size);
 		offset /= tile_size;
 
-		if (!is_surface_linear(fb->modifier, i)) {
+		if (!is_surface_linear(fb, i)) {
+			struct intel_remapped_plane_info plane_info;
 			unsigned int tile_width, tile_height;
-			unsigned int pitch_tiles;
-			struct drm_rect r;
 
 			intel_tile_dims(fb, i, &tile_width, &tile_height);
 
-			rot_info->plane[i].offset = offset;
-			rot_info->plane[i].stride = DIV_ROUND_UP(fb->pitches[i], tile_width * cpp);
-			rot_info->plane[i].width = DIV_ROUND_UP(x + width, tile_width);
-			rot_info->plane[i].height = DIV_ROUND_UP(y + height, tile_height);
-
-			intel_fb->rotated[i].pitch =
-				rot_info->plane[i].height * tile_height;
+			plane_info.offset = offset;
+			plane_info.stride = DIV_ROUND_UP(fb->pitches[i],
+							 tile_width * cpp);
+			plane_info.width = DIV_ROUND_UP(x + width, tile_width);
+			plane_info.height = DIV_ROUND_UP(y + height,
+							 tile_height);
 
 			/* how many tiles does this plane need */
-			size = rot_info->plane[i].stride * rot_info->plane[i].height;
+			size = plane_info.stride * plane_info.height;
 			/*
 			 * If the plane isn't horizontally tile aligned,
 			 * we need one more tile.
@@ -2763,36 +3027,13 @@ intel_fill_fb_info(struct drm_i915_private *dev_priv,
 			if (x != 0)
 				size++;
 
-			/* rotate the x/y offsets to match the GTT view */
-			drm_rect_init(&r, x, y, width, height);
-			drm_rect_rotate(&r,
-					rot_info->plane[i].width * tile_width,
-					rot_info->plane[i].height * tile_height,
-					DRM_MODE_ROTATE_270);
-			x = r.x1;
-			y = r.y1;
-
-			/* rotate the tile dimensions to match the GTT view */
-			pitch_tiles = intel_fb->rotated[i].pitch / tile_height;
-			swap(tile_width, tile_height);
-
-			/*
-			 * We only keep the x/y offsets, so push all of the
-			 * gtt offset into the x/y offsets.
-			 */
-			intel_adjust_tile_offset(&x, &y,
-						 tile_width, tile_height,
-						 tile_size, pitch_tiles,
-						 gtt_offset_rotated * tile_size, 0);
-
-			gtt_offset_rotated += rot_info->plane[i].width * rot_info->plane[i].height;
-
-			/*
-			 * First pixel of the framebuffer from
-			 * the start of the rotated gtt mapping.
-			 */
-			intel_fb->rotated[i].x = x;
-			intel_fb->rotated[i].y = y;
+			gtt_offset_rotated +=
+				setup_fb_rotation(i, &plane_info,
+						  gtt_offset_rotated,
+						  x, y, width, height,
+						  tile_size,
+						  tile_width, tile_height,
+						  fb);
 		} else {
 			size = DIV_ROUND_UP((y + height) * fb->pitches[i] +
 					    x * cpp, tile_size);
@@ -2815,11 +3056,11 @@ static void
 intel_plane_remap_gtt(struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	struct drm_framebuffer *fb = plane_state->base.fb;
+		to_i915(plane_state->uapi.plane->dev);
+	struct drm_framebuffer *fb = plane_state->hw.fb;
 	struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb);
 	struct intel_rotation_info *info = &plane_state->view.rotated;
-	unsigned int rotation = plane_state->base.rotation;
+	unsigned int rotation = plane_state->hw.rotation;
 	int i, num_planes = fb->format->num_planes;
 	unsigned int tile_size = intel_tile_size(dev_priv);
 	unsigned int src_x, src_y;
@@ -2830,20 +3071,20 @@ intel_plane_remap_gtt(struct intel_plane_state *plane_state)
 	plane_state->view.type = drm_rotation_90_or_270(rotation) ?
 		I915_GGTT_VIEW_ROTATED : I915_GGTT_VIEW_REMAPPED;
 
-	src_x = plane_state->base.src.x1 >> 16;
-	src_y = plane_state->base.src.y1 >> 16;
-	src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	src_h = drm_rect_height(&plane_state->base.src) >> 16;
+	src_x = plane_state->uapi.src.x1 >> 16;
+	src_y = plane_state->uapi.src.y1 >> 16;
+	src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
 
 	WARN_ON(is_ccs_modifier(fb->modifier));
 
 	/* Make src coordinates relative to the viewport */
-	drm_rect_translate(&plane_state->base.src,
+	drm_rect_translate(&plane_state->uapi.src,
 			   -(src_x << 16), -(src_y << 16));
 
 	/* Rotate src coordinates to match rotated GTT view */
 	if (drm_rotation_90_or_270(rotation))
-		drm_rect_rotate(&plane_state->base.src,
+		drm_rect_rotate(&plane_state->uapi.src,
 				src_w << 16, src_h << 16,
 				DRM_MODE_ROTATE_270);
 
@@ -2876,6 +3117,7 @@ intel_plane_remap_gtt(struct intel_plane_state *plane_state)
 						      DRM_MODE_ROTATE_0, tile_size);
 		offset /= tile_size;
 
+		WARN_ON(i >= ARRAY_SIZE(info->plane));
 		info->plane[i].offset = offset;
 		info->plane[i].stride = DIV_ROUND_UP(fb->pitches[i],
 						     tile_width * cpp);
@@ -2925,8 +3167,8 @@ static int
 intel_plane_compute_gtt(struct intel_plane_state *plane_state)
 {
 	const struct intel_framebuffer *fb =
-		to_intel_framebuffer(plane_state->base.fb);
-	unsigned int rotation = plane_state->base.rotation;
+		to_intel_framebuffer(plane_state->hw.fb);
+	unsigned int rotation = plane_state->hw.rotation;
 	int i, num_planes;
 
 	if (!fb)
@@ -2963,7 +3205,7 @@ intel_plane_compute_gtt(struct intel_plane_state *plane_state)
 
 	/* Rotate src coordinates to match rotated GTT view */
 	if (drm_rotation_90_or_270(rotation))
-		drm_rect_rotate(&plane_state->base.src,
+		drm_rect_rotate(&plane_state->uapi.src,
 				fb->base.width << 16, fb->base.height << 16,
 				DRM_MODE_ROTATE_270);
 
@@ -2975,6 +3217,8 @@ static int i9xx_format_to_fourcc(int format)
 	switch (format) {
 	case DISPPLANE_8BPP:
 		return DRM_FORMAT_C8;
+	case DISPPLANE_BGRA555:
+		return DRM_FORMAT_ARGB1555;
 	case DISPPLANE_BGRX555:
 		return DRM_FORMAT_XRGB1555;
 	case DISPPLANE_BGRX565:
@@ -2984,10 +3228,18 @@ static int i9xx_format_to_fourcc(int format)
 		return DRM_FORMAT_XRGB8888;
 	case DISPPLANE_RGBX888:
 		return DRM_FORMAT_XBGR8888;
+	case DISPPLANE_BGRA888:
+		return DRM_FORMAT_ARGB8888;
+	case DISPPLANE_RGBA888:
+		return DRM_FORMAT_ABGR8888;
 	case DISPPLANE_BGRX101010:
 		return DRM_FORMAT_XRGB2101010;
 	case DISPPLANE_RGBX101010:
 		return DRM_FORMAT_XBGR2101010;
+	case DISPPLANE_BGRA101010:
+		return DRM_FORMAT_ARGB2101010;
+	case DISPPLANE_RGBA101010:
+		return DRM_FORMAT_ABGR2101010;
 	case DISPPLANE_RGBX161616:
 		return DRM_FORMAT_XBGR16161616F;
 	}
@@ -3032,10 +3284,17 @@ int skl_format_to_fourcc(int format, bool rgb_order, bool alpha)
 				return DRM_FORMAT_XRGB8888;
 		}
 	case PLANE_CTL_FORMAT_XRGB_2101010:
-		if (rgb_order)
-			return DRM_FORMAT_XBGR2101010;
-		else
-			return DRM_FORMAT_XRGB2101010;
+		if (rgb_order) {
+			if (alpha)
+				return DRM_FORMAT_ABGR2101010;
+			else
+				return DRM_FORMAT_XBGR2101010;
+		} else {
+			if (alpha)
+				return DRM_FORMAT_ARGB2101010;
+			else
+				return DRM_FORMAT_XRGB2101010;
+		}
 	case PLANE_CTL_FORMAT_XRGB_16161616F:
 		if (rgb_order) {
 			if (alpha)
@@ -3131,19 +3390,19 @@ intel_set_plane_visible(struct intel_crtc_state *crtc_state,
 			struct intel_plane_state *plane_state,
 			bool visible)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 
-	plane_state->base.visible = visible;
+	plane_state->uapi.visible = visible;
 
 	if (visible)
-		crtc_state->base.plane_mask |= drm_plane_mask(&plane->base);
+		crtc_state->uapi.plane_mask |= drm_plane_mask(&plane->base);
 	else
-		crtc_state->base.plane_mask &= ~drm_plane_mask(&plane->base);
+		crtc_state->uapi.plane_mask &= ~drm_plane_mask(&plane->base);
 }
 
 static void fixup_active_planes(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	struct drm_plane *plane;
 
 	/*
@@ -3154,13 +3413,14 @@ static void fixup_active_planes(struct intel_crtc_state *crtc_state)
 	crtc_state->active_planes = 0;
 
 	drm_for_each_plane_mask(plane, &dev_priv->drm,
-				crtc_state->base.plane_mask)
+				crtc_state->uapi.plane_mask)
 		crtc_state->active_planes |= BIT(to_intel_plane(plane)->id);
 }
 
 static void intel_plane_disable_noatomic(struct intel_crtc *crtc,
 					 struct intel_plane *plane)
 {
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_crtc_state *crtc_state =
 		to_intel_crtc_state(crtc->base.state);
 	struct intel_plane_state *plane_state =
@@ -3176,7 +3436,27 @@ static void intel_plane_disable_noatomic(struct intel_crtc *crtc,
 	crtc_state->min_cdclk[plane->id] = 0;
 
 	if (plane->id == PLANE_PRIMARY)
-		intel_pre_disable_primary_noatomic(&crtc->base);
+		hsw_disable_ips(crtc_state);
+
+	/*
+	 * Vblank time updates from the shadow to live plane control register
+	 * are blocked if the memory self-refresh mode is active at that
+	 * moment. So to make sure the plane gets truly disabled, disable
+	 * first the self-refresh mode. The self-refresh enable bit in turn
+	 * will be checked/applied by the HW only at the next frame start
+	 * event which is after the vblank start event, so we need to have a
+	 * wait-for-vblank between disabling the plane and the pipe.
+	 */
+	if (HAS_GMCH(dev_priv) &&
+	    intel_set_memory_cxsr(dev_priv, false))
+		intel_wait_for_vblank(dev_priv, crtc->pipe);
+
+	/*
+	 * Gen2 reports pipe underruns whenever all planes are disabled.
+	 * So disable underrun reporting before all the planes get disabled.
+	 */
+	if (IS_GEN(dev_priv, 2) && !crtc_state->active_planes)
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, false);
 
 	intel_disable_plane(plane, crtc_state);
 }
@@ -3229,7 +3509,7 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc,
 			continue;
 
 		if (intel_plane_ggtt_offset(state) == plane_config->base) {
-			fb = state->base.fb;
+			fb = state->hw.fb;
 			drm_framebuffer_get(fb);
 			goto valid_fb;
 		}
@@ -3247,11 +3527,11 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc,
 	return;
 
 valid_fb:
-	intel_state->base.rotation = plane_config->rotation;
+	intel_state->hw.rotation = plane_config->rotation;
 	intel_fill_fb_ggtt_view(&intel_state->view, fb,
-				intel_state->base.rotation);
+				intel_state->hw.rotation);
 	intel_state->color_plane[0].stride =
-		intel_fb_pitch(fb, 0, intel_state->base.rotation);
+		intel_fb_pitch(fb, 0, intel_state->hw.rotation);
 
 	intel_state->vma =
 		intel_pin_and_fence_fb_obj(fb,
@@ -3279,14 +3559,15 @@ valid_fb:
 	plane_state->crtc_w = fb->width;
 	plane_state->crtc_h = fb->height;
 
-	intel_state->base.src = drm_plane_state_src(plane_state);
-	intel_state->base.dst = drm_plane_state_dest(plane_state);
+	intel_state->uapi.src = drm_plane_state_src(plane_state);
+	intel_state->uapi.dst = drm_plane_state_dest(plane_state);
 
 	if (plane_config->tiling)
 		dev_priv->preserve_bios_swizzle = true;
 
 	plane_state->fb = fb;
 	plane_state->crtc = &intel_crtc->base;
+	intel_plane_copy_uapi_to_hw_state(intel_state, intel_state);
 
 	atomic_or(to_intel_plane(primary)->frontbuffer_bit,
 		  &to_intel_frontbuffer(fb)->bits);
@@ -3378,14 +3659,16 @@ static int icl_max_plane_height(void)
 static bool skl_check_main_ccs_coordinates(struct intel_plane_state *plane_state,
 					   int main_x, int main_y, u32 main_offset)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	int hsub = fb->format->hsub;
-	int vsub = fb->format->vsub;
-	int aux_x = plane_state->color_plane[1].x;
-	int aux_y = plane_state->color_plane[1].y;
-	u32 aux_offset = plane_state->color_plane[1].offset;
-	u32 alignment = intel_surf_alignment(fb, 1);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	int ccs_plane = main_to_ccs_plane(fb, 0);
+	int aux_x = plane_state->color_plane[ccs_plane].x;
+	int aux_y = plane_state->color_plane[ccs_plane].y;
+	u32 aux_offset = plane_state->color_plane[ccs_plane].offset;
+	u32 alignment = intel_surf_alignment(fb, ccs_plane);
+	int hsub;
+	int vsub;
 
+	intel_fb_plane_get_subsampling(&hsub, &vsub, fb, ccs_plane);
 	while (aux_offset >= main_offset && aux_y <= main_y) {
 		int x, y;
 
@@ -3397,8 +3680,12 @@ static bool skl_check_main_ccs_coordinates(struct intel_plane_state *plane_state
 
 		x = aux_x / hsub;
 		y = aux_y / vsub;
-		aux_offset = intel_plane_adjust_aligned_offset(&x, &y, plane_state, 1,
-							       aux_offset, aux_offset - alignment);
+		aux_offset = intel_plane_adjust_aligned_offset(&x, &y,
+							       plane_state,
+							       ccs_plane,
+							       aux_offset,
+							       aux_offset -
+								alignment);
 		aux_x = x * hsub + aux_x % hsub;
 		aux_y = y * vsub + aux_y % vsub;
 	}
@@ -3406,25 +3693,28 @@ static bool skl_check_main_ccs_coordinates(struct intel_plane_state *plane_state
 	if (aux_x != main_x || aux_y != main_y)
 		return false;
 
-	plane_state->color_plane[1].offset = aux_offset;
-	plane_state->color_plane[1].x = aux_x;
-	plane_state->color_plane[1].y = aux_y;
+	plane_state->color_plane[ccs_plane].offset = aux_offset;
+	plane_state->color_plane[ccs_plane].x = aux_x;
+	plane_state->color_plane[ccs_plane].y = aux_y;
 
 	return true;
 }
 
 static int skl_check_main_surface(struct intel_plane_state *plane_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
-	int x = plane_state->base.src.x1 >> 16;
-	int y = plane_state->base.src.y1 >> 16;
-	int w = drm_rect_width(&plane_state->base.src) >> 16;
-	int h = drm_rect_height(&plane_state->base.src) >> 16;
+	struct drm_i915_private *dev_priv = to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
+	int x = plane_state->uapi.src.x1 >> 16;
+	int y = plane_state->uapi.src.y1 >> 16;
+	int w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	int h = drm_rect_height(&plane_state->uapi.src) >> 16;
 	int max_width;
 	int max_height;
-	u32 alignment, offset, aux_offset = plane_state->color_plane[1].offset;
+	u32 alignment;
+	u32 offset;
+	int aux_plane = intel_main_to_aux_plane(fb, 0);
+	u32 aux_offset = plane_state->color_plane[aux_plane].offset;
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		max_width = icl_max_plane_width(fb, 0, rotation);
@@ -3490,7 +3780,8 @@ static int skl_check_main_surface(struct intel_plane_state *plane_state)
 								   offset, offset - alignment);
 		}
 
-		if (x != plane_state->color_plane[1].x || y != plane_state->color_plane[1].y) {
+		if (x != plane_state->color_plane[aux_plane].x ||
+		    y != plane_state->color_plane[aux_plane].y) {
 			DRM_DEBUG_KMS("Unable to find suitable display surface offset due to CCS\n");
 			return -EINVAL;
 		}
@@ -3504,7 +3795,7 @@ static int skl_check_main_surface(struct intel_plane_state *plane_state)
 	 * Put the final coordinates back so that the src
 	 * coordinate checks will see the right values.
 	 */
-	drm_rect_translate_to(&plane_state->base.src,
+	drm_rect_translate_to(&plane_state->uapi.src,
 			      x << 16, y << 16);
 
 	return 0;
@@ -3512,14 +3803,14 @@ static int skl_check_main_surface(struct intel_plane_state *plane_state)
 
 static int skl_check_nv12_aux_surface(struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	int max_width = skl_max_plane_width(fb, 1, rotation);
 	int max_height = 4096;
-	int x = plane_state->base.src.x1 >> 17;
-	int y = plane_state->base.src.y1 >> 17;
-	int w = drm_rect_width(&plane_state->base.src) >> 17;
-	int h = drm_rect_height(&plane_state->base.src) >> 17;
+	int x = plane_state->uapi.src.x1 >> 17;
+	int y = plane_state->uapi.src.y1 >> 17;
+	int w = drm_rect_width(&plane_state->uapi.src) >> 17;
+	int h = drm_rect_height(&plane_state->uapi.src) >> 17;
 	u32 offset;
 
 	intel_add_fb_offsets(&x, &y, plane_state, 1);
@@ -3541,15 +3832,18 @@ static int skl_check_nv12_aux_surface(struct intel_plane_state *plane_state)
 
 static int skl_check_ccs_aux_surface(struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	int src_x = plane_state->base.src.x1 >> 16;
-	int src_y = plane_state->base.src.y1 >> 16;
-	int hsub = fb->format->hsub;
-	int vsub = fb->format->vsub;
-	int x = src_x / hsub;
-	int y = src_y / vsub;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	int src_x = plane_state->uapi.src.x1 >> 16;
+	int src_y = plane_state->uapi.src.y1 >> 16;
+	int hsub;
+	int vsub;
+	int x;
+	int y;
 	u32 offset;
 
+	intel_fb_plane_get_subsampling(&hsub, &vsub, fb, 1);
+	x = src_x / hsub;
+	y = src_y / vsub;
 	intel_add_fb_offsets(&x, &y, plane_state, 1);
 	offset = intel_plane_compute_aligned_offset(&x, &y, plane_state, 1);
 
@@ -3562,21 +3856,22 @@ static int skl_check_ccs_aux_surface(struct intel_plane_state *plane_state)
 
 int skl_check_plane_surface(struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int ret;
 
 	ret = intel_plane_compute_gtt(plane_state);
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	/*
 	 * Handle the AUX surface first since
 	 * the main surface setup depends on it.
 	 */
-	if (drm_format_info_is_yuv_semiplanar(fb->format)) {
+	if (intel_format_info_is_yuv_semiplanar(fb->format,
+						fb->modifier)) {
 		ret = skl_check_nv12_aux_surface(plane_state);
 		if (ret)
 			return ret;
@@ -3601,7 +3896,7 @@ static void i9xx_plane_ratio(const struct intel_crtc_state *crtc_state,
 			     const struct intel_plane_state *plane_state,
 			     unsigned int *num, unsigned int *den)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int cpp = fb->format->cpp[0];
 
 	/*
@@ -3673,7 +3968,7 @@ i9xx_plane_max_stride(struct intel_plane *plane,
 
 static u32 i9xx_plane_ctl_crtc(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 dspcntr = 0;
 
@@ -3693,9 +3988,9 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	u32 dspcntr;
 
 	dspcntr = DISPLAY_PLANE_ENABLE;
@@ -3711,6 +4006,9 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XRGB1555:
 		dspcntr |= DISPPLANE_BGRX555;
 		break;
+	case DRM_FORMAT_ARGB1555:
+		dspcntr |= DISPPLANE_BGRA555;
+		break;
 	case DRM_FORMAT_RGB565:
 		dspcntr |= DISPPLANE_BGRX565;
 		break;
@@ -3720,12 +4018,24 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XBGR8888:
 		dspcntr |= DISPPLANE_RGBX888;
 		break;
+	case DRM_FORMAT_ARGB8888:
+		dspcntr |= DISPPLANE_BGRA888;
+		break;
+	case DRM_FORMAT_ABGR8888:
+		dspcntr |= DISPPLANE_RGBA888;
+		break;
 	case DRM_FORMAT_XRGB2101010:
 		dspcntr |= DISPPLANE_BGRX101010;
 		break;
 	case DRM_FORMAT_XBGR2101010:
 		dspcntr |= DISPPLANE_RGBX101010;
 		break;
+	case DRM_FORMAT_ARGB2101010:
+		dspcntr |= DISPPLANE_BGRA101010;
+		break;
+	case DRM_FORMAT_ABGR2101010:
+		dspcntr |= DISPPLANE_RGBA101010;
+		break;
 	case DRM_FORMAT_XBGR16161616F:
 		dspcntr |= DISPPLANE_RGBX161616;
 		break;
@@ -3750,8 +4060,8 @@ static u32 i9xx_plane_ctl(const struct intel_crtc_state *crtc_state,
 int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int src_x, src_y, src_w;
 	u32 offset;
 	int ret;
@@ -3760,12 +4070,12 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
-	src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	src_x = plane_state->base.src.x1 >> 16;
-	src_y = plane_state->base.src.y1 >> 16;
+	src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	src_x = plane_state->uapi.src.x1 >> 16;
+	src_y = plane_state->uapi.src.y1 >> 16;
 
 	/* Undocumented hardware limit on i965/g4x/vlv/chv */
 	if (HAS_GMCH(dev_priv) && fb->format->cpp[0] == 8 && src_w > 2048)
@@ -3783,14 +4093,14 @@ int i9xx_check_plane_surface(struct intel_plane_state *plane_state)
 	 * Put the final coordinates back so that the src
 	 * coordinate checks will see the right values.
 	 */
-	drm_rect_translate_to(&plane_state->base.src,
+	drm_rect_translate_to(&plane_state->uapi.src,
 			      src_x << 16, src_y << 16);
 
 	/* HSW/BDW do this automagically in hardware */
 	if (!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv)) {
-		unsigned int rotation = plane_state->base.rotation;
-		int src_w = drm_rect_width(&plane_state->base.src) >> 16;
-		int src_h = drm_rect_height(&plane_state->base.src) >> 16;
+		unsigned int rotation = plane_state->hw.rotation;
+		int src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+		int src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
 
 		if (rotation & DRM_MODE_ROTATE_180) {
 			src_x += src_w - 1;
@@ -3827,15 +4137,15 @@ static int
 i9xx_plane_check(struct intel_crtc_state *crtc_state,
 		 struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	int ret;
 
 	ret = chv_plane_check_rotation(plane_state);
 	if (ret)
 		return ret;
 
-	ret = drm_atomic_helper_check_plane_state(&plane_state->base,
-						  &crtc_state->base,
+	ret = drm_atomic_helper_check_plane_state(&plane_state->uapi,
+						  &crtc_state->uapi,
 						  DRM_PLANE_HELPER_NO_SCALING,
 						  DRM_PLANE_HELPER_NO_SCALING,
 						  i9xx_plane_has_windowing(plane),
@@ -3847,7 +4157,7 @@ i9xx_plane_check(struct intel_crtc_state *crtc_state,
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	ret = intel_plane_check_src_coordinates(plane_state);
@@ -3868,10 +4178,10 @@ static void i9xx_update_plane(struct intel_plane *plane,
 	u32 linear_offset;
 	int x = plane_state->color_plane[0].x;
 	int y = plane_state->color_plane[0].y;
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_y = plane_state->base.dst.y1;
-	int crtc_w = drm_rect_width(&plane_state->base.dst);
-	int crtc_h = drm_rect_height(&plane_state->base.dst);
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_y = plane_state->uapi.dst.y1;
+	int crtc_w = drm_rect_width(&plane_state->uapi.dst);
+	int crtc_h = drm_rect_height(&plane_state->uapi.dst);
 	unsigned long irqflags;
 	u32 dspaddr_offset;
 	u32 dspcntr;
@@ -4011,7 +4321,7 @@ static void skl_detach_scaler(struct intel_crtc *intel_crtc, int id)
  */
 static void skl_detach_scalers(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	const struct intel_crtc_scaler_state *scaler_state =
 		&crtc_state->scaler_state;
 	int i;
@@ -4030,7 +4340,7 @@ static unsigned int skl_plane_stride_mult(const struct drm_framebuffer *fb,
 	 * The stride is either expressed as a multiple of 64 bytes chunks for
 	 * linear buffers or in number of tiles for tiled buffers.
 	 */
-	if (fb->modifier == DRM_FORMAT_MOD_LINEAR)
+	if (is_surface_linear(fb, color_plane))
 		return 64;
 	else if (drm_rotation_90_or_270(rotation))
 		return intel_tile_height(fb, color_plane);
@@ -4041,8 +4351,8 @@ static unsigned int skl_plane_stride_mult(const struct drm_framebuffer *fb,
 u32 skl_plane_stride(const struct intel_plane_state *plane_state,
 		     int color_plane)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	u32 stride = plane_state->color_plane[color_plane].stride;
 
 	if (color_plane >= fb->format->num_planes)
@@ -4065,8 +4375,10 @@ static u32 skl_plane_ctl_format(u32 pixel_format)
 	case DRM_FORMAT_ARGB8888:
 		return PLANE_CTL_FORMAT_XRGB_8888;
 	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ABGR2101010:
 		return PLANE_CTL_FORMAT_XRGB_2101010 | PLANE_CTL_ORDER_RGBX;
 	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_ARGB2101010:
 		return PLANE_CTL_FORMAT_XRGB_2101010;
 	case DRM_FORMAT_XBGR16161616F:
 	case DRM_FORMAT_ABGR16161616F:
@@ -4111,10 +4423,10 @@ static u32 skl_plane_ctl_format(u32 pixel_format)
 
 static u32 skl_plane_ctl_alpha(const struct intel_plane_state *plane_state)
 {
-	if (!plane_state->base.fb->format->has_alpha)
+	if (!plane_state->hw.fb->format->has_alpha)
 		return PLANE_CTL_ALPHA_DISABLE;
 
-	switch (plane_state->base.pixel_blend_mode) {
+	switch (plane_state->hw.pixel_blend_mode) {
 	case DRM_MODE_BLEND_PIXEL_NONE:
 		return PLANE_CTL_ALPHA_DISABLE;
 	case DRM_MODE_BLEND_PREMULTI:
@@ -4122,17 +4434,17 @@ static u32 skl_plane_ctl_alpha(const struct intel_plane_state *plane_state)
 	case DRM_MODE_BLEND_COVERAGE:
 		return PLANE_CTL_ALPHA_HW_PREMULTIPLY;
 	default:
-		MISSING_CASE(plane_state->base.pixel_blend_mode);
+		MISSING_CASE(plane_state->hw.pixel_blend_mode);
 		return PLANE_CTL_ALPHA_DISABLE;
 	}
 }
 
 static u32 glk_plane_color_ctl_alpha(const struct intel_plane_state *plane_state)
 {
-	if (!plane_state->base.fb->format->has_alpha)
+	if (!plane_state->hw.fb->format->has_alpha)
 		return PLANE_COLOR_ALPHA_DISABLE;
 
-	switch (plane_state->base.pixel_blend_mode) {
+	switch (plane_state->hw.pixel_blend_mode) {
 	case DRM_MODE_BLEND_PIXEL_NONE:
 		return PLANE_COLOR_ALPHA_DISABLE;
 	case DRM_MODE_BLEND_PREMULTI:
@@ -4140,7 +4452,7 @@ static u32 glk_plane_color_ctl_alpha(const struct intel_plane_state *plane_state
 	case DRM_MODE_BLEND_COVERAGE:
 		return PLANE_COLOR_ALPHA_HW_PREMULTIPLY;
 	default:
-		MISSING_CASE(plane_state->base.pixel_blend_mode);
+		MISSING_CASE(plane_state->hw.pixel_blend_mode);
 		return PLANE_COLOR_ALPHA_DISABLE;
 	}
 }
@@ -4156,6 +4468,10 @@ static u32 skl_plane_ctl_tiling(u64 fb_modifier)
 		return PLANE_CTL_TILED_Y;
 	case I915_FORMAT_MOD_Y_TILED_CCS:
 		return PLANE_CTL_TILED_Y | PLANE_CTL_RENDER_DECOMPRESSION_ENABLE;
+	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
+		return PLANE_CTL_TILED_Y |
+		       PLANE_CTL_RENDER_DECOMPRESSION_ENABLE |
+		       PLANE_CTL_CLEAR_COLOR_DISABLE;
 	case I915_FORMAT_MOD_Yf_TILED:
 		return PLANE_CTL_TILED_YF;
 	case I915_FORMAT_MOD_Yf_TILED_CCS:
@@ -4206,7 +4522,7 @@ static u32 cnl_plane_ctl_flip(unsigned int reflect)
 
 u32 skl_plane_ctl_crtc(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	u32 plane_ctl = 0;
 
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
@@ -4225,9 +4541,9 @@ u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state,
 		  const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 	u32 plane_ctl;
 
@@ -4237,10 +4553,10 @@ u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state,
 		plane_ctl |= skl_plane_ctl_alpha(plane_state);
 		plane_ctl |= PLANE_CTL_PLANE_GAMMA_DISABLE;
 
-		if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
+		if (plane_state->hw.color_encoding == DRM_COLOR_YCBCR_BT709)
 			plane_ctl |= PLANE_CTL_YUV_TO_RGB_CSC_FORMAT_BT709;
 
-		if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
+		if (plane_state->hw.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
 			plane_ctl |= PLANE_CTL_YUV_RANGE_CORRECTION_DISABLE;
 	}
 
@@ -4262,7 +4578,7 @@ u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state,
 
 u32 glk_plane_color_ctl_crtc(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	u32 plane_color_ctl = 0;
 
 	if (INTEL_GEN(dev_priv) >= 11)
@@ -4281,21 +4597,21 @@ u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 			const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	u32 plane_color_ctl = 0;
 
 	plane_color_ctl |= PLANE_COLOR_PLANE_GAMMA_DISABLE;
 	plane_color_ctl |= glk_plane_color_ctl_alpha(plane_state);
 
 	if (fb->format->is_yuv && !icl_is_hdr_plane(dev_priv, plane->id)) {
-		if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
+		if (plane_state->hw.color_encoding == DRM_COLOR_YCBCR_BT709)
 			plane_color_ctl |= PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709;
 		else
 			plane_color_ctl |= PLANE_COLOR_CSC_MODE_YUV601_TO_RGB709;
 
-		if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
+		if (plane_state->hw.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
 			plane_color_ctl |= PLANE_COLOR_YUV_RANGE_CORRECTION_DISABLE;
 	} else if (fb->format->is_yuv) {
 		plane_color_ctl |= PLANE_COLOR_INPUT_CSC_ENABLE;
@@ -4483,7 +4799,7 @@ static void icl_set_pipe_chicken(struct intel_crtc *crtc)
 
 static void icl_enable_trans_port_sync(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 trans_ddi_func_ctl2_val;
 	u8 master_select;
@@ -4511,25 +4827,6 @@ static void icl_enable_trans_port_sync(const struct intel_crtc_state *crtc_state
 		   trans_ddi_func_ctl2_val);
 }
 
-static void icl_disable_transcoder_port_sync(const struct intel_crtc_state *old_crtc_state)
-{
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	i915_reg_t reg;
-	u32 trans_ddi_func_ctl2_val;
-
-	if (old_crtc_state->master_transcoder == INVALID_TRANSCODER)
-		return;
-
-	DRM_DEBUG_KMS("Disabling Transcoder Port Sync on Slave Transcoder %s\n",
-		      transcoder_name(old_crtc_state->cpu_transcoder));
-
-	reg = TRANS_DDI_FUNC_CTL2(old_crtc_state->cpu_transcoder);
-	trans_ddi_func_ctl2_val = ~(PORT_SYNC_MODE_ENABLE |
-				    PORT_SYNC_MODE_MASTER_SELECT_MASK);
-	I915_WRITE(reg, trans_ddi_func_ctl2_val);
-}
-
 static void intel_fdi_normal_train(struct intel_crtc *crtc)
 {
 	struct drm_device *dev = crtc->base.dev;
@@ -4582,7 +4879,7 @@ static void ironlake_fdi_link_train(struct intel_crtc *crtc,
 	u32 temp, tries;
 
 	/* FDI needs bits from pipe first */
-	assert_pipe_enabled(dev_priv, pipe);
+	assert_pipe_enabled(dev_priv, crtc_state->cpu_transcoder);
 
 	/* Train 1: umask FDI RX Interrupt symbol_lock and bit_lock bit
 	   for train result */
@@ -4927,7 +5224,7 @@ train_done:
 
 static void ironlake_fdi_pll_enable(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
 	enum pipe pipe = intel_crtc->pipe;
 	i915_reg_t reg;
@@ -4992,12 +5289,10 @@ static void ironlake_fdi_pll_disable(struct intel_crtc *intel_crtc)
 	udelay(100);
 }
 
-static void ironlake_fdi_disable(struct drm_crtc *crtc)
+static void ironlake_fdi_disable(struct intel_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 	i915_reg_t reg;
 	u32 temp;
 
@@ -5088,9 +5383,9 @@ void lpt_disable_iclkip(struct drm_i915_private *dev_priv)
 /* Program iCLKIP clock to the desired frequency */
 static void lpt_program_iclkip(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	int clock = crtc_state->base.adjusted_mode.crtc_clock;
+	int clock = crtc_state->hw.adjusted_mode.crtc_clock;
 	u32 divsel, phaseinc, auxdiv, phasedir = 0;
 	u32 temp;
 
@@ -5204,7 +5499,7 @@ int lpt_get_iclkip(struct drm_i915_private *dev_priv)
 static void ironlake_pch_transcoder_set_timings(const struct intel_crtc_state *crtc_state,
 						enum pipe pch_transcoder)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 
@@ -5247,7 +5542,7 @@ static void cpt_set_fdi_bc_bifurcation(struct drm_i915_private *dev_priv, bool e
 
 static void ivybridge_update_fdi_bc_bifurcation(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	switch (crtc->pipe) {
@@ -5277,7 +5572,7 @@ static struct intel_encoder *
 intel_get_crtc_new_encoder(const struct intel_atomic_state *state,
 			   const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	const struct drm_connector_state *connector_state;
 	const struct drm_connector *connector;
 	struct intel_encoder *encoder = NULL;
@@ -5309,7 +5604,7 @@ intel_get_crtc_new_encoder(const struct intel_atomic_state *state,
 static void ironlake_pch_enable(const struct intel_atomic_state *state,
 				const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	enum pipe pipe = crtc->pipe;
@@ -5363,7 +5658,7 @@ static void ironlake_pch_enable(const struct intel_atomic_state *state,
 	if (HAS_PCH_CPT(dev_priv) &&
 	    intel_crtc_has_dp_encoder(crtc_state)) {
 		const struct drm_display_mode *adjusted_mode =
-			&crtc_state->base.adjusted_mode;
+			&crtc_state->hw.adjusted_mode;
 		u32 bpc = (I915_READ(PIPECONF(pipe)) & PIPECONF_BPC_MASK) >> 5;
 		i915_reg_t reg = TRANS_DP_CTL(pipe);
 		enum port port;
@@ -5393,7 +5688,7 @@ static void ironlake_pch_enable(const struct intel_atomic_state *state,
 static void lpt_pch_enable(const struct intel_atomic_state *state,
 			   const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 
@@ -5407,9 +5702,9 @@ static void lpt_pch_enable(const struct intel_atomic_state *state,
 	lpt_enable_pch_transcoder(dev_priv, cpu_transcoder);
 }
 
-static void cpt_verify_modeset(struct drm_device *dev, enum pipe pipe)
+static void cpt_verify_modeset(struct drm_i915_private *dev_priv,
+			       enum pipe pipe)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
 	i915_reg_t dslreg = PIPEDSL(pipe);
 	u32 temp;
 
@@ -5505,15 +5800,16 @@ static int
 skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
 		  unsigned int scaler_user, int *scaler_id,
 		  int src_w, int src_h, int dst_w, int dst_h,
-		  const struct drm_format_info *format, bool need_scaler)
+		  const struct drm_format_info *format,
+		  u64 modifier, bool need_scaler)
 {
 	struct intel_crtc_scaler_state *scaler_state =
 		&crtc_state->scaler_state;
 	struct intel_crtc *intel_crtc =
-		to_intel_crtc(crtc_state->base.crtc);
+		to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 
 	/*
 	 * Src coordinates are already rotated by 270 degrees for
@@ -5529,7 +5825,7 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
 	 * Once NV12 is enabled, handle it here while allocating scaler
 	 * for NV12.
 	 */
-	if (INTEL_GEN(dev_priv) >= 9 && crtc_state->base.enable &&
+	if (INTEL_GEN(dev_priv) >= 9 && crtc_state->hw.enable &&
 	    need_scaler && adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) {
 		DRM_DEBUG_KMS("Pipe/Plane scaling not supported with IF-ID mode\n");
 		return -EINVAL;
@@ -5559,7 +5855,7 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
 		return 0;
 	}
 
-	if (format && drm_format_info_is_yuv_semiplanar(format) &&
+	if (format && intel_format_info_is_yuv_semiplanar(format, modifier) &&
 	    (src_h < SKL_MIN_YUV_420_SRC_H || src_w < SKL_MIN_YUV_420_SRC_W)) {
 		DRM_DEBUG_KMS("Planar YUV: src dimensions not met\n");
 		return -EINVAL;
@@ -5601,17 +5897,18 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach,
  */
 int skl_update_scaler_crtc(struct intel_crtc_state *state)
 {
-	const struct drm_display_mode *adjusted_mode = &state->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &state->hw.adjusted_mode;
 	bool need_scaler = false;
 
 	if (state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420)
 		need_scaler = true;
 
-	return skl_update_scaler(state, !state->base.active, SKL_CRTC_INDEX,
+	return skl_update_scaler(state, !state->hw.active, SKL_CRTC_INDEX,
 				 &state->scaler_state.scaler_id,
 				 state->pipe_src_w, state->pipe_src_h,
 				 adjusted_mode->crtc_hdisplay,
-				 adjusted_mode->crtc_vdisplay, NULL, need_scaler);
+				 adjusted_mode->crtc_vdisplay, NULL, 0,
+				 need_scaler);
 }
 
 /**
@@ -5627,26 +5924,28 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 				   struct intel_plane_state *plane_state)
 {
 	struct intel_plane *intel_plane =
-		to_intel_plane(plane_state->base.plane);
+		to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(intel_plane->base.dev);
-	struct drm_framebuffer *fb = plane_state->base.fb;
+	struct drm_framebuffer *fb = plane_state->hw.fb;
 	int ret;
-	bool force_detach = !fb || !plane_state->base.visible;
+	bool force_detach = !fb || !plane_state->uapi.visible;
 	bool need_scaler = false;
 
 	/* Pre-gen11 and SDR planes always need a scaler for planar formats. */
 	if (!icl_is_hdr_plane(dev_priv, intel_plane->id) &&
-	    fb && drm_format_info_is_yuv_semiplanar(fb->format))
+	    fb && intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier))
 		need_scaler = true;
 
 	ret = skl_update_scaler(crtc_state, force_detach,
 				drm_plane_index(&intel_plane->base),
 				&plane_state->scaler_id,
-				drm_rect_width(&plane_state->base.src) >> 16,
-				drm_rect_height(&plane_state->base.src) >> 16,
-				drm_rect_width(&plane_state->base.dst),
-				drm_rect_height(&plane_state->base.dst),
-				fb ? fb->format : NULL, need_scaler);
+				drm_rect_width(&plane_state->uapi.src) >> 16,
+				drm_rect_height(&plane_state->uapi.src) >> 16,
+				drm_rect_width(&plane_state->uapi.dst),
+				drm_rect_height(&plane_state->uapi.dst),
+				fb ? fb->format : NULL,
+				fb ? fb->modifier : 0,
+				need_scaler);
 
 	if (ret || plane_state->scaler_id < 0)
 		return ret;
@@ -5668,6 +5967,8 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_ARGB8888:
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_ABGR2101010:
 	case DRM_FORMAT_YUYV:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
@@ -5700,8 +6001,9 @@ static int skl_update_scaler_plane(struct intel_crtc_state *crtc_state,
 	return 0;
 }
 
-static void skylake_scaler_disable(struct intel_crtc *crtc)
+void skylake_scaler_disable(const struct intel_crtc_state *old_crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	int i;
 
 	for (i = 0; i < crtc->num_scalers; i++)
@@ -5710,7 +6012,7 @@ static void skylake_scaler_disable(struct intel_crtc *crtc)
 
 static void skylake_pfit_enable(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	const struct intel_crtc_scaler_state *scaler_state =
@@ -5747,7 +6049,7 @@ static void skylake_pfit_enable(const struct intel_crtc_state *crtc_state)
 
 static void ironlake_pfit_enable(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
@@ -5768,7 +6070,7 @@ static void ironlake_pfit_enable(const struct intel_crtc_state *crtc_state)
 
 void hsw_enable_ips(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
@@ -5804,7 +6106,7 @@ void hsw_enable_ips(const struct intel_crtc_state *crtc_state)
 
 void hsw_disable_ips(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 
@@ -5839,77 +6141,10 @@ static void intel_crtc_dpms_overlay_disable(struct intel_crtc *intel_crtc)
 	 */
 }
 
-/**
- * intel_post_enable_primary - Perform operations after enabling primary plane
- * @crtc: the CRTC whose primary plane was just enabled
- * @new_crtc_state: the enabling state
- *
- * Performs potentially sleeping operations that must be done after the primary
- * plane is enabled, such as updating FBC and IPS.  Note that this may be
- * called due to an explicit primary plane update, or due to an implicit
- * re-enable that is caused when a sprite plane is updated to no longer
- * completely hide the primary plane.
- */
-static void
-intel_post_enable_primary(struct drm_crtc *crtc,
-			  const struct intel_crtc_state *new_crtc_state)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
-
-	/*
-	 * Gen2 reports pipe underruns whenever all planes are disabled.
-	 * So don't enable underrun reporting before at least some planes
-	 * are enabled.
-	 * FIXME: Need to fix the logic to work when we turn off all planes
-	 * but leave the pipe running.
-	 */
-	if (IS_GEN(dev_priv, 2))
-		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
-
-	/* Underruns don't always raise interrupts, so check manually. */
-	intel_check_cpu_fifo_underruns(dev_priv);
-	intel_check_pch_fifo_underruns(dev_priv);
-}
-
-/* FIXME get rid of this and use pre_plane_update */
-static void
-intel_pre_disable_primary_noatomic(struct drm_crtc *crtc)
-{
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
-
-	/*
-	 * Gen2 reports pipe underruns whenever all planes are disabled.
-	 * So disable underrun reporting before all the planes get disabled.
-	 */
-	if (IS_GEN(dev_priv, 2))
-		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
-
-	hsw_disable_ips(to_intel_crtc_state(crtc->state));
-
-	/*
-	 * Vblank time updates from the shadow to live plane control register
-	 * are blocked if the memory self-refresh mode is active at that
-	 * moment. So to make sure the plane gets truly disabled, disable
-	 * first the self-refresh mode. The self-refresh enable bit in turn
-	 * will be checked/applied by the HW only at the next frame start
-	 * event which is after the vblank start event, so we need to have a
-	 * wait-for-vblank between disabling the plane and the pipe.
-	 */
-	if (HAS_GMCH(dev_priv) &&
-	    intel_set_memory_cxsr(dev_priv, false))
-		intel_wait_for_vblank(dev_priv, pipe);
-}
-
 static bool hsw_pre_update_disable_ips(const struct intel_crtc_state *old_crtc_state,
 				       const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (!old_crtc_state->ips_enabled)
@@ -5925,7 +6160,7 @@ static bool hsw_pre_update_disable_ips(const struct intel_crtc_state *old_crtc_s
 	 * Disable IPS before we program the LUT.
 	 */
 	if (IS_HASWELL(dev_priv) &&
-	    (new_crtc_state->base.color_mgmt_changed ||
+	    (new_crtc_state->uapi.color_mgmt_changed ||
 	     new_crtc_state->update_pipe) &&
 	    new_crtc_state->gamma_mode == GAMMA_MODE_MODE_SPLIT)
 		return true;
@@ -5936,7 +6171,7 @@ static bool hsw_pre_update_disable_ips(const struct intel_crtc_state *old_crtc_s
 static bool hsw_post_update_enable_ips(const struct intel_crtc_state *old_crtc_state,
 				       const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (!new_crtc_state->ips_enabled)
@@ -5952,7 +6187,7 @@ static bool hsw_post_update_enable_ips(const struct intel_crtc_state *old_crtc_s
 	 * Re-enable IPS after the LUT has been programmed.
 	 */
 	if (IS_HASWELL(dev_priv) &&
-	    (new_crtc_state->base.color_mgmt_changed ||
+	    (new_crtc_state->uapi.color_mgmt_changed ||
 	     new_crtc_state->update_pipe) &&
 	    new_crtc_state->gamma_mode == GAMMA_MODE_MODE_SPLIT)
 		return true;
@@ -5962,15 +6197,16 @@ static bool hsw_post_update_enable_ips(const struct intel_crtc_state *old_crtc_s
 	 * forcibly enable IPS on the first fastset.
 	 */
 	if (new_crtc_state->update_pipe &&
-	    old_crtc_state->base.adjusted_mode.private_flags & I915_MODE_FLAG_INHERITED)
+	    old_crtc_state->hw.adjusted_mode.private_flags & I915_MODE_FLAG_INHERITED)
 		return true;
 
 	return !old_crtc_state->ips_enabled;
 }
 
-static bool needs_nv12_wa(struct drm_i915_private *dev_priv,
-			  const struct intel_crtc_state *crtc_state)
+static bool needs_nv12_wa(const struct intel_crtc_state *crtc_state)
 {
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
+
 	if (!crtc_state->nv12_planes)
 		return false;
 
@@ -5981,9 +6217,10 @@ static bool needs_nv12_wa(struct drm_i915_private *dev_priv,
 	return false;
 }
 
-static bool needs_scalerclk_wa(struct drm_i915_private *dev_priv,
-			       const struct intel_crtc_state *crtc_state)
+static bool needs_scalerclk_wa(const struct intel_crtc_state *crtc_state)
 {
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
+
 	/* Wa_2006604312:icl */
 	if (crtc_state->scaler_state.scaler_users > 0 && IS_ICELAKE(dev_priv))
 		return true;
@@ -5991,89 +6228,82 @@ static bool needs_scalerclk_wa(struct drm_i915_private *dev_priv,
 	return false;
 }
 
-static void intel_post_plane_update(struct intel_crtc_state *old_crtc_state)
+static bool planes_enabling(const struct intel_crtc_state *old_crtc_state,
+			    const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct drm_atomic_state *state = old_crtc_state->base.state;
-	struct intel_crtc_state *pipe_config =
-		intel_atomic_get_new_crtc_state(to_intel_atomic_state(state),
-						crtc);
-	struct drm_plane *primary = crtc->base.primary;
-	struct drm_plane_state *old_primary_state =
-		drm_atomic_get_old_plane_state(state, primary);
+	return (!old_crtc_state->active_planes || needs_modeset(new_crtc_state)) &&
+		new_crtc_state->active_planes;
+}
 
-	intel_frontbuffer_flip(to_i915(crtc->base.dev), pipe_config->fb_bits);
+static bool planes_disabling(const struct intel_crtc_state *old_crtc_state,
+			     const struct intel_crtc_state *new_crtc_state)
+{
+	return old_crtc_state->active_planes &&
+		(!new_crtc_state->active_planes || needs_modeset(new_crtc_state));
+}
 
-	if (pipe_config->update_wm_post && pipe_config->base.active)
-		intel_update_watermarks(crtc);
+static void intel_post_plane_update(struct intel_atomic_state *state,
+				    struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	struct intel_plane *primary = to_intel_plane(crtc->base.primary);
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	const struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct intel_plane_state *new_primary_state =
+		intel_atomic_get_new_plane_state(state, primary);
+	enum pipe pipe = crtc->pipe;
+
+	intel_frontbuffer_flip(dev_priv, new_crtc_state->fb_bits);
 
-	if (hsw_post_update_enable_ips(old_crtc_state, pipe_config))
-		hsw_enable_ips(pipe_config);
+	if (new_crtc_state->update_wm_post && new_crtc_state->hw.active)
+		intel_update_watermarks(crtc);
 
-	if (old_primary_state) {
-		struct drm_plane_state *new_primary_state =
-			drm_atomic_get_new_plane_state(state, primary);
+	if (hsw_post_update_enable_ips(old_crtc_state, new_crtc_state))
+		hsw_enable_ips(new_crtc_state);
 
+	if (new_primary_state)
 		intel_fbc_post_update(crtc);
 
-		if (new_primary_state->visible &&
-		    (needs_modeset(pipe_config) ||
-		     !old_primary_state->visible))
-			intel_post_enable_primary(&crtc->base, pipe_config);
-	}
-
-	if (needs_nv12_wa(dev_priv, old_crtc_state) &&
-	    !needs_nv12_wa(dev_priv, pipe_config))
-		skl_wa_827(dev_priv, crtc->pipe, false);
+	if (needs_nv12_wa(old_crtc_state) &&
+	    !needs_nv12_wa(new_crtc_state))
+		skl_wa_827(dev_priv, pipe, false);
 
-	if (needs_scalerclk_wa(dev_priv, old_crtc_state) &&
-	    !needs_scalerclk_wa(dev_priv, pipe_config))
-		icl_wa_scalerclkgating(dev_priv, crtc->pipe, false);
+	if (needs_scalerclk_wa(old_crtc_state) &&
+	    !needs_scalerclk_wa(new_crtc_state))
+		icl_wa_scalerclkgating(dev_priv, pipe, false);
 }
 
-static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
-				   struct intel_crtc_state *pipe_config)
+static void intel_pre_plane_update(struct intel_atomic_state *state,
+				   struct intel_crtc *crtc)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct drm_atomic_state *state = old_crtc_state->base.state;
-	struct drm_plane *primary = crtc->base.primary;
-	struct drm_plane_state *old_primary_state =
-		drm_atomic_get_old_plane_state(state, primary);
-	bool modeset = needs_modeset(pipe_config);
-	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(state);
+	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
+	struct intel_plane *primary = to_intel_plane(crtc->base.primary);
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	const struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct intel_plane_state *new_primary_state =
+		intel_atomic_get_new_plane_state(state, primary);
+	enum pipe pipe = crtc->pipe;
 
-	if (hsw_pre_update_disable_ips(old_crtc_state, pipe_config))
+	if (hsw_pre_update_disable_ips(old_crtc_state, new_crtc_state))
 		hsw_disable_ips(old_crtc_state);
 
-	if (old_primary_state) {
-		struct intel_plane_state *new_primary_state =
-			intel_atomic_get_new_plane_state(intel_state,
-							 to_intel_plane(primary));
-
-		intel_fbc_pre_update(crtc, pipe_config, new_primary_state);
-		/*
-		 * Gen2 reports pipe underruns whenever all planes are disabled.
-		 * So disable underrun reporting before all the planes get disabled.
-		 */
-		if (IS_GEN(dev_priv, 2) && old_primary_state->visible &&
-		    (modeset || !new_primary_state->base.visible))
-			intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, false);
-	}
+	if (new_primary_state &&
+	    intel_fbc_pre_update(crtc, new_crtc_state, new_primary_state))
+		intel_wait_for_vblank(dev_priv, pipe);
 
 	/* Display WA 827 */
-	if (!needs_nv12_wa(dev_priv, old_crtc_state) &&
-	    needs_nv12_wa(dev_priv, pipe_config))
-		skl_wa_827(dev_priv, crtc->pipe, true);
+	if (!needs_nv12_wa(old_crtc_state) &&
+	    needs_nv12_wa(new_crtc_state))
+		skl_wa_827(dev_priv, pipe, true);
 
 	/* Wa_2006604312:icl */
-	if (!needs_scalerclk_wa(dev_priv, old_crtc_state) &&
-	    needs_scalerclk_wa(dev_priv, pipe_config))
-		icl_wa_scalerclkgating(dev_priv, crtc->pipe, true);
+	if (!needs_scalerclk_wa(old_crtc_state) &&
+	    needs_scalerclk_wa(new_crtc_state))
+		icl_wa_scalerclkgating(dev_priv, pipe, true);
 
 	/*
 	 * Vblank time updates from the shadow to live plane control register
@@ -6084,9 +6314,9 @@ static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
 	 * event which is after the vblank start event, so we need to have a
 	 * wait-for-vblank between disabling the plane and the pipe.
 	 */
-	if (HAS_GMCH(dev_priv) && old_crtc_state->base.active &&
-	    pipe_config->disable_cxsr && intel_set_memory_cxsr(dev_priv, false))
-		intel_wait_for_vblank(dev_priv, crtc->pipe);
+	if (HAS_GMCH(dev_priv) && old_crtc_state->hw.active &&
+	    new_crtc_state->disable_cxsr && intel_set_memory_cxsr(dev_priv, false))
+		intel_wait_for_vblank(dev_priv, pipe);
 
 	/*
 	 * IVB workaround: must disable low power watermarks for at least
@@ -6095,36 +6325,45 @@ static void intel_pre_plane_update(struct intel_crtc_state *old_crtc_state,
 	 *
 	 * WaCxSRDisabledForSpriteScaling:ivb
 	 */
-	if (pipe_config->disable_lp_wm && ilk_disable_lp_wm(dev) &&
-	    old_crtc_state->base.active)
-		intel_wait_for_vblank(dev_priv, crtc->pipe);
+	if (old_crtc_state->hw.active &&
+	    new_crtc_state->disable_lp_wm && ilk_disable_lp_wm(dev_priv))
+		intel_wait_for_vblank(dev_priv, pipe);
 
 	/*
-	 * If we're doing a modeset, we're done.  No need to do any pre-vblank
-	 * watermark programming here.
+	 * If we're doing a modeset we don't need to do any
+	 * pre-vblank watermark programming here.
 	 */
-	if (needs_modeset(pipe_config))
-		return;
+	if (!needs_modeset(new_crtc_state)) {
+		/*
+		 * For platforms that support atomic watermarks, program the
+		 * 'intermediate' watermarks immediately.  On pre-gen9 platforms, these
+		 * will be the intermediate values that are safe for both pre- and
+		 * post- vblank; when vblank happens, the 'active' values will be set
+		 * to the final 'target' values and we'll do this again to get the
+		 * optimal watermarks.  For gen9+ platforms, the values we program here
+		 * will be the final target values which will get automatically latched
+		 * at vblank time; no further programming will be necessary.
+		 *
+		 * If a platform hasn't been transitioned to atomic watermarks yet,
+		 * we'll continue to update watermarks the old way, if flags tell
+		 * us to.
+		 */
+		if (dev_priv->display.initial_watermarks)
+			dev_priv->display.initial_watermarks(state, crtc);
+		else if (new_crtc_state->update_wm_pre)
+			intel_update_watermarks(crtc);
+	}
 
 	/*
-	 * For platforms that support atomic watermarks, program the
-	 * 'intermediate' watermarks immediately.  On pre-gen9 platforms, these
-	 * will be the intermediate values that are safe for both pre- and
-	 * post- vblank; when vblank happens, the 'active' values will be set
-	 * to the final 'target' values and we'll do this again to get the
-	 * optimal watermarks.  For gen9+ platforms, the values we program here
-	 * will be the final target values which will get automatically latched
-	 * at vblank time; no further programming will be necessary.
+	 * Gen2 reports pipe underruns whenever all planes are disabled.
+	 * So disable underrun reporting before all the planes get disabled.
 	 *
-	 * If a platform hasn't been transitioned to atomic watermarks yet,
-	 * we'll continue to update watermarks the old way, if flags tell
-	 * us to.
+	 * We do this after .initial_watermarks() so that we have a
+	 * chance of catching underruns with the intermediate watermarks
+	 * vs. the old plane configuration.
 	 */
-	if (dev_priv->display.initial_watermarks != NULL)
-		dev_priv->display.initial_watermarks(intel_state,
-						     pipe_config);
-	else if (pipe_config->update_wm_pre)
-		intel_update_watermarks(crtc);
+	if (IS_GEN(dev_priv, 2) && planes_disabling(old_crtc_state, new_crtc_state))
+		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 }
 
 static void intel_crtc_disable_planes(struct intel_atomic_state *state,
@@ -6148,7 +6387,7 @@ static void intel_crtc_disable_planes(struct intel_atomic_state *state,
 
 		intel_disable_plane(plane, new_crtc_state);
 
-		if (old_plane_state->base.visible)
+		if (old_plane_state->uapi.visible)
 			fb_bits |= plane->frontbuffer_bit;
 	}
 
@@ -6247,11 +6486,12 @@ static void intel_encoders_update_complete(struct intel_atomic_state *state)
 	}
 }
 
-static void intel_encoders_pre_pll_enable(struct intel_crtc *crtc,
-					  struct intel_crtc_state *crtc_state,
-					  struct intel_atomic_state *state)
+static void intel_encoders_pre_pll_enable(struct intel_atomic_state *state,
+					  struct intel_crtc *crtc)
 {
-	struct drm_connector_state *conn_state;
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct drm_connector_state *conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6267,11 +6507,12 @@ static void intel_encoders_pre_pll_enable(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_encoders_pre_enable(struct intel_crtc *crtc,
-				      struct intel_crtc_state *crtc_state,
-				      struct intel_atomic_state *state)
+static void intel_encoders_pre_enable(struct intel_atomic_state *state,
+				      struct intel_crtc *crtc)
 {
-	struct drm_connector_state *conn_state;
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct drm_connector_state *conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6287,11 +6528,12 @@ static void intel_encoders_pre_enable(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_encoders_enable(struct intel_crtc *crtc,
-				  struct intel_crtc_state *crtc_state,
-				  struct intel_atomic_state *state)
+static void intel_encoders_enable(struct intel_atomic_state *state,
+				  struct intel_crtc *crtc)
 {
-	struct drm_connector_state *conn_state;
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct drm_connector_state *conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6308,11 +6550,12 @@ static void intel_encoders_enable(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_encoders_disable(struct intel_crtc *crtc,
-				   struct intel_crtc_state *old_crtc_state,
-				   struct intel_atomic_state *state)
+static void intel_encoders_disable(struct intel_atomic_state *state,
+				   struct intel_crtc *crtc)
 {
-	struct drm_connector_state *old_conn_state;
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	const struct drm_connector_state *old_conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6329,11 +6572,12 @@ static void intel_encoders_disable(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_encoders_post_disable(struct intel_crtc *crtc,
-					struct intel_crtc_state *old_crtc_state,
-					struct intel_atomic_state *state)
+static void intel_encoders_post_disable(struct intel_atomic_state *state,
+					struct intel_crtc *crtc)
 {
-	struct drm_connector_state *old_conn_state;
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	const struct drm_connector_state *old_conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6349,11 +6593,12 @@ static void intel_encoders_post_disable(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_encoders_post_pll_disable(struct intel_crtc *crtc,
-					    struct intel_crtc_state *old_crtc_state,
-					    struct intel_atomic_state *state)
+static void intel_encoders_post_pll_disable(struct intel_atomic_state *state,
+					    struct intel_crtc *crtc)
 {
-	struct drm_connector_state *old_conn_state;
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	const struct drm_connector_state *old_conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6369,11 +6614,12 @@ static void intel_encoders_post_pll_disable(struct intel_crtc *crtc,
 	}
 }
 
-static void intel_encoders_update_pipe(struct intel_crtc *crtc,
-				       struct intel_crtc_state *crtc_state,
-				       struct intel_atomic_state *state)
+static void intel_encoders_update_pipe(struct intel_atomic_state *state,
+				       struct intel_crtc *crtc)
 {
-	struct drm_connector_state *conn_state;
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct drm_connector_state *conn_state;
 	struct drm_connector *conn;
 	int i;
 
@@ -6391,22 +6637,21 @@ static void intel_encoders_update_pipe(struct intel_crtc *crtc,
 
 static void intel_disable_primary_plane(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_plane *plane = to_intel_plane(crtc->base.primary);
 
 	plane->disable_plane(plane, crtc_state);
 }
 
-static void ironlake_crtc_enable(struct intel_crtc_state *pipe_config,
-				 struct intel_atomic_state *state)
+static void ironlake_crtc_enable(struct intel_atomic_state *state,
+				 struct intel_crtc *crtc)
 {
-	struct drm_crtc *crtc = pipe_config->base.crtc;
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
+	const struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 
-	if (WARN_ON(intel_crtc->active))
+	if (WARN_ON(crtc->active))
 		return;
 
 	/*
@@ -6422,61 +6667,59 @@ static void ironlake_crtc_enable(struct intel_crtc_state *pipe_config,
 	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 	intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false);
 
-	if (pipe_config->has_pch_encoder)
-		intel_prepare_shared_dpll(pipe_config);
+	if (new_crtc_state->has_pch_encoder)
+		intel_prepare_shared_dpll(new_crtc_state);
 
-	if (intel_crtc_has_dp_encoder(pipe_config))
-		intel_dp_set_m_n(pipe_config, M1_N1);
+	if (intel_crtc_has_dp_encoder(new_crtc_state))
+		intel_dp_set_m_n(new_crtc_state, M1_N1);
 
-	intel_set_pipe_timings(pipe_config);
-	intel_set_pipe_src_size(pipe_config);
+	intel_set_pipe_timings(new_crtc_state);
+	intel_set_pipe_src_size(new_crtc_state);
 
-	if (pipe_config->has_pch_encoder) {
-		intel_cpu_transcoder_set_m_n(pipe_config,
-					     &pipe_config->fdi_m_n, NULL);
-	}
+	if (new_crtc_state->has_pch_encoder)
+		intel_cpu_transcoder_set_m_n(new_crtc_state,
+					     &new_crtc_state->fdi_m_n, NULL);
 
-	ironlake_set_pipeconf(pipe_config);
+	ironlake_set_pipeconf(new_crtc_state);
 
-	intel_crtc->active = true;
+	crtc->active = true;
 
-	intel_encoders_pre_enable(intel_crtc, pipe_config, state);
+	intel_encoders_pre_enable(state, crtc);
 
-	if (pipe_config->has_pch_encoder) {
+	if (new_crtc_state->has_pch_encoder) {
 		/* Note: FDI PLL enabling _must_ be done before we enable the
 		 * cpu pipes, hence this is separate from all the other fdi/pch
 		 * enabling. */
-		ironlake_fdi_pll_enable(pipe_config);
+		ironlake_fdi_pll_enable(new_crtc_state);
 	} else {
 		assert_fdi_tx_disabled(dev_priv, pipe);
 		assert_fdi_rx_disabled(dev_priv, pipe);
 	}
 
-	ironlake_pfit_enable(pipe_config);
+	ironlake_pfit_enable(new_crtc_state);
 
 	/*
 	 * On ILK+ LUT must be loaded before the pipe is running but with
 	 * clocks enabled
 	 */
-	intel_color_load_luts(pipe_config);
-	intel_color_commit(pipe_config);
+	intel_color_load_luts(new_crtc_state);
+	intel_color_commit(new_crtc_state);
 	/* update DSPCNTR to configure gamma for pipe bottom color */
-	intel_disable_primary_plane(pipe_config);
+	intel_disable_primary_plane(new_crtc_state);
 
-	if (dev_priv->display.initial_watermarks != NULL)
-		dev_priv->display.initial_watermarks(state, pipe_config);
-	intel_enable_pipe(pipe_config);
+	if (dev_priv->display.initial_watermarks)
+		dev_priv->display.initial_watermarks(state, crtc);
+	intel_enable_pipe(new_crtc_state);
 
-	if (pipe_config->has_pch_encoder)
-		ironlake_pch_enable(state, pipe_config);
+	if (new_crtc_state->has_pch_encoder)
+		ironlake_pch_enable(state, new_crtc_state);
 
-	assert_vblank_disabled(crtc);
-	intel_crtc_vblank_on(pipe_config);
+	intel_crtc_vblank_on(new_crtc_state);
 
-	intel_encoders_enable(intel_crtc, pipe_config, state);
+	intel_encoders_enable(state, crtc);
 
 	if (HAS_PCH_CPT(dev_priv))
-		cpt_verify_modeset(dev, intel_crtc->pipe);
+		cpt_verify_modeset(dev_priv, pipe);
 
 	/*
 	 * Must wait for vblank to avoid spurious PCH FIFO underruns.
@@ -6484,7 +6727,7 @@ static void ironlake_crtc_enable(struct intel_crtc_state *pipe_config,
 	 * some interlaced HDMI modes. Let's do the double wait always
 	 * in case there are more corner cases we don't know about.
 	 */
-	if (pipe_config->has_pch_encoder) {
+	if (new_crtc_state->has_pch_encoder) {
 		intel_wait_for_vblank(dev_priv, pipe);
 		intel_wait_for_vblank(dev_priv, pipe);
 	}
@@ -6531,103 +6774,112 @@ static void icl_pipe_mbus_enable(struct intel_crtc *crtc)
 	I915_WRITE(PIPE_MBUS_DBOX_CTL(pipe), val);
 }
 
-static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
-				struct intel_atomic_state *state)
+static void hsw_set_frame_start_delay(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_crtc *crtc = pipe_config->base.crtc;
-	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe, hsw_workaround_pipe;
-	enum transcoder cpu_transcoder = pipe_config->cpu_transcoder;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	i915_reg_t reg = CHICKEN_TRANS(crtc_state->cpu_transcoder);
+	u32 val;
+
+	val = I915_READ(reg);
+	val &= ~HSW_FRAME_START_DELAY_MASK;
+	val |= HSW_FRAME_START_DELAY(0);
+	I915_WRITE(reg, val);
+}
+
+static void haswell_crtc_enable(struct intel_atomic_state *state,
+				struct intel_crtc *crtc)
+{
+	const struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe, hsw_workaround_pipe;
+	enum transcoder cpu_transcoder = new_crtc_state->cpu_transcoder;
 	bool psl_clkgate_wa;
 
-	if (WARN_ON(intel_crtc->active))
+	if (WARN_ON(crtc->active))
 		return;
 
-	intel_encoders_pre_pll_enable(intel_crtc, pipe_config, state);
+	intel_encoders_pre_pll_enable(state, crtc);
 
-	if (pipe_config->shared_dpll)
-		intel_enable_shared_dpll(pipe_config);
+	if (new_crtc_state->shared_dpll)
+		intel_enable_shared_dpll(new_crtc_state);
 
-	intel_encoders_pre_enable(intel_crtc, pipe_config, state);
+	intel_encoders_pre_enable(state, crtc);
 
-	if (intel_crtc_has_dp_encoder(pipe_config))
-		intel_dp_set_m_n(pipe_config, M1_N1);
+	if (intel_crtc_has_dp_encoder(new_crtc_state))
+		intel_dp_set_m_n(new_crtc_state, M1_N1);
 
 	if (!transcoder_is_dsi(cpu_transcoder))
-		intel_set_pipe_timings(pipe_config);
+		intel_set_pipe_timings(new_crtc_state);
 
 	if (INTEL_GEN(dev_priv) >= 11)
-		icl_enable_trans_port_sync(pipe_config);
+		icl_enable_trans_port_sync(new_crtc_state);
 
-	intel_set_pipe_src_size(pipe_config);
+	intel_set_pipe_src_size(new_crtc_state);
 
 	if (cpu_transcoder != TRANSCODER_EDP &&
-	    !transcoder_is_dsi(cpu_transcoder)) {
+	    !transcoder_is_dsi(cpu_transcoder))
 		I915_WRITE(PIPE_MULT(cpu_transcoder),
-			   pipe_config->pixel_multiplier - 1);
-	}
+			   new_crtc_state->pixel_multiplier - 1);
 
-	if (pipe_config->has_pch_encoder) {
-		intel_cpu_transcoder_set_m_n(pipe_config,
-					     &pipe_config->fdi_m_n, NULL);
-	}
+	if (new_crtc_state->has_pch_encoder)
+		intel_cpu_transcoder_set_m_n(new_crtc_state,
+					     &new_crtc_state->fdi_m_n, NULL);
 
-	if (!transcoder_is_dsi(cpu_transcoder))
-		haswell_set_pipeconf(pipe_config);
+	if (!transcoder_is_dsi(cpu_transcoder)) {
+		hsw_set_frame_start_delay(new_crtc_state);
+		haswell_set_pipeconf(new_crtc_state);
+	}
 
 	if (INTEL_GEN(dev_priv) >= 9 || IS_BROADWELL(dev_priv))
-		bdw_set_pipemisc(pipe_config);
+		bdw_set_pipemisc(new_crtc_state);
 
-	intel_crtc->active = true;
+	crtc->active = true;
 
 	/* Display WA #1180: WaDisableScalarClockGating: glk, cnl */
 	psl_clkgate_wa = (IS_GEMINILAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) &&
-			 pipe_config->pch_pfit.enabled;
+		new_crtc_state->pch_pfit.enabled;
 	if (psl_clkgate_wa)
 		glk_pipe_scaler_clock_gating_wa(dev_priv, pipe, true);
 
 	if (INTEL_GEN(dev_priv) >= 9)
-		skylake_pfit_enable(pipe_config);
+		skylake_pfit_enable(new_crtc_state);
 	else
-		ironlake_pfit_enable(pipe_config);
+		ironlake_pfit_enable(new_crtc_state);
 
 	/*
 	 * On ILK+ LUT must be loaded before the pipe is running but with
 	 * clocks enabled
 	 */
-	intel_color_load_luts(pipe_config);
-	intel_color_commit(pipe_config);
+	intel_color_load_luts(new_crtc_state);
+	intel_color_commit(new_crtc_state);
 	/* update DSPCNTR to configure gamma/csc for pipe bottom color */
 	if (INTEL_GEN(dev_priv) < 9)
-		intel_disable_primary_plane(pipe_config);
+		intel_disable_primary_plane(new_crtc_state);
 
 	if (INTEL_GEN(dev_priv) >= 11)
-		icl_set_pipe_chicken(intel_crtc);
+		icl_set_pipe_chicken(crtc);
 
 	if (!transcoder_is_dsi(cpu_transcoder))
-		intel_ddi_enable_transcoder_func(pipe_config);
+		intel_ddi_enable_transcoder_func(new_crtc_state);
 
-	if (dev_priv->display.initial_watermarks != NULL)
-		dev_priv->display.initial_watermarks(state, pipe_config);
+	if (dev_priv->display.initial_watermarks)
+		dev_priv->display.initial_watermarks(state, crtc);
 
 	if (INTEL_GEN(dev_priv) >= 11)
-		icl_pipe_mbus_enable(intel_crtc);
+		icl_pipe_mbus_enable(crtc);
 
 	/* XXX: Do the pipe assertions at the right place for BXT DSI. */
 	if (!transcoder_is_dsi(cpu_transcoder))
-		intel_enable_pipe(pipe_config);
-
-	if (pipe_config->has_pch_encoder)
-		lpt_pch_enable(state, pipe_config);
+		intel_enable_pipe(new_crtc_state);
 
-	if (intel_crtc_has_type(pipe_config, INTEL_OUTPUT_DP_MST))
-		intel_ddi_set_vc_payload_alloc(pipe_config, true);
+	if (new_crtc_state->has_pch_encoder)
+		lpt_pch_enable(state, new_crtc_state);
 
-	assert_vblank_disabled(crtc);
-	intel_crtc_vblank_on(pipe_config);
+	intel_crtc_vblank_on(new_crtc_state);
 
-	intel_encoders_enable(intel_crtc, pipe_config, state);
+	intel_encoders_enable(state, crtc);
 
 	if (psl_clkgate_wa) {
 		intel_wait_for_vblank(dev_priv, pipe);
@@ -6636,16 +6888,16 @@ static void haswell_crtc_enable(struct intel_crtc_state *pipe_config,
 
 	/* If we change the relative order between pipe/planes enabling, we need
 	 * to change the workaround. */
-	hsw_workaround_pipe = pipe_config->hsw_workaround_pipe;
+	hsw_workaround_pipe = new_crtc_state->hsw_workaround_pipe;
 	if (IS_HASWELL(dev_priv) && hsw_workaround_pipe != INVALID_PIPE) {
 		intel_wait_for_vblank(dev_priv, hsw_workaround_pipe);
 		intel_wait_for_vblank(dev_priv, hsw_workaround_pipe);
 	}
 }
 
-static void ironlake_pfit_disable(const struct intel_crtc_state *old_crtc_state)
+void ironlake_pfit_disable(const struct intel_crtc_state *old_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
@@ -6658,14 +6910,13 @@ static void ironlake_pfit_disable(const struct intel_crtc_state *old_crtc_state)
 	}
 }
 
-static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state,
-				  struct intel_atomic_state *state)
+static void ironlake_crtc_disable(struct intel_atomic_state *state,
+				  struct intel_crtc *crtc)
 {
-	struct drm_crtc *crtc = old_crtc_state->base.crtc;
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
+	const struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 
 	/*
 	 * Sometimes spurious CPU pipe underruns happen when the
@@ -6675,10 +6926,9 @@ static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state,
 	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 	intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, false);
 
-	intel_encoders_disable(intel_crtc, old_crtc_state, state);
+	intel_encoders_disable(state, crtc);
 
-	drm_crtc_vblank_off(crtc);
-	assert_vblank_disabled(crtc);
+	intel_crtc_vblank_off(old_crtc_state);
 
 	intel_disable_pipe(old_crtc_state);
 
@@ -6687,7 +6937,7 @@ static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state,
 	if (old_crtc_state->has_pch_encoder)
 		ironlake_fdi_disable(crtc);
 
-	intel_encoders_post_disable(intel_crtc, old_crtc_state, state);
+	intel_encoders_post_disable(state, crtc);
 
 	if (old_crtc_state->has_pch_encoder) {
 		ironlake_disable_pch_transcoder(dev_priv, pipe);
@@ -6710,54 +6960,27 @@ static void ironlake_crtc_disable(struct intel_crtc_state *old_crtc_state,
 			I915_WRITE(PCH_DPLL_SEL, temp);
 		}
 
-		ironlake_fdi_pll_disable(intel_crtc);
+		ironlake_fdi_pll_disable(crtc);
 	}
 
 	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 	intel_set_pch_fifo_underrun_reporting(dev_priv, pipe, true);
 }
 
-static void haswell_crtc_disable(struct intel_crtc_state *old_crtc_state,
-				 struct intel_atomic_state *state)
+static void haswell_crtc_disable(struct intel_atomic_state *state,
+				 struct intel_crtc *crtc)
 {
-	struct drm_crtc *crtc = old_crtc_state->base.crtc;
-	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum transcoder cpu_transcoder = old_crtc_state->cpu_transcoder;
-
-	intel_encoders_disable(intel_crtc, old_crtc_state, state);
-
-	drm_crtc_vblank_off(crtc);
-	assert_vblank_disabled(crtc);
-
-	/* XXX: Do the pipe assertions at the right place for BXT DSI. */
-	if (!transcoder_is_dsi(cpu_transcoder))
-		intel_disable_pipe(old_crtc_state);
-
-	if (intel_crtc_has_type(old_crtc_state, INTEL_OUTPUT_DP_MST))
-		intel_ddi_set_vc_payload_alloc(old_crtc_state, false);
-
-	if (INTEL_GEN(dev_priv) >= 11)
-		icl_disable_transcoder_port_sync(old_crtc_state);
-
-	if (!transcoder_is_dsi(cpu_transcoder))
-		intel_ddi_disable_transcoder_func(old_crtc_state);
-
-	intel_dsc_disable(old_crtc_state);
-
-	if (INTEL_GEN(dev_priv) >= 9)
-		skylake_scaler_disable(intel_crtc);
-	else
-		ironlake_pfit_disable(old_crtc_state);
-
-	intel_encoders_post_disable(intel_crtc, old_crtc_state, state);
-
-	intel_encoders_post_pll_disable(intel_crtc, old_crtc_state, state);
+	/*
+	 * FIXME collapse everything to one hook.
+	 * Need care with mst->ddi interactions.
+	 */
+	intel_encoders_disable(state, crtc);
+	intel_encoders_post_disable(state, crtc);
 }
 
 static void i9xx_pfit_enable(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (!crtc_state->gmch_pfit.control)
@@ -6768,7 +6991,7 @@ static void i9xx_pfit_enable(const struct intel_crtc_state *crtc_state)
 	 * according to register description and PRM.
 	 */
 	WARN_ON(I915_READ(PFIT_CONTROL) & PFIT_ENABLE);
-	assert_pipe_disabled(dev_priv, crtc->pipe);
+	assert_pipe_disabled(dev_priv, crtc_state->cpu_transcoder);
 
 	I915_WRITE(PFIT_PGM_RATIOS, crtc_state->gmch_pfit.pgm_ratios);
 	I915_WRITE(PFIT_CONTROL, crtc_state->gmch_pfit.control);
@@ -6893,14 +7116,14 @@ intel_aux_power_domain(struct intel_digital_port *dig_port)
 
 static u64 get_crtc_power_domains(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct drm_encoder *encoder;
 	enum pipe pipe = crtc->pipe;
 	u64 mask;
 	enum transcoder transcoder = crtc_state->cpu_transcoder;
 
-	if (!crtc_state->base.active)
+	if (!crtc_state->hw.active)
 		return 0;
 
 	mask = BIT_ULL(POWER_DOMAIN_PIPE(pipe));
@@ -6910,7 +7133,7 @@ static u64 get_crtc_power_domains(struct intel_crtc_state *crtc_state)
 		mask |= BIT_ULL(POWER_DOMAIN_PIPE_PANEL_FITTER(pipe));
 
 	drm_for_each_encoder_mask(encoder, &dev_priv->drm,
-				  crtc_state->base.encoder_mask) {
+				  crtc_state->uapi.encoder_mask) {
 		struct intel_encoder *intel_encoder = to_intel_encoder(encoder);
 
 		mask |= BIT_ULL(intel_encoder->power_domain);
@@ -6928,7 +7151,7 @@ static u64 get_crtc_power_domains(struct intel_crtc_state *crtc_state)
 static u64
 modeset_get_crtc_power_domains(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum intel_display_power_domain domain;
 	u64 domains, new_domains, old_domains;
@@ -6954,146 +7177,140 @@ static void modeset_put_power_domains(struct drm_i915_private *dev_priv,
 		intel_display_power_put_unchecked(dev_priv, domain);
 }
 
-static void valleyview_crtc_enable(struct intel_crtc_state *pipe_config,
-				   struct intel_atomic_state *state)
+static void valleyview_crtc_enable(struct intel_atomic_state *state,
+				   struct intel_crtc *crtc)
 {
-	struct drm_crtc *crtc = pipe_config->base.crtc;
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
+	const struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 
-	if (WARN_ON(intel_crtc->active))
+	if (WARN_ON(crtc->active))
 		return;
 
-	if (intel_crtc_has_dp_encoder(pipe_config))
-		intel_dp_set_m_n(pipe_config, M1_N1);
+	if (intel_crtc_has_dp_encoder(new_crtc_state))
+		intel_dp_set_m_n(new_crtc_state, M1_N1);
 
-	intel_set_pipe_timings(pipe_config);
-	intel_set_pipe_src_size(pipe_config);
+	intel_set_pipe_timings(new_crtc_state);
+	intel_set_pipe_src_size(new_crtc_state);
 
 	if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) {
 		I915_WRITE(CHV_BLEND(pipe), CHV_BLEND_LEGACY);
 		I915_WRITE(CHV_CANVAS(pipe), 0);
 	}
 
-	i9xx_set_pipeconf(pipe_config);
+	i9xx_set_pipeconf(new_crtc_state);
 
-	intel_crtc->active = true;
+	crtc->active = true;
 
 	intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 
-	intel_encoders_pre_pll_enable(intel_crtc, pipe_config, state);
+	intel_encoders_pre_pll_enable(state, crtc);
 
 	if (IS_CHERRYVIEW(dev_priv)) {
-		chv_prepare_pll(intel_crtc, pipe_config);
-		chv_enable_pll(intel_crtc, pipe_config);
+		chv_prepare_pll(crtc, new_crtc_state);
+		chv_enable_pll(crtc, new_crtc_state);
 	} else {
-		vlv_prepare_pll(intel_crtc, pipe_config);
-		vlv_enable_pll(intel_crtc, pipe_config);
+		vlv_prepare_pll(crtc, new_crtc_state);
+		vlv_enable_pll(crtc, new_crtc_state);
 	}
 
-	intel_encoders_pre_enable(intel_crtc, pipe_config, state);
+	intel_encoders_pre_enable(state, crtc);
 
-	i9xx_pfit_enable(pipe_config);
+	i9xx_pfit_enable(new_crtc_state);
 
-	intel_color_load_luts(pipe_config);
-	intel_color_commit(pipe_config);
+	intel_color_load_luts(new_crtc_state);
+	intel_color_commit(new_crtc_state);
 	/* update DSPCNTR to configure gamma for pipe bottom color */
-	intel_disable_primary_plane(pipe_config);
+	intel_disable_primary_plane(new_crtc_state);
 
-	dev_priv->display.initial_watermarks(state, pipe_config);
-	intel_enable_pipe(pipe_config);
+	dev_priv->display.initial_watermarks(state, crtc);
+	intel_enable_pipe(new_crtc_state);
 
-	assert_vblank_disabled(crtc);
-	intel_crtc_vblank_on(pipe_config);
+	intel_crtc_vblank_on(new_crtc_state);
 
-	intel_encoders_enable(intel_crtc, pipe_config, state);
+	intel_encoders_enable(state, crtc);
 }
 
 static void i9xx_set_pll_dividers(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	I915_WRITE(FP0(crtc->pipe), crtc_state->dpll_hw_state.fp0);
 	I915_WRITE(FP1(crtc->pipe), crtc_state->dpll_hw_state.fp1);
 }
 
-static void i9xx_crtc_enable(struct intel_crtc_state *pipe_config,
-			     struct intel_atomic_state *state)
+static void i9xx_crtc_enable(struct intel_atomic_state *state,
+			     struct intel_crtc *crtc)
 {
-	struct drm_crtc *crtc = pipe_config->base.crtc;
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
+	const struct intel_crtc_state *new_crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 
-	if (WARN_ON(intel_crtc->active))
+	if (WARN_ON(crtc->active))
 		return;
 
-	i9xx_set_pll_dividers(pipe_config);
+	i9xx_set_pll_dividers(new_crtc_state);
 
-	if (intel_crtc_has_dp_encoder(pipe_config))
-		intel_dp_set_m_n(pipe_config, M1_N1);
+	if (intel_crtc_has_dp_encoder(new_crtc_state))
+		intel_dp_set_m_n(new_crtc_state, M1_N1);
 
-	intel_set_pipe_timings(pipe_config);
-	intel_set_pipe_src_size(pipe_config);
+	intel_set_pipe_timings(new_crtc_state);
+	intel_set_pipe_src_size(new_crtc_state);
 
-	i9xx_set_pipeconf(pipe_config);
+	i9xx_set_pipeconf(new_crtc_state);
 
-	intel_crtc->active = true;
+	crtc->active = true;
 
 	if (!IS_GEN(dev_priv, 2))
 		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, true);
 
-	intel_encoders_pre_enable(intel_crtc, pipe_config, state);
+	intel_encoders_pre_enable(state, crtc);
 
-	i9xx_enable_pll(intel_crtc, pipe_config);
+	i9xx_enable_pll(crtc, new_crtc_state);
 
-	i9xx_pfit_enable(pipe_config);
+	i9xx_pfit_enable(new_crtc_state);
 
-	intel_color_load_luts(pipe_config);
-	intel_color_commit(pipe_config);
+	intel_color_load_luts(new_crtc_state);
+	intel_color_commit(new_crtc_state);
 	/* update DSPCNTR to configure gamma for pipe bottom color */
-	intel_disable_primary_plane(pipe_config);
+	intel_disable_primary_plane(new_crtc_state);
 
-	if (dev_priv->display.initial_watermarks != NULL)
-		dev_priv->display.initial_watermarks(state,
-						     pipe_config);
+	if (dev_priv->display.initial_watermarks)
+		dev_priv->display.initial_watermarks(state, crtc);
 	else
-		intel_update_watermarks(intel_crtc);
-	intel_enable_pipe(pipe_config);
+		intel_update_watermarks(crtc);
+	intel_enable_pipe(new_crtc_state);
 
-	assert_vblank_disabled(crtc);
-	intel_crtc_vblank_on(pipe_config);
+	intel_crtc_vblank_on(new_crtc_state);
 
-	intel_encoders_enable(intel_crtc, pipe_config, state);
+	intel_encoders_enable(state, crtc);
 }
 
 static void i9xx_pfit_disable(const struct intel_crtc_state *old_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (!old_crtc_state->gmch_pfit.control)
 		return;
 
-	assert_pipe_disabled(dev_priv, crtc->pipe);
+	assert_pipe_disabled(dev_priv, old_crtc_state->cpu_transcoder);
 
 	DRM_DEBUG_KMS("disabling pfit, current: 0x%08x\n",
 		      I915_READ(PFIT_CONTROL));
 	I915_WRITE(PFIT_CONTROL, 0);
 }
 
-static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state,
-			      struct intel_atomic_state *state)
+static void i9xx_crtc_disable(struct intel_atomic_state *state,
+			      struct intel_crtc *crtc)
 {
-	struct drm_crtc *crtc = old_crtc_state->base.crtc;
-	struct drm_device *dev = crtc->dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	enum pipe pipe = intel_crtc->pipe;
+	struct intel_crtc_state *old_crtc_state =
+		intel_atomic_get_old_crtc_state(state, crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 
 	/*
 	 * On gen2 planes are double buffered but the pipe isn't, so we must
@@ -7102,16 +7319,15 @@ static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state,
 	if (IS_GEN(dev_priv, 2))
 		intel_wait_for_vblank(dev_priv, pipe);
 
-	intel_encoders_disable(intel_crtc, old_crtc_state, state);
+	intel_encoders_disable(state, crtc);
 
-	drm_crtc_vblank_off(crtc);
-	assert_vblank_disabled(crtc);
+	intel_crtc_vblank_off(old_crtc_state);
 
 	intel_disable_pipe(old_crtc_state);
 
 	i9xx_pfit_disable(old_crtc_state);
 
-	intel_encoders_post_disable(intel_crtc, old_crtc_state, state);
+	intel_encoders_post_disable(state, crtc);
 
 	if (!intel_crtc_has_type(old_crtc_state, INTEL_OUTPUT_DSI)) {
 		if (IS_CHERRYVIEW(dev_priv))
@@ -7122,92 +7338,97 @@ static void i9xx_crtc_disable(struct intel_crtc_state *old_crtc_state,
 			i9xx_disable_pll(old_crtc_state);
 	}
 
-	intel_encoders_post_pll_disable(intel_crtc, old_crtc_state, state);
+	intel_encoders_post_pll_disable(state, crtc);
 
 	if (!IS_GEN(dev_priv, 2))
 		intel_set_cpu_fifo_underrun_reporting(dev_priv, pipe, false);
 
 	if (!dev_priv->display.initial_watermarks)
-		intel_update_watermarks(intel_crtc);
+		intel_update_watermarks(crtc);
 
 	/* clock the pipe down to 640x480@60 to potentially save power */
 	if (IS_I830(dev_priv))
 		i830_enable_pipe(dev_priv, pipe);
 }
 
-static void intel_crtc_disable_noatomic(struct drm_crtc *crtc,
+static void intel_crtc_disable_noatomic(struct intel_crtc *crtc,
 					struct drm_modeset_acquire_ctx *ctx)
 {
 	struct intel_encoder *encoder;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
-	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_bw_state *bw_state =
 		to_intel_bw_state(dev_priv->bw_obj.state);
+	struct intel_crtc_state *crtc_state =
+		to_intel_crtc_state(crtc->base.state);
 	enum intel_display_power_domain domain;
 	struct intel_plane *plane;
-	u64 domains;
 	struct drm_atomic_state *state;
-	struct intel_crtc_state *crtc_state;
+	struct intel_crtc_state *temp_crtc_state;
+	enum pipe pipe = crtc->pipe;
+	u64 domains;
 	int ret;
 
-	if (!intel_crtc->active)
+	if (!crtc_state->hw.active)
 		return;
 
-	for_each_intel_plane_on_crtc(&dev_priv->drm, intel_crtc, plane) {
+	for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) {
 		const struct intel_plane_state *plane_state =
 			to_intel_plane_state(plane->base.state);
 
-		if (plane_state->base.visible)
-			intel_plane_disable_noatomic(intel_crtc, plane);
+		if (plane_state->uapi.visible)
+			intel_plane_disable_noatomic(crtc, plane);
 	}
 
-	state = drm_atomic_state_alloc(crtc->dev);
+	state = drm_atomic_state_alloc(&dev_priv->drm);
 	if (!state) {
 		DRM_DEBUG_KMS("failed to disable [CRTC:%d:%s], out of memory",
-			      crtc->base.id, crtc->name);
+			      crtc->base.base.id, crtc->base.name);
 		return;
 	}
 
 	state->acquire_ctx = ctx;
 
 	/* Everything's already locked, -EDEADLK can't happen. */
-	crtc_state = intel_atomic_get_crtc_state(state, intel_crtc);
-	ret = drm_atomic_add_affected_connectors(state, crtc);
+	temp_crtc_state = intel_atomic_get_crtc_state(state, crtc);
+	ret = drm_atomic_add_affected_connectors(state, &crtc->base);
 
-	WARN_ON(IS_ERR(crtc_state) || ret);
+	WARN_ON(IS_ERR(temp_crtc_state) || ret);
 
-	dev_priv->display.crtc_disable(crtc_state, to_intel_atomic_state(state));
+	dev_priv->display.crtc_disable(to_intel_atomic_state(state), crtc);
 
 	drm_atomic_state_put(state);
 
 	DRM_DEBUG_KMS("[CRTC:%d:%s] hw state adjusted, was enabled, now disabled\n",
-		      crtc->base.id, crtc->name);
+		      crtc->base.base.id, crtc->base.name);
+
+	crtc->active = false;
+	crtc->base.enabled = false;
 
-	WARN_ON(drm_atomic_set_mode_for_crtc(crtc->state, NULL) < 0);
-	crtc->state->active = false;
-	intel_crtc->active = false;
-	crtc->enabled = false;
-	crtc->state->connector_mask = 0;
-	crtc->state->encoder_mask = 0;
+	WARN_ON(drm_atomic_set_mode_for_crtc(&crtc_state->uapi, NULL) < 0);
+	crtc_state->uapi.active = false;
+	crtc_state->uapi.connector_mask = 0;
+	crtc_state->uapi.encoder_mask = 0;
+	intel_crtc_free_hw_state(crtc_state);
+	memset(&crtc_state->hw, 0, sizeof(crtc_state->hw));
 
-	for_each_encoder_on_crtc(crtc->dev, crtc, encoder)
+	for_each_encoder_on_crtc(&dev_priv->drm, &crtc->base, encoder)
 		encoder->base.crtc = NULL;
 
-	intel_fbc_disable(intel_crtc);
-	intel_update_watermarks(intel_crtc);
-	intel_disable_shared_dpll(to_intel_crtc_state(crtc->state));
+	intel_fbc_disable(crtc);
+	intel_update_watermarks(crtc);
+	intel_disable_shared_dpll(crtc_state);
 
-	domains = intel_crtc->enabled_power_domains;
+	domains = crtc->enabled_power_domains;
 	for_each_power_domain(domain, domains)
 		intel_display_power_put_unchecked(dev_priv, domain);
-	intel_crtc->enabled_power_domains = 0;
+	crtc->enabled_power_domains = 0;
 
-	dev_priv->active_pipes &= ~BIT(intel_crtc->pipe);
-	dev_priv->min_cdclk[intel_crtc->pipe] = 0;
-	dev_priv->min_voltage_level[intel_crtc->pipe] = 0;
+	dev_priv->active_pipes &= ~BIT(pipe);
+	dev_priv->min_cdclk[pipe] = 0;
+	dev_priv->min_voltage_level[pipe] = 0;
 
-	bw_state->data_rate[intel_crtc->pipe] = 0;
-	bw_state->num_active_planes[intel_crtc->pipe] = 0;
+	bw_state->data_rate[pipe] = 0;
+	bw_state->num_active_planes[pipe] = 0;
 }
 
 /*
@@ -7257,8 +7478,8 @@ static void intel_connector_verify_state(struct intel_crtc_state *crtc_state,
 		if (!crtc_state)
 			return;
 
-		I915_STATE_WARN(!crtc_state->base.active,
-		      "connector is active, but attached crtc isn't\n");
+		I915_STATE_WARN(!crtc_state->hw.active,
+				"connector is active, but attached crtc isn't\n");
 
 		if (!encoder || encoder->type == INTEL_OUTPUT_DP_MST)
 			return;
@@ -7269,8 +7490,8 @@ static void intel_connector_verify_state(struct intel_crtc_state *crtc_state,
 		I915_STATE_WARN(conn_state->crtc != encoder->base.crtc,
 			"attached encoder crtc differs from connector crtc\n");
 	} else {
-		I915_STATE_WARN(crtc_state && crtc_state->base.active,
-			"attached crtc is active, but connector isn't\n");
+		I915_STATE_WARN(crtc_state && crtc_state->hw.active,
+				"attached crtc is active, but connector isn't\n");
 		I915_STATE_WARN(!crtc_state && conn_state->best_encoder,
 			"best encoder set without crtc!\n");
 	}
@@ -7278,7 +7499,7 @@ static void intel_connector_verify_state(struct intel_crtc_state *crtc_state,
 
 static int pipe_required_fdi_lanes(struct intel_crtc_state *crtc_state)
 {
-	if (crtc_state->base.enable && crtc_state->has_pch_encoder)
+	if (crtc_state->hw.enable && crtc_state->has_pch_encoder)
 		return crtc_state->fdi_lanes;
 
 	return 0;
@@ -7288,7 +7509,7 @@ static int ironlake_check_fdi_lanes(struct drm_device *dev, enum pipe pipe,
 				     struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct drm_atomic_state *state = pipe_config->base.state;
+	struct drm_atomic_state *state = pipe_config->uapi.state;
 	struct intel_crtc *other_crtc;
 	struct intel_crtc_state *other_crtc_state;
 
@@ -7361,7 +7582,7 @@ static int ironlake_fdi_compute_config(struct intel_crtc *intel_crtc,
 				       struct intel_crtc_state *pipe_config)
 {
 	struct drm_device *dev = intel_crtc->base.dev;
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	int lane, link_bw, fdi_dotclock, ret;
 	bool needs_recompute = false;
 
@@ -7407,7 +7628,7 @@ retry:
 
 bool hsw_crtc_state_ips_capable(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	/* IPS only exists on ULT machines and is tied to pipe A. */
@@ -7437,9 +7658,9 @@ bool hsw_crtc_state_ips_capable(const struct intel_crtc_state *crtc_state)
 static bool hsw_compute_ips_config(struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(crtc_state->base.crtc->dev);
+		to_i915(crtc_state->uapi.crtc->dev);
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(crtc_state->base.state);
+		to_intel_atomic_state(crtc_state->uapi.state);
 
 	if (!hsw_crtc_state_ips_capable(crtc_state))
 		return false;
@@ -7478,7 +7699,7 @@ static u32 ilk_pipe_pixel_rate(const struct intel_crtc_state *pipe_config)
 {
 	u32 pixel_rate;
 
-	pixel_rate = pipe_config->base.adjusted_mode.crtc_clock;
+	pixel_rate = pipe_config->hw.adjusted_mode.crtc_clock;
 
 	/*
 	 * We only use IF-ID interlacing. If we ever use
@@ -7511,12 +7732,12 @@ static u32 ilk_pipe_pixel_rate(const struct intel_crtc_state *pipe_config)
 
 static void intel_crtc_compute_pixel_rate(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	if (HAS_GMCH(dev_priv))
 		/* FIXME calculate proper pipe pixel rate for GMCH pfit */
 		crtc_state->pixel_rate =
-			crtc_state->base.adjusted_mode.crtc_clock;
+			crtc_state->hw.adjusted_mode.crtc_clock;
 	else
 		crtc_state->pixel_rate =
 			ilk_pipe_pixel_rate(crtc_state);
@@ -7526,7 +7747,7 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc,
 				     struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	int clock_limit = dev_priv->max_dotclk_freq;
 
 	if (INTEL_GEN(dev_priv) < 4) {
@@ -7552,7 +7773,7 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc,
 
 	if ((pipe_config->output_format == INTEL_OUTPUT_FORMAT_YCBCR420 ||
 	     pipe_config->output_format == INTEL_OUTPUT_FORMAT_YCBCR444) &&
-	     pipe_config->base.ctm) {
+	     pipe_config->hw.ctm) {
 		/*
 		 * There is only one pipe CSC unit per pipe, and we need that
 		 * for output conversion from RGB->YCBCR. So if CTM is already
@@ -7746,7 +7967,7 @@ static void vlv_pllb_recal_opamp(struct drm_i915_private *dev_priv, enum pipe
 static void intel_pch_transcoder_set_m_n(const struct intel_crtc_state *crtc_state,
 					 const struct intel_link_m_n *m_n)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
@@ -7773,7 +7994,7 @@ static void intel_cpu_transcoder_set_m_n(const struct intel_crtc_state *crtc_sta
 					 const struct intel_link_m_n *m_n,
 					 const struct intel_link_m_n *m2_n2)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	enum transcoder transcoder = crtc_state->cpu_transcoder;
@@ -8082,11 +8303,11 @@ int vlv_force_pll_on(struct drm_i915_private *dev_priv, enum pipe pipe,
 	struct intel_crtc *crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
 	struct intel_crtc_state *pipe_config;
 
-	pipe_config = kzalloc(sizeof(*pipe_config), GFP_KERNEL);
+	pipe_config = intel_crtc_state_alloc(crtc);
 	if (!pipe_config)
 		return -ENOMEM;
 
-	pipe_config->base.crtc = &crtc->base;
+	pipe_config->cpu_transcoder = (enum transcoder)pipe;
 	pipe_config->pixel_multiplier = 1;
 	pipe_config->dpll = *dpll;
 
@@ -8246,11 +8467,11 @@ static void i8xx_compute_dpll(struct intel_crtc *crtc,
 
 static void intel_set_pipe_timings(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
-	const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &crtc_state->hw.adjusted_mode;
 	u32 crtc_vtotal, crtc_vblank_end;
 	int vsyncshift = 0;
 
@@ -8308,7 +8529,7 @@ static void intel_set_pipe_timings(const struct intel_crtc_state *crtc_state)
 
 static void intel_set_pipe_src_size(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 
@@ -8322,7 +8543,7 @@ static void intel_set_pipe_src_size(const struct intel_crtc_state *crtc_state)
 
 static bool intel_pipe_is_interlaced(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 
 	if (IS_GEN(dev_priv, 2))
@@ -8344,39 +8565,39 @@ static void intel_get_pipe_timings(struct intel_crtc *crtc,
 	u32 tmp;
 
 	tmp = I915_READ(HTOTAL(cpu_transcoder));
-	pipe_config->base.adjusted_mode.crtc_hdisplay = (tmp & 0xffff) + 1;
-	pipe_config->base.adjusted_mode.crtc_htotal = ((tmp >> 16) & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_hdisplay = (tmp & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_htotal = ((tmp >> 16) & 0xffff) + 1;
 
 	if (!transcoder_is_dsi(cpu_transcoder)) {
 		tmp = I915_READ(HBLANK(cpu_transcoder));
-		pipe_config->base.adjusted_mode.crtc_hblank_start =
+		pipe_config->hw.adjusted_mode.crtc_hblank_start =
 							(tmp & 0xffff) + 1;
-		pipe_config->base.adjusted_mode.crtc_hblank_end =
+		pipe_config->hw.adjusted_mode.crtc_hblank_end =
 						((tmp >> 16) & 0xffff) + 1;
 	}
 	tmp = I915_READ(HSYNC(cpu_transcoder));
-	pipe_config->base.adjusted_mode.crtc_hsync_start = (tmp & 0xffff) + 1;
-	pipe_config->base.adjusted_mode.crtc_hsync_end = ((tmp >> 16) & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_hsync_start = (tmp & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_hsync_end = ((tmp >> 16) & 0xffff) + 1;
 
 	tmp = I915_READ(VTOTAL(cpu_transcoder));
-	pipe_config->base.adjusted_mode.crtc_vdisplay = (tmp & 0xffff) + 1;
-	pipe_config->base.adjusted_mode.crtc_vtotal = ((tmp >> 16) & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_vdisplay = (tmp & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_vtotal = ((tmp >> 16) & 0xffff) + 1;
 
 	if (!transcoder_is_dsi(cpu_transcoder)) {
 		tmp = I915_READ(VBLANK(cpu_transcoder));
-		pipe_config->base.adjusted_mode.crtc_vblank_start =
+		pipe_config->hw.adjusted_mode.crtc_vblank_start =
 							(tmp & 0xffff) + 1;
-		pipe_config->base.adjusted_mode.crtc_vblank_end =
+		pipe_config->hw.adjusted_mode.crtc_vblank_end =
 						((tmp >> 16) & 0xffff) + 1;
 	}
 	tmp = I915_READ(VSYNC(cpu_transcoder));
-	pipe_config->base.adjusted_mode.crtc_vsync_start = (tmp & 0xffff) + 1;
-	pipe_config->base.adjusted_mode.crtc_vsync_end = ((tmp >> 16) & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_vsync_start = (tmp & 0xffff) + 1;
+	pipe_config->hw.adjusted_mode.crtc_vsync_end = ((tmp >> 16) & 0xffff) + 1;
 
 	if (intel_pipe_is_interlaced(pipe_config)) {
-		pipe_config->base.adjusted_mode.flags |= DRM_MODE_FLAG_INTERLACE;
-		pipe_config->base.adjusted_mode.crtc_vtotal += 1;
-		pipe_config->base.adjusted_mode.crtc_vblank_end += 1;
+		pipe_config->hw.adjusted_mode.flags |= DRM_MODE_FLAG_INTERLACE;
+		pipe_config->hw.adjusted_mode.crtc_vtotal += 1;
+		pipe_config->hw.adjusted_mode.crtc_vblank_end += 1;
 	}
 }
 
@@ -8391,27 +8612,27 @@ static void intel_get_pipe_src_size(struct intel_crtc *crtc,
 	pipe_config->pipe_src_h = (tmp & 0xffff) + 1;
 	pipe_config->pipe_src_w = ((tmp >> 16) & 0xffff) + 1;
 
-	pipe_config->base.mode.vdisplay = pipe_config->pipe_src_h;
-	pipe_config->base.mode.hdisplay = pipe_config->pipe_src_w;
+	pipe_config->hw.mode.vdisplay = pipe_config->pipe_src_h;
+	pipe_config->hw.mode.hdisplay = pipe_config->pipe_src_w;
 }
 
 void intel_mode_from_pipe_config(struct drm_display_mode *mode,
 				 struct intel_crtc_state *pipe_config)
 {
-	mode->hdisplay = pipe_config->base.adjusted_mode.crtc_hdisplay;
-	mode->htotal = pipe_config->base.adjusted_mode.crtc_htotal;
-	mode->hsync_start = pipe_config->base.adjusted_mode.crtc_hsync_start;
-	mode->hsync_end = pipe_config->base.adjusted_mode.crtc_hsync_end;
+	mode->hdisplay = pipe_config->hw.adjusted_mode.crtc_hdisplay;
+	mode->htotal = pipe_config->hw.adjusted_mode.crtc_htotal;
+	mode->hsync_start = pipe_config->hw.adjusted_mode.crtc_hsync_start;
+	mode->hsync_end = pipe_config->hw.adjusted_mode.crtc_hsync_end;
 
-	mode->vdisplay = pipe_config->base.adjusted_mode.crtc_vdisplay;
-	mode->vtotal = pipe_config->base.adjusted_mode.crtc_vtotal;
-	mode->vsync_start = pipe_config->base.adjusted_mode.crtc_vsync_start;
-	mode->vsync_end = pipe_config->base.adjusted_mode.crtc_vsync_end;
+	mode->vdisplay = pipe_config->hw.adjusted_mode.crtc_vdisplay;
+	mode->vtotal = pipe_config->hw.adjusted_mode.crtc_vtotal;
+	mode->vsync_start = pipe_config->hw.adjusted_mode.crtc_vsync_start;
+	mode->vsync_end = pipe_config->hw.adjusted_mode.crtc_vsync_end;
 
-	mode->flags = pipe_config->base.adjusted_mode.flags;
+	mode->flags = pipe_config->hw.adjusted_mode.flags;
 	mode->type = DRM_MODE_TYPE_DRIVER;
 
-	mode->clock = pipe_config->base.adjusted_mode.crtc_clock;
+	mode->clock = pipe_config->hw.adjusted_mode.crtc_clock;
 
 	mode->hsync = drm_mode_hsync(mode);
 	mode->vrefresh = drm_mode_vrefresh(mode);
@@ -8420,7 +8641,7 @@ void intel_mode_from_pipe_config(struct drm_display_mode *mode,
 
 static void i9xx_set_pipeconf(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 pipeconf;
 
@@ -8457,7 +8678,7 @@ static void i9xx_set_pipeconf(const struct intel_crtc_state *crtc_state)
 		}
 	}
 
-	if (crtc_state->base.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE) {
+	if (crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE) {
 		if (INTEL_GEN(dev_priv) < 4 ||
 		    intel_crtc_has_type(crtc_state, INTEL_OUTPUT_SDVO))
 			pipeconf |= PIPECONF_INTERLACE_W_FIELD_INDICATION;
@@ -8473,6 +8694,8 @@ static void i9xx_set_pipeconf(const struct intel_crtc_state *crtc_state)
 
 	pipeconf |= PIPECONF_GAMMA_MODE(crtc_state->gamma_mode);
 
+	pipeconf |= PIPECONF_FRAME_START_DELAY(0);
+
 	I915_WRITE(PIPECONF(crtc->pipe), pipeconf);
 	POSTING_READ(PIPECONF(crtc->pipe));
 }
@@ -8866,7 +9089,7 @@ bdw_get_pipemisc_output_format(struct intel_crtc *crtc)
 
 static void i9xx_get_pipe_color_config(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_plane *plane = to_intel_plane(crtc->base.primary);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum i9xx_plane_id i9xx_plane = plane->i9xx_plane;
@@ -8990,7 +9213,7 @@ static bool i9xx_get_pipe_config(struct intel_crtc *crtc,
 	 * but in case the pipe is enabled w/o any ports we need a sane
 	 * default.
 	 */
-	pipe_config->base.adjusted_mode.crtc_clock =
+	pipe_config->hw.adjusted_mode.crtc_clock =
 		pipe_config->port_clock / pipe_config->pixel_multiplier;
 
 	ret = true;
@@ -9506,7 +9729,7 @@ void intel_init_pch_refclk(struct drm_i915_private *dev_priv)
 
 static void ironlake_set_pipeconf(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	u32 val;
@@ -9534,7 +9757,7 @@ static void ironlake_set_pipeconf(const struct intel_crtc_state *crtc_state)
 	if (crtc_state->dither)
 		val |= (PIPECONF_DITHER_EN | PIPECONF_DITHER_TYPE_SP);
 
-	if (crtc_state->base.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
+	if (crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
 		val |= PIPECONF_INTERLACED_ILK;
 	else
 		val |= PIPECONF_PROGRESSIVE;
@@ -9554,13 +9777,15 @@ static void ironlake_set_pipeconf(const struct intel_crtc_state *crtc_state)
 
 	val |= PIPECONF_GAMMA_MODE(crtc_state->gamma_mode);
 
+	val |= PIPECONF_FRAME_START_DELAY(0);
+
 	I915_WRITE(PIPECONF(pipe), val);
 	POSTING_READ(PIPECONF(pipe));
 }
 
 static void haswell_set_pipeconf(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 val = 0;
@@ -9568,7 +9793,7 @@ static void haswell_set_pipeconf(const struct intel_crtc_state *crtc_state)
 	if (IS_HASWELL(dev_priv) && crtc_state->dither)
 		val |= (PIPECONF_DITHER_EN | PIPECONF_DITHER_TYPE_SP);
 
-	if (crtc_state->base.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
+	if (crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
 		val |= PIPECONF_INTERLACED_ILK;
 	else
 		val |= PIPECONF_PROGRESSIVE;
@@ -9583,7 +9808,7 @@ static void haswell_set_pipeconf(const struct intel_crtc_state *crtc_state)
 
 static void bdw_set_pipemisc(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 val = 0;
 
@@ -9769,7 +9994,7 @@ static int ironlake_crtc_compute_clock(struct intel_crtc *crtc,
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(crtc_state->base.state);
+		to_intel_atomic_state(crtc_state->uapi.state);
 	const struct intel_limit *limit;
 	int refclk = 120000;
 
@@ -9982,7 +10207,9 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc,
 	case PLANE_CTL_TILED_Y:
 		plane_config->tiling = I915_TILING_Y;
 		if (val & PLANE_CTL_RENDER_DECOMPRESSION_ENABLE)
-			fb->modifier = I915_FORMAT_MOD_Y_TILED_CCS;
+			fb->modifier = INTEL_GEN(dev_priv) >= 12 ?
+				I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS :
+				I915_FORMAT_MOD_Y_TILED_CCS;
 		else
 			fb->modifier = I915_FORMAT_MOD_Y_TILED;
 		break;
@@ -10195,7 +10422,7 @@ static int haswell_crtc_compute_clock(struct intel_crtc *crtc,
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(crtc_state->base.state);
+		to_intel_atomic_state(crtc_state->uapi.state);
 
 	if (!intel_crtc_has_type(crtc_state, INTEL_OUTPUT_DSI) ||
 	    INTEL_GEN(dev_priv) >= 11) {
@@ -10408,6 +10635,9 @@ static bool hsw_get_transcoder_state(struct intel_crtc *crtc,
 		case TRANS_DDI_EDP_INPUT_C_ONOFF:
 			trans_pipe = PIPE_C;
 			break;
+		case TRANS_DDI_EDP_INPUT_D_ONOFF:
+			trans_pipe = PIPE_D;
+			break;
 		}
 
 		if (trans_pipe == crtc->pipe) {
@@ -10496,16 +10726,21 @@ static void haswell_get_ddi_port_state(struct intel_crtc *crtc,
 				       struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	enum transcoder cpu_transcoder = pipe_config->cpu_transcoder;
 	struct intel_shared_dpll *pll;
 	enum port port;
 	u32 tmp;
 
-	tmp = I915_READ(TRANS_DDI_FUNC_CTL(pipe_config->cpu_transcoder));
-
-	if (INTEL_GEN(dev_priv) >= 12)
-		port = TGL_TRANS_DDI_FUNC_CTL_VAL_TO_PORT(tmp);
-	else
-		port = TRANS_DDI_FUNC_CTL_VAL_TO_PORT(tmp);
+	if (transcoder_is_dsi(cpu_transcoder)) {
+		port = (cpu_transcoder == TRANSCODER_DSI_A) ?
+						PORT_A : PORT_B;
+	} else {
+		tmp = I915_READ(TRANS_DDI_FUNC_CTL(cpu_transcoder));
+		if (INTEL_GEN(dev_priv) >= 12)
+			port = TGL_TRANS_DDI_FUNC_CTL_VAL_TO_PORT(tmp);
+		else
+			port = TRANS_DDI_FUNC_CTL_VAL_TO_PORT(tmp);
+	}
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		icelake_get_ddi_pll(dev_priv, port, pipe_config);
@@ -10561,7 +10796,7 @@ static enum transcoder transcoder_master_readout(struct drm_i915_private *dev_pr
 
 static void icelake_get_trans_port_sync_config(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	u32 transcoders;
 	enum transcoder cpu_transcoder;
 
@@ -10603,8 +10838,6 @@ static bool haswell_get_pipe_config(struct intel_crtc *crtc,
 	u64 power_domain_mask;
 	bool active;
 
-	intel_crtc_init_scalers(crtc, pipe_config);
-
 	pipe_config->master_transcoder = INVALID_TRANSCODER;
 
 	power_domain = POWER_DOMAIN_PIPE(crtc->pipe);
@@ -10730,8 +10963,8 @@ out:
 static u32 intel_cursor_base(const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	const struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	u32 base;
 
@@ -10745,8 +10978,8 @@ static u32 intel_cursor_base(const struct intel_plane_state *plane_state)
 
 static u32 intel_cursor_position(const struct intel_plane_state *plane_state)
 {
-	int x = plane_state->base.dst.x1;
-	int y = plane_state->base.dst.y1;
+	int x = plane_state->uapi.dst.x1;
+	int y = plane_state->uapi.dst.y1;
 	u32 pos = 0;
 
 	if (x < 0) {
@@ -10767,9 +11000,9 @@ static u32 intel_cursor_position(const struct intel_plane_state *plane_state)
 static bool intel_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
 	const struct drm_mode_config *config =
-		&plane_state->base.plane->dev->mode_config;
-	int width = drm_rect_width(&plane_state->base.dst);
-	int height = drm_rect_height(&plane_state->base.dst);
+		&plane_state->uapi.plane->dev->mode_config;
+	int width = drm_rect_width(&plane_state->uapi.dst);
+	int height = drm_rect_height(&plane_state->uapi.dst);
 
 	return width > 0 && width <= config->cursor_width &&
 		height > 0 && height <= config->cursor_height;
@@ -10778,8 +11011,8 @@ static bool intel_cursor_size_ok(const struct intel_plane_state *plane_state)
 static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	unsigned int rotation = plane_state->base.rotation;
+		to_i915(plane_state->uapi.plane->dev);
+	unsigned int rotation = plane_state->hw.rotation;
 	int src_x, src_y;
 	u32 offset;
 	int ret;
@@ -10788,11 +11021,11 @@ static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
-	src_x = plane_state->base.src.x1 >> 16;
-	src_y = plane_state->base.src.y1 >> 16;
+	src_x = plane_state->uapi.src.x1 >> 16;
+	src_y = plane_state->uapi.src.y1 >> 16;
 
 	intel_add_fb_offsets(&src_x, &src_y, plane_state, 0);
 	offset = intel_plane_compute_aligned_offset(&src_x, &src_y,
@@ -10807,14 +11040,14 @@ static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 	 * Put the final coordinates back so that the src
 	 * coordinate checks will see the right values.
 	 */
-	drm_rect_translate_to(&plane_state->base.src,
+	drm_rect_translate_to(&plane_state->uapi.src,
 			      src_x << 16, src_y << 16);
 
 	/* ILK+ do this automagically in hardware */
 	if (HAS_GMCH(dev_priv) && rotation & DRM_MODE_ROTATE_180) {
-		const struct drm_framebuffer *fb = plane_state->base.fb;
-		int src_w = drm_rect_width(&plane_state->base.src) >> 16;
-		int src_h = drm_rect_height(&plane_state->base.src) >> 16;
+		const struct drm_framebuffer *fb = plane_state->hw.fb;
+		int src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+		int src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
 
 		offset += (src_h * src_w - 1) * fb->format->cpp[0];
 	}
@@ -10829,7 +11062,7 @@ static int intel_cursor_check_surface(struct intel_plane_state *plane_state)
 static int intel_check_cursor(struct intel_crtc_state *crtc_state,
 			      struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int ret;
 
 	if (fb && fb->modifier != DRM_FORMAT_MOD_LINEAR) {
@@ -10837,8 +11070,8 @@ static int intel_check_cursor(struct intel_crtc_state *crtc_state,
 		return -EINVAL;
 	}
 
-	ret = drm_atomic_helper_check_plane_state(&plane_state->base,
-						  &crtc_state->base,
+	ret = drm_atomic_helper_check_plane_state(&plane_state->uapi,
+						  &crtc_state->uapi,
 						  DRM_PLANE_HELPER_NO_SCALING,
 						  DRM_PLANE_HELPER_NO_SCALING,
 						  true, true);
@@ -10846,14 +11079,14 @@ static int intel_check_cursor(struct intel_crtc_state *crtc_state,
 		return ret;
 
 	/* Use the unclipped src/dst rectangles, which we program to hw */
-	plane_state->base.src = drm_plane_state_src(&plane_state->base);
-	plane_state->base.dst = drm_plane_state_dest(&plane_state->base);
+	plane_state->uapi.src = drm_plane_state_src(&plane_state->uapi);
+	plane_state->uapi.dst = drm_plane_state_dest(&plane_state->uapi);
 
 	ret = intel_cursor_check_surface(plane_state);
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	ret = intel_plane_check_src_coordinates(plane_state);
@@ -10891,7 +11124,7 @@ static u32 i845_cursor_ctl(const struct intel_crtc_state *crtc_state,
 
 static bool i845_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
-	int width = drm_rect_width(&plane_state->base.dst);
+	int width = drm_rect_width(&plane_state->uapi.dst);
 
 	/*
 	 * 845g/865g are only limited by the width of their cursors,
@@ -10903,7 +11136,7 @@ static bool i845_cursor_size_ok(const struct intel_plane_state *plane_state)
 static int i845_check_cursor(struct intel_crtc_state *crtc_state,
 			     struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int ret;
 
 	ret = intel_check_cursor(crtc_state, plane_state);
@@ -10917,12 +11150,12 @@ static int i845_check_cursor(struct intel_crtc_state *crtc_state,
 	/* Check for which cursor types we support */
 	if (!i845_cursor_size_ok(plane_state)) {
 		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
-			  drm_rect_width(&plane_state->base.dst),
-			  drm_rect_height(&plane_state->base.dst));
+			  drm_rect_width(&plane_state->uapi.dst),
+			  drm_rect_height(&plane_state->uapi.dst));
 		return -EINVAL;
 	}
 
-	WARN_ON(plane_state->base.visible &&
+	WARN_ON(plane_state->uapi.visible &&
 		plane_state->color_plane[0].stride != fb->pitches[0]);
 
 	switch (fb->pitches[0]) {
@@ -10950,9 +11183,9 @@ static void i845_update_cursor(struct intel_plane *plane,
 	u32 cntl = 0, base = 0, pos = 0, size = 0;
 	unsigned long irqflags;
 
-	if (plane_state && plane_state->base.visible) {
-		unsigned int width = drm_rect_width(&plane_state->base.dst);
-		unsigned int height = drm_rect_height(&plane_state->base.dst);
+	if (plane_state && plane_state->uapi.visible) {
+		unsigned int width = drm_rect_width(&plane_state->uapi.dst);
+		unsigned int height = drm_rect_height(&plane_state->uapi.dst);
 
 		cntl = plane_state->ctl |
 			i845_cursor_ctl_crtc(crtc_state);
@@ -11025,7 +11258,7 @@ i9xx_cursor_max_stride(struct intel_plane *plane,
 
 static u32 i9xx_cursor_ctl_crtc(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 cntl = 0;
 
@@ -11048,13 +11281,13 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 			   const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
+		to_i915(plane_state->uapi.plane->dev);
 	u32 cntl = 0;
 
 	if (IS_GEN(dev_priv, 6) || IS_IVYBRIDGE(dev_priv))
 		cntl |= MCURSOR_TRICKLE_FEED_DISABLE;
 
-	switch (drm_rect_width(&plane_state->base.dst)) {
+	switch (drm_rect_width(&plane_state->uapi.dst)) {
 	case 64:
 		cntl |= MCURSOR_MODE_64_ARGB_AX;
 		break;
@@ -11065,11 +11298,11 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 		cntl |= MCURSOR_MODE_256_ARGB_AX;
 		break;
 	default:
-		MISSING_CASE(drm_rect_width(&plane_state->base.dst));
+		MISSING_CASE(drm_rect_width(&plane_state->uapi.dst));
 		return 0;
 	}
 
-	if (plane_state->base.rotation & DRM_MODE_ROTATE_180)
+	if (plane_state->hw.rotation & DRM_MODE_ROTATE_180)
 		cntl |= MCURSOR_ROTATE_180;
 
 	return cntl;
@@ -11078,9 +11311,9 @@ static u32 i9xx_cursor_ctl(const struct intel_crtc_state *crtc_state,
 static bool i9xx_cursor_size_ok(const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	int width = drm_rect_width(&plane_state->base.dst);
-	int height = drm_rect_height(&plane_state->base.dst);
+		to_i915(plane_state->uapi.plane->dev);
+	int width = drm_rect_width(&plane_state->uapi.dst);
+	int height = drm_rect_height(&plane_state->uapi.dst);
 
 	if (!intel_cursor_size_ok(plane_state))
 		return false;
@@ -11102,7 +11335,7 @@ static bool i9xx_cursor_size_ok(const struct intel_plane_state *plane_state)
 	 * cursors.
 	 */
 	if (HAS_CUR_FBC(dev_priv) &&
-	    plane_state->base.rotation & DRM_MODE_ROTATE_0) {
+	    plane_state->hw.rotation & DRM_MODE_ROTATE_0) {
 		if (height < 8 || height > width)
 			return false;
 	} else {
@@ -11116,9 +11349,9 @@ static bool i9xx_cursor_size_ok(const struct intel_plane_state *plane_state)
 static int i9xx_check_cursor(struct intel_crtc_state *crtc_state,
 			     struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum pipe pipe = plane->pipe;
 	int ret;
 
@@ -11133,19 +11366,19 @@ static int i9xx_check_cursor(struct intel_crtc_state *crtc_state,
 	/* Check for which cursor types we support */
 	if (!i9xx_cursor_size_ok(plane_state)) {
 		DRM_DEBUG("Cursor dimension %dx%d not supported\n",
-			  drm_rect_width(&plane_state->base.dst),
-			  drm_rect_height(&plane_state->base.dst));
+			  drm_rect_width(&plane_state->uapi.dst),
+			  drm_rect_height(&plane_state->uapi.dst));
 		return -EINVAL;
 	}
 
-	WARN_ON(plane_state->base.visible &&
+	WARN_ON(plane_state->uapi.visible &&
 		plane_state->color_plane[0].stride != fb->pitches[0]);
 
 	if (fb->pitches[0] !=
-	    drm_rect_width(&plane_state->base.dst) * fb->format->cpp[0]) {
+	    drm_rect_width(&plane_state->uapi.dst) * fb->format->cpp[0]) {
 		DRM_DEBUG_KMS("Invalid cursor stride (%u) (cursor width %d)\n",
 			      fb->pitches[0],
-			      drm_rect_width(&plane_state->base.dst));
+			      drm_rect_width(&plane_state->uapi.dst));
 		return -EINVAL;
 	}
 
@@ -11160,7 +11393,7 @@ static int i9xx_check_cursor(struct intel_crtc_state *crtc_state,
 	 * Refuse the put the cursor into that compromised position.
 	 */
 	if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_C &&
-	    plane_state->base.visible && plane_state->base.dst.x1 < 0) {
+	    plane_state->uapi.visible && plane_state->uapi.dst.x1 < 0) {
 		DRM_DEBUG_KMS("CHV cursor C not allowed to straddle the left screen edge\n");
 		return -EINVAL;
 	}
@@ -11179,9 +11412,9 @@ static void i9xx_update_cursor(struct intel_plane *plane,
 	u32 cntl = 0, base = 0, pos = 0, fbc_ctl = 0;
 	unsigned long irqflags;
 
-	if (plane_state && plane_state->base.visible) {
-		unsigned width = drm_rect_width(&plane_state->base.dst);
-		unsigned height = drm_rect_height(&plane_state->base.dst);
+	if (plane_state && plane_state->uapi.visible) {
+		unsigned width = drm_rect_width(&plane_state->uapi.dst);
+		unsigned height = drm_rect_height(&plane_state->uapi.dst);
 
 		cntl = plane_state->ctl |
 			i9xx_cursor_ctl_crtc(crtc_state);
@@ -11436,9 +11669,9 @@ found:
 		goto fail;
 	}
 
-	crtc_state->base.active = crtc_state->base.enable = true;
+	crtc_state->uapi.active = true;
 
-	ret = drm_atomic_set_mode_for_crtc(&crtc_state->base,
+	ret = drm_atomic_set_mode_for_crtc(&crtc_state->uapi,
 					   &load_detect_mode);
 	if (ret)
 		goto fail;
@@ -11647,11 +11880,37 @@ static void ironlake_pch_clock_get(struct intel_crtc *crtc,
 	 * we may need some idea for the dotclock anyway.
 	 * Calculate one based on the FDI configuration.
 	 */
-	pipe_config->base.adjusted_mode.crtc_clock =
+	pipe_config->hw.adjusted_mode.crtc_clock =
 		intel_dotclock_calculate(intel_fdi_link_freq(dev_priv, pipe_config),
 					 &pipe_config->fdi_m_n);
 }
 
+static void intel_crtc_state_reset(struct intel_crtc_state *crtc_state,
+				   struct intel_crtc *crtc)
+{
+	memset(crtc_state, 0, sizeof(*crtc_state));
+
+	__drm_atomic_helper_crtc_state_reset(&crtc_state->uapi, &crtc->base);
+
+	crtc_state->cpu_transcoder = INVALID_TRANSCODER;
+	crtc_state->master_transcoder = INVALID_TRANSCODER;
+	crtc_state->hsw_workaround_pipe = INVALID_PIPE;
+	crtc_state->output_format = INTEL_OUTPUT_FORMAT_INVALID;
+	crtc_state->scaler_state.scaler_id = -1;
+}
+
+static struct intel_crtc_state *intel_crtc_state_alloc(struct intel_crtc *crtc)
+{
+	struct intel_crtc_state *crtc_state;
+
+	crtc_state = kmalloc(sizeof(*crtc_state), GFP_KERNEL);
+
+	if (crtc_state)
+		intel_crtc_state_reset(crtc_state, crtc);
+
+	return crtc_state;
+}
+
 /* Returns the currently programmed mode of the given encoder. */
 struct drm_display_mode *
 intel_encoder_current_mode(struct intel_encoder *encoder)
@@ -11671,14 +11930,12 @@ intel_encoder_current_mode(struct intel_encoder *encoder)
 	if (!mode)
 		return NULL;
 
-	crtc_state = kzalloc(sizeof(*crtc_state), GFP_KERNEL);
+	crtc_state = intel_crtc_state_alloc(crtc);
 	if (!crtc_state) {
 		kfree(mode);
 		return NULL;
 	}
 
-	crtc_state->base.crtc = &crtc->base;
-
 	if (!dev_priv->display.get_pipe_config(crtc, crtc_state)) {
 		kfree(crtc_state);
 		kfree(mode);
@@ -11716,18 +11973,18 @@ static bool intel_wm_need_update(const struct intel_plane_state *cur,
 				 struct intel_plane_state *new)
 {
 	/* Update watermarks on tiling or size changes. */
-	if (new->base.visible != cur->base.visible)
+	if (new->uapi.visible != cur->uapi.visible)
 		return true;
 
-	if (!cur->base.fb || !new->base.fb)
+	if (!cur->hw.fb || !new->hw.fb)
 		return false;
 
-	if (cur->base.fb->modifier != new->base.fb->modifier ||
-	    cur->base.rotation != new->base.rotation ||
-	    drm_rect_width(&new->base.src) != drm_rect_width(&cur->base.src) ||
-	    drm_rect_height(&new->base.src) != drm_rect_height(&cur->base.src) ||
-	    drm_rect_width(&new->base.dst) != drm_rect_width(&cur->base.dst) ||
-	    drm_rect_height(&new->base.dst) != drm_rect_height(&cur->base.dst))
+	if (cur->hw.fb->modifier != new->hw.fb->modifier ||
+	    cur->hw.rotation != new->hw.rotation ||
+	    drm_rect_width(&new->uapi.src) != drm_rect_width(&cur->uapi.src) ||
+	    drm_rect_height(&new->uapi.src) != drm_rect_height(&cur->uapi.src) ||
+	    drm_rect_width(&new->uapi.dst) != drm_rect_width(&cur->uapi.dst) ||
+	    drm_rect_height(&new->uapi.dst) != drm_rect_height(&cur->uapi.dst))
 		return true;
 
 	return false;
@@ -11735,10 +11992,10 @@ static bool intel_wm_need_update(const struct intel_plane_state *cur,
 
 static bool needs_scaling(const struct intel_plane_state *state)
 {
-	int src_w = drm_rect_width(&state->base.src) >> 16;
-	int src_h = drm_rect_height(&state->base.src) >> 16;
-	int dst_w = drm_rect_width(&state->base.dst);
-	int dst_h = drm_rect_height(&state->base.dst);
+	int src_w = drm_rect_width(&state->uapi.src) >> 16;
+	int src_h = drm_rect_height(&state->uapi.src) >> 16;
+	int dst_w = drm_rect_width(&state->uapi.dst);
+	int dst_h = drm_rect_height(&state->uapi.dst);
 
 	return (src_w != dst_w || src_h != dst_h);
 }
@@ -11748,12 +12005,12 @@ int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_stat
 				    const struct intel_plane_state *old_plane_state,
 				    struct intel_plane_state *plane_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	bool mode_changed = needs_modeset(crtc_state);
-	bool was_crtc_enabled = old_crtc_state->base.active;
-	bool is_crtc_enabled = crtc_state->base.active;
+	bool was_crtc_enabled = old_crtc_state->hw.active;
+	bool is_crtc_enabled = crtc_state->hw.active;
 	bool turn_off, turn_on, visible, was_visible;
 	int ret;
 
@@ -11763,8 +12020,8 @@ int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_stat
 			return ret;
 	}
 
-	was_visible = old_plane_state->base.visible;
-	visible = plane_state->base.visible;
+	was_visible = old_plane_state->uapi.visible;
+	visible = plane_state->uapi.visible;
 
 	if (!was_crtc_enabled && WARN_ON(was_visible))
 		was_visible = false;
@@ -11780,7 +12037,7 @@ int intel_plane_atomic_calc_changes(const struct intel_crtc_state *old_crtc_stat
 	 * only combine the results from all planes in the current place?
 	 */
 	if (!is_crtc_enabled) {
-		plane_state->base.visible = visible = false;
+		plane_state->uapi.visible = visible = false;
 		crtc_state->active_planes &= ~BIT(plane->id);
 		crtc_state->data_rate[plane->id] = 0;
 		crtc_state->min_cdclk[plane->id] = 0;
@@ -11921,9 +12178,9 @@ static int icl_add_linked_planes(struct intel_atomic_state *state)
 
 static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	struct intel_atomic_state *state = to_intel_atomic_state(crtc_state->base.state);
+	struct intel_atomic_state *state = to_intel_atomic_state(crtc_state->uapi.state);
 	struct intel_plane *plane, *linked;
 	struct intel_plane_state *plane_state;
 	int i;
@@ -11940,7 +12197,7 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state)
 			continue;
 
 		plane_state->planar_linked_plane = NULL;
-		if (plane_state->planar_slave && !plane_state->base.visible) {
+		if (plane_state->planar_slave && !plane_state->uapi.visible) {
 			crtc_state->active_planes &= ~BIT(plane->id);
 			crtc_state->update_planes |= BIT(plane->id);
 		}
@@ -11986,6 +12243,25 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state)
 		crtc_state->active_planes |= BIT(linked->id);
 		crtc_state->update_planes |= BIT(linked->id);
 		DRM_DEBUG_KMS("Using %s as Y plane for %s\n", linked->base.name, plane->base.name);
+
+		/* Copy parameters to slave plane */
+		linked_state->ctl = plane_state->ctl | PLANE_CTL_YUV420_Y_PLANE;
+		linked_state->color_ctl = plane_state->color_ctl;
+		memcpy(linked_state->color_plane, plane_state->color_plane,
+		       sizeof(linked_state->color_plane));
+
+		intel_plane_copy_uapi_to_hw_state(linked_state, plane_state);
+		linked_state->uapi.src = plane_state->uapi.src;
+		linked_state->uapi.dst = plane_state->uapi.dst;
+
+		if (icl_is_hdr_plane(dev_priv, plane->id)) {
+			if (linked->id == PLANE_SPRITE5)
+				plane_state->cus_ctl |= PLANE_CUS_PLANE_7;
+			else if (linked->id == PLANE_SPRITE4)
+				plane_state->cus_ctl |= PLANE_CUS_PLANE_6;
+			else
+				MISSING_CASE(linked->id);
+		}
 	}
 
 	return 0;
@@ -11993,9 +12269,9 @@ static int icl_check_nv12_planes(struct intel_crtc_state *crtc_state)
 
 static bool c8_planes_changed(const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(state, crtc);
 
@@ -12004,9 +12280,9 @@ static bool c8_planes_changed(const struct intel_crtc_state *new_crtc_state)
 
 static int icl_add_sync_mode_crtcs(struct intel_crtc_state *crtc_state)
 {
-	struct drm_crtc *crtc = crtc_state->base.crtc;
-	struct intel_atomic_state *state = to_intel_atomic_state(crtc_state->base.state);
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_crtc *crtc = crtc_state->uapi.crtc;
+	struct intel_atomic_state *state = to_intel_atomic_state(crtc_state->uapi.state);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	struct drm_connector *master_connector, *connector;
 	struct drm_connector_state *connector_state;
 	struct drm_connector_list_iter conn_iter;
@@ -12030,8 +12306,8 @@ static int icl_add_sync_mode_crtcs(struct intel_crtc_state *crtc_state)
 			continue;
 		if (!connector->has_tile)
 			continue;
-		if (crtc_state->base.mode.hdisplay != connector->tile_h_size ||
-		    crtc_state->base.mode.vdisplay != connector->tile_v_size)
+		if (crtc_state->hw.mode.hdisplay != connector->tile_h_size ||
+		    crtc_state->hw.mode.vdisplay != connector->tile_v_size)
 			return 0;
 		if (connector->tile_h_loc == connector->num_h_tile - 1 &&
 		    connector->tile_v_loc == connector->num_v_tile - 1)
@@ -12080,7 +12356,7 @@ static int icl_add_sync_mode_crtcs(struct intel_crtc_state *crtc_state)
 			BIT(crtc_state->cpu_transcoder);
 		DRM_DEBUG_KMS("Master Transcoder = %s added for Slave CRTC = %d, slave transcoder bitmask = %d\n",
 			      transcoder_name(crtc_state->master_transcoder),
-			      crtc_state->base.crtc->base.id,
+			      crtc_state->uapi.crtc->base.id,
 			      master_pipe_config->sync_mode_slaves_mask);
 	}
 
@@ -12097,10 +12373,10 @@ static int intel_crtc_atomic_check(struct intel_atomic_state *state,
 	int ret;
 
 	if (INTEL_GEN(dev_priv) < 5 && !IS_G4X(dev_priv) &&
-	    mode_changed && !crtc_state->base.active)
+	    mode_changed && !crtc_state->hw.active)
 		crtc_state->update_wm_post = true;
 
-	if (mode_changed && crtc_state->base.enable &&
+	if (mode_changed && crtc_state->hw.enable &&
 	    dev_priv->display.crtc_compute_clock &&
 	    !WARN_ON(crtc_state->shared_dpll)) {
 		ret = dev_priv->display.crtc_compute_clock(crtc, crtc_state);
@@ -12113,10 +12389,10 @@ static int intel_crtc_atomic_check(struct intel_atomic_state *state,
 	 * when C8 planes are getting enabled/disabled.
 	 */
 	if (c8_planes_changed(crtc_state))
-		crtc_state->base.color_mgmt_changed = true;
+		crtc_state->uapi.color_mgmt_changed = true;
 
 	if (mode_changed || crtc_state->update_pipe ||
-	    crtc_state->base.color_mgmt_changed) {
+	    crtc_state->uapi.color_mgmt_changed) {
 		ret = intel_color_check(crtc_state);
 		if (ret)
 			return ret;
@@ -12229,7 +12505,7 @@ compute_baseline_pipe_bpp(struct intel_crtc *crtc,
 			  struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	struct drm_atomic_state *state = pipe_config->base.state;
+	struct drm_atomic_state *state = pipe_config->uapi.state;
 	struct drm_connector *connector;
 	struct drm_connector_state *connector_state;
 	int bpp, i;
@@ -12354,14 +12630,14 @@ static const char *output_formats(enum intel_output_format format)
 
 static void intel_dump_plane_state(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	struct drm_format_name_buf format_name;
 
 	if (!fb) {
 		DRM_DEBUG_KMS("[PLANE:%d:%s] fb: [NOFB], visible: %s\n",
 			      plane->base.base.id, plane->base.name,
-			      yesno(plane_state->base.visible));
+			      yesno(plane_state->uapi.visible));
 		return;
 	}
 
@@ -12369,20 +12645,20 @@ static void intel_dump_plane_state(const struct intel_plane_state *plane_state)
 		      plane->base.base.id, plane->base.name,
 		      fb->base.id, fb->width, fb->height,
 		      drm_get_format_name(fb->format->format, &format_name),
-		      yesno(plane_state->base.visible));
+		      yesno(plane_state->uapi.visible));
 	DRM_DEBUG_KMS("\trotation: 0x%x, scaler: %d\n",
-		      plane_state->base.rotation, plane_state->scaler_id);
-	if (plane_state->base.visible)
+		      plane_state->hw.rotation, plane_state->scaler_id);
+	if (plane_state->uapi.visible)
 		DRM_DEBUG_KMS("\tsrc: " DRM_RECT_FP_FMT " dst: " DRM_RECT_FMT "\n",
-			      DRM_RECT_FP_ARG(&plane_state->base.src),
-			      DRM_RECT_ARG(&plane_state->base.dst));
+			      DRM_RECT_FP_ARG(&plane_state->uapi.src),
+			      DRM_RECT_ARG(&plane_state->uapi.dst));
 }
 
 static void intel_dump_pipe_config(const struct intel_crtc_state *pipe_config,
 				   struct intel_atomic_state *state,
 				   const char *context)
 {
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	const struct intel_plane_state *plane_state;
 	struct intel_plane *plane;
@@ -12391,14 +12667,14 @@ static void intel_dump_pipe_config(const struct intel_crtc_state *pipe_config,
 
 	DRM_DEBUG_KMS("[CRTC:%d:%s] enable: %s %s\n",
 		      crtc->base.base.id, crtc->base.name,
-		      yesno(pipe_config->base.enable), context);
+		      yesno(pipe_config->hw.enable), context);
 
-	if (!pipe_config->base.enable)
+	if (!pipe_config->hw.enable)
 		goto dump_planes;
 
 	snprintf_output_types(buf, sizeof(buf), pipe_config->output_types);
 	DRM_DEBUG_KMS("active: %s, output_types: %s (0x%x), output format: %s\n",
-		      yesno(pipe_config->base.active),
+		      yesno(pipe_config->hw.active),
 		      buf, pipe_config->output_types,
 		      output_formats(pipe_config->output_format));
 
@@ -12438,10 +12714,10 @@ static void intel_dump_pipe_config(const struct intel_crtc_state *pipe_config,
 		intel_dump_infoframe(dev_priv, &pipe_config->infoframes.hdmi);
 
 	DRM_DEBUG_KMS("requested mode:\n");
-	drm_mode_debug_printmodeline(&pipe_config->base.mode);
+	drm_mode_debug_printmodeline(&pipe_config->hw.mode);
 	DRM_DEBUG_KMS("adjusted mode:\n");
-	drm_mode_debug_printmodeline(&pipe_config->base.adjusted_mode);
-	intel_dump_crtc_timings(&pipe_config->base.adjusted_mode);
+	drm_mode_debug_printmodeline(&pipe_config->hw.adjusted_mode);
+	intel_dump_crtc_timings(&pipe_config->hw.adjusted_mode);
 	DRM_DEBUG_KMS("port clock: %d, pipe src size: %dx%d, pixel rate %d\n",
 		      pipe_config->port_clock,
 		      pipe_config->pipe_src_w, pipe_config->pipe_src_h,
@@ -12561,22 +12837,59 @@ static bool check_digital_port_conflicts(struct intel_atomic_state *state)
 	return ret;
 }
 
+static void
+intel_crtc_copy_uapi_to_hw_state_nomodeset(struct intel_crtc_state *crtc_state)
+{
+	intel_crtc_copy_color_blobs(crtc_state);
+}
+
+static void
+intel_crtc_copy_uapi_to_hw_state(struct intel_crtc_state *crtc_state)
+{
+	crtc_state->hw.enable = crtc_state->uapi.enable;
+	crtc_state->hw.active = crtc_state->uapi.active;
+	crtc_state->hw.mode = crtc_state->uapi.mode;
+	crtc_state->hw.adjusted_mode = crtc_state->uapi.adjusted_mode;
+	intel_crtc_copy_uapi_to_hw_state_nomodeset(crtc_state);
+}
+
+static void intel_crtc_copy_hw_to_uapi_state(struct intel_crtc_state *crtc_state)
+{
+	crtc_state->uapi.enable = crtc_state->hw.enable;
+	crtc_state->uapi.active = crtc_state->hw.active;
+	WARN_ON(drm_atomic_set_mode_for_crtc(&crtc_state->uapi, &crtc_state->hw.mode) < 0);
+
+	crtc_state->uapi.adjusted_mode = crtc_state->hw.adjusted_mode;
+
+	/* copy color blobs to uapi */
+	drm_property_replace_blob(&crtc_state->uapi.degamma_lut,
+				  crtc_state->hw.degamma_lut);
+	drm_property_replace_blob(&crtc_state->uapi.gamma_lut,
+				  crtc_state->hw.gamma_lut);
+	drm_property_replace_blob(&crtc_state->uapi.ctm,
+				  crtc_state->hw.ctm);
+}
+
 static int
-clear_intel_crtc_state(struct intel_crtc_state *crtc_state)
+intel_crtc_prepare_cleared_state(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv =
-		to_i915(crtc_state->base.crtc->dev);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_crtc_state *saved_state;
 
-	saved_state = kzalloc(sizeof(*saved_state), GFP_KERNEL);
+	saved_state = intel_crtc_state_alloc(crtc);
 	if (!saved_state)
 		return -ENOMEM;
 
+	/* free the old crtc_state->hw members */
+	intel_crtc_free_hw_state(crtc_state);
+
 	/* FIXME: before the switch to atomic started, a new pipe_config was
 	 * kzalloc'd. Code that depends on any field being zero should be
 	 * fixed, so that the crtc_state can be safely duplicated. For now,
 	 * only fields that are know to not cause problems are preserved. */
 
+	saved_state->uapi = crtc_state->uapi;
 	saved_state->scaler_state = crtc_state->scaler_state;
 	saved_state->shared_dpll = crtc_state->shared_dpll;
 	saved_state->dpll_hw_state = crtc_state->dpll_hw_state;
@@ -12594,20 +12907,19 @@ clear_intel_crtc_state(struct intel_crtc_state *crtc_state)
 		saved_state->sync_mode_slaves_mask =
 			crtc_state->sync_mode_slaves_mask;
 
-	/* Keep base drm_crtc_state intact, only clear our extended struct */
-	BUILD_BUG_ON(offsetof(struct intel_crtc_state, base));
-	memcpy(&crtc_state->base + 1, &saved_state->base + 1,
-	       sizeof(*crtc_state) - sizeof(crtc_state->base));
-
+	memcpy(crtc_state, saved_state, sizeof(*crtc_state));
 	kfree(saved_state);
+
+	intel_crtc_copy_uapi_to_hw_state(crtc_state);
+
 	return 0;
 }
 
 static int
 intel_modeset_pipe_config(struct intel_crtc_state *pipe_config)
 {
-	struct drm_crtc *crtc = pipe_config->base.crtc;
-	struct drm_atomic_state *state = pipe_config->base.state;
+	struct drm_crtc *crtc = pipe_config->uapi.crtc;
+	struct drm_atomic_state *state = pipe_config->uapi.state;
 	struct intel_encoder *encoder;
 	struct drm_connector *connector;
 	struct drm_connector_state *connector_state;
@@ -12615,10 +12927,6 @@ intel_modeset_pipe_config(struct intel_crtc_state *pipe_config)
 	int i;
 	bool retry = true;
 
-	ret = clear_intel_crtc_state(pipe_config);
-	if (ret)
-		return ret;
-
 	pipe_config->cpu_transcoder =
 		(enum transcoder) to_intel_crtc(crtc)->pipe;
 
@@ -12627,13 +12935,13 @@ intel_modeset_pipe_config(struct intel_crtc_state *pipe_config)
 	 * positive or negative polarity is requested, treat this as meaning
 	 * negative polarity.
 	 */
-	if (!(pipe_config->base.adjusted_mode.flags &
+	if (!(pipe_config->hw.adjusted_mode.flags &
 	      (DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_NHSYNC)))
-		pipe_config->base.adjusted_mode.flags |= DRM_MODE_FLAG_NHSYNC;
+		pipe_config->hw.adjusted_mode.flags |= DRM_MODE_FLAG_NHSYNC;
 
-	if (!(pipe_config->base.adjusted_mode.flags &
+	if (!(pipe_config->hw.adjusted_mode.flags &
 	      (DRM_MODE_FLAG_PVSYNC | DRM_MODE_FLAG_NVSYNC)))
-		pipe_config->base.adjusted_mode.flags |= DRM_MODE_FLAG_NVSYNC;
+		pipe_config->hw.adjusted_mode.flags |= DRM_MODE_FLAG_NVSYNC;
 
 	ret = compute_baseline_pipe_bpp(to_intel_crtc(crtc),
 					pipe_config);
@@ -12650,7 +12958,7 @@ intel_modeset_pipe_config(struct intel_crtc_state *pipe_config)
 	 * computation to clearly distinguish it from the adjusted mode, which
 	 * can be changed by the connectors in the below retry loop.
 	 */
-	drm_mode_get_hv_timing(&pipe_config->base.mode,
+	drm_mode_get_hv_timing(&pipe_config->hw.mode,
 			       &pipe_config->pipe_src_w,
 			       &pipe_config->pipe_src_h);
 
@@ -12683,7 +12991,7 @@ encoder_retry:
 	pipe_config->pixel_multiplier = 1;
 
 	/* Fill in default crtc timings, allow encoders to overwrite them. */
-	drm_mode_set_crtcinfo(&pipe_config->base.adjusted_mode,
+	drm_mode_set_crtcinfo(&pipe_config->hw.adjusted_mode,
 			      CRTC_STEREO_DOUBLE);
 
 	/* Set the crtc_state defaults for trans_port_sync */
@@ -12717,7 +13025,7 @@ encoder_retry:
 	/* Set default port clock if not overwritten by the encoder. Needs to be
 	 * done afterwards in case the encoder adjusts the mode. */
 	if (!pipe_config->port_clock)
-		pipe_config->port_clock = pipe_config->base.adjusted_mode.crtc_clock
+		pipe_config->port_clock = pipe_config->hw.adjusted_mode.crtc_clock
 			* pipe_config->pixel_multiplier;
 
 	ret = intel_crtc_compute_config(to_intel_crtc(crtc), pipe_config);
@@ -12746,6 +13054,12 @@ encoder_retry:
 	DRM_DEBUG_KMS("hw max bpp: %i, pipe bpp: %i, dithering: %i\n",
 		      base_bpp, pipe_config->pipe_bpp, pipe_config->dither);
 
+	/*
+	 * Make drm_calc_timestamping_constants in
+	 * drm_atomic_helper_update_legacy_modeset_state() happy
+	 */
+	pipe_config->uapi.adjusted_mode = pipe_config->hw.adjusted_mode;
+
 	return 0;
 }
 
@@ -12884,13 +13198,13 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 			  const struct intel_crtc_state *pipe_config,
 			  bool fastset)
 {
-	struct drm_i915_private *dev_priv = to_i915(current_config->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(current_config->uapi.crtc->dev);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	bool ret = true;
 	u32 bp_gamma = 0;
 	bool fixup_inherited = fastset &&
-		(current_config->base.mode.private_flags & I915_MODE_FLAG_INHERITED) &&
-		!(pipe_config->base.mode.private_flags & I915_MODE_FLAG_INHERITED);
+		(current_config->hw.mode.private_flags & I915_MODE_FLAG_INHERITED) &&
+		!(pipe_config->hw.mode.private_flags & I915_MODE_FLAG_INHERITED);
 
 	if (fixup_inherited && !fastboot_enabled(dev_priv)) {
 		DRM_DEBUG_KMS("initial modeset and fastboot not set\n");
@@ -13079,19 +13393,19 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 	PIPE_CONF_CHECK_X(output_types);
 
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_hdisplay);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_htotal);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_hblank_start);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_hblank_end);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_hsync_start);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_hsync_end);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_hdisplay);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_htotal);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_hblank_start);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_hblank_end);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_hsync_start);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_hsync_end);
 
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_vdisplay);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_vtotal);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_vblank_start);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_vblank_end);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_vsync_start);
-	PIPE_CONF_CHECK_I(base.adjusted_mode.crtc_vsync_end);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_vdisplay);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_vtotal);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_vblank_start);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_vblank_end);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_vsync_start);
+	PIPE_CONF_CHECK_I(hw.adjusted_mode.crtc_vsync_end);
 
 	PIPE_CONF_CHECK_I(pixel_multiplier);
 	PIPE_CONF_CHECK_I(output_format);
@@ -13108,17 +13422,17 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 	PIPE_CONF_CHECK_BOOL_INCOMPLETE(has_audio);
 
-	PIPE_CONF_CHECK_FLAGS(base.adjusted_mode.flags,
+	PIPE_CONF_CHECK_FLAGS(hw.adjusted_mode.flags,
 			      DRM_MODE_FLAG_INTERLACE);
 
 	if (!PIPE_CONF_QUIRK(PIPE_CONFIG_QUIRK_MODE_SYNC_FLAGS)) {
-		PIPE_CONF_CHECK_FLAGS(base.adjusted_mode.flags,
+		PIPE_CONF_CHECK_FLAGS(hw.adjusted_mode.flags,
 				      DRM_MODE_FLAG_PHSYNC);
-		PIPE_CONF_CHECK_FLAGS(base.adjusted_mode.flags,
+		PIPE_CONF_CHECK_FLAGS(hw.adjusted_mode.flags,
 				      DRM_MODE_FLAG_NHSYNC);
-		PIPE_CONF_CHECK_FLAGS(base.adjusted_mode.flags,
+		PIPE_CONF_CHECK_FLAGS(hw.adjusted_mode.flags,
 				      DRM_MODE_FLAG_PVSYNC);
-		PIPE_CONF_CHECK_FLAGS(base.adjusted_mode.flags,
+		PIPE_CONF_CHECK_FLAGS(hw.adjusted_mode.flags,
 				      DRM_MODE_FLAG_NVSYNC);
 	}
 
@@ -13157,7 +13471,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 
 		bp_gamma = intel_color_get_gamma_bit_precision(pipe_config);
 		if (bp_gamma)
-			PIPE_CONF_CHECK_COLOR_LUT(gamma_mode, base.gamma_lut, bp_gamma);
+			PIPE_CONF_CHECK_COLOR_LUT(gamma_mode, hw.gamma_lut, bp_gamma);
 
 	}
 
@@ -13202,7 +13516,7 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	if (IS_G4X(dev_priv) || INTEL_GEN(dev_priv) >= 5)
 		PIPE_CONF_CHECK_I(pipe_bpp);
 
-	PIPE_CONF_CHECK_CLOCK_FUZZY(base.adjusted_mode.crtc_clock);
+	PIPE_CONF_CHECK_CLOCK_FUZZY(hw.adjusted_mode.crtc_clock);
 	PIPE_CONF_CHECK_CLOCK_FUZZY(port_clock);
 
 	PIPE_CONF_CHECK_I(min_voltage_level);
@@ -13217,6 +13531,10 @@ intel_pipe_config_compare(const struct intel_crtc_state *current_config,
 	PIPE_CONF_CHECK_I(sync_mode_slaves_mask);
 	PIPE_CONF_CHECK_I(master_transcoder);
 
+	PIPE_CONF_CHECK_I(dsc.compression_enable);
+	PIPE_CONF_CHECK_I(dsc.dsc_split);
+	PIPE_CONF_CHECK_I(dsc.compressed_bpp);
+
 #undef PIPE_CONF_CHECK_X
 #undef PIPE_CONF_CHECK_I
 #undef PIPE_CONF_CHECK_BOOL
@@ -13236,7 +13554,7 @@ static void intel_pipe_config_sanity_check(struct drm_i915_private *dev_priv,
 	if (pipe_config->has_pch_encoder) {
 		int fdi_dotclock = intel_dotclock_calculate(intel_fdi_link_freq(dev_priv, pipe_config),
 							    &pipe_config->fdi_m_n);
-		int dotclock = pipe_config->base.adjusted_mode.crtc_clock;
+		int dotclock = pipe_config->hw.adjusted_mode.crtc_clock;
 
 		/*
 		 * FDI already provided one idea for the dotclock.
@@ -13264,7 +13582,7 @@ static void verify_wm_state(struct intel_crtc *crtc,
 	const enum pipe pipe = crtc->pipe;
 	int plane, level, max_level = ilk_wm_max_level(dev_priv);
 
-	if (INTEL_GEN(dev_priv) < 9 || !new_crtc_state->base.active)
+	if (INTEL_GEN(dev_priv) < 9 || !new_crtc_state->hw.active)
 		return;
 
 	hw = kzalloc(sizeof(*hw), GFP_KERNEL);
@@ -13469,16 +13787,14 @@ verify_crtc_state(struct intel_crtc *crtc,
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_encoder *encoder;
-	struct intel_crtc_state *pipe_config;
-	struct drm_atomic_state *state;
+	struct intel_crtc_state *pipe_config = old_crtc_state;
+	struct drm_atomic_state *state = old_crtc_state->uapi.state;
 	bool active;
 
-	state = old_crtc_state->base.state;
-	__drm_atomic_helper_crtc_destroy_state(&old_crtc_state->base);
-	pipe_config = old_crtc_state;
-	memset(pipe_config, 0, sizeof(*pipe_config));
-	pipe_config->base.crtc = &crtc->base;
-	pipe_config->base.state = state;
+	__drm_atomic_helper_crtc_destroy_state(&old_crtc_state->uapi);
+	intel_crtc_free_hw_state(old_crtc_state);
+	intel_crtc_state_reset(old_crtc_state, crtc);
+	old_crtc_state->uapi.state = state;
 
 	DRM_DEBUG_KMS("[CRTC:%d:%s]\n", crtc->base.base.id, crtc->base.name);
 
@@ -13486,23 +13802,26 @@ verify_crtc_state(struct intel_crtc *crtc,
 
 	/* we keep both pipes enabled on 830 */
 	if (IS_I830(dev_priv))
-		active = new_crtc_state->base.active;
+		active = new_crtc_state->hw.active;
 
-	I915_STATE_WARN(new_crtc_state->base.active != active,
-	     "crtc active state doesn't match with hw state "
-	     "(expected %i, found %i)\n", new_crtc_state->base.active, active);
+	I915_STATE_WARN(new_crtc_state->hw.active != active,
+			"crtc active state doesn't match with hw state "
+			"(expected %i, found %i)\n",
+			new_crtc_state->hw.active, active);
 
-	I915_STATE_WARN(crtc->active != new_crtc_state->base.active,
-	     "transitional active state does not match atomic hw state "
-	     "(expected %i, found %i)\n", new_crtc_state->base.active, crtc->active);
+	I915_STATE_WARN(crtc->active != new_crtc_state->hw.active,
+			"transitional active state does not match atomic hw state "
+			"(expected %i, found %i)\n",
+			new_crtc_state->hw.active, crtc->active);
 
 	for_each_encoder_on_crtc(dev, &crtc->base, encoder) {
 		enum pipe pipe;
 
 		active = encoder->get_hw_state(encoder, &pipe);
-		I915_STATE_WARN(active != new_crtc_state->base.active,
-			"[ENCODER:%i] active %i with crtc active %i\n",
-			encoder->base.base.id, active, new_crtc_state->base.active);
+		I915_STATE_WARN(active != new_crtc_state->hw.active,
+				"[ENCODER:%i] active %i with crtc active %i\n",
+				encoder->base.base.id, active,
+				new_crtc_state->hw.active);
 
 		I915_STATE_WARN(active && crtc->pipe != pipe,
 				"Encoder connected to wrong pipe %c\n",
@@ -13514,7 +13833,7 @@ verify_crtc_state(struct intel_crtc *crtc,
 
 	intel_crtc_compute_pixel_rate(pipe_config);
 
-	if (!new_crtc_state->base.active)
+	if (!new_crtc_state->hw.active)
 		return;
 
 	intel_pipe_config_sanity_check(dev_priv, pipe_config);
@@ -13537,7 +13856,7 @@ intel_verify_planes(struct intel_atomic_state *state)
 	for_each_new_intel_plane_in_state(state, plane,
 					  plane_state, i)
 		assert_plane(plane, plane_state->planar_slave ||
-			     plane_state->base.visible);
+			     plane_state->uapi.visible);
 }
 
 static void
@@ -13576,7 +13895,7 @@ verify_single_dpll_state(struct drm_i915_private *dev_priv,
 
 	crtc_mask = drm_crtc_mask(&crtc->base);
 
-	if (new_crtc_state->base.active)
+	if (new_crtc_state->hw.active)
 		I915_STATE_WARN(!(pll->active_mask & crtc_mask),
 				"pll active mismatch (expected pipe %c in active mask 0x%02x)\n",
 				pipe_name(drm_crtc_index(&crtc->base)), pll->active_mask);
@@ -13655,10 +13974,10 @@ intel_modeset_verify_disabled(struct drm_i915_private *dev_priv,
 static void
 intel_crtc_update_active_timings(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 
 	drm_calc_timestamping_constants(&crtc->base, adjusted_mode);
 
@@ -13740,7 +14059,7 @@ static int haswell_mode_set_planes_workaround(struct intel_atomic_state *state)
 
 	/* look at all crtc's that are going to be enabled in during modeset */
 	for_each_new_intel_crtc_in_state(state, crtc, crtc_state, i) {
-		if (!crtc_state->base.active ||
+		if (!crtc_state->hw.active ||
 		    !needs_modeset(crtc_state))
 			continue;
 
@@ -13765,7 +14084,7 @@ static int haswell_mode_set_planes_workaround(struct intel_atomic_state *state)
 
 		crtc_state->hsw_workaround_pipe = INVALID_PIPE;
 
-		if (!crtc_state->base.active ||
+		if (!crtc_state->hw.active ||
 		    needs_modeset(crtc_state))
 			continue;
 
@@ -13802,12 +14121,12 @@ static int intel_modeset_checks(struct intel_atomic_state *state)
 
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
 					    new_crtc_state, i) {
-		if (new_crtc_state->base.active)
+		if (new_crtc_state->hw.active)
 			state->active_pipes |= BIT(crtc->pipe);
 		else
 			state->active_pipes &= ~BIT(crtc->pipe);
 
-		if (old_crtc_state->base.active != new_crtc_state->base.active)
+		if (old_crtc_state->hw.active != new_crtc_state->hw.active)
 			state->active_pipe_changes |= BIT(crtc->pipe);
 	}
 
@@ -13852,7 +14171,7 @@ static void intel_crtc_check_fastset(const struct intel_crtc_state *old_crtc_sta
 	if (!intel_pipe_config_compare(old_crtc_state, new_crtc_state, true))
 		return;
 
-	new_crtc_state->base.mode_changed = false;
+	new_crtc_state->uapi.mode_changed = false;
 	new_crtc_state->update_pipe = true;
 
 	/*
@@ -13995,9 +14314,9 @@ static int intel_atomic_check(struct drm_device *dev,
 	/* Catch I915_MODE_FLAG_INHERITED */
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
 					    new_crtc_state, i) {
-		if (new_crtc_state->base.mode.private_flags !=
-		    old_crtc_state->base.mode.private_flags)
-			new_crtc_state->base.mode_changed = true;
+		if (new_crtc_state->hw.mode.private_flags !=
+		    old_crtc_state->hw.mode.private_flags)
+			new_crtc_state->uapi.mode_changed = true;
 	}
 
 	ret = drm_atomic_helper_check_modeset(dev, &state->base);
@@ -14006,14 +14325,24 @@ static int intel_atomic_check(struct drm_device *dev,
 
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
 					    new_crtc_state, i) {
-		if (!needs_modeset(new_crtc_state))
+		if (!needs_modeset(new_crtc_state)) {
+			/* Light copy */
+			intel_crtc_copy_uapi_to_hw_state_nomodeset(new_crtc_state);
+
 			continue;
+		}
+
+		if (!new_crtc_state->uapi.enable) {
+			intel_crtc_copy_uapi_to_hw_state(new_crtc_state);
 
-		if (!new_crtc_state->base.enable) {
 			any_ms = true;
 			continue;
 		}
 
+		ret = intel_crtc_prepare_cleared_state(new_crtc_state);
+		if (ret)
+			goto fail;
+
 		ret = intel_modeset_pipe_config(new_crtc_state);
 		if (ret)
 			goto fail;
@@ -14111,7 +14440,7 @@ void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc,
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
-	if (!IS_GEN(dev_priv, 2))
+	if (!IS_GEN(dev_priv, 2) || crtc_state->active_planes)
 		intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, true);
 
 	if (crtc_state->has_pch_encoder) {
@@ -14125,7 +14454,7 @@ void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc,
 static void intel_pipe_fastset(const struct intel_crtc_state *old_crtc_state,
 			       const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	/*
@@ -14159,6 +14488,7 @@ static void commit_pipe_config(struct intel_atomic_state *state,
 			       struct intel_crtc_state *old_crtc_state,
 			       struct intel_crtc_state *new_crtc_state)
 {
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	bool modeset = needs_modeset(new_crtc_state);
 
@@ -14167,7 +14497,7 @@ static void commit_pipe_config(struct intel_atomic_state *state,
 	 * CRTC was enabled.
 	 */
 	if (!modeset) {
-		if (new_crtc_state->base.color_mgmt_changed ||
+		if (new_crtc_state->uapi.color_mgmt_changed ||
 		    new_crtc_state->update_pipe)
 			intel_color_commit(new_crtc_state);
 
@@ -14182,8 +14512,7 @@ static void commit_pipe_config(struct intel_atomic_state *state,
 	}
 
 	if (dev_priv->display.atomic_update_watermarks)
-		dev_priv->display.atomic_update_watermarks(state,
-							   new_crtc_state);
+		dev_priv->display.atomic_update_watermarks(state, crtc);
 }
 
 static void intel_update_crtc(struct intel_crtc *crtc,
@@ -14200,20 +14529,20 @@ static void intel_update_crtc(struct intel_crtc *crtc,
 	if (modeset) {
 		intel_crtc_update_active_timings(new_crtc_state);
 
-		dev_priv->display.crtc_enable(new_crtc_state, state);
+		dev_priv->display.crtc_enable(state, crtc);
 
 		/* vblanks work again, re-enable pipe CRC. */
 		intel_crtc_enable_pipe_crc(crtc);
 	} else {
 		if (new_crtc_state->preload_luts &&
-		    (new_crtc_state->base.color_mgmt_changed ||
+		    (new_crtc_state->uapi.color_mgmt_changed ||
 		     new_crtc_state->update_pipe))
 			intel_color_load_luts(new_crtc_state);
 
-		intel_pre_plane_update(old_crtc_state, new_crtc_state);
+		intel_pre_plane_update(state, crtc);
 
 		if (new_crtc_state->update_pipe)
-			intel_encoders_update_pipe(crtc, new_crtc_state, state);
+			intel_encoders_update_pipe(state, crtc);
 	}
 
 	if (new_crtc_state->update_pipe && !new_crtc_state->enable_fbc)
@@ -14240,13 +14569,13 @@ static void intel_update_crtc(struct intel_crtc *crtc,
 	 * of enabling them on the CRTC's first fastset.
 	 */
 	if (new_crtc_state->update_pipe && !modeset &&
-	    old_crtc_state->base.mode.private_flags & I915_MODE_FLAG_INHERITED)
+	    old_crtc_state->hw.mode.private_flags & I915_MODE_FLAG_INHERITED)
 		intel_crtc_arm_fifo_underrun(crtc, new_crtc_state);
 }
 
 static struct intel_crtc *intel_get_slave_crtc(const struct intel_crtc_state *new_crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(new_crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(new_crtc_state->uapi.crtc->dev);
 	enum transcoder slave_transcoder;
 
 	WARN_ON(!is_power_of_2(new_crtc_state->sync_mode_slaves_mask));
@@ -14271,97 +14600,59 @@ static void intel_old_crtc_state_disables(struct intel_atomic_state *state,
 	 */
 	intel_crtc_disable_pipe_crc(crtc);
 
-	dev_priv->display.crtc_disable(old_crtc_state, state);
+	dev_priv->display.crtc_disable(state, crtc);
 	crtc->active = false;
 	intel_fbc_disable(crtc);
 	intel_disable_shared_dpll(old_crtc_state);
 
-	/*
-	 * Underruns don't always raise interrupts,
-	 * so check manually.
-	 */
-	intel_check_cpu_fifo_underruns(dev_priv);
-	intel_check_pch_fifo_underruns(dev_priv);
-
 	/* FIXME unify this for all platforms */
-	if (!new_crtc_state->base.active &&
+	if (!new_crtc_state->hw.active &&
 	    !HAS_GMCH(dev_priv) &&
 	    dev_priv->display.initial_watermarks)
-		dev_priv->display.initial_watermarks(state,
-						     new_crtc_state);
-}
-
-static void intel_trans_port_sync_modeset_disables(struct intel_atomic_state *state,
-						   struct intel_crtc *crtc,
-						   struct intel_crtc_state *old_crtc_state,
-						   struct intel_crtc_state *new_crtc_state)
-{
-	struct intel_crtc *slave_crtc = intel_get_slave_crtc(new_crtc_state);
-	struct intel_crtc_state *new_slave_crtc_state =
-		intel_atomic_get_new_crtc_state(state, slave_crtc);
-	struct intel_crtc_state *old_slave_crtc_state =
-		intel_atomic_get_old_crtc_state(state, slave_crtc);
-
-	WARN_ON(!slave_crtc || !new_slave_crtc_state ||
-		!old_slave_crtc_state);
-
-	/* Disable Slave first */
-	intel_pre_plane_update(old_slave_crtc_state, new_slave_crtc_state);
-	if (old_slave_crtc_state->base.active)
-		intel_old_crtc_state_disables(state,
-					      old_slave_crtc_state,
-					      new_slave_crtc_state,
-					      slave_crtc);
-
-	/* Disable Master */
-	intel_pre_plane_update(old_crtc_state, new_crtc_state);
-	if (old_crtc_state->base.active)
-		intel_old_crtc_state_disables(state,
-					      old_crtc_state,
-					      new_crtc_state,
-					      crtc);
+		dev_priv->display.initial_watermarks(state, crtc);
 }
 
 static void intel_commit_modeset_disables(struct intel_atomic_state *state)
 {
 	struct intel_crtc_state *new_crtc_state, *old_crtc_state;
 	struct intel_crtc *crtc;
+	u32 handled = 0;
 	int i;
 
-	/*
-	 * Disable CRTC/pipes in reverse order because some features(MST in
-	 * TGL+) requires master and slave relationship between pipes, so it
-	 * should always pick the lowest pipe as master as it will be enabled
-	 * first and disable in the reverse order so the master will be the
-	 * last one to be disabled.
-	 */
-	for_each_oldnew_intel_crtc_in_state_reverse(state, crtc, old_crtc_state,
-						    new_crtc_state, i) {
+	/* Only disable port sync slaves */
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+					    new_crtc_state, i) {
 		if (!needs_modeset(new_crtc_state))
 			continue;
 
+		if (!old_crtc_state->hw.active)
+			continue;
+
 		/* In case of Transcoder port Sync master slave CRTCs can be
 		 * assigned in any order and we need to make sure that
 		 * slave CRTCs are disabled first and then master CRTC since
 		 * Slave vblanks are masked till Master Vblanks.
 		 */
-		if (is_trans_port_sync_mode(new_crtc_state)) {
-			if (is_trans_port_sync_master(new_crtc_state))
-				intel_trans_port_sync_modeset_disables(state,
-								       crtc,
-								       old_crtc_state,
-								       new_crtc_state);
-			else
-				continue;
-		} else {
-			intel_pre_plane_update(old_crtc_state, new_crtc_state);
+		if (!is_trans_port_sync_slave(old_crtc_state))
+			continue;
 
-			if (old_crtc_state->base.active)
-				intel_old_crtc_state_disables(state,
-							      old_crtc_state,
-							      new_crtc_state,
-							      crtc);
-		}
+		intel_pre_plane_update(state, crtc);
+		intel_old_crtc_state_disables(state, old_crtc_state,
+					      new_crtc_state, crtc);
+		handled |= BIT(crtc->pipe);
+	}
+
+	/* Disable everything else left on */
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+					    new_crtc_state, i) {
+		if (!needs_modeset(new_crtc_state) ||
+		    (handled & BIT(crtc->pipe)))
+			continue;
+
+		intel_pre_plane_update(state, crtc);
+		if (old_crtc_state->hw.active)
+			intel_old_crtc_state_disables(state, old_crtc_state,
+						      new_crtc_state, crtc);
 	}
 }
 
@@ -14372,7 +14663,7 @@ static void intel_commit_modeset_enables(struct intel_atomic_state *state)
 	int i;
 
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		if (!new_crtc_state->base.active)
+		if (!new_crtc_state->hw.active)
 			continue;
 
 		intel_update_crtc(crtc, state, old_crtc_state,
@@ -14387,7 +14678,7 @@ static void intel_crtc_enable_trans_port_sync(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 
 	intel_crtc_update_active_timings(new_crtc_state);
-	dev_priv->display.crtc_enable(new_crtc_state, state);
+	dev_priv->display.crtc_enable(state, crtc);
 	intel_crtc_enable_pipe_crc(crtc);
 }
 
@@ -14437,7 +14728,7 @@ static void intel_post_crtc_enable_updates(struct intel_crtc *crtc,
 	 * of enabling them on the CRTC's first fastset.
 	 */
 	if (new_crtc_state->update_pipe && !modeset &&
-	    old_crtc_state->base.mode.private_flags & I915_MODE_FLAG_INHERITED)
+	    old_crtc_state->hw.mode.private_flags & I915_MODE_FLAG_INHERITED)
 		intel_crtc_arm_fifo_underrun(crtc, new_crtc_state);
 }
 
@@ -14492,17 +14783,19 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
 	struct intel_crtc *crtc;
 	struct intel_crtc_state *old_crtc_state, *new_crtc_state;
-	unsigned int updated = 0;
-	bool progress;
-	int i;
 	u8 hw_enabled_slices = dev_priv->wm.skl_hw.ddb.enabled_slices;
 	u8 required_slices = state->wm_results.ddb.enabled_slices;
 	struct skl_ddb_entry entries[I915_MAX_PIPES] = {};
+	u8 dirty_pipes = 0;
+	int i;
 
-	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i)
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
 		/* ignore allocations for crtc's that have been turned off. */
-		if (new_crtc_state->base.active)
+		if (!needs_modeset(new_crtc_state) && new_crtc_state->hw.active)
 			entries[i] = old_crtc_state->wm.skl.ddb;
+		if (new_crtc_state->hw.active)
+			dirty_pipes |= BIT(crtc->pipe);
+	}
 
 	/* If 2nd DBuf slice required, enable it here */
 	if (INTEL_GEN(dev_priv) >= 11 && required_slices > hw_enabled_slices)
@@ -14514,15 +14807,13 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 	 * never overlap with eachother inbetween CRTC updates. Otherwise we'll
 	 * cause pipe underruns and other bad stuff.
 	 */
-	do {
-		progress = false;
-
-		for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+	while (dirty_pipes) {
+		for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+						    new_crtc_state, i) {
 			enum pipe pipe = crtc->pipe;
-			bool vbl_wait = false;
 			bool modeset = needs_modeset(new_crtc_state);
 
-			if (updated & BIT(crtc->pipe) || !new_crtc_state->base.active)
+			if ((dirty_pipes & BIT(pipe)) == 0)
 				continue;
 
 			if (skl_ddb_allocation_overlaps(&new_crtc_state->wm.skl.ddb,
@@ -14530,20 +14821,8 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 							INTEL_NUM_PIPES(dev_priv), i))
 				continue;
 
-			updated |= BIT(pipe);
 			entries[i] = new_crtc_state->wm.skl.ddb;
-
-			/*
-			 * If this is an already active pipe, it's DDB changed,
-			 * and this isn't the last pipe that needs updating
-			 * then we need to wait for a vblank to pass for the
-			 * new ddb allocation to take effect.
-			 */
-			if (!skl_ddb_entry_equal(&new_crtc_state->wm.skl.ddb,
-						 &old_crtc_state->wm.skl.ddb) &&
-			    !modeset &&
-			    state->wm_results.dirty_pipes != updated)
-				vbl_wait = true;
+			dirty_pipes &= ~BIT(pipe);
 
 			if (modeset && is_trans_port_sync_mode(new_crtc_state)) {
 				if (is_trans_port_sync_master(new_crtc_state))
@@ -14558,12 +14837,18 @@ static void skl_commit_modeset_enables(struct intel_atomic_state *state)
 						  new_crtc_state);
 			}
 
-			if (vbl_wait)
+			/*
+			 * If this is an already active pipe, it's DDB changed,
+			 * and this isn't the last pipe that needs updating
+			 * then we need to wait for a vblank to pass for the
+			 * new ddb allocation to take effect.
+			 */
+			if (!skl_ddb_entry_equal(&new_crtc_state->wm.skl.ddb,
+						 &old_crtc_state->wm.skl.ddb) &&
+			    !modeset && dirty_pipes)
 				intel_wait_for_vblank(dev_priv, pipe);
-
-			progress = true;
 		}
-	} while (progress);
+	}
 
 	/* If 2nd DBuf slice is no more required disable it */
 	if (INTEL_GEN(dev_priv) >= 11 && required_slices < hw_enabled_slices)
@@ -14684,12 +14969,13 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 		bool modeset = needs_modeset(new_crtc_state);
 
 		/* Complete events for now disable pipes here. */
-		if (modeset && !new_crtc_state->base.active && new_crtc_state->base.event) {
+		if (modeset && !new_crtc_state->hw.active && new_crtc_state->uapi.event) {
 			spin_lock_irq(&dev->event_lock);
-			drm_crtc_send_vblank_event(&crtc->base, new_crtc_state->base.event);
+			drm_crtc_send_vblank_event(&crtc->base,
+						   new_crtc_state->uapi.event);
 			spin_unlock_irq(&dev->event_lock);
 
-			new_crtc_state->base.event = NULL;
+			new_crtc_state->uapi.event = NULL;
 		}
 	}
 
@@ -14720,10 +15006,10 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 	drm_atomic_helper_wait_for_flip_done(dev, &state->base);
 
 	for_each_new_intel_crtc_in_state(state, crtc, new_crtc_state, i) {
-		if (new_crtc_state->base.active &&
+		if (new_crtc_state->hw.active &&
 		    !needs_modeset(new_crtc_state) &&
 		    !new_crtc_state->preload_luts &&
-		    (new_crtc_state->base.color_mgmt_changed ||
+		    (new_crtc_state->uapi.color_mgmt_changed ||
 		     new_crtc_state->update_pipe))
 			intel_color_load_luts(new_crtc_state);
 	}
@@ -14735,14 +15021,25 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 	 *
 	 * TODO: Move this (and other cleanup) to an async worker eventually.
 	 */
-	for_each_new_intel_crtc_in_state(state, crtc, new_crtc_state, i) {
+	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state,
+					    new_crtc_state, i) {
+		/*
+		 * Gen2 reports pipe underruns whenever all planes are disabled.
+		 * So re-enable underrun reporting after some planes get enabled.
+		 *
+		 * We do this before .optimize_watermarks() so that we have a
+		 * chance of catching underruns with the intermediate watermarks
+		 * vs. the new plane configuration.
+		 */
+		if (IS_GEN(dev_priv, 2) && planes_enabling(old_crtc_state, new_crtc_state))
+			intel_set_cpu_fifo_underrun_reporting(dev_priv, crtc->pipe, true);
+
 		if (dev_priv->display.optimize_watermarks)
-			dev_priv->display.optimize_watermarks(state,
-							      new_crtc_state);
+			dev_priv->display.optimize_watermarks(state, crtc);
 	}
 
 	for_each_oldnew_intel_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		intel_post_plane_update(old_crtc_state);
+		intel_post_plane_update(state, crtc);
 
 		if (put_domains[i])
 			modeset_put_power_domains(dev_priv, put_domains[i]);
@@ -14750,6 +15047,10 @@ static void intel_atomic_commit_tail(struct intel_atomic_state *state)
 		intel_modeset_verify_crtc(crtc, state, old_crtc_state, new_crtc_state);
 	}
 
+	/* Underruns don't always raise interrupts, so check manually */
+	intel_check_cpu_fifo_underruns(dev_priv);
+	intel_check_pch_fifo_underruns(dev_priv);
+
 	if (state->modeset)
 		intel_verify_planes(state);
 
@@ -14823,8 +15124,8 @@ static void intel_atomic_track_fbs(struct intel_atomic_state *state)
 
 	for_each_oldnew_intel_plane_in_state(state, plane, old_plane_state,
 					     new_plane_state, i)
-		intel_frontbuffer_track(to_intel_frontbuffer(old_plane_state->base.fb),
-					to_intel_frontbuffer(new_plane_state->base.fb),
+		intel_frontbuffer_track(to_intel_frontbuffer(old_plane_state->hw.fb),
+					to_intel_frontbuffer(new_plane_state->hw.fb),
 					plane->frontbuffer_bit);
 }
 
@@ -14991,9 +15292,9 @@ static void add_rps_boost_after_vblank(struct drm_crtc *crtc,
 
 static int intel_plane_pin_fb(struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	struct drm_framebuffer *fb = plane_state->base.fb;
+	struct drm_framebuffer *fb = plane_state->hw.fb;
 	struct i915_vma *vma;
 
 	if (plane->id == PLANE_CURSOR &&
@@ -15056,9 +15357,9 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 	struct intel_plane_state *new_plane_state =
 		to_intel_plane_state(_new_plane_state);
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(new_plane_state->base.state);
+		to_intel_atomic_state(new_plane_state->uapi.state);
 	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-	struct drm_framebuffer *fb = new_plane_state->base.fb;
+	struct drm_framebuffer *fb = new_plane_state->hw.fb;
 	struct drm_i915_gem_object *obj = intel_fb_obj(fb);
 	struct drm_i915_gem_object *old_obj = intel_fb_obj(plane->state->fb);
 	int ret;
@@ -15089,9 +15390,9 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 		}
 	}
 
-	if (new_plane_state->base.fence) { /* explicit fencing */
+	if (new_plane_state->uapi.fence) { /* explicit fencing */
 		ret = i915_sw_fence_await_dma_fence(&intel_state->commit_ready,
-						    new_plane_state->base.fence,
+						    new_plane_state->uapi.fence,
 						    I915_FENCE_TIMEOUT,
 						    GFP_KERNEL);
 		if (ret < 0)
@@ -15112,9 +15413,9 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 		return ret;
 
 	fb_obj_bump_render_priority(obj);
-	intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_DIRTYFB);
+	i915_gem_object_flush_frontbuffer(obj, ORIGIN_DIRTYFB);
 
-	if (!new_plane_state->base.fence) { /* implicit fencing */
+	if (!new_plane_state->uapi.fence) { /* implicit fencing */
 		struct dma_fence *fence;
 
 		ret = i915_sw_fence_await_reservation(&intel_state->commit_ready,
@@ -15126,13 +15427,13 @@ intel_prepare_plane_fb(struct drm_plane *plane,
 
 		fence = dma_resv_get_excl_rcu(obj->base.resv);
 		if (fence) {
-			add_rps_boost_after_vblank(new_plane_state->base.crtc,
+			add_rps_boost_after_vblank(new_plane_state->hw.crtc,
 						   fence);
 			dma_fence_put(fence);
 		}
 	} else {
-		add_rps_boost_after_vblank(new_plane_state->base.crtc,
-					   new_plane_state->base.fence);
+		add_rps_boost_after_vblank(new_plane_state->hw.crtc,
+					   new_plane_state->uapi.fence);
 	}
 
 	/*
@@ -15165,7 +15466,7 @@ intel_cleanup_plane_fb(struct drm_plane *plane,
 	struct intel_plane_state *old_plane_state =
 		to_intel_plane_state(_old_plane_state);
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(old_plane_state->base.state);
+		to_intel_atomic_state(old_plane_state->uapi.state);
 	struct drm_i915_private *dev_priv = to_i915(plane->dev);
 
 	if (intel_state->rps_interactive) {
@@ -15229,8 +15530,12 @@ static bool i965_plane_format_mod_supported(struct drm_plane *_plane,
 	case DRM_FORMAT_RGB565:
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_ABGR8888:
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_ABGR2101010:
 	case DRM_FORMAT_XBGR16161616F:
 		return modifier == DRM_FORMAT_MOD_LINEAR ||
 			modifier == I915_FORMAT_MOD_X_TILED;
@@ -15288,7 +15593,7 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 	 * When crtc is inactive or there is a modeset pending,
 	 * wait for it to complete in the slowpath
 	 */
-	if (!crtc_state->base.active || needs_modeset(crtc_state) ||
+	if (!crtc_state->hw.active || needs_modeset(crtc_state) ||
 	    crtc_state->update_pipe)
 		goto slow;
 
@@ -15297,8 +15602,8 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 	 * the plane.  This prevents our async update's changes from getting
 	 * overridden by a previous synchronous update's state.
 	 */
-	if (old_plane_state->base.commit &&
-	    !try_wait_for_completion(&old_plane_state->base.commit->hw_done))
+	if (old_plane_state->uapi.commit &&
+	    !try_wait_for_completion(&old_plane_state->uapi.commit->hw_done))
 		goto slow;
 
 	/*
@@ -15306,12 +15611,12 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 	 * take the slowpath. Only changing fb or position should be
 	 * in the fastpath.
 	 */
-	if (old_plane_state->base.crtc != &crtc->base ||
-	    old_plane_state->base.src_w != src_w ||
-	    old_plane_state->base.src_h != src_h ||
-	    old_plane_state->base.crtc_w != crtc_w ||
-	    old_plane_state->base.crtc_h != crtc_h ||
-	    !old_plane_state->base.fb != !fb)
+	if (old_plane_state->uapi.crtc != &crtc->base ||
+	    old_plane_state->uapi.src_w != src_w ||
+	    old_plane_state->uapi.src_h != src_h ||
+	    old_plane_state->uapi.crtc_w != crtc_w ||
+	    old_plane_state->uapi.crtc_h != crtc_h ||
+	    !old_plane_state->uapi.fb != !fb)
 		goto slow;
 
 	new_plane_state = to_intel_plane_state(intel_plane_duplicate_state(&plane->base));
@@ -15324,16 +15629,16 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 		goto out_free;
 	}
 
-	drm_atomic_set_fb_for_plane(&new_plane_state->base, fb);
+	drm_atomic_set_fb_for_plane(&new_plane_state->uapi, fb);
 
-	new_plane_state->base.src_x = src_x;
-	new_plane_state->base.src_y = src_y;
-	new_plane_state->base.src_w = src_w;
-	new_plane_state->base.src_h = src_h;
-	new_plane_state->base.crtc_x = crtc_x;
-	new_plane_state->base.crtc_y = crtc_y;
-	new_plane_state->base.crtc_w = crtc_w;
-	new_plane_state->base.crtc_h = crtc_h;
+	new_plane_state->uapi.src_x = src_x;
+	new_plane_state->uapi.src_y = src_y;
+	new_plane_state->uapi.src_w = src_w;
+	new_plane_state->uapi.src_h = src_h;
+	new_plane_state->uapi.crtc_x = crtc_x;
+	new_plane_state->uapi.crtc_y = crtc_y;
+	new_plane_state->uapi.crtc_w = crtc_w;
+	new_plane_state->uapi.crtc_h = crtc_h;
 
 	ret = intel_plane_atomic_check_with_state(crtc_state, new_crtc_state,
 						  old_plane_state, new_plane_state);
@@ -15344,13 +15649,14 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 	if (ret)
 		goto out_free;
 
-	intel_frontbuffer_flush(to_intel_frontbuffer(new_plane_state->base.fb), ORIGIN_FLIP);
-	intel_frontbuffer_track(to_intel_frontbuffer(old_plane_state->base.fb),
-				to_intel_frontbuffer(new_plane_state->base.fb),
+	intel_frontbuffer_flush(to_intel_frontbuffer(new_plane_state->hw.fb),
+				ORIGIN_FLIP);
+	intel_frontbuffer_track(to_intel_frontbuffer(old_plane_state->hw.fb),
+				to_intel_frontbuffer(new_plane_state->hw.fb),
 				plane->frontbuffer_bit);
 
 	/* Swap plane state */
-	plane->base.state = &new_plane_state->base;
+	plane->base.state = &new_plane_state->uapi;
 
 	/*
 	 * We cannot swap crtc_state as it may be in use by an atomic commit or
@@ -15364,7 +15670,7 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 	 */
 	crtc_state->active_planes = new_crtc_state->active_planes;
 
-	if (new_plane_state->base.visible)
+	if (new_plane_state->uapi.visible)
 		intel_update_plane(plane, crtc_state, new_plane_state);
 	else
 		intel_disable_plane(plane, crtc_state);
@@ -15373,11 +15679,11 @@ intel_legacy_cursor_update(struct drm_plane *_plane,
 
 out_free:
 	if (new_crtc_state)
-		intel_crtc_destroy_state(&crtc->base, &new_crtc_state->base);
+		intel_crtc_destroy_state(&crtc->base, &new_crtc_state->uapi);
 	if (ret)
-		intel_plane_destroy_state(&plane->base, &new_plane_state->base);
+		intel_plane_destroy_state(&plane->base, &new_plane_state->uapi);
 	else
-		intel_plane_destroy_state(&plane->base, &old_plane_state->base);
+		intel_plane_destroy_state(&plane->base, &old_plane_state->uapi);
 	return ret;
 
 slow:
@@ -15419,7 +15725,6 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	const struct drm_plane_funcs *plane_funcs;
 	unsigned int supported_rotations;
 	unsigned int possible_crtcs;
-	const u64 *modifiers;
 	const u32 *formats;
 	int num_formats;
 	int ret, zpos;
@@ -15451,7 +15756,10 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 		fbc->possible_framebuffer_bits |= plane->frontbuffer_bit;
 	}
 
-	if (INTEL_GEN(dev_priv) >= 4) {
+	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
+		formats = vlv_primary_formats;
+		num_formats = ARRAY_SIZE(vlv_primary_formats);
+	} else if (INTEL_GEN(dev_priv) >= 4) {
 		/*
 		 * WaFP16GammaEnabling:ivb
 		 * "Workaround : When using the 64-bit format, the plane
@@ -15472,51 +15780,45 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 			formats = i965_primary_formats;
 			num_formats = ARRAY_SIZE(i965_primary_formats);
 		}
-		modifiers = i9xx_format_modifiers;
-
-		plane->max_stride = i9xx_plane_max_stride;
-		plane->update_plane = i9xx_update_plane;
-		plane->disable_plane = i9xx_disable_plane;
-		plane->get_hw_state = i9xx_plane_get_hw_state;
-		plane->check_plane = i9xx_plane_check;
-
-		if (IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv))
-			plane->min_cdclk = hsw_plane_min_cdclk;
-		else if (IS_IVYBRIDGE(dev_priv))
-			plane->min_cdclk = ivb_plane_min_cdclk;
-		else if (IS_CHERRYVIEW(dev_priv) || IS_VALLEYVIEW(dev_priv))
-			plane->min_cdclk = vlv_plane_min_cdclk;
-		else
-			plane->min_cdclk = i9xx_plane_min_cdclk;
-
-		plane_funcs = &i965_plane_funcs;
 	} else {
 		formats = i8xx_primary_formats;
 		num_formats = ARRAY_SIZE(i8xx_primary_formats);
-		modifiers = i9xx_format_modifiers;
+	}
+
+	if (INTEL_GEN(dev_priv) >= 4)
+		plane_funcs = &i965_plane_funcs;
+	else
+		plane_funcs = &i8xx_plane_funcs;
 
-		plane->max_stride = i9xx_plane_max_stride;
-		plane->update_plane = i9xx_update_plane;
-		plane->disable_plane = i9xx_disable_plane;
-		plane->get_hw_state = i9xx_plane_get_hw_state;
-		plane->check_plane = i9xx_plane_check;
+	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
+		plane->min_cdclk = vlv_plane_min_cdclk;
+	else if (IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv))
+		plane->min_cdclk = hsw_plane_min_cdclk;
+	else if (IS_IVYBRIDGE(dev_priv))
+		plane->min_cdclk = ivb_plane_min_cdclk;
+	else
 		plane->min_cdclk = i9xx_plane_min_cdclk;
 
-		plane_funcs = &i8xx_plane_funcs;
-	}
+	plane->max_stride = i9xx_plane_max_stride;
+	plane->update_plane = i9xx_update_plane;
+	plane->disable_plane = i9xx_disable_plane;
+	plane->get_hw_state = i9xx_plane_get_hw_state;
+	plane->check_plane = i9xx_plane_check;
 
 	possible_crtcs = BIT(pipe);
 
 	if (INTEL_GEN(dev_priv) >= 5 || IS_G4X(dev_priv))
 		ret = drm_universal_plane_init(&dev_priv->drm, &plane->base,
 					       possible_crtcs, plane_funcs,
-					       formats, num_formats, modifiers,
+					       formats, num_formats,
+					       i9xx_format_modifiers,
 					       DRM_PLANE_TYPE_PRIMARY,
 					       "primary %c", pipe_name(pipe));
 	else
 		ret = drm_universal_plane_init(&dev_priv->drm, &plane->base,
 					       possible_crtcs, plane_funcs,
-					       formats, num_formats, modifiers,
+					       formats, num_formats,
+					       i9xx_format_modifiers,
 					       DRM_PLANE_TYPE_PRIMARY,
 					       "plane %c",
 					       plane_name(plane->i9xx_plane));
@@ -15620,28 +15922,6 @@ fail:
 	return ERR_PTR(ret);
 }
 
-static void intel_crtc_init_scalers(struct intel_crtc *crtc,
-				    struct intel_crtc_state *crtc_state)
-{
-	struct intel_crtc_scaler_state *scaler_state =
-		&crtc_state->scaler_state;
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	int i;
-
-	crtc->num_scalers = RUNTIME_INFO(dev_priv)->num_scalers[crtc->pipe];
-	if (!crtc->num_scalers)
-		return;
-
-	for (i = 0; i < crtc->num_scalers; i++) {
-		struct intel_scaler *scaler = &scaler_state->scalers[i];
-
-		scaler->in_use = 0;
-		scaler->mode = 0;
-	}
-
-	scaler_state->scaler_id = -1;
-}
-
 #define INTEL_CRTC_FUNCS \
 	.gamma_set = drm_atomic_helper_legacy_gamma_set, \
 	.set_config = drm_atomic_helper_set_config, \
@@ -15709,33 +15989,53 @@ static const struct drm_crtc_funcs i8xx_crtc_funcs = {
 	.disable_vblank = i8xx_disable_vblank,
 };
 
+static struct intel_crtc *intel_crtc_alloc(void)
+{
+	struct intel_crtc_state *crtc_state;
+	struct intel_crtc *crtc;
+
+	crtc = kzalloc(sizeof(*crtc), GFP_KERNEL);
+	if (!crtc)
+		return ERR_PTR(-ENOMEM);
+
+	crtc_state = intel_crtc_state_alloc(crtc);
+	if (!crtc_state) {
+		kfree(crtc);
+		return ERR_PTR(-ENOMEM);
+	}
+
+	crtc->base.state = &crtc_state->uapi;
+	crtc->config = crtc_state;
+
+	return crtc;
+}
+
+static void intel_crtc_free(struct intel_crtc *crtc)
+{
+	intel_crtc_destroy_state(&crtc->base, crtc->base.state);
+	kfree(crtc);
+}
+
 static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
+	struct intel_plane *primary, *cursor;
 	const struct drm_crtc_funcs *funcs;
-	struct intel_crtc *intel_crtc;
-	struct intel_crtc_state *crtc_state = NULL;
-	struct intel_plane *primary = NULL;
-	struct intel_plane *cursor = NULL;
+	struct intel_crtc *crtc;
 	int sprite, ret;
 
-	intel_crtc = kzalloc(sizeof(*intel_crtc), GFP_KERNEL);
-	if (!intel_crtc)
-		return -ENOMEM;
+	crtc = intel_crtc_alloc();
+	if (IS_ERR(crtc))
+		return PTR_ERR(crtc);
 
-	crtc_state = kzalloc(sizeof(*crtc_state), GFP_KERNEL);
-	if (!crtc_state) {
-		ret = -ENOMEM;
-		goto fail;
-	}
-	__drm_atomic_helper_crtc_reset(&intel_crtc->base, &crtc_state->base);
-	intel_crtc->config = crtc_state;
+	crtc->pipe = pipe;
+	crtc->num_scalers = RUNTIME_INFO(dev_priv)->num_scalers[pipe];
 
 	primary = intel_primary_plane_create(dev_priv, pipe);
 	if (IS_ERR(primary)) {
 		ret = PTR_ERR(primary);
 		goto fail;
 	}
-	intel_crtc->plane_ids_mask |= BIT(primary->id);
+	crtc->plane_ids_mask |= BIT(primary->id);
 
 	for_each_sprite(dev_priv, pipe, sprite) {
 		struct intel_plane *plane;
@@ -15745,7 +16045,7 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 			ret = PTR_ERR(plane);
 			goto fail;
 		}
-		intel_crtc->plane_ids_mask |= BIT(plane->id);
+		crtc->plane_ids_mask |= BIT(plane->id);
 	}
 
 	cursor = intel_cursor_plane_create(dev_priv, pipe);
@@ -15753,7 +16053,7 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 		ret = PTR_ERR(cursor);
 		goto fail;
 	}
-	intel_crtc->plane_ids_mask |= BIT(cursor->id);
+	crtc->plane_ids_mask |= BIT(cursor->id);
 
 	if (HAS_GMCH(dev_priv)) {
 		if (IS_CHERRYVIEW(dev_priv) ||
@@ -15774,42 +16074,32 @@ static int intel_crtc_init(struct drm_i915_private *dev_priv, enum pipe pipe)
 			funcs = &ilk_crtc_funcs;
 	}
 
-	ret = drm_crtc_init_with_planes(&dev_priv->drm, &intel_crtc->base,
+	ret = drm_crtc_init_with_planes(&dev_priv->drm, &crtc->base,
 					&primary->base, &cursor->base,
 					funcs, "pipe %c", pipe_name(pipe));
 	if (ret)
 		goto fail;
 
-	intel_crtc->pipe = pipe;
-
-	/* initialize shared scalers */
-	intel_crtc_init_scalers(intel_crtc, crtc_state);
-
 	BUG_ON(pipe >= ARRAY_SIZE(dev_priv->pipe_to_crtc_mapping) ||
 	       dev_priv->pipe_to_crtc_mapping[pipe] != NULL);
-	dev_priv->pipe_to_crtc_mapping[pipe] = intel_crtc;
+	dev_priv->pipe_to_crtc_mapping[pipe] = crtc;
 
 	if (INTEL_GEN(dev_priv) < 9) {
 		enum i9xx_plane_id i9xx_plane = primary->i9xx_plane;
 
 		BUG_ON(i9xx_plane >= ARRAY_SIZE(dev_priv->plane_to_crtc_mapping) ||
 		       dev_priv->plane_to_crtc_mapping[i9xx_plane] != NULL);
-		dev_priv->plane_to_crtc_mapping[i9xx_plane] = intel_crtc;
+		dev_priv->plane_to_crtc_mapping[i9xx_plane] = crtc;
 	}
 
-	intel_color_init(intel_crtc);
+	intel_color_init(crtc);
 
-	WARN_ON(drm_crtc_index(&intel_crtc->base) != intel_crtc->pipe);
+	WARN_ON(drm_crtc_index(&crtc->base) != crtc->pipe);
 
 	return 0;
 
 fail:
-	/*
-	 * drm_mode_config_cleanup() will free up any
-	 * crtcs/planes already initialized.
-	 */
-	kfree(crtc_state);
-	kfree(intel_crtc);
+	intel_crtc_free(crtc);
 
 	return ret;
 }
@@ -16310,26 +16600,23 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb,
 		}
 
 		stride_alignment = intel_fb_stride_alignment(fb, i);
-
-		/*
-		 * Display WA #0531: skl,bxt,kbl,glk
-		 *
-		 * Render decompression and plane width > 3840
-		 * combined with horizontal panning requires the
-		 * plane stride to be a multiple of 4. We'll just
-		 * require the entire fb to accommodate that to avoid
-		 * potential runtime errors at plane configuration time.
-		 */
-		if (IS_GEN(dev_priv, 9) && i == 0 && fb->width > 3840 &&
-		    is_ccs_modifier(fb->modifier))
-			stride_alignment *= 4;
-
 		if (fb->pitches[i] & (stride_alignment - 1)) {
 			DRM_DEBUG_KMS("plane %d pitch (%d) must be at least %u byte aligned\n",
 				      i, fb->pitches[i], stride_alignment);
 			goto err;
 		}
 
+		if (is_gen12_ccs_plane(fb, i)) {
+			int ccs_aux_stride = gen12_ccs_aux_stride(fb, i);
+
+			if (fb->pitches[i] != ccs_aux_stride) {
+				DRM_DEBUG_KMS("ccs aux plane %d pitch (%d) must be %d\n",
+					      i,
+					      fb->pitches[i], ccs_aux_stride);
+				goto err;
+			}
+		}
+
 		fb->obj[i] = &obj->base;
 	}
 
@@ -16601,8 +16888,6 @@ void intel_init_display_hooks(struct drm_i915_private *dev_priv)
 	} else if (IS_IVYBRIDGE(dev_priv)) {
 		/* FIXME: detect B0+ stepping and use auto training */
 		dev_priv->display.fdi_link_train = ivb_manual_fdi_link_train;
-	} else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) {
-		dev_priv->display.fdi_link_train = hsw_fdi_link_train;
 	}
 
 	if (INTEL_GEN(dev_priv) >= 9)
@@ -16692,7 +16977,7 @@ retry:
 	/* Write calculated watermark values back */
 	for_each_new_intel_crtc_in_state(intel_state, crtc, crtc_state, i) {
 		crtc_state->wm.need_postvbl_update = true;
-		dev_priv->display.optimize_watermarks(intel_state, crtc_state);
+		dev_priv->display.optimize_watermarks(intel_state, crtc);
 
 		to_intel_crtc_state(crtc->base.state)->wm = crtc_state->wm;
 	}
@@ -16724,8 +17009,7 @@ static int intel_initial_commit(struct drm_device *dev)
 {
 	struct drm_atomic_state *state = NULL;
 	struct drm_modeset_acquire_ctx ctx;
-	struct drm_crtc *crtc;
-	struct drm_crtc_state *crtc_state;
+	struct intel_crtc *crtc;
 	int ret = 0;
 
 	state = drm_atomic_state_alloc(dev);
@@ -16737,15 +17021,17 @@ static int intel_initial_commit(struct drm_device *dev)
 retry:
 	state->acquire_ctx = &ctx;
 
-	drm_for_each_crtc(crtc, dev) {
-		crtc_state = drm_atomic_get_crtc_state(state, crtc);
+	for_each_intel_crtc(dev, crtc) {
+		struct intel_crtc_state *crtc_state =
+			intel_atomic_get_crtc_state(state, crtc);
+
 		if (IS_ERR(crtc_state)) {
 			ret = PTR_ERR(crtc_state);
 			goto out;
 		}
 
-		if (crtc_state->active) {
-			ret = drm_atomic_add_affected_planes(state, crtc);
+		if (crtc_state->hw.active) {
+			ret = drm_atomic_add_affected_planes(state, &crtc->base);
 			if (ret)
 				goto out;
 
@@ -16755,7 +17041,7 @@ retry:
 			 * having a proper LUT loaded. Remove once we
 			 * have readout for pipe gamma enable.
 			 */
-			crtc_state->color_mgmt_changed = true;
+			crtc_state->uapi.color_mgmt_changed = true;
 		}
 	}
 
@@ -17078,31 +17364,76 @@ static bool has_pch_trancoder(struct drm_i915_private *dev_priv,
 		(HAS_PCH_LPT_H(dev_priv) && pch_transcoder == PIPE_A);
 }
 
-static void intel_sanitize_crtc(struct intel_crtc *crtc,
-				struct drm_modeset_acquire_ctx *ctx)
+static void intel_sanitize_frame_start_delay(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_device *dev = crtc->base.dev;
-	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc_state *crtc_state = to_intel_crtc_state(crtc->base.state);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 
-	/* Clear any frame start delays used for debugging left by the BIOS */
-	if (crtc->active && !transcoder_is_dsi(cpu_transcoder)) {
+	if (INTEL_GEN(dev_priv) >= 9 ||
+	    IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv)) {
+		i915_reg_t reg = CHICKEN_TRANS(cpu_transcoder);
+		u32 val;
+
+		if (transcoder_is_dsi(cpu_transcoder))
+			return;
+
+		val = I915_READ(reg);
+		val &= ~HSW_FRAME_START_DELAY_MASK;
+		val |= HSW_FRAME_START_DELAY(0);
+		I915_WRITE(reg, val);
+	} else {
 		i915_reg_t reg = PIPECONF(cpu_transcoder);
+		u32 val;
+
+		val = I915_READ(reg);
+		val &= ~PIPECONF_FRAME_START_DELAY_MASK;
+		val |= PIPECONF_FRAME_START_DELAY(0);
+		I915_WRITE(reg, val);
+	}
+
+	if (!crtc_state->has_pch_encoder)
+		return;
+
+	if (HAS_PCH_IBX(dev_priv)) {
+		i915_reg_t reg = PCH_TRANSCONF(crtc->pipe);
+		u32 val;
+
+		val = I915_READ(reg);
+		val &= ~TRANS_FRAME_START_DELAY_MASK;
+		val |= TRANS_FRAME_START_DELAY(0);
+		I915_WRITE(reg, val);
+	} else {
+		enum pipe pch_transcoder = intel_crtc_pch_transcoder(crtc);
+		i915_reg_t reg = TRANS_CHICKEN2(pch_transcoder);
+		u32 val;
 
-		I915_WRITE(reg,
-			   I915_READ(reg) & ~PIPECONF_FRAME_START_DELAY_MASK);
+		val = I915_READ(reg);
+		val &= ~TRANS_CHICKEN2_FRAME_START_DELAY_MASK;
+		val |= TRANS_CHICKEN2_FRAME_START_DELAY(0);
+		I915_WRITE(reg, val);
 	}
+}
 
-	if (crtc_state->base.active) {
+static void intel_sanitize_crtc(struct intel_crtc *crtc,
+				struct drm_modeset_acquire_ctx *ctx)
+{
+	struct drm_device *dev = crtc->base.dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct intel_crtc_state *crtc_state = to_intel_crtc_state(crtc->base.state);
+
+	if (crtc_state->hw.active) {
 		struct intel_plane *plane;
 
+		/* Clear any frame start delays used for debugging left by the BIOS */
+		intel_sanitize_frame_start_delay(crtc_state);
+
 		/* Disable everything but the primary plane */
 		for_each_intel_plane_on_crtc(dev, crtc, plane) {
 			const struct intel_plane_state *plane_state =
 				to_intel_plane_state(plane->base.state);
 
-			if (plane_state->base.visible &&
+			if (plane_state->uapi.visible &&
 			    plane->base.type != DRM_PLANE_TYPE_PRIMARY)
 				intel_plane_disable_noatomic(crtc, plane);
 		}
@@ -17119,10 +17450,10 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc,
 
 	/* Adjust the state of the output pipe according to whether we
 	 * have active connectors/encoders. */
-	if (crtc_state->base.active && !intel_crtc_has_encoders(crtc))
-		intel_crtc_disable_noatomic(&crtc->base, ctx);
+	if (crtc_state->hw.active && !intel_crtc_has_encoders(crtc))
+		intel_crtc_disable_noatomic(crtc, ctx);
 
-	if (crtc_state->base.active || HAS_GMCH(dev_priv)) {
+	if (crtc_state->hw.active || HAS_GMCH(dev_priv)) {
 		/*
 		 * We start out with underrun reporting disabled to avoid races.
 		 * For correct bookkeeping mark this on active crtcs.
@@ -17153,7 +17484,7 @@ static void intel_sanitize_crtc(struct intel_crtc *crtc,
 
 static bool has_bogus_dpll_config(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	/*
 	 * Some SNB BIOSen (eg. ASUS K53SV) are known to misprogram
@@ -17166,7 +17497,7 @@ static bool has_bogus_dpll_config(const struct intel_crtc_state *crtc_state)
 	 * road.
 	 */
 	return IS_GEN(dev_priv, 6) &&
-		crtc_state->base.active &&
+		crtc_state->hw.active &&
 		crtc_state->shared_dpll &&
 		crtc_state->port_clock == 0;
 }
@@ -17183,7 +17514,7 @@ static void intel_sanitize_encoder(struct intel_encoder *encoder)
 	 * encoder is active and trying to read from a pipe) and the
 	 * pipe itself being active. */
 	bool has_active_crtc = crtc_state &&
-		crtc_state->base.active;
+		crtc_state->hw.active;
 
 	if (crtc_state && has_bogus_dpll_config(crtc_state)) {
 		DRM_DEBUG_KMS("BIOS has misprogrammed the hardware. Disabling pipe %c\n",
@@ -17287,22 +17618,22 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		struct intel_crtc_state *crtc_state =
 			to_intel_crtc_state(crtc->base.state);
 
-		__drm_atomic_helper_crtc_destroy_state(&crtc_state->base);
-		memset(crtc_state, 0, sizeof(*crtc_state));
-		__drm_atomic_helper_crtc_reset(&crtc->base, &crtc_state->base);
+		__drm_atomic_helper_crtc_destroy_state(&crtc_state->uapi);
+		intel_crtc_free_hw_state(crtc_state);
+		intel_crtc_state_reset(crtc_state, crtc);
 
-		crtc_state->base.active = crtc_state->base.enable =
+		crtc_state->hw.active = crtc_state->hw.enable =
 			dev_priv->display.get_pipe_config(crtc, crtc_state);
 
-		crtc->base.enabled = crtc_state->base.enable;
-		crtc->active = crtc_state->base.active;
+		crtc->base.enabled = crtc_state->hw.enable;
+		crtc->active = crtc_state->hw.active;
 
-		if (crtc_state->base.active)
+		if (crtc_state->hw.active)
 			dev_priv->active_pipes |= BIT(crtc->pipe);
 
 		DRM_DEBUG_KMS("[CRTC:%d:%s] hw state readout: %s\n",
 			      crtc->base.base.id, crtc->base.name,
-			      enableddisabled(crtc_state->base.active));
+			      enableddisabled(crtc_state->hw.active));
 	}
 
 	readout_plane_state(dev_priv);
@@ -17324,7 +17655,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			struct intel_crtc_state *crtc_state =
 				to_intel_crtc_state(crtc->base.state);
 
-			if (crtc_state->base.active &&
+			if (crtc_state->hw.active &&
 			    crtc_state->shared_dpll == pll)
 				pll->state.crtc_mask |= 1 << crtc->pipe;
 		}
@@ -17369,15 +17700,15 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			crtc = to_intel_crtc(encoder->base.crtc);
 			crtc_state = crtc ? to_intel_crtc_state(crtc->base.state) : NULL;
 
-			if (crtc_state && crtc_state->base.active) {
+			if (crtc_state && crtc_state->hw.active) {
 				/*
 				 * This has to be done during hardware readout
 				 * because anything calling .crtc_disable may
 				 * rely on the connector_mask being accurate.
 				 */
-				crtc_state->base.connector_mask |=
+				crtc_state->uapi.connector_mask |=
 					drm_connector_mask(&connector->base);
-				crtc_state->base.encoder_mask |=
+				crtc_state->uapi.encoder_mask |=
 					drm_encoder_mask(&encoder->base);
 			}
 		} else {
@@ -17398,16 +17729,15 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 		struct intel_plane *plane;
 		int min_cdclk = 0;
 
-		if (crtc_state->base.active) {
-			struct drm_display_mode mode;
+		if (crtc_state->hw.active) {
+			struct drm_display_mode *mode = &crtc_state->hw.mode;
 
-			intel_mode_from_pipe_config(&crtc_state->base.adjusted_mode,
+			intel_mode_from_pipe_config(&crtc_state->hw.adjusted_mode,
 						    crtc_state);
 
-			mode = crtc_state->base.adjusted_mode;
-			mode.hdisplay = crtc_state->pipe_src_w;
-			mode.vdisplay = crtc_state->pipe_src_h;
-			WARN_ON(drm_atomic_set_mode_for_crtc(&crtc_state->base, &mode));
+			*mode = crtc_state->hw.adjusted_mode;
+			mode->hdisplay = crtc_state->pipe_src_w;
+			mode->vdisplay = crtc_state->pipe_src_h;
 
 			/*
 			 * The initial mode needs to be set in order to keep
@@ -17418,11 +17748,13 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			 * set a flag to indicate that a full recalculation is
 			 * needed on the next commit.
 			 */
-			crtc_state->base.mode.private_flags = I915_MODE_FLAG_INHERITED;
+			mode->private_flags = I915_MODE_FLAG_INHERITED;
 
 			intel_crtc_compute_pixel_rate(crtc_state);
 
 			intel_crtc_update_active_timings(crtc_state);
+
+			intel_crtc_copy_hw_to_uapi_state(crtc_state);
 		}
 
 		for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) {
@@ -17433,14 +17765,14 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 			 * FIXME don't have the fb yet, so can't
 			 * use intel_plane_data_rate() :(
 			 */
-			if (plane_state->base.visible)
+			if (plane_state->uapi.visible)
 				crtc_state->data_rate[plane->id] =
 					4 * crtc_state->pixel_rate;
 			/*
 			 * FIXME don't have the fb yet, so can't
 			 * use plane->min_cdclk() :(
 			 */
-			if (plane_state->base.visible && plane->min_cdclk) {
+			if (plane_state->uapi.visible && plane->min_cdclk) {
 				if (crtc_state->double_wide ||
 				    INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 					crtc_state->min_cdclk[plane->id] =
@@ -17454,7 +17786,7 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 				      crtc_state->min_cdclk[plane->id]);
 		}
 
-		if (crtc_state->base.active) {
+		if (crtc_state->hw.active) {
 			min_cdclk = intel_crtc_compute_min_cdclk(crtc_state);
 			if (WARN_ON(min_cdclk < 0))
 				min_cdclk = 0;
@@ -17577,7 +17909,6 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
 			     struct drm_modeset_acquire_ctx *ctx)
 {
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc_state *crtc_state;
 	struct intel_encoder *encoder;
 	struct intel_crtc *crtc;
 	intel_wakeref_t wakeref;
@@ -17610,11 +17941,12 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
 	 * waits, so we need vblank interrupts restored beforehand.
 	 */
 	for_each_intel_crtc(&dev_priv->drm, crtc) {
-		crtc_state = to_intel_crtc_state(crtc->base.state);
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
 
 		drm_crtc_vblank_reset(&crtc->base);
 
-		if (crtc_state->base.active)
+		if (crtc_state->hw.active)
 			intel_crtc_vblank_on(crtc_state);
 	}
 
@@ -17624,7 +17956,9 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
 		intel_sanitize_encoder(encoder);
 
 	for_each_intel_crtc(&dev_priv->drm, crtc) {
-		crtc_state = to_intel_crtc_state(crtc->base.state);
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
+
 		intel_sanitize_crtc(crtc, ctx);
 		intel_dump_pipe_config(crtc_state, NULL, "[setup_hw_state]");
 	}
@@ -17657,17 +17991,16 @@ intel_modeset_setup_hw_state(struct drm_device *dev,
 	}
 
 	for_each_intel_crtc(dev, crtc) {
+		struct intel_crtc_state *crtc_state =
+			to_intel_crtc_state(crtc->base.state);
 		u64 put_domains;
 
-		crtc_state = to_intel_crtc_state(crtc->base.state);
 		put_domains = modeset_get_crtc_power_domains(crtc_state);
 		if (WARN_ON(put_domains))
 			modeset_put_power_domains(dev_priv, put_domains);
 	}
 
 	intel_display_power_put(dev_priv, POWER_DOMAIN_INIT, wakeref);
-
-	intel_fbc_init_pipe_state(dev_priv);
 }
 
 void intel_display_resume(struct drm_device *dev)
@@ -17743,6 +18076,13 @@ void intel_modeset_driver_remove(struct drm_i915_private *i915)
 	 */
 	intel_hpd_poll_fini(i915);
 
+	/*
+	 * MST topology needs to be suspended so we don't have any calls to
+	 * fbdev after it's finalized. MST will be destroyed later as part of
+	 * drm_mode_config_cleanup()
+	 */
+	intel_dp_mst_suspend(i915);
+
 	/* poll work can call into fbdev, hence clean that up afterwards */
 	intel_fbdev_fini(i915);
 
diff --git a/drivers/gpu/drm/i915/display/intel_display.h b/drivers/gpu/drm/i915/display/intel_display.h
index f417e0948001..0fef9263cddc 100644
--- a/drivers/gpu/drm/i915/display/intel_display.h
+++ b/drivers/gpu/drm/i915/display/intel_display.h
@@ -332,8 +332,11 @@ enum phy_fia {
 	     (__s) < RUNTIME_INFO(__dev_priv)->num_sprites[(__p)];	\
 	     (__s)++)
 
-#define for_each_port_masked(__port, __ports_mask) \
-	for ((__port) = PORT_A; (__port) < I915_MAX_PORTS; (__port)++)	\
+#define for_each_port(__port) \
+	for ((__port) = PORT_A; (__port) < I915_MAX_PORTS; (__port)++)
+
+#define for_each_port_masked(__port, __ports_mask)			\
+	for_each_port(__port)						\
 		for_each_if((__ports_mask) & BIT(__port))
 
 #define for_each_phy_masked(__phy, __phys_mask) \
@@ -377,6 +380,13 @@ enum phy_fia {
 			    &(dev)->mode_config.encoder_list,	\
 			    base.head)
 
+#define for_each_intel_encoder_mask(dev, intel_encoder, encoder_mask)	\
+	list_for_each_entry(intel_encoder,				\
+			    &(dev)->mode_config.encoder_list,		\
+			    base.head)					\
+		for_each_if((encoder_mask) &				\
+			    drm_encoder_mask(&intel_encoder->base))
+
 #define for_each_intel_dp(dev, intel_encoder)			\
 	for_each_intel_encoder(dev, intel_encoder)		\
 		for_each_if(intel_encoder_is_dp(intel_encoder))
@@ -446,10 +456,18 @@ enum phy_fia {
 #define intel_atomic_crtc_state_for_each_plane_state( \
 		  plane, plane_state, \
 		  crtc_state) \
-	for_each_intel_plane_mask(((crtc_state)->base.state->dev), (plane), \
-				((crtc_state)->base.plane_mask)) \
+	for_each_intel_plane_mask(((crtc_state)->uapi.state->dev), (plane), \
+				((crtc_state)->uapi.plane_mask)) \
 		for_each_if ((plane_state = \
-			      to_intel_plane_state(__drm_atomic_get_current_plane_state((crtc_state)->base.state, &plane->base))))
+			      to_intel_plane_state(__drm_atomic_get_current_plane_state((crtc_state)->uapi.state, &plane->base))))
+
+#define for_each_new_intel_connector_in_state(__state, connector, new_connector_state, __i) \
+	for ((__i) = 0; \
+	     (__i) < (__state)->base.num_connector; \
+	     (__i)++) \
+		for_each_if ((__state)->base.connectors[__i].ptr && \
+			     ((connector) = to_intel_connector((__state)->base.connectors[__i].ptr), \
+			     (new_connector_state) = to_intel_digital_connector_state((__state)->base.connectors[__i].new_state), 1))
 
 void intel_link_compute_m_n(u16 bpp, int nlanes,
 			    int pixel_clock, int link_clock,
@@ -467,6 +485,7 @@ enum phy intel_port_to_phy(struct drm_i915_private *i915, enum port port);
 bool is_trans_port_sync_mode(const struct intel_crtc_state *state);
 
 void intel_plane_destroy(struct drm_plane *plane);
+void intel_disable_pipe(const struct intel_crtc_state *old_crtc_state);
 void i830_enable_pipe(struct drm_i915_private *dev_priv, enum pipe pipe);
 void i830_disable_pipe(struct drm_i915_private *dev_priv, enum pipe pipe);
 enum pipe intel_crtc_pch_transcoder(struct intel_crtc *crtc);
@@ -499,9 +518,8 @@ enum tc_port intel_port_to_tc(struct drm_i915_private *dev_priv,
 			      enum port port);
 int intel_get_pipe_from_crtc_id_ioctl(struct drm_device *dev, void *data,
 				      struct drm_file *file_priv);
-enum transcoder intel_pipe_to_cpu_transcoder(struct drm_i915_private *dev_priv,
-					     enum pipe pipe);
 u32 intel_crtc_get_vblank_counter(struct intel_crtc *crtc);
+void intel_crtc_vblank_off(const struct intel_crtc_state *crtc_state);
 
 int ironlake_get_lanes_required(int target_clock, int link_bw, int bpp);
 void vlv_wait_port_ready(struct drm_i915_private *dev_priv,
@@ -547,7 +565,6 @@ bool bxt_find_best_dpll(struct intel_crtc_state *crtc_state,
 			struct dpll *best_clock);
 int chv_calc_dpll_params(int refclk, struct dpll *pll_clock);
 
-bool intel_crtc_active(struct intel_crtc *crtc);
 bool hsw_crtc_state_ips_capable(const struct intel_crtc_state *crtc_state);
 void hsw_enable_ips(const struct intel_crtc_state *crtc_state);
 void hsw_disable_ips(const struct intel_crtc_state *crtc_state);
@@ -561,6 +578,8 @@ void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc,
 
 u16 skl_scaler_calc_phase(int sub, int scale, bool chroma_center);
 int skl_update_scaler_crtc(struct intel_crtc_state *crtc_state);
+void skylake_scaler_disable(const struct intel_crtc_state *old_crtc_state);
+void ironlake_pfit_disable(const struct intel_crtc_state *old_crtc_state);
 u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 			const struct intel_plane_state *plane_state);
 u32 glk_plane_color_ctl_crtc(const struct intel_crtc_state *crtc_state);
@@ -582,6 +601,10 @@ intel_display_capture_error_state(struct drm_i915_private *dev_priv);
 void intel_display_print_error_state(struct drm_i915_error_state_buf *e,
 				     struct intel_display_error_state *error);
 
+bool
+intel_format_info_is_yuv_semiplanar(const struct drm_format_info *info,
+				    uint64_t modifier);
+
 /* modesetting */
 void intel_modeset_init_hw(struct drm_i915_private *i915);
 int intel_modeset_init(struct drm_i915_private *i915);
@@ -603,9 +626,10 @@ void assert_fdi_rx_pll(struct drm_i915_private *dev_priv,
 		       enum pipe pipe, bool state);
 #define assert_fdi_rx_pll_enabled(d, p) assert_fdi_rx_pll(d, p, true)
 #define assert_fdi_rx_pll_disabled(d, p) assert_fdi_rx_pll(d, p, false)
-void assert_pipe(struct drm_i915_private *dev_priv, enum pipe pipe, bool state);
-#define assert_pipe_enabled(d, p) assert_pipe(d, p, true)
-#define assert_pipe_disabled(d, p) assert_pipe(d, p, false)
+void assert_pipe(struct drm_i915_private *dev_priv,
+		 enum transcoder cpu_transcoder, bool state);
+#define assert_pipe_enabled(d, t) assert_pipe(d, t, true)
+#define assert_pipe_disabled(d, t) assert_pipe(d, t, false)
 
 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and
  * WARN_ON()) for hw state sanity checks to check for unexpected conditions
diff --git a/drivers/gpu/drm/i915/display/intel_display_power.c b/drivers/gpu/drm/i915/display/intel_display_power.c
index ce1b64f4dd44..679457156797 100644
--- a/drivers/gpu/drm/i915/display/intel_display_power.c
+++ b/drivers/gpu/drm/i915/display/intel_display_power.c
@@ -418,7 +418,8 @@ icl_combo_phy_aux_power_well_enable(struct drm_i915_private *dev_priv,
 	int pw_idx = power_well->desc->hsw.idx;
 	enum phy phy = ICL_AUX_PW_TO_PHY(pw_idx);
 	u32 val;
-	int wa_idx_max;
+
+	WARN_ON(!IS_ICELAKE(dev_priv));
 
 	val = I915_READ(regs->driver);
 	I915_WRITE(regs->driver, val | HSW_PWR_WELL_CTL_REQ(pw_idx));
@@ -430,14 +431,8 @@ icl_combo_phy_aux_power_well_enable(struct drm_i915_private *dev_priv,
 
 	hsw_wait_for_power_well_enable(dev_priv, power_well);
 
-	/* Display WA #1178: icl, tgl */
-	if (IS_TIGERLAKE(dev_priv))
-		wa_idx_max = ICL_PW_CTL_IDX_AUX_C;
-	else
-		wa_idx_max = ICL_PW_CTL_IDX_AUX_B;
-
-	if (!IS_ELKHARTLAKE(dev_priv) &&
-	    pw_idx >= ICL_PW_CTL_IDX_AUX_A && pw_idx <= wa_idx_max &&
+	/* Display WA #1178: icl */
+	if (pw_idx >= ICL_PW_CTL_IDX_AUX_A && pw_idx <= ICL_PW_CTL_IDX_AUX_B &&
 	    !intel_bios_is_port_edp(dev_priv, (enum port)phy)) {
 		val = I915_READ(ICL_AUX_ANAOVRD1(pw_idx));
 		val |= ICL_AUX_ANAOVRD1_ENABLE | ICL_AUX_ANAOVRD1_LDO_BYPASS;
@@ -454,10 +449,10 @@ icl_combo_phy_aux_power_well_disable(struct drm_i915_private *dev_priv,
 	enum phy phy = ICL_AUX_PW_TO_PHY(pw_idx);
 	u32 val;
 
-	if (INTEL_GEN(dev_priv) < 12) {
-		val = I915_READ(ICL_PORT_CL_DW12(phy));
-		I915_WRITE(ICL_PORT_CL_DW12(phy), val & ~ICL_LANE_ENABLE_AUX);
-	}
+	WARN_ON(!IS_ICELAKE(dev_priv));
+
+	val = I915_READ(ICL_PORT_CL_DW12(phy));
+	I915_WRITE(ICL_PORT_CL_DW12(phy), val & ~ICL_LANE_ENABLE_AUX);
 
 	val = I915_READ(regs->driver);
 	I915_WRITE(regs->driver, val & ~HSW_PWR_WELL_CTL_REQ(pw_idx));
@@ -3688,6 +3683,151 @@ static const struct i915_power_well_desc icl_power_wells[] = {
 	},
 };
 
+static const struct i915_power_well_desc ehl_power_wells[] = {
+	{
+		.name = "always-on",
+		.always_on = true,
+		.domains = POWER_DOMAIN_MASK,
+		.ops = &i9xx_always_on_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+	},
+	{
+		.name = "power well 1",
+		/* Handled by the DMC firmware */
+		.always_on = true,
+		.domains = 0,
+		.ops = &hsw_power_well_ops,
+		.id = SKL_DISP_PW_1,
+		{
+			.hsw.regs = &hsw_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_PW_1,
+			.hsw.has_fuses = true,
+		},
+	},
+	{
+		.name = "DC off",
+		.domains = ICL_DISPLAY_DC_OFF_POWER_DOMAINS,
+		.ops = &gen9_dc_off_power_well_ops,
+		.id = SKL_DISP_DC_OFF,
+	},
+	{
+		.name = "power well 2",
+		.domains = ICL_PW_2_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = SKL_DISP_PW_2,
+		{
+			.hsw.regs = &hsw_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_PW_2,
+			.hsw.has_fuses = true,
+		},
+	},
+	{
+		.name = "power well 3",
+		.domains = ICL_PW_3_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &hsw_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_PW_3,
+			.hsw.irq_pipe_mask = BIT(PIPE_B),
+			.hsw.has_vga = true,
+			.hsw.has_fuses = true,
+		},
+	},
+	{
+		.name = "DDI A IO",
+		.domains = ICL_DDI_IO_A_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_ddi_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_DDI_A,
+		},
+	},
+	{
+		.name = "DDI B IO",
+		.domains = ICL_DDI_IO_B_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_ddi_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_DDI_B,
+		},
+	},
+	{
+		.name = "DDI C IO",
+		.domains = ICL_DDI_IO_C_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_ddi_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_DDI_C,
+		},
+	},
+	{
+		.name = "DDI D IO",
+		.domains = ICL_DDI_IO_D_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_ddi_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_DDI_D,
+		},
+	},
+	{
+		.name = "AUX A",
+		.domains = ICL_AUX_A_IO_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_aux_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_AUX_A,
+		},
+	},
+	{
+		.name = "AUX B",
+		.domains = ICL_AUX_B_IO_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_aux_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_AUX_B,
+		},
+	},
+	{
+		.name = "AUX C",
+		.domains = ICL_AUX_C_TC1_IO_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_aux_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_AUX_C,
+		},
+	},
+	{
+		.name = "AUX D",
+		.domains = ICL_AUX_D_TC2_IO_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &icl_aux_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_AUX_D,
+		},
+	},
+	{
+		.name = "power well 4",
+		.domains = ICL_PW_4_POWER_DOMAINS,
+		.ops = &hsw_power_well_ops,
+		.id = DISP_PW_ID_NONE,
+		{
+			.hsw.regs = &hsw_power_well_regs,
+			.hsw.idx = ICL_PW_CTL_IDX_PW_4,
+			.hsw.has_fuses = true,
+			.hsw.irq_pipe_mask = BIT(PIPE_C),
+		},
+	},
+};
+
 static const struct i915_power_well_desc tgl_power_wells[] = {
 	{
 		.name = "always-on",
@@ -3832,7 +3972,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 	{
 		.name = "AUX A",
 		.domains = TGL_AUX_A_IO_POWER_DOMAINS,
-		.ops = &icl_combo_phy_aux_power_well_ops,
+		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
 			.hsw.regs = &icl_aux_power_well_regs,
@@ -3842,7 +3982,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 	{
 		.name = "AUX B",
 		.domains = TGL_AUX_B_IO_POWER_DOMAINS,
-		.ops = &icl_combo_phy_aux_power_well_ops,
+		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
 			.hsw.regs = &icl_aux_power_well_regs,
@@ -3852,7 +3992,7 @@ static const struct i915_power_well_desc tgl_power_wells[] = {
 	{
 		.name = "AUX C",
 		.domains = TGL_AUX_C_IO_POWER_DOMAINS,
-		.ops = &icl_combo_phy_aux_power_well_ops,
+		.ops = &hsw_power_well_ops,
 		.id = DISP_PW_ID_NONE,
 		{
 			.hsw.regs = &icl_aux_power_well_regs,
@@ -4162,6 +4302,8 @@ int intel_power_domains_init(struct drm_i915_private *dev_priv)
 	 */
 	if (IS_GEN(dev_priv, 12)) {
 		err = set_power_wells(power_domains, tgl_power_wells);
+	} else if (IS_ELKHARTLAKE(dev_priv)) {
+		err = set_power_wells(power_domains, ehl_power_wells);
 	} else if (IS_GEN(dev_priv, 11)) {
 		err = set_power_wells(power_domains, icl_power_wells);
 	} else if (IS_CANNONLAKE(dev_priv)) {
@@ -4781,6 +4923,56 @@ static void cnl_display_core_uninit(struct drm_i915_private *dev_priv)
 	intel_combo_phy_uninit(dev_priv);
 }
 
+struct buddy_page_mask {
+	u32 page_mask;
+	u8 type;
+	u8 num_channels;
+};
+
+static const struct buddy_page_mask tgl_buddy_page_masks[] = {
+	{ .num_channels = 1, .type = INTEL_DRAM_LPDDR4, .page_mask = 0xE },
+	{ .num_channels = 1, .type = INTEL_DRAM_DDR4,   .page_mask = 0xF },
+	{ .num_channels = 2, .type = INTEL_DRAM_LPDDR4, .page_mask = 0x1C },
+	{ .num_channels = 2, .type = INTEL_DRAM_DDR4,   .page_mask = 0x1F },
+	{}
+};
+
+static const struct buddy_page_mask wa_1409767108_buddy_page_masks[] = {
+	{ .num_channels = 1, .type = INTEL_DRAM_LPDDR4, .page_mask = 0x1 },
+	{ .num_channels = 1, .type = INTEL_DRAM_DDR4,   .page_mask = 0x1 },
+	{ .num_channels = 2, .type = INTEL_DRAM_LPDDR4, .page_mask = 0x3 },
+	{ .num_channels = 2, .type = INTEL_DRAM_DDR4,   .page_mask = 0x3 },
+	{}
+};
+
+static void tgl_bw_buddy_init(struct drm_i915_private *dev_priv)
+{
+	enum intel_dram_type type = dev_priv->dram_info.type;
+	u8 num_channels = dev_priv->dram_info.num_channels;
+	const struct buddy_page_mask *table;
+	int i;
+
+	if (IS_TGL_REVID(dev_priv, TGL_REVID_A0, TGL_REVID_A0))
+		/* Wa_1409767108: tgl */
+		table = wa_1409767108_buddy_page_masks;
+	else
+		table = tgl_buddy_page_masks;
+
+	for (i = 0; table[i].page_mask != 0; i++)
+		if (table[i].num_channels == num_channels &&
+		    table[i].type == type)
+			break;
+
+	if (table[i].page_mask == 0) {
+		DRM_DEBUG_DRIVER("Unknown memory configuration; disabling address buddy logic.\n");
+		I915_WRITE(BW_BUDDY1_CTL, BW_BUDDY_DISABLE);
+		I915_WRITE(BW_BUDDY2_CTL, BW_BUDDY_DISABLE);
+	} else {
+		I915_WRITE(BW_BUDDY1_PAGE_MASK, table[i].page_mask);
+		I915_WRITE(BW_BUDDY2_PAGE_MASK, table[i].page_mask);
+	}
+}
+
 static void icl_display_core_init(struct drm_i915_private *dev_priv,
 				  bool resume)
 {
@@ -4813,6 +5005,10 @@ static void icl_display_core_init(struct drm_i915_private *dev_priv,
 	/* 6. Setup MBUS. */
 	icl_mbus_init(dev_priv);
 
+	/* 7. Program arbiter BW_BUDDY registers */
+	if (INTEL_GEN(dev_priv) >= 12)
+		tgl_bw_buddy_init(dev_priv);
+
 	if (resume && dev_priv->csr.dmc_payload)
 		intel_csr_load_program(dev_priv);
 }
diff --git a/drivers/gpu/drm/i915/display/intel_display_power.h b/drivers/gpu/drm/i915/display/intel_display_power.h
index 1da04f3e0fb3..2608a65af7fa 100644
--- a/drivers/gpu/drm/i915/display/intel_display_power.h
+++ b/drivers/gpu/drm/i915/display/intel_display_power.h
@@ -28,7 +28,7 @@ enum intel_display_power_domain {
 	POWER_DOMAIN_TRANSCODER_C,
 	POWER_DOMAIN_TRANSCODER_D,
 	POWER_DOMAIN_TRANSCODER_EDP,
-	/* VDSC/joining for TRANSCODER_EDP (ICL) or TRANSCODER_A (TGL) */
+	/* VDSC/joining for eDP/DSI transcoder (ICL) or pipe A (TGL) */
 	POWER_DOMAIN_TRANSCODER_VDSC_PW2,
 	POWER_DOMAIN_TRANSCODER_DSI_A,
 	POWER_DOMAIN_TRANSCODER_DSI_C,
diff --git a/drivers/gpu/drm/i915/display/intel_display_types.h b/drivers/gpu/drm/i915/display/intel_display_types.h
index 1a7334dbe802..83ea04149b77 100644
--- a/drivers/gpu/drm/i915/display/intel_display_types.h
+++ b/drivers/gpu/drm/i915/display/intel_display_types.h
@@ -523,7 +523,24 @@ struct intel_atomic_state {
 };
 
 struct intel_plane_state {
-	struct drm_plane_state base;
+	struct drm_plane_state uapi;
+
+	/*
+	 * actual hardware state, the state we program to the hardware.
+	 * The following members are used to verify the hardware state:
+	 * During initial hw readout, they need to be copied from uapi.
+	 */
+	struct {
+		struct drm_crtc *crtc;
+		struct drm_framebuffer *fb;
+
+		u16 alpha;
+		uint16_t pixel_blend_mode;
+		unsigned int rotation;
+		enum drm_color_encoding color_encoding;
+		enum drm_color_range color_range;
+	} hw;
+
 	struct i915_ggtt_view view;
 	struct i915_vma *vma;
 	unsigned long flags;
@@ -546,6 +563,9 @@ struct intel_plane_state {
 	/* plane color control register */
 	u32 color_ctl;
 
+	/* chroma upsampler control register */
+	u32 cus_ctl;
+
 	/*
 	 * scaler_id
 	 *    = -1 : not using a scaler
@@ -757,7 +777,33 @@ enum intel_output_format {
 };
 
 struct intel_crtc_state {
-	struct drm_crtc_state base;
+	/*
+	 * uapi (drm) state. This is the software state shown to userspace.
+	 * In particular, the following members are used for bookkeeping:
+	 * - crtc
+	 * - state
+	 * - *_changed
+	 * - event
+	 * - commit
+	 * - mode_blob
+	 */
+	struct drm_crtc_state uapi;
+
+	/*
+	 * actual hardware state, the state we program to the hardware.
+	 * The following members are used to verify the hardware state:
+	 * - enable
+	 * - active
+	 * - mode / adjusted_mode
+	 * - color property blobs.
+	 *
+	 * During initial hw readout, they need to be copied to uapi.
+	 */
+	struct {
+		bool active, enable;
+		struct drm_property_blob *degamma_lut, *gamma_lut, *ctm;
+		struct drm_display_mode mode, adjusted_mode;
+	} hw;
 
 	/**
 	 * quirks - bitfield with hw state readout quirks
@@ -1080,9 +1126,6 @@ struct intel_plane {
 	void (*update_plane)(struct intel_plane *plane,
 			     const struct intel_crtc_state *crtc_state,
 			     const struct intel_plane_state *plane_state);
-	void (*update_slave)(struct intel_plane *plane,
-			     const struct intel_crtc_state *crtc_state,
-			     const struct intel_plane_state *plane_state);
 	void (*disable_plane)(struct intel_plane *plane,
 			      const struct intel_crtc_state *crtc_state);
 	bool (*get_hw_state)(struct intel_plane *plane, enum pipe *pipe);
@@ -1113,12 +1156,12 @@ struct cxsr_latency {
 
 #define to_intel_atomic_state(x) container_of(x, struct intel_atomic_state, base)
 #define to_intel_crtc(x) container_of(x, struct intel_crtc, base)
-#define to_intel_crtc_state(x) container_of(x, struct intel_crtc_state, base)
+#define to_intel_crtc_state(x) container_of(x, struct intel_crtc_state, uapi)
 #define to_intel_connector(x) container_of(x, struct intel_connector, base)
 #define to_intel_encoder(x) container_of(x, struct intel_encoder, base)
 #define to_intel_framebuffer(x) container_of(x, struct intel_framebuffer, base)
 #define to_intel_plane(x) container_of(x, struct intel_plane, base)
-#define to_intel_plane_state(x) container_of(x, struct intel_plane_state, base)
+#define to_intel_plane_state(x) container_of(x, struct intel_plane_state, uapi)
 #define intel_fb_obj(x) ((x) ? to_intel_bo((x)->obj[0]) : NULL)
 
 struct intel_hdmi {
@@ -1528,6 +1571,24 @@ intel_atomic_get_new_crtc_state(struct intel_atomic_state *state,
 								 &crtc->base));
 }
 
+static inline struct intel_digital_connector_state *
+intel_atomic_get_new_connector_state(struct intel_atomic_state *state,
+				     struct intel_connector *connector)
+{
+	return to_intel_digital_connector_state(
+			drm_atomic_get_new_connector_state(&state->base,
+			&connector->base));
+}
+
+static inline struct intel_digital_connector_state *
+intel_atomic_get_old_connector_state(struct intel_atomic_state *state,
+				     struct intel_connector *connector)
+{
+	return to_intel_digital_connector_state(
+			drm_atomic_get_old_connector_state(&state->base,
+			&connector->base));
+}
+
 /* intel_display.c */
 static inline bool
 intel_crtc_has_type(const struct intel_crtc_state *crtc_state,
diff --git a/drivers/gpu/drm/i915/display/intel_dp.c b/drivers/gpu/drm/i915/display/intel_dp.c
index 7cbdac14d34b..2f31d226c6eb 100644
--- a/drivers/gpu/drm/i915/display/intel_dp.c
+++ b/drivers/gpu/drm/i915/display/intel_dp.c
@@ -1889,32 +1889,15 @@ static bool intel_dp_supports_fec(struct intel_dp *intel_dp,
 		drm_dp_sink_supports_fec(intel_dp->fec_capable);
 }
 
-static bool intel_dp_source_supports_dsc(struct intel_dp *intel_dp,
-					 const struct intel_crtc_state *pipe_config)
-{
-	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
-
-	if (!INTEL_INFO(dev_priv)->display.has_dsc)
-		return false;
-
-	/* On TGL, DSC is supported on all Pipes */
-	if (INTEL_GEN(dev_priv) >= 12)
-		return true;
-
-	if (INTEL_GEN(dev_priv) >= 10 &&
-	    pipe_config->cpu_transcoder != TRANSCODER_A)
-		return true;
-
-	return false;
-}
-
 static bool intel_dp_supports_dsc(struct intel_dp *intel_dp,
-				  const struct intel_crtc_state *pipe_config)
+				  const struct intel_crtc_state *crtc_state)
 {
-	if (!intel_dp_is_edp(intel_dp) && !pipe_config->fec_enable)
+	struct intel_encoder *encoder = &dp_to_dig_port(intel_dp)->base;
+
+	if (!intel_dp_is_edp(intel_dp) && !crtc_state->fec_enable)
 		return false;
 
-	return intel_dp_source_supports_dsc(intel_dp, pipe_config) &&
+	return intel_dsc_source_support(encoder, crtc_state) &&
 		drm_dp_sink_supports_dsc(intel_dp->dsc_dpcd);
 }
 
@@ -1999,7 +1982,7 @@ intel_dp_compute_link_config_wide(struct intel_dp *intel_dp,
 				  struct intel_crtc_state *pipe_config,
 				  const struct link_config_limits *limits)
 {
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	int bpp, clock, lane_count;
 	int mode_rate, link_clock, link_avail;
 
@@ -2046,6 +2029,63 @@ static int intel_dp_dsc_compute_bpp(struct intel_dp *intel_dp, u8 dsc_max_bpc)
 	return 0;
 }
 
+#define DSC_SUPPORTED_VERSION_MIN		1
+
+static int intel_dp_dsc_compute_params(struct intel_encoder *encoder,
+				       struct intel_crtc_state *crtc_state)
+{
+	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
+	struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
+	u8 line_buf_depth;
+	int ret;
+
+	ret = intel_dsc_compute_params(encoder, crtc_state);
+	if (ret)
+		return ret;
+
+	/*
+	 * Slice Height of 8 works for all currently available panels. So start
+	 * with that if pic_height is an integral multiple of 8. Eventually add
+	 * logic to try multiple slice heights.
+	 */
+	if (vdsc_cfg->pic_height % 8 == 0)
+		vdsc_cfg->slice_height = 8;
+	else if (vdsc_cfg->pic_height % 4 == 0)
+		vdsc_cfg->slice_height = 4;
+	else
+		vdsc_cfg->slice_height = 2;
+
+	vdsc_cfg->dsc_version_major =
+		(intel_dp->dsc_dpcd[DP_DSC_REV - DP_DSC_SUPPORT] &
+		 DP_DSC_MAJOR_MASK) >> DP_DSC_MAJOR_SHIFT;
+	vdsc_cfg->dsc_version_minor =
+		min(DSC_SUPPORTED_VERSION_MIN,
+		    (intel_dp->dsc_dpcd[DP_DSC_REV - DP_DSC_SUPPORT] &
+		     DP_DSC_MINOR_MASK) >> DP_DSC_MINOR_SHIFT);
+
+	vdsc_cfg->convert_rgb = intel_dp->dsc_dpcd[DP_DSC_DEC_COLOR_FORMAT_CAP - DP_DSC_SUPPORT] &
+		DP_DSC_RGB;
+
+	line_buf_depth = drm_dp_dsc_sink_line_buf_depth(intel_dp->dsc_dpcd);
+	if (!line_buf_depth) {
+		DRM_DEBUG_KMS("DSC Sink Line Buffer Depth invalid\n");
+		return -EINVAL;
+	}
+
+	if (vdsc_cfg->dsc_version_minor == 2)
+		vdsc_cfg->line_buf_depth = (line_buf_depth == DSC_1_2_MAX_LINEBUF_DEPTH_BITS) ?
+			DSC_1_2_MAX_LINEBUF_DEPTH_VAL : line_buf_depth;
+	else
+		vdsc_cfg->line_buf_depth = (line_buf_depth > DSC_1_1_MAX_LINEBUF_DEPTH_BITS) ?
+			DSC_1_1_MAX_LINEBUF_DEPTH_BITS : line_buf_depth;
+
+	vdsc_cfg->block_pred_enable =
+		intel_dp->dsc_dpcd[DP_DSC_BLK_PREDICTION_SUPPORT - DP_DSC_SUPPORT] &
+		DP_DSC_BLK_PREDICTION_IS_SUPPORTED;
+
+	return drm_dsc_compute_rc_parameters(vdsc_cfg);
+}
+
 static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 				       struct intel_crtc_state *pipe_config,
 				       struct drm_connector_state *conn_state,
@@ -2053,7 +2093,7 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 {
 	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
 	struct drm_i915_private *dev_priv = to_i915(dig_port->base.base.dev);
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	u8 dsc_max_bpc;
 	int pipe_bpp;
 	int ret;
@@ -2132,7 +2172,7 @@ static int intel_dp_dsc_compute_config(struct intel_dp *intel_dp,
 		}
 	}
 
-	ret = intel_dp_compute_dsc_params(intel_dp, pipe_config);
+	ret = intel_dp_dsc_compute_params(&dig_port->base, pipe_config);
 	if (ret < 0) {
 		DRM_DEBUG_KMS("Cannot compute valid DSC parameters for Input Bpp = %d "
 			      "Compressed BPP = %d\n",
@@ -2164,7 +2204,7 @@ intel_dp_compute_link_config(struct intel_encoder *encoder,
 			     struct intel_crtc_state *pipe_config,
 			     struct drm_connector_state *conn_state)
 {
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct link_config_limits limits;
 	int common_len;
@@ -2252,8 +2292,8 @@ intel_dp_ycbcr420_config(struct intel_dp *intel_dp,
 {
 	const struct drm_display_info *info = &connector->display_info;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+		&crtc_state->hw.adjusted_mode;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	int ret;
 
 	if (!drm_mode_is_420_only(info, adjusted_mode) ||
@@ -2281,7 +2321,7 @@ bool intel_dp_limited_color_range(const struct intel_crtc_state *crtc_state,
 	const struct intel_digital_connector_state *intel_conn_state =
 		to_intel_digital_connector_state(conn_state);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 
 	/*
 	 * Our YCbCr output is always limited range.
@@ -2308,17 +2348,28 @@ bool intel_dp_limited_color_range(const struct intel_crtc_state *crtc_state,
 	}
 }
 
+static bool intel_dp_port_has_audio(struct drm_i915_private *dev_priv,
+				    enum port port)
+{
+	if (IS_G4X(dev_priv))
+		return false;
+	if (INTEL_GEN(dev_priv) < 12 && port == PORT_A)
+		return false;
+
+	return true;
+}
+
 int
 intel_dp_compute_config(struct intel_encoder *encoder,
 			struct intel_crtc_state *pipe_config,
 			struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct intel_lspcon *lspcon = enc_to_intel_lspcon(&encoder->base);
 	enum port port = encoder->port;
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
 	struct intel_digital_connector_state *intel_conn_state =
 		to_intel_digital_connector_state(conn_state);
@@ -2341,7 +2392,7 @@ intel_dp_compute_config(struct intel_encoder *encoder,
 		return ret;
 
 	pipe_config->has_drrs = false;
-	if (IS_G4X(dev_priv) || port == PORT_A)
+	if (!intel_dp_port_has_audio(dev_priv, port))
 		pipe_config->has_audio = false;
 	else if (intel_conn_state->force_audio == HDMI_AUDIO_AUTO)
 		pipe_config->has_audio = intel_dp->has_audio;
@@ -2433,8 +2484,8 @@ static void intel_dp_prepare(struct intel_encoder *encoder,
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	enum port port = encoder->port;
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 
 	intel_dp_set_link_params(intel_dp, pipe_config->port_clock,
 				 pipe_config->lane_count,
@@ -3031,7 +3082,7 @@ static void assert_edp_pll(struct drm_i915_private *dev_priv, bool state)
 static void ironlake_edp_pll_on(struct intel_dp *intel_dp,
 				const struct intel_crtc_state *pipe_config)
 {
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	assert_pipe_disabled(dev_priv, crtc->pipe);
@@ -3071,7 +3122,7 @@ static void ironlake_edp_pll_on(struct intel_dp *intel_dp,
 static void ironlake_edp_pll_off(struct intel_dp *intel_dp,
 				 const struct intel_crtc_state *old_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	assert_pipe_disabled(dev_priv, crtc->pipe);
@@ -3231,7 +3282,7 @@ static void intel_dp_get_config(struct intel_encoder *encoder,
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	u32 tmp, flags = 0;
 	enum port port = encoder->port;
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 
 	if (encoder->type == INTEL_OUTPUT_EDP)
 		pipe_config->output_types |= BIT(INTEL_OUTPUT_EDP);
@@ -3266,7 +3317,7 @@ static void intel_dp_get_config(struct intel_encoder *encoder,
 			flags |= DRM_MODE_FLAG_NVSYNC;
 	}
 
-	pipe_config->base.adjusted_mode.flags |= flags;
+	pipe_config->hw.adjusted_mode.flags |= flags;
 
 	if (IS_G4X(dev_priv) && tmp & DP_COLOR_RANGE_16_235)
 		pipe_config->limited_color_range = true;
@@ -3283,7 +3334,7 @@ static void intel_dp_get_config(struct intel_encoder *encoder,
 			pipe_config->port_clock = 270000;
 	}
 
-	pipe_config->base.adjusted_mode.crtc_clock =
+	pipe_config->hw.adjusted_mode.crtc_clock =
 		intel_dotclock_calculate(pipe_config->port_clock,
 					 &pipe_config->dp_m_n);
 
@@ -3498,7 +3549,7 @@ static void intel_enable_dp(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	u32 dp_reg = I915_READ(intel_dp->output_reg);
 	enum pipe pipe = crtc->pipe;
 	intel_wakeref_t wakeref;
@@ -3631,7 +3682,7 @@ static void vlv_init_panel_power_sequencer(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 
 	lockdep_assert_held(&dev_priv->pps_mutex);
 
@@ -4153,7 +4204,7 @@ intel_dp_link_down(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	enum port port = encoder->port;
 	u32 DP = intel_dp->DP;
 
@@ -5076,7 +5127,7 @@ int intel_dp_retrain_link(struct intel_encoder *encoder,
 
 	WARN_ON(!intel_crtc_has_dp_encoder(crtc_state));
 
-	if (!crtc_state->base.active)
+	if (!crtc_state->hw.active)
 		return 0;
 
 	if (conn_state->commit &&
@@ -5482,7 +5533,7 @@ static bool intel_combo_phy_connected(struct drm_i915_private *dev_priv,
 	return I915_READ(SDEISR) & SDE_DDI_HOTPLUG_ICP(phy);
 }
 
-static bool icl_digital_port_connected(struct intel_encoder *encoder)
+static bool icp_digital_port_connected(struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_digital_port *dig_port = enc_to_dig_port(&encoder->base);
@@ -5520,9 +5571,9 @@ static bool __intel_digital_port_connected(struct intel_encoder *encoder)
 			return g4x_digital_port_connected(encoder);
 	}
 
-	if (INTEL_GEN(dev_priv) >= 11)
-		return icl_digital_port_connected(encoder);
-	else if (IS_GEN(dev_priv, 10) || IS_GEN9_BC(dev_priv))
+	if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP)
+		return icp_digital_port_connected(encoder);
+	else if (INTEL_PCH_TYPE(dev_priv) >= PCH_SPT)
 		return spt_digital_port_connected(encoder);
 	else if (IS_GEN9_LP(dev_priv))
 		return bxt_digital_port_connected(encoder);
@@ -6909,7 +6960,7 @@ static void intel_dp_set_drrs_state(struct drm_i915_private *dev_priv,
 				    int refresh_rate)
 {
 	struct intel_dp *intel_dp = dev_priv->drrs.dp;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum drrs_refresh_rate_type index = DRRS_HIGH_RR;
 
 	if (refresh_rate <= 0) {
@@ -6942,7 +6993,7 @@ static void intel_dp_set_drrs_state(struct drm_i915_private *dev_priv,
 		return;
 	}
 
-	if (!crtc_state->base.active) {
+	if (!crtc_state->hw.active) {
 		DRM_DEBUG_KMS("eDP encoder disabled. CRTC not Active\n");
 		return;
 	}
diff --git a/drivers/gpu/drm/i915/display/intel_dp_mst.c b/drivers/gpu/drm/i915/display/intel_dp_mst.c
index 03d1cba0b696..7aa0975c33b7 100644
--- a/drivers/gpu/drm/i915/display/intel_dp_mst.c
+++ b/drivers/gpu/drm/i915/display/intel_dp_mst.c
@@ -42,13 +42,13 @@ static int intel_dp_mst_compute_link_config(struct intel_encoder *encoder,
 					    struct drm_connector_state *conn_state,
 					    struct link_config_limits *limits)
 {
-	struct drm_atomic_state *state = crtc_state->base.state;
+	struct drm_atomic_state *state = crtc_state->uapi.state;
 	struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
 	struct intel_dp *intel_dp = &intel_mst->primary->dp;
 	struct intel_connector *connector =
 		to_intel_connector(conn_state->connector);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	void *port = connector->port;
 	bool constant_n = drm_dp_has_quirk(&intel_dp->desc,
 					   DP_DPCD_QUIRK_CONSTANT_N);
@@ -99,7 +99,7 @@ static int intel_dp_mst_compute_config(struct intel_encoder *encoder,
 	struct intel_digital_connector_state *intel_conn_state =
 		to_intel_digital_connector_state(conn_state);
 	const struct drm_display_mode *adjusted_mode =
-		&pipe_config->base.adjusted_mode;
+		&pipe_config->hw.adjusted_mode;
 	void *port = connector->port;
 	struct link_config_limits limits;
 	int ret;
@@ -168,7 +168,6 @@ intel_dp_mst_atomic_check(struct drm_connector *connector,
 	struct intel_connector *intel_connector =
 		to_intel_connector(connector);
 	struct drm_crtc *new_crtc = new_conn_state->crtc;
-	struct drm_crtc_state *crtc_state;
 	struct drm_dp_mst_topology_mgr *mgr;
 	int ret;
 
@@ -183,11 +182,16 @@ intel_dp_mst_atomic_check(struct drm_connector *connector,
 	 * connector
 	 */
 	if (new_crtc) {
-		crtc_state = drm_atomic_get_new_crtc_state(state, new_crtc);
+		struct intel_atomic_state *intel_state =
+			to_intel_atomic_state(state);
+		struct intel_crtc *intel_crtc = to_intel_crtc(new_crtc);
+		struct intel_crtc_state *crtc_state =
+			intel_atomic_get_new_crtc_state(intel_state,
+							intel_crtc);
 
 		if (!crtc_state ||
-		    !drm_atomic_crtc_needs_modeset(crtc_state) ||
-		    crtc_state->enable)
+		    !drm_atomic_crtc_needs_modeset(&crtc_state->uapi) ||
+		    crtc_state->uapi.enable)
 			return 0;
 	}
 
@@ -231,8 +235,32 @@ static void intel_mst_post_disable_dp(struct intel_encoder *encoder,
 	struct intel_dp *intel_dp = &intel_dig_port->dp;
 	struct intel_connector *connector =
 		to_intel_connector(old_conn_state->connector);
+	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
+	bool last_mst_stream;
 
-	intel_ddi_disable_pipe_clock(old_crtc_state);
+	intel_dp->active_mst_links--;
+	last_mst_stream = intel_dp->active_mst_links == 0;
+
+	intel_crtc_vblank_off(old_crtc_state);
+
+	intel_disable_pipe(old_crtc_state);
+
+	intel_ddi_disable_transcoder_func(old_crtc_state);
+
+	if (INTEL_GEN(dev_priv) >= 9)
+		skylake_scaler_disable(old_crtc_state);
+	else
+		ironlake_pfit_disable(old_crtc_state);
+
+	/*
+	 * From TGL spec: "If multi-stream slave transcoder: Configure
+	 * Transcoder Clock Select to direct no clock to the transcoder"
+	 *
+	 * From older GENs spec: "Configure Transcoder Clock Select to direct
+	 * no clock to the transcoder"
+	 */
+	if (INTEL_GEN(dev_priv) < 12 || !last_mst_stream)
+		intel_ddi_disable_pipe_clock(old_crtc_state);
 
 	/* this can fail */
 	drm_dp_check_act_status(&intel_dp->mst_mgr);
@@ -248,14 +276,10 @@ static void intel_mst_post_disable_dp(struct intel_encoder *encoder,
 	drm_dp_send_power_updown_phy(&intel_dp->mst_mgr, connector->port,
 				     false);
 
-	intel_dp->active_mst_links--;
-
 	intel_mst->connector = NULL;
-	if (intel_dp->active_mst_links == 0) {
-		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_OFF);
+	if (last_mst_stream)
 		intel_dig_port->base.post_disable(&intel_dig_port->base,
 						  old_crtc_state, NULL);
-	}
 
 	DRM_DEBUG_KMS("active links %d\n", intel_dp->active_mst_links);
 }
@@ -273,20 +297,6 @@ static void intel_mst_pre_pll_enable_dp(struct intel_encoder *encoder,
 						    pipe_config, NULL);
 }
 
-static void intel_mst_post_pll_disable_dp(struct intel_encoder *encoder,
-					  const struct intel_crtc_state *old_crtc_state,
-					  const struct drm_connector_state *old_conn_state)
-{
-	struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base);
-	struct intel_digital_port *intel_dig_port = intel_mst->primary;
-	struct intel_dp *intel_dp = &intel_dig_port->dp;
-
-	if (intel_dp->active_mst_links == 0)
-		intel_dig_port->base.post_pll_disable(&intel_dig_port->base,
-						      old_crtc_state,
-						      old_conn_state);
-}
-
 static void intel_mst_pre_enable_dp(struct intel_encoder *encoder,
 				    const struct intel_crtc_state *pipe_config,
 				    const struct drm_connector_state *conn_state)
@@ -299,21 +309,23 @@ static void intel_mst_pre_enable_dp(struct intel_encoder *encoder,
 		to_intel_connector(conn_state->connector);
 	int ret;
 	u32 temp;
+	bool first_mst_stream;
 
 	/* MST encoders are bound to a crtc, not to a connector,
 	 * force the mapping here for get_hw_state.
 	 */
 	connector->encoder = encoder;
 	intel_mst->connector = connector;
+	first_mst_stream = intel_dp->active_mst_links == 0;
 
 	DRM_DEBUG_KMS("active links %d\n", intel_dp->active_mst_links);
 
-	if (intel_dp->active_mst_links == 0)
+	if (first_mst_stream)
 		intel_dp_sink_dpms(intel_dp, DRM_MODE_DPMS_ON);
 
 	drm_dp_send_power_updown_phy(&intel_dp->mst_mgr, connector->port, true);
 
-	if (intel_dp->active_mst_links == 0)
+	if (first_mst_stream)
 		intel_dig_port->base.pre_enable(&intel_dig_port->base,
 						pipe_config, NULL);
 
@@ -330,7 +342,15 @@ static void intel_mst_pre_enable_dp(struct intel_encoder *encoder,
 
 	ret = drm_dp_update_payload_part1(&intel_dp->mst_mgr);
 
-	intel_ddi_enable_pipe_clock(pipe_config);
+	/*
+	 * Before Gen 12 this is not done as part of
+	 * intel_dig_port->base.pre_enable() and should be done here. For
+	 * Gen 12+ the step in which this should be done is different for the
+	 * first MST stream, so it's done on the DDI for the first stream and
+	 * here for the following ones.
+	 */
+	if (INTEL_GEN(dev_priv) < 12 || !first_mst_stream)
+		intel_ddi_enable_pipe_clock(pipe_config);
 
 	intel_ddi_set_dp_msa(pipe_config, conn_state);
 }
@@ -633,7 +653,6 @@ intel_dp_create_fake_mst_encoder(struct intel_digital_port *intel_dig_port, enum
 	intel_encoder->disable = intel_mst_disable_dp;
 	intel_encoder->post_disable = intel_mst_post_disable_dp;
 	intel_encoder->pre_pll_enable = intel_mst_pre_pll_enable_dp;
-	intel_encoder->post_pll_disable = intel_mst_post_pll_disable_dp;
 	intel_encoder->pre_enable = intel_mst_pre_enable_dp;
 	intel_encoder->enable = intel_mst_enable_dp;
 	intel_encoder->get_hw_state = intel_dp_mst_enc_get_hw_state;
diff --git a/drivers/gpu/drm/i915/display/intel_dpio_phy.c b/drivers/gpu/drm/i915/display/intel_dpio_phy.c
index 556d1b30f06a..704f38681c4b 100644
--- a/drivers/gpu/drm/i915/display/intel_dpio_phy.c
+++ b/drivers/gpu/drm/i915/display/intel_dpio_phy.c
@@ -739,7 +739,7 @@ void chv_data_lane_soft_reset(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	enum dpio_channel ch = vlv_dport_to_channel(enc_to_dig_port(&encoder->base));
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum pipe pipe = crtc->pipe;
 	u32 val;
 
@@ -783,7 +783,7 @@ void chv_phy_pre_pll_enable(struct intel_encoder *encoder,
 {
 	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum dpio_channel ch = vlv_dport_to_channel(dport);
 	enum pipe pipe = crtc->pipe;
 	unsigned int lane_mask =
@@ -864,7 +864,7 @@ void chv_phy_pre_encoder_enable(struct intel_encoder *encoder,
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum dpio_channel ch = vlv_dport_to_channel(dport);
 	enum pipe pipe = crtc->pipe;
 	int data, i, stagger;
@@ -953,7 +953,7 @@ void chv_phy_post_pll_disable(struct intel_encoder *encoder,
 			      const struct intel_crtc_state *old_crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum pipe pipe = to_intel_crtc(old_crtc_state->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(old_crtc_state->uapi.crtc)->pipe;
 	u32 val;
 
 	vlv_dpio_get(dev_priv);
@@ -1016,7 +1016,7 @@ void vlv_phy_pre_pll_enable(struct intel_encoder *encoder,
 {
 	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum dpio_channel port = vlv_dport_to_channel(dport);
 	enum pipe pipe = crtc->pipe;
 
@@ -1046,7 +1046,7 @@ void vlv_phy_pre_encoder_enable(struct intel_encoder *encoder,
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct intel_digital_port *dport = dp_to_dig_port(intel_dp);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	enum dpio_channel port = vlv_dport_to_channel(dport);
 	enum pipe pipe = crtc->pipe;
 	u32 val;
@@ -1075,7 +1075,7 @@ void vlv_phy_reset_lanes(struct intel_encoder *encoder,
 {
 	struct intel_digital_port *dport = enc_to_dig_port(&encoder->base);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	enum dpio_channel port = vlv_dport_to_channel(dport);
 	enum pipe pipe = crtc->pipe;
 
diff --git a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
index 3ce0a023eee0..728a4b045de7 100644
--- a/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/display/intel_dpll_mgr.c
@@ -136,7 +136,7 @@ void assert_shared_dpll(struct drm_i915_private *dev_priv,
  */
 void intel_prepare_shared_dpll(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_shared_dpll *pll = crtc_state->shared_dpll;
 
@@ -163,7 +163,7 @@ void intel_prepare_shared_dpll(const struct intel_crtc_state *crtc_state)
  */
 void intel_enable_shared_dpll(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_shared_dpll *pll = crtc_state->shared_dpll;
 	unsigned int crtc_mask = drm_crtc_mask(&crtc->base);
@@ -208,7 +208,7 @@ out:
  */
 void intel_disable_shared_dpll(const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_shared_dpll *pll = crtc_state->shared_dpll;
 	unsigned int crtc_mask = drm_crtc_mask(&crtc->base);
@@ -842,7 +842,7 @@ hsw_ddi_hdmi_get_dpll(struct intel_atomic_state *state,
 static struct intel_shared_dpll *
 hsw_ddi_dp_get_dpll(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	struct intel_shared_dpll *pll;
 	enum intel_dpll_id pll_id;
 	int clock = crtc_state->port_clock;
@@ -1751,7 +1751,7 @@ static bool
 bxt_ddi_hdmi_pll_dividers(struct intel_crtc_state *crtc_state,
 			  struct bxt_clk_div *clk_div)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct dpll best_clock;
 
 	/* Calculate HDMI div */
@@ -2274,7 +2274,7 @@ static bool
 cnl_ddi_calculate_wrpll(struct intel_crtc_state *crtc_state,
 			struct skl_wrpll_params *wrpll_params)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	u32 afe_clock = crtc_state->port_clock * 5;
 	u32 ref_clock;
 	u32 dco_min = 7998000;
@@ -2553,7 +2553,7 @@ static const struct skl_wrpll_params tgl_tbt_pll_24MHz_values = {
 static bool icl_calc_dp_combo_pll(struct intel_crtc_state *crtc_state,
 				  struct skl_wrpll_params *pll_params)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	const struct icl_combo_pll_params *params =
 		dev_priv->cdclk.hw.ref == 24000 ?
 		icl_dp_combo_pll_24MHz_values :
@@ -2575,7 +2575,7 @@ static bool icl_calc_dp_combo_pll(struct intel_crtc_state *crtc_state,
 static bool icl_calc_tbt_pll(struct intel_crtc_state *crtc_state,
 			     struct skl_wrpll_params *pll_params)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	if (INTEL_GEN(dev_priv) >= 12) {
 		switch (dev_priv->cdclk.hw.ref) {
@@ -2612,7 +2612,7 @@ static bool icl_calc_dpll_state(struct intel_crtc_state *crtc_state,
 				struct intel_encoder *encoder,
 				struct intel_dpll_hw_state *pll_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	u32 cfgcr0, cfgcr1;
 	struct skl_wrpll_params pll_params = { 0 };
 	bool ret;
@@ -2744,7 +2744,7 @@ static bool icl_mg_pll_find_divisors(int clock_khz, bool is_dp, bool use_ssc,
 static bool icl_calc_mg_pll_state(struct intel_crtc_state *crtc_state,
 				  struct intel_dpll_hw_state *pll_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	int refclk_khz = dev_priv->cdclk.hw.ref;
 	int clock = crtc_state->port_clock;
 	u32 dco_khz, m1div, m2div_int, m2div_rem, m2div_frac;
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.c b/drivers/gpu/drm/i915/display/intel_dsb.c
index bb5a0e91b370..ada006a690df 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.c
+++ b/drivers/gpu/drm/i915/display/intel_dsb.c
@@ -102,43 +102,50 @@ intel_dsb_get(struct intel_crtc *crtc)
 	struct intel_dsb *dsb = &crtc->dsb;
 	struct drm_i915_gem_object *obj;
 	struct i915_vma *vma;
+	u32 *buf;
 	intel_wakeref_t wakeref;
 
 	if (!HAS_DSB(i915))
 		return dsb;
 
-	if (atomic_add_return(1, &dsb->refcount) != 1)
+	if (dsb->refcount++ != 0)
 		return dsb;
 
-	dsb->id = DSB1;
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 
 	obj = i915_gem_object_create_internal(i915, DSB_BUF_SIZE);
 	if (IS_ERR(obj)) {
 		DRM_ERROR("Gem object creation failed\n");
-		goto err;
+		goto out;
 	}
 
-	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, PIN_MAPPABLE);
+	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
 	if (IS_ERR(vma)) {
 		DRM_ERROR("Vma creation failed\n");
 		i915_gem_object_put(obj);
-		atomic_dec(&dsb->refcount);
-		goto err;
+		goto out;
 	}
 
-	dsb->cmd_buf = i915_gem_object_pin_map(vma->obj, I915_MAP_WC);
-	if (IS_ERR(dsb->cmd_buf)) {
+	buf = i915_gem_object_pin_map(vma->obj, I915_MAP_WC);
+	if (IS_ERR(buf)) {
 		DRM_ERROR("Command buffer creation failed\n");
-		i915_vma_unpin_and_release(&vma, 0);
-		dsb->cmd_buf = NULL;
-		atomic_dec(&dsb->refcount);
-		goto err;
+		goto out;
 	}
+
+	dsb->id = DSB1;
 	dsb->vma = vma;
+	dsb->cmd_buf = buf;
+
+out:
+	/*
+	 * On error dsb->cmd_buf will continue to be NULL, making the writes
+	 * pass-through. Leave the dangling ref to be removed later by the
+	 * corresponding intel_dsb_put(): the important error message will
+	 * already be logged above.
+	 */
 
-err:
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+
 	return dsb;
 }
 
@@ -158,10 +165,10 @@ void intel_dsb_put(struct intel_dsb *dsb)
 	if (!HAS_DSB(i915))
 		return;
 
-	if (WARN_ON(atomic_read(&dsb->refcount) == 0))
+	if (WARN_ON(dsb->refcount == 0))
 		return;
 
-	if (atomic_dec_and_test(&dsb->refcount)) {
+	if (--dsb->refcount == 0) {
 		i915_vma_unpin_and_release(&dsb->vma, I915_VMA_RELEASE_MAP);
 		dsb->cmd_buf = NULL;
 		dsb->free_pos = 0;
diff --git a/drivers/gpu/drm/i915/display/intel_dsb.h b/drivers/gpu/drm/i915/display/intel_dsb.h
index 6f95c8e909e6..395ef9ce558e 100644
--- a/drivers/gpu/drm/i915/display/intel_dsb.h
+++ b/drivers/gpu/drm/i915/display/intel_dsb.h
@@ -22,7 +22,7 @@ enum dsb_id {
 };
 
 struct intel_dsb {
-	atomic_t refcount;
+	long refcount;
 	enum dsb_id id;
 	u32 *cmd_buf;
 	struct i915_vma *vma;
diff --git a/drivers/gpu/drm/i915/display/intel_dvo.c b/drivers/gpu/drm/i915/display/intel_dvo.c
index bcfbcb743e7d..a74dc5b915d1 100644
--- a/drivers/gpu/drm/i915/display/intel_dvo.c
+++ b/drivers/gpu/drm/i915/display/intel_dvo.c
@@ -178,9 +178,9 @@ static void intel_dvo_get_config(struct intel_encoder *encoder,
 	else
 		flags |= DRM_MODE_FLAG_NVSYNC;
 
-	pipe_config->base.adjusted_mode.flags |= flags;
+	pipe_config->hw.adjusted_mode.flags |= flags;
 
-	pipe_config->base.adjusted_mode.crtc_clock = pipe_config->port_clock;
+	pipe_config->hw.adjusted_mode.crtc_clock = pipe_config->port_clock;
 }
 
 static void intel_disable_dvo(struct intel_encoder *encoder,
@@ -207,8 +207,8 @@ static void intel_enable_dvo(struct intel_encoder *encoder,
 	u32 temp = I915_READ(dvo_reg);
 
 	intel_dvo->dev.dev_ops->mode_set(&intel_dvo->dev,
-					 &pipe_config->base.mode,
-					 &pipe_config->base.adjusted_mode);
+					 &pipe_config->hw.mode,
+					 &pipe_config->hw.adjusted_mode);
 
 	I915_WRITE(dvo_reg, temp | DVO_ENABLE);
 	I915_READ(dvo_reg);
@@ -253,7 +253,7 @@ static int intel_dvo_compute_config(struct intel_encoder *encoder,
 	struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
 	const struct drm_display_mode *fixed_mode =
 		intel_dvo->attached_connector->panel.fixed_mode;
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 
 	/*
 	 * If we have timings from the BIOS for the panel, put them in
@@ -277,8 +277,8 @@ static void intel_dvo_pre_enable(struct intel_encoder *encoder,
 				 const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	struct intel_dvo *intel_dvo = enc_to_dvo(encoder);
 	enum pipe pipe = crtc->pipe;
 	u32 dvo_val;
diff --git a/drivers/gpu/drm/i915/display/intel_fbc.c b/drivers/gpu/drm/i915/display/intel_fbc.c
index 20616639b8ab..a1048ece541e 100644
--- a/drivers/gpu/drm/i915/display/intel_fbc.c
+++ b/drivers/gpu/drm/i915/display/intel_fbc.c
@@ -50,11 +50,6 @@ static inline bool fbc_supported(struct drm_i915_private *dev_priv)
 	return HAS_FBC(dev_priv);
 }
 
-static inline bool no_fbc_on_multiple_pipes(struct drm_i915_private *dev_priv)
-{
-	return INTEL_GEN(dev_priv) <= 3;
-}
-
 /*
  * In some platforms where the CRTC's x:0/y:0 coordinates doesn't match the
  * frontbuffer's x:0/y:0 coordinates we lie to the hardware about the plane's
@@ -73,7 +68,7 @@ static unsigned int get_crtc_fence_y_offset(struct intel_fbc *fbc)
  * write to the PLANE_SIZE register. For BDW-, the hardware looks at the value
  * we wrote to PIPESRC.
  */
-static void intel_fbc_get_plane_source_size(struct intel_fbc_state_cache *cache,
+static void intel_fbc_get_plane_source_size(const struct intel_fbc_state_cache *cache,
 					    int *width, int *height)
 {
 	if (width)
@@ -83,7 +78,7 @@ static void intel_fbc_get_plane_source_size(struct intel_fbc_state_cache *cache,
 }
 
 static int intel_fbc_calculate_cfb_size(struct drm_i915_private *dev_priv,
-					struct intel_fbc_state_cache *cache)
+					const struct intel_fbc_state_cache *cache)
 {
 	int lines;
 
@@ -143,8 +138,10 @@ static void i8xx_fbc_activate(struct drm_i915_private *dev_priv)
 		u32 fbc_ctl2;
 
 		/* Set it up... */
-		fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM | FBC_CTL_CPU_FENCE;
+		fbc_ctl2 = FBC_CTL_FENCE_DBL | FBC_CTL_IDLE_IMM;
 		fbc_ctl2 |= FBC_CTL_PLANE(params->crtc.i9xx_plane);
+		if (params->fence_id >= 0)
+			fbc_ctl2 |= FBC_CTL_CPU_FENCE;
 		I915_WRITE(FBC_CONTROL2, fbc_ctl2);
 		I915_WRITE(FBC_FENCE_OFF, params->crtc.fence_y_offset);
 	}
@@ -156,7 +153,8 @@ static void i8xx_fbc_activate(struct drm_i915_private *dev_priv)
 	if (IS_I945GM(dev_priv))
 		fbc_ctl |= FBC_CTL_C3_IDLE; /* 945 needs special SR handling */
 	fbc_ctl |= (cfb_pitch & 0xff) << FBC_CTL_STRIDE_SHIFT;
-	fbc_ctl |= params->vma->fence->id;
+	if (params->fence_id >= 0)
+		fbc_ctl |= params->fence_id;
 	I915_WRITE(FBC_CONTROL, fbc_ctl);
 }
 
@@ -176,8 +174,8 @@ static void g4x_fbc_activate(struct drm_i915_private *dev_priv)
 	else
 		dpfc_ctl |= DPFC_CTL_LIMIT_1X;
 
-	if (params->flags & PLANE_HAS_FENCE) {
-		dpfc_ctl |= DPFC_CTL_FENCE_EN | params->vma->fence->id;
+	if (params->fence_id >= 0) {
+		dpfc_ctl |= DPFC_CTL_FENCE_EN | params->fence_id;
 		I915_WRITE(DPFC_FENCE_YOFF, params->crtc.fence_y_offset);
 	} else {
 		I915_WRITE(DPFC_FENCE_YOFF, 0);
@@ -234,14 +232,14 @@ static void ilk_fbc_activate(struct drm_i915_private *dev_priv)
 		break;
 	}
 
-	if (params->flags & PLANE_HAS_FENCE) {
+	if (params->fence_id >= 0) {
 		dpfc_ctl |= DPFC_CTL_FENCE_EN;
 		if (IS_GEN(dev_priv, 5))
-			dpfc_ctl |= params->vma->fence->id;
+			dpfc_ctl |= params->fence_id;
 		if (IS_GEN(dev_priv, 6)) {
 			I915_WRITE(SNB_DPFC_CTL_SA,
 				   SNB_CPU_FENCE_ENABLE |
-				   params->vma->fence->id);
+				   params->fence_id);
 			I915_WRITE(DPFC_CPU_FENCE_OFFSET,
 				   params->crtc.fence_y_offset);
 		}
@@ -253,8 +251,6 @@ static void ilk_fbc_activate(struct drm_i915_private *dev_priv)
 	}
 
 	I915_WRITE(ILK_DPFC_FENCE_YOFF, params->crtc.fence_y_offset);
-	I915_WRITE(ILK_FBC_RT_BASE,
-		   i915_ggtt_offset(params->vma) | ILK_FBC_RT_VALID);
 	/* enable it... */
 	I915_WRITE(ILK_DPFC_CONTROL, dpfc_ctl | DPFC_CTL_EN);
 
@@ -285,13 +281,12 @@ static void gen7_fbc_activate(struct drm_i915_private *dev_priv)
 	int threshold = dev_priv->fbc.threshold;
 
 	/* Display WA #0529: skl, kbl, bxt. */
-	if (IS_GEN(dev_priv, 9) && !IS_GEMINILAKE(dev_priv)) {
+	if (IS_GEN9_BC(dev_priv) || IS_BROXTON(dev_priv)) {
 		u32 val = I915_READ(CHICKEN_MISC_4);
 
 		val &= ~(FBC_STRIDE_OVERRIDE | FBC_STRIDE_MASK);
 
-		if (i915_gem_object_get_tiling(params->vma->obj) !=
-		    I915_TILING_X)
+		if (params->gen9_wa_cfb_stride)
 			val |= FBC_STRIDE_OVERRIDE | params->gen9_wa_cfb_stride;
 
 		I915_WRITE(CHICKEN_MISC_4, val);
@@ -317,11 +312,11 @@ static void gen7_fbc_activate(struct drm_i915_private *dev_priv)
 		break;
 	}
 
-	if (params->flags & PLANE_HAS_FENCE) {
+	if (params->fence_id >= 0) {
 		dpfc_ctl |= IVB_DPFC_CTL_FENCE_EN;
 		I915_WRITE(SNB_DPFC_CTL_SA,
 			   SNB_CPU_FENCE_ENABLE |
-			   params->vma->fence->id);
+			   params->fence_id);
 		I915_WRITE(DPFC_CPU_FENCE_OFFSET, params->crtc.fence_y_offset);
 	} else {
 		I915_WRITE(SNB_DPFC_CTL_SA,0);
@@ -367,6 +362,7 @@ static void intel_fbc_hw_activate(struct drm_i915_private *dev_priv)
 	struct intel_fbc *fbc = &dev_priv->fbc;
 
 	fbc->active = true;
+	fbc->activated = true;
 
 	if (INTEL_GEN(dev_priv) >= 7)
 		gen7_fbc_activate(dev_priv);
@@ -419,29 +415,10 @@ static void intel_fbc_deactivate(struct drm_i915_private *dev_priv,
 	fbc->no_fbc_reason = reason;
 }
 
-static bool multiple_pipes_ok(struct intel_crtc *crtc,
-			      struct intel_plane_state *plane_state)
-{
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	struct intel_fbc *fbc = &dev_priv->fbc;
-	enum pipe pipe = crtc->pipe;
-
-	/* Don't even bother tracking anything we don't need. */
-	if (!no_fbc_on_multiple_pipes(dev_priv))
-		return true;
-
-	if (plane_state->base.visible)
-		fbc->visible_pipes_mask |= (1 << pipe);
-	else
-		fbc->visible_pipes_mask &= ~(1 << pipe);
-
-	return (fbc->visible_pipes_mask & ~(1 << pipe)) != 0;
-}
-
 static int find_compression_threshold(struct drm_i915_private *dev_priv,
 				      struct drm_mm_node *node,
-				      int size,
-				      int fb_cpp)
+				      unsigned int size,
+				      unsigned int fb_cpp)
 {
 	int compression_threshold = 1;
 	int ret;
@@ -487,18 +464,15 @@ again:
 	}
 }
 
-static int intel_fbc_alloc_cfb(struct intel_crtc *crtc)
+static int intel_fbc_alloc_cfb(struct drm_i915_private *dev_priv,
+			       unsigned int size, unsigned int fb_cpp)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_fbc *fbc = &dev_priv->fbc;
 	struct drm_mm_node *uninitialized_var(compressed_llb);
-	int size, fb_cpp, ret;
+	int ret;
 
 	WARN_ON(drm_mm_node_allocated(&fbc->compressed_fb));
 
-	size = intel_fbc_calculate_cfb_size(dev_priv, &fbc->state_cache);
-	fb_cpp = fbc->state_cache.fb.format->cpp[0];
-
 	ret = find_compression_threshold(dev_priv, &fbc->compressed_fb,
 					 size, fb_cpp);
 	if (!ret)
@@ -656,46 +630,55 @@ static bool intel_fbc_hw_tracking_covers_screen(struct intel_crtc *crtc)
 }
 
 static void intel_fbc_update_state_cache(struct intel_crtc *crtc,
-					 struct intel_crtc_state *crtc_state,
-					 struct intel_plane_state *plane_state)
+					 const struct intel_crtc_state *crtc_state,
+					 const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_fbc *fbc = &dev_priv->fbc;
 	struct intel_fbc_state_cache *cache = &fbc->state_cache;
-	struct drm_framebuffer *fb = plane_state->base.fb;
+	struct drm_framebuffer *fb = plane_state->hw.fb;
 
-	cache->vma = NULL;
-	cache->flags = 0;
+	cache->plane.visible = plane_state->uapi.visible;
+	if (!cache->plane.visible)
+		return;
 
-	cache->crtc.mode_flags = crtc_state->base.adjusted_mode.flags;
+	cache->crtc.mode_flags = crtc_state->hw.adjusted_mode.flags;
 	if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv))
 		cache->crtc.hsw_bdw_pixel_rate = crtc_state->pixel_rate;
 
-	cache->plane.rotation = plane_state->base.rotation;
+	cache->plane.rotation = plane_state->hw.rotation;
 	/*
 	 * Src coordinates are already rotated by 270 degrees for
 	 * the 90/270 degree plane rotation cases (to match the
 	 * GTT mapping), hence no need to account for rotation here.
 	 */
-	cache->plane.src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	cache->plane.src_h = drm_rect_height(&plane_state->base.src) >> 16;
-	cache->plane.visible = plane_state->base.visible;
+	cache->plane.src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	cache->plane.src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
 	cache->plane.adjusted_x = plane_state->color_plane[0].x;
 	cache->plane.adjusted_y = plane_state->color_plane[0].y;
-	cache->plane.y = plane_state->base.src.y1 >> 16;
+	cache->plane.y = plane_state->uapi.src.y1 >> 16;
 
-	cache->plane.pixel_blend_mode = plane_state->base.pixel_blend_mode;
-
-	if (!cache->plane.visible)
-		return;
+	cache->plane.pixel_blend_mode = plane_state->hw.pixel_blend_mode;
 
 	cache->fb.format = fb->format;
 	cache->fb.stride = fb->pitches[0];
 
-	cache->vma = plane_state->vma;
-	cache->flags = plane_state->flags;
-	if (WARN_ON(cache->flags & PLANE_HAS_FENCE && !cache->vma->fence))
-		cache->flags &= ~PLANE_HAS_FENCE;
+	WARN_ON(plane_state->flags & PLANE_HAS_FENCE &&
+		!plane_state->vma->fence);
+
+	if (plane_state->flags & PLANE_HAS_FENCE &&
+	    plane_state->vma->fence)
+		cache->fence_id = plane_state->vma->fence->id;
+	else
+		cache->fence_id = -1;
+}
+
+static bool intel_fbc_cfb_size_changed(struct drm_i915_private *dev_priv)
+{
+	struct intel_fbc *fbc = &dev_priv->fbc;
+
+	return intel_fbc_calculate_cfb_size(dev_priv, &fbc->state_cache) >
+		fbc->compressed_fb.size * fbc->threshold;
 }
 
 static bool intel_fbc_can_activate(struct intel_crtc *crtc)
@@ -704,6 +687,11 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 	struct intel_fbc *fbc = &dev_priv->fbc;
 	struct intel_fbc_state_cache *cache = &fbc->state_cache;
 
+	if (!cache->plane.visible) {
+		fbc->no_fbc_reason = "primary plane not visible";
+		return false;
+	}
+
 	/* We don't need to use a state cache here since this information is
 	 * global for all CRTC.
 	 */
@@ -712,11 +700,6 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 		return false;
 	}
 
-	if (!cache->vma) {
-		fbc->no_fbc_reason = "primary plane not visible";
-		return false;
-	}
-
 	if (cache->crtc.mode_flags & DRM_MODE_FLAG_INTERLACE) {
 		fbc->no_fbc_reason = "incompatible mode";
 		return false;
@@ -740,7 +723,7 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 	 * For now this will effecively disable FBC with 90/270 degree
 	 * rotation.
 	 */
-	if (!(cache->flags & PLANE_HAS_FENCE)) {
+	if (cache->fence_id < 0) {
 		fbc->no_fbc_reason = "framebuffer not tiled or fenced";
 		return false;
 	}
@@ -783,8 +766,7 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 	 * we didn't get any invalidate/deactivate calls, but this would require
 	 * a lot of tracking just for a specific case. If we conclude it's an
 	 * important case, we can implement it later. */
-	if (intel_fbc_calculate_cfb_size(dev_priv, &fbc->state_cache) >
-	    fbc->compressed_fb.size * fbc->threshold) {
+	if (intel_fbc_cfb_size_changed(dev_priv)) {
 		fbc->no_fbc_reason = "CFB requirements changed";
 		return false;
 	}
@@ -794,7 +776,7 @@ static bool intel_fbc_can_activate(struct intel_crtc *crtc)
 	 * having a Y offset that isn't divisible by 4 causes FIFO underrun
 	 * and screen flicker.
 	 */
-	if (IS_GEN_RANGE(dev_priv, 9, 10) &&
+	if (INTEL_GEN(dev_priv) >= 9 &&
 	    (fbc->state_cache.plane.adjusted_y & 3)) {
 		fbc->no_fbc_reason = "plane Y offset is misaligned";
 		return false;
@@ -837,8 +819,7 @@ static void intel_fbc_get_reg_params(struct intel_crtc *crtc,
 	 * zero. */
 	memset(params, 0, sizeof(*params));
 
-	params->vma = cache->vma;
-	params->flags = cache->flags;
+	params->fence_id = cache->fence_id;
 
 	params->crtc.pipe = crtc->pipe;
 	params->crtc.i9xx_plane = to_intel_plane(crtc->base.primary)->i9xx_plane;
@@ -849,39 +830,88 @@ static void intel_fbc_get_reg_params(struct intel_crtc *crtc,
 
 	params->cfb_size = intel_fbc_calculate_cfb_size(dev_priv, cache);
 
-	if (IS_GEN(dev_priv, 9) && !IS_GEMINILAKE(dev_priv))
-		params->gen9_wa_cfb_stride = DIV_ROUND_UP(cache->plane.src_w,
-						32 * fbc->threshold) * 8;
+	params->gen9_wa_cfb_stride = cache->gen9_wa_cfb_stride;
+
+	params->plane_visible = cache->plane.visible;
+}
+
+static bool intel_fbc_can_flip_nuke(const struct intel_crtc_state *crtc_state)
+{
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_fbc *fbc = &dev_priv->fbc;
+	const struct intel_fbc_state_cache *cache = &fbc->state_cache;
+	const struct intel_fbc_reg_params *params = &fbc->params;
+
+	if (drm_atomic_crtc_needs_modeset(&crtc_state->uapi))
+		return false;
+
+	if (!params->plane_visible)
+		return false;
+
+	if (!intel_fbc_can_activate(crtc))
+		return false;
+
+	if (params->fb.format != cache->fb.format)
+		return false;
+
+	if (params->fb.stride != cache->fb.stride)
+		return false;
+
+	if (params->cfb_size != intel_fbc_calculate_cfb_size(dev_priv, cache))
+		return false;
+
+	if (params->gen9_wa_cfb_stride != cache->gen9_wa_cfb_stride)
+		return false;
+
+	return true;
 }
 
-void intel_fbc_pre_update(struct intel_crtc *crtc,
-			  struct intel_crtc_state *crtc_state,
-			  struct intel_plane_state *plane_state)
+bool intel_fbc_pre_update(struct intel_crtc *crtc,
+			  const struct intel_crtc_state *crtc_state,
+			  const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_fbc *fbc = &dev_priv->fbc;
 	const char *reason = "update pending";
+	bool need_vblank_wait = false;
 
 	if (!fbc_supported(dev_priv))
-		return;
+		return need_vblank_wait;
 
 	mutex_lock(&fbc->lock);
 
-	if (!multiple_pipes_ok(crtc, plane_state)) {
-		reason = "more than one pipe active";
-		goto deactivate;
-	}
-
-	if (!fbc->enabled || fbc->crtc != crtc)
+	if (fbc->crtc != crtc)
 		goto unlock;
 
 	intel_fbc_update_state_cache(crtc, crtc_state, plane_state);
 	fbc->flip_pending = true;
 
-deactivate:
-	intel_fbc_deactivate(dev_priv, reason);
+	if (!intel_fbc_can_flip_nuke(crtc_state)) {
+		intel_fbc_deactivate(dev_priv, reason);
+
+		/*
+		 * Display WA #1198: glk+
+		 * Need an extra vblank wait between FBC disable and most plane
+		 * updates. Bspec says this is only needed for plane disable, but
+		 * that is not true. Touching most plane registers will cause the
+		 * corruption to appear. Also SKL/derivatives do not seem to be
+		 * affected.
+		 *
+		 * TODO: could optimize this a bit by sampling the frame
+		 * counter when we disable FBC (if it was already done earlier)
+		 * and skipping the extra vblank wait before the plane update
+		 * if at least one frame has already passed.
+		 */
+		if (fbc->activated &&
+		    (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)))
+			need_vblank_wait = true;
+		fbc->activated = false;
+	}
 unlock:
 	mutex_unlock(&fbc->lock);
+
+	return need_vblank_wait;
 }
 
 /**
@@ -897,14 +927,13 @@ static void __intel_fbc_disable(struct drm_i915_private *dev_priv)
 	struct intel_crtc *crtc = fbc->crtc;
 
 	WARN_ON(!mutex_is_locked(&fbc->lock));
-	WARN_ON(!fbc->enabled);
+	WARN_ON(!fbc->crtc);
 	WARN_ON(fbc->active);
 
 	DRM_DEBUG_KMS("Disabling FBC on pipe %c\n", pipe_name(crtc->pipe));
 
 	__intel_fbc_cleanup_cfb(dev_priv);
 
-	fbc->enabled = false;
 	fbc->crtc = NULL;
 }
 
@@ -915,11 +944,10 @@ static void __intel_fbc_post_update(struct intel_crtc *crtc)
 
 	WARN_ON(!mutex_is_locked(&fbc->lock));
 
-	if (!fbc->enabled || fbc->crtc != crtc)
+	if (fbc->crtc != crtc)
 		return;
 
 	fbc->flip_pending = false;
-	WARN_ON(fbc->active);
 
 	if (!i915_modparams.enable_fbc) {
 		intel_fbc_deactivate(dev_priv, "disabled at runtime per module param");
@@ -933,10 +961,9 @@ static void __intel_fbc_post_update(struct intel_crtc *crtc)
 	if (!intel_fbc_can_activate(crtc))
 		return;
 
-	if (!fbc->busy_bits) {
-		intel_fbc_deactivate(dev_priv, "FBC enabled (active or scheduled)");
+	if (!fbc->busy_bits)
 		intel_fbc_hw_activate(dev_priv);
-	} else
+	else
 		intel_fbc_deactivate(dev_priv, "frontbuffer write");
 }
 
@@ -955,7 +982,7 @@ void intel_fbc_post_update(struct intel_crtc *crtc)
 
 static unsigned int intel_fbc_get_frontbuffer_bit(struct intel_fbc *fbc)
 {
-	if (fbc->enabled)
+	if (fbc->crtc)
 		return to_intel_plane(fbc->crtc->base.primary)->frontbuffer_bit;
 	else
 		return fbc->possible_framebuffer_bits;
@@ -977,7 +1004,7 @@ void intel_fbc_invalidate(struct drm_i915_private *dev_priv,
 
 	fbc->busy_bits |= intel_fbc_get_frontbuffer_bit(fbc) & frontbuffer_bits;
 
-	if (fbc->enabled && fbc->busy_bits)
+	if (fbc->crtc && fbc->busy_bits)
 		intel_fbc_deactivate(dev_priv, "frontbuffer write");
 
 	mutex_unlock(&fbc->lock);
@@ -998,7 +1025,7 @@ void intel_fbc_flush(struct drm_i915_private *dev_priv,
 	if (origin == ORIGIN_GTT || origin == ORIGIN_FLIP)
 		goto out;
 
-	if (!fbc->busy_bits && fbc->enabled &&
+	if (!fbc->busy_bits && fbc->crtc &&
 	    (frontbuffer_bits & intel_fbc_get_frontbuffer_bit(fbc))) {
 		if (fbc->active)
 			intel_fbc_recompress(dev_priv);
@@ -1047,12 +1074,12 @@ void intel_fbc_choose_crtc(struct drm_i915_private *dev_priv,
 	 * to pipe or plane A. */
 	for_each_new_intel_plane_in_state(state, plane, plane_state, i) {
 		struct intel_crtc_state *crtc_state;
-		struct intel_crtc *crtc = to_intel_crtc(plane_state->base.crtc);
+		struct intel_crtc *crtc = to_intel_crtc(plane_state->hw.crtc);
 
 		if (!plane->has_fbc)
 			continue;
 
-		if (!plane_state->base.visible)
+		if (!plane_state->uapi.visible)
 			continue;
 
 		crtc_state = intel_atomic_get_new_crtc_state(state, crtc);
@@ -1081,42 +1108,53 @@ out:
  * intel_fbc_disable in the middle, as long as it is deactivated.
  */
 void intel_fbc_enable(struct intel_crtc *crtc,
-		      struct intel_crtc_state *crtc_state,
-		      struct intel_plane_state *plane_state)
+		      const struct intel_crtc_state *crtc_state,
+		      const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_fbc *fbc = &dev_priv->fbc;
+	struct intel_fbc_state_cache *cache = &fbc->state_cache;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 
 	if (!fbc_supported(dev_priv))
 		return;
 
 	mutex_lock(&fbc->lock);
 
-	if (fbc->enabled) {
-		WARN_ON(fbc->crtc == NULL);
-		if (fbc->crtc == crtc) {
-			WARN_ON(!crtc_state->enable_fbc);
-			WARN_ON(fbc->active);
-		}
-		goto out;
-	}
+	if (fbc->crtc) {
+		if (fbc->crtc != crtc ||
+		    !intel_fbc_cfb_size_changed(dev_priv))
+			goto out;
 
-	if (!crtc_state->enable_fbc)
-		goto out;
+		__intel_fbc_disable(dev_priv);
+	}
 
 	WARN_ON(fbc->active);
-	WARN_ON(fbc->crtc != NULL);
 
 	intel_fbc_update_state_cache(crtc, crtc_state, plane_state);
-	if (intel_fbc_alloc_cfb(crtc)) {
+
+	/* FIXME crtc_state->enable_fbc lies :( */
+	if (!cache->plane.visible)
+		goto out;
+
+	if (intel_fbc_alloc_cfb(dev_priv,
+				intel_fbc_calculate_cfb_size(dev_priv, cache),
+				fb->format->cpp[0])) {
+		cache->plane.visible = false;
 		fbc->no_fbc_reason = "not enough stolen memory";
 		goto out;
 	}
 
+	if ((IS_GEN9_BC(dev_priv) || IS_BROXTON(dev_priv)) &&
+	    fb->modifier != I915_FORMAT_MOD_X_TILED)
+		cache->gen9_wa_cfb_stride =
+			DIV_ROUND_UP(cache->plane.src_w, 32 * fbc->threshold) * 8;
+	else
+		cache->gen9_wa_cfb_stride = 0;
+
 	DRM_DEBUG_KMS("Enabling FBC on pipe %c\n", pipe_name(crtc->pipe));
 	fbc->no_fbc_reason = "FBC enabled but not active yet\n";
 
-	fbc->enabled = true;
 	fbc->crtc = crtc;
 out:
 	mutex_unlock(&fbc->lock);
@@ -1156,7 +1194,7 @@ void intel_fbc_global_disable(struct drm_i915_private *dev_priv)
 		return;
 
 	mutex_lock(&fbc->lock);
-	if (fbc->enabled) {
+	if (fbc->crtc) {
 		WARN_ON(fbc->crtc->active);
 		__intel_fbc_disable(dev_priv);
 	}
@@ -1172,7 +1210,7 @@ static void intel_fbc_underrun_work_fn(struct work_struct *work)
 	mutex_lock(&fbc->lock);
 
 	/* Maybe we were scheduled twice. */
-	if (fbc->underrun_detected || !fbc->enabled)
+	if (fbc->underrun_detected || !fbc->crtc)
 		goto out;
 
 	DRM_DEBUG_KMS("Disabling FBC due to FIFO underrun.\n");
@@ -1244,28 +1282,6 @@ void intel_fbc_handle_fifo_underrun_irq(struct drm_i915_private *dev_priv)
 	schedule_work(&fbc->underrun_work);
 }
 
-/**
- * intel_fbc_init_pipe_state - initialize FBC's CRTC visibility tracking
- * @dev_priv: i915 device instance
- *
- * The FBC code needs to track CRTC visibility since the older platforms can't
- * have FBC enabled while multiple pipes are used. This function does the
- * initial setup at driver load to make sure FBC is matching the real hardware.
- */
-void intel_fbc_init_pipe_state(struct drm_i915_private *dev_priv)
-{
-	struct intel_crtc *crtc;
-
-	/* Don't even bother tracking anything if we don't need. */
-	if (!no_fbc_on_multiple_pipes(dev_priv))
-		return;
-
-	for_each_intel_crtc(&dev_priv->drm, crtc)
-		if (intel_crtc_active(crtc) &&
-		    crtc->base.primary->state->visible)
-			dev_priv->fbc.visible_pipes_mask |= (1 << crtc->pipe);
-}
-
 /*
  * The DDX driver changes its behavior depending on the value it reads from
  * i915.enable_fbc, so sanitize it by translating the default value into either
@@ -1283,10 +1299,6 @@ static int intel_sanitize_fbc_option(struct drm_i915_private *dev_priv)
 	if (!HAS_FBC(dev_priv))
 		return 0;
 
-	/* https://bugs.freedesktop.org/show_bug.cgi?id=108085 */
-	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
-		return 0;
-
 	if (IS_BROADWELL(dev_priv) || INTEL_GEN(dev_priv) >= 9)
 		return 1;
 
@@ -1317,7 +1329,6 @@ void intel_fbc_init(struct drm_i915_private *dev_priv)
 
 	INIT_WORK(&fbc->underrun_work, intel_fbc_underrun_work_fn);
 	mutex_init(&fbc->lock);
-	fbc->enabled = false;
 	fbc->active = false;
 
 	if (!drm_mm_initialized(&dev_priv->mm.stolen))
diff --git a/drivers/gpu/drm/i915/display/intel_fbc.h b/drivers/gpu/drm/i915/display/intel_fbc.h
index 50272eda8d43..c8a5e5098687 100644
--- a/drivers/gpu/drm/i915/display/intel_fbc.h
+++ b/drivers/gpu/drm/i915/display/intel_fbc.h
@@ -19,15 +19,14 @@ struct intel_plane_state;
 void intel_fbc_choose_crtc(struct drm_i915_private *dev_priv,
 			   struct intel_atomic_state *state);
 bool intel_fbc_is_active(struct drm_i915_private *dev_priv);
-void intel_fbc_pre_update(struct intel_crtc *crtc,
-			  struct intel_crtc_state *crtc_state,
-			  struct intel_plane_state *plane_state);
+bool intel_fbc_pre_update(struct intel_crtc *crtc,
+			  const struct intel_crtc_state *crtc_state,
+			  const struct intel_plane_state *plane_state);
 void intel_fbc_post_update(struct intel_crtc *crtc);
 void intel_fbc_init(struct drm_i915_private *dev_priv);
-void intel_fbc_init_pipe_state(struct drm_i915_private *dev_priv);
 void intel_fbc_enable(struct intel_crtc *crtc,
-		      struct intel_crtc_state *crtc_state,
-		      struct intel_plane_state *plane_state);
+		      const struct intel_crtc_state *crtc_state,
+		      const struct intel_plane_state *plane_state);
 void intel_fbc_disable(struct intel_crtc *crtc);
 void intel_fbc_global_disable(struct drm_i915_private *dev_priv);
 void intel_fbc_invalidate(struct drm_i915_private *dev_priv,
diff --git a/drivers/gpu/drm/i915/display/intel_frontbuffer.c b/drivers/gpu/drm/i915/display/intel_frontbuffer.c
index 84b164f31895..6cb02c912acc 100644
--- a/drivers/gpu/drm/i915/display/intel_frontbuffer.c
+++ b/drivers/gpu/drm/i915/display/intel_frontbuffer.c
@@ -229,11 +229,11 @@ static void frontbuffer_release(struct kref *ref)
 		vma->display_alignment = I915_GTT_MIN_ALIGNMENT;
 	spin_unlock(&obj->vma.lock);
 
-	obj->frontbuffer = NULL;
+	RCU_INIT_POINTER(obj->frontbuffer, NULL);
 	spin_unlock(&to_i915(obj->base.dev)->fb_tracking.lock);
 
 	i915_gem_object_put(obj);
-	kfree(front);
+	kfree_rcu(front, rcu);
 }
 
 struct intel_frontbuffer *
@@ -242,11 +242,7 @@ intel_frontbuffer_get(struct drm_i915_gem_object *obj)
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	struct intel_frontbuffer *front;
 
-	spin_lock(&i915->fb_tracking.lock);
-	front = obj->frontbuffer;
-	if (front)
-		kref_get(&front->ref);
-	spin_unlock(&i915->fb_tracking.lock);
+	front = __intel_frontbuffer_get(obj);
 	if (front)
 		return front;
 
@@ -262,13 +258,13 @@ intel_frontbuffer_get(struct drm_i915_gem_object *obj)
 			 i915_active_may_sleep(frontbuffer_retire));
 
 	spin_lock(&i915->fb_tracking.lock);
-	if (obj->frontbuffer) {
+	if (rcu_access_pointer(obj->frontbuffer)) {
 		kfree(front);
-		front = obj->frontbuffer;
+		front = rcu_dereference_protected(obj->frontbuffer, true);
 		kref_get(&front->ref);
 	} else {
 		i915_gem_object_get(obj);
-		obj->frontbuffer = front;
+		rcu_assign_pointer(obj->frontbuffer, front);
 	}
 	spin_unlock(&i915->fb_tracking.lock);
 
diff --git a/drivers/gpu/drm/i915/display/intel_frontbuffer.h b/drivers/gpu/drm/i915/display/intel_frontbuffer.h
index adc64d61a4a5..6d41f5394425 100644
--- a/drivers/gpu/drm/i915/display/intel_frontbuffer.h
+++ b/drivers/gpu/drm/i915/display/intel_frontbuffer.h
@@ -27,10 +27,10 @@
 #include <linux/atomic.h>
 #include <linux/kref.h>
 
+#include "gem/i915_gem_object_types.h"
 #include "i915_active.h"
 
 struct drm_i915_private;
-struct drm_i915_gem_object;
 
 enum fb_op_origin {
 	ORIGIN_GTT,
@@ -45,6 +45,7 @@ struct intel_frontbuffer {
 	atomic_t bits;
 	struct i915_active write;
 	struct drm_i915_gem_object *obj;
+	struct rcu_head rcu;
 };
 
 void intel_frontbuffer_flip_prepare(struct drm_i915_private *i915,
@@ -54,6 +55,35 @@ void intel_frontbuffer_flip_complete(struct drm_i915_private *i915,
 void intel_frontbuffer_flip(struct drm_i915_private *i915,
 			    unsigned frontbuffer_bits);
 
+void intel_frontbuffer_put(struct intel_frontbuffer *front);
+
+static inline struct intel_frontbuffer *
+__intel_frontbuffer_get(const struct drm_i915_gem_object *obj)
+{
+	struct intel_frontbuffer *front;
+
+	if (likely(!rcu_access_pointer(obj->frontbuffer)))
+		return NULL;
+
+	rcu_read_lock();
+	do {
+		front = rcu_dereference(obj->frontbuffer);
+		if (!front)
+			break;
+
+		if (unlikely(!kref_get_unless_zero(&front->ref)))
+			continue;
+
+		if (likely(front == rcu_access_pointer(obj->frontbuffer)))
+			break;
+
+		intel_frontbuffer_put(front);
+	} while (1);
+	rcu_read_unlock();
+
+	return front;
+}
+
 struct intel_frontbuffer *
 intel_frontbuffer_get(struct drm_i915_gem_object *obj);
 
@@ -119,6 +149,4 @@ void intel_frontbuffer_track(struct intel_frontbuffer *old,
 			     struct intel_frontbuffer *new,
 			     unsigned int frontbuffer_bits);
 
-void intel_frontbuffer_put(struct intel_frontbuffer *front);
-
 #endif /* __INTEL_FRONTBUFFER_H__ */
diff --git a/drivers/gpu/drm/i915/display/intel_hdcp.c b/drivers/gpu/drm/i915/display/intel_hdcp.c
index a448815d8fc2..0fdbd39f6641 100644
--- a/drivers/gpu/drm/i915/display/intel_hdcp.c
+++ b/drivers/gpu/drm/i915/display/intel_hdcp.c
@@ -1870,7 +1870,7 @@ static bool is_hdcp2_supported(struct drm_i915_private *dev_priv)
 		return false;
 
 	return (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv) ||
-		IS_KABYLAKE(dev_priv));
+		IS_KABYLAKE(dev_priv) || IS_COFFEELAKE(dev_priv));
 }
 
 void intel_hdcp_component_init(struct drm_i915_private *dev_priv)
diff --git a/drivers/gpu/drm/i915/display/intel_hdmi.c b/drivers/gpu/drm/i915/display/intel_hdmi.c
index f56fffc474fa..685589064d10 100644
--- a/drivers/gpu/drm/i915/display/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/display/intel_hdmi.c
@@ -285,7 +285,7 @@ static void ibx_write_infoframe(struct intel_encoder *encoder,
 {
 	const u32 *data = frame;
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	i915_reg_t reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
 	int i;
@@ -321,7 +321,7 @@ static void ibx_read_infoframe(struct intel_encoder *encoder,
 			       void *frame, ssize_t len)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	u32 val, *data = frame;
 	int i;
 
@@ -340,7 +340,7 @@ static u32 ibx_infoframes_enabled(struct intel_encoder *encoder,
 				  const struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum pipe pipe = to_intel_crtc(pipe_config->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(pipe_config->uapi.crtc)->pipe;
 	i915_reg_t reg = TVIDEO_DIP_CTL(pipe);
 	u32 val = I915_READ(reg);
 
@@ -362,7 +362,7 @@ static void cpt_write_infoframe(struct intel_encoder *encoder,
 {
 	const u32 *data = frame;
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	i915_reg_t reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
 	int i;
@@ -401,7 +401,7 @@ static void cpt_read_infoframe(struct intel_encoder *encoder,
 			       void *frame, ssize_t len)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	u32 val, *data = frame;
 	int i;
 
@@ -420,7 +420,7 @@ static u32 cpt_infoframes_enabled(struct intel_encoder *encoder,
 				  const struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum pipe pipe = to_intel_crtc(pipe_config->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(pipe_config->uapi.crtc)->pipe;
 	u32 val = I915_READ(TVIDEO_DIP_CTL(pipe));
 
 	if ((val & VIDEO_DIP_ENABLE) == 0)
@@ -438,7 +438,7 @@ static void vlv_write_infoframe(struct intel_encoder *encoder,
 {
 	const u32 *data = frame;
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	i915_reg_t reg = VLV_TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
 	int i;
@@ -474,7 +474,7 @@ static void vlv_read_infoframe(struct intel_encoder *encoder,
 			       void *frame, ssize_t len)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	u32 val, *data = frame;
 	int i;
 
@@ -493,7 +493,7 @@ static u32 vlv_infoframes_enabled(struct intel_encoder *encoder,
 				  const struct intel_crtc_state *pipe_config)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	enum pipe pipe = to_intel_crtc(pipe_config->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(pipe_config->uapi.crtc)->pipe;
 	u32 val = I915_READ(VLV_TVIDEO_DIP_CTL(pipe));
 
 	if ((val & VIDEO_DIP_ENABLE) == 0)
@@ -708,7 +708,7 @@ intel_hdmi_compute_avi_infoframe(struct intel_encoder *encoder,
 {
 	struct hdmi_avi_infoframe *frame = &crtc_state->infoframes.avi.avi;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	struct drm_connector *connector = conn_state->connector;
 	int ret;
 
@@ -804,7 +804,7 @@ intel_hdmi_compute_hdmi_infoframe(struct intel_encoder *encoder,
 
 	ret = drm_hdmi_vendor_infoframe_from_display_mode(frame,
 							  conn_state->connector,
-							  &crtc_state->base.adjusted_mode);
+							  &crtc_state->hw.adjusted_mode);
 	if (WARN_ON(ret))
 		return false;
 
@@ -965,7 +965,7 @@ static bool intel_hdmi_set_gcp_infoframe(struct intel_encoder *encoder,
 					 const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	i915_reg_t reg;
 
 	if ((crtc_state->infoframes.enable &
@@ -990,7 +990,7 @@ void intel_hdmi_read_gcp_infoframe(struct intel_encoder *encoder,
 				   struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	i915_reg_t reg;
 
 	if ((crtc_state->infoframes.enable &
@@ -1027,7 +1027,7 @@ static void intel_hdmi_compute_gcp_infoframe(struct intel_encoder *encoder,
 
 	/* Enable default_phase whenever the display mode is suitably aligned */
 	if (gcp_default_phase_possible(crtc_state->pipe_bpp,
-				       &crtc_state->base.adjusted_mode))
+				       &crtc_state->hw.adjusted_mode))
 		crtc_state->infoframes.gcp |= GCP_DEFAULT_PHASE_ENABLE;
 }
 
@@ -1037,7 +1037,7 @@ static void ibx_set_infoframes(struct intel_encoder *encoder,
 			       const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_digital_port *intel_dig_port = enc_to_dig_port(&encoder->base);
 	struct intel_hdmi *intel_hdmi = &intel_dig_port->hdmi;
 	i915_reg_t reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
@@ -1096,7 +1096,7 @@ static void cpt_set_infoframes(struct intel_encoder *encoder,
 			       const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
 	i915_reg_t reg = TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
@@ -1145,7 +1145,7 @@ static void vlv_set_infoframes(struct intel_encoder *encoder,
 			       const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
 	i915_reg_t reg = VLV_TVIDEO_DIP_CTL(intel_crtc->pipe);
 	u32 val = I915_READ(reg);
@@ -1736,9 +1736,9 @@ static void intel_hdmi_prepare(struct intel_encoder *encoder,
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
-	const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &crtc_state->hw.adjusted_mode;
 	u32 hdmi_val;
 
 	intel_dp_dual_mode_set_tmds_output(intel_hdmi, true);
@@ -1829,7 +1829,7 @@ static void intel_hdmi_get_config(struct intel_encoder *encoder,
 	    tmp & HDMI_COLOR_RANGE_16_235)
 		pipe_config->limited_color_range = true;
 
-	pipe_config->base.adjusted_mode.flags |= flags;
+	pipe_config->hw.adjusted_mode.flags |= flags;
 
 	if ((tmp & SDVO_COLOR_FORMAT_MASK) == HDMI_COLOR_FORMAT_12bpc)
 		dotclock = pipe_config->port_clock * 2 / 3;
@@ -1839,7 +1839,7 @@ static void intel_hdmi_get_config(struct intel_encoder *encoder,
 	if (pipe_config->pixel_multiplier)
 		dotclock /= pipe_config->pixel_multiplier;
 
-	pipe_config->base.adjusted_mode.crtc_clock = dotclock;
+	pipe_config->hw.adjusted_mode.crtc_clock = dotclock;
 
 	pipe_config->lane_count = 4;
 
@@ -1860,7 +1860,7 @@ static void intel_enable_hdmi_audio(struct intel_encoder *encoder,
 				    const struct intel_crtc_state *pipe_config,
 				    const struct drm_connector_state *conn_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 
 	WARN_ON(!pipe_config->has_hdmi_sink);
 	DRM_DEBUG_DRIVER("Enabling HDMI audio on pipe %c\n",
@@ -1946,7 +1946,7 @@ static void cpt_enable_hdmi(struct intel_encoder *encoder,
 {
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
 	enum pipe pipe = crtc->pipe;
 	u32 temp;
@@ -2010,7 +2010,7 @@ static void intel_disable_hdmi(struct intel_encoder *encoder,
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
 	struct intel_digital_port *intel_dig_port =
 		hdmi_to_dig_port(intel_hdmi);
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	u32 temp;
 
 	temp = I915_READ(intel_hdmi->hdmi_reg);
@@ -2210,12 +2210,12 @@ static bool hdmi_deep_color_possible(const struct intel_crtc_state *crtc_state,
 				     int bpc)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(crtc_state->base.crtc->dev);
-	struct drm_atomic_state *state = crtc_state->base.state;
+		to_i915(crtc_state->uapi.crtc->dev);
+	struct drm_atomic_state *state = crtc_state->uapi.state;
 	struct drm_connector_state *connector_state;
 	struct drm_connector *connector;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	int i;
 
 	if (HAS_GMCH(dev_priv))
@@ -2240,7 +2240,7 @@ static bool hdmi_deep_color_possible(const struct intel_crtc_state *crtc_state,
 	for_each_new_connector_in_state(state, connector, connector_state, i) {
 		const struct drm_display_info *info = &connector->display_info;
 
-		if (connector_state->crtc != crtc_state->base.crtc)
+		if (connector_state->crtc != crtc_state->uapi.crtc)
 			continue;
 
 		if (crtc_state->output_format == INTEL_OUTPUT_FORMAT_YCBCR420) {
@@ -2281,7 +2281,7 @@ static bool
 intel_hdmi_ycbcr420_config(struct drm_connector *connector,
 			   struct intel_crtc_state *config)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(config->uapi.crtc);
 
 	if (!connector->ycbcr_420_allowed) {
 		DRM_ERROR("Platform doesn't support YCBCR420 output\n");
@@ -2336,7 +2336,7 @@ static int intel_hdmi_compute_clock(struct intel_encoder *encoder,
 {
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	int bpc, clock = adjusted_mode->crtc_clock;
 
 	if (adjusted_mode->flags & DRM_MODE_FLAG_DBLCLK)
@@ -2378,7 +2378,7 @@ static bool intel_hdmi_limited_color_range(const struct intel_crtc_state *crtc_s
 	const struct intel_digital_connector_state *intel_conn_state =
 		to_intel_digital_connector_state(conn_state);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 
 	/*
 	 * Our YCbCr output is always limited range.
@@ -2406,7 +2406,7 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
 {
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(&encoder->base);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	struct drm_connector *connector = conn_state->connector;
 	struct drm_scdc *scdc = &connector->display_info.hdmi.scdc;
 	struct intel_digital_connector_state *intel_conn_state =
@@ -2451,8 +2451,9 @@ int intel_hdmi_compute_config(struct intel_encoder *encoder,
 	if (ret)
 		return ret;
 
-	/* Set user selected PAR to incoming mode's member */
-	adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio;
+	if (conn_state->picture_aspect_ratio)
+		adjusted_mode->picture_aspect_ratio =
+			conn_state->picture_aspect_ratio;
 
 	pipe_config->lane_count = 4;
 
@@ -2872,7 +2873,6 @@ intel_hdmi_add_properties(struct intel_hdmi *intel_hdmi, struct drm_connector *c
 		intel_attach_colorspace_property(connector);
 
 	drm_connector_attach_content_type_property(connector);
-	connector->state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
 
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 		drm_object_attach_property(&connector->base,
@@ -3131,20 +3131,29 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 	struct intel_encoder *intel_encoder = &intel_dig_port->base;
 	struct drm_device *dev = intel_encoder->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct i2c_adapter *ddc;
 	enum port port = intel_encoder->port;
 	struct cec_connector_info conn_info;
 
 	DRM_DEBUG_KMS("Adding HDMI connector on [ENCODER:%d:%s]\n",
 		      intel_encoder->base.base.id, intel_encoder->base.name);
 
+	if (INTEL_GEN(dev_priv) < 12 && WARN_ON(port == PORT_A))
+		return;
+
 	if (WARN(intel_dig_port->max_lanes < 4,
 		 "Not enough lanes (%d) for HDMI on [ENCODER:%d:%s]\n",
 		 intel_dig_port->max_lanes, intel_encoder->base.base.id,
 		 intel_encoder->base.name))
 		return;
 
-	drm_connector_init(dev, connector, &intel_hdmi_connector_funcs,
-			   DRM_MODE_CONNECTOR_HDMIA);
+	intel_hdmi->ddc_bus = intel_hdmi_ddc_pin(dev_priv, port);
+	ddc = intel_gmbus_get_adapter(dev_priv, intel_hdmi->ddc_bus);
+
+	drm_connector_init_with_ddc(dev, connector,
+				    &intel_hdmi_connector_funcs,
+				    DRM_MODE_CONNECTOR_HDMIA,
+				    ddc);
 	drm_connector_helper_add(connector, &intel_hdmi_connector_helper_funcs);
 
 	connector->interlace_allowed = 1;
@@ -3154,10 +3163,6 @@ void intel_hdmi_init_connector(struct intel_digital_port *intel_dig_port,
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 		connector->ycbcr_420_allowed = true;
 
-	intel_hdmi->ddc_bus = intel_hdmi_ddc_pin(dev_priv, port);
-
-	if (WARN_ON(port == PORT_A))
-		return;
 	intel_encoder->hpd_pin = intel_hpd_pin_default(dev_priv, port);
 
 	if (HAS_DDI(dev_priv))
diff --git a/drivers/gpu/drm/i915/display/intel_lspcon.c b/drivers/gpu/drm/i915/display/intel_lspcon.c
index f8f1308643a9..5145ff8b962b 100644
--- a/drivers/gpu/drm/i915/display/intel_lspcon.c
+++ b/drivers/gpu/drm/i915/display/intel_lspcon.c
@@ -189,7 +189,7 @@ void lspcon_ycbcr420_config(struct drm_connector *connector,
 {
 	const struct drm_display_info *info = &connector->display_info;
 	const struct drm_display_mode *adjusted_mode =
-					&crtc_state->base.adjusted_mode;
+					&crtc_state->hw.adjusted_mode;
 
 	if (drm_mode_is_420_only(info, adjusted_mode) &&
 	    connector->ycbcr_420_allowed) {
@@ -475,7 +475,7 @@ void lspcon_set_infoframes(struct intel_encoder *encoder,
 	struct intel_digital_port *dig_port = enc_to_dig_port(&encoder->base);
 	struct intel_lspcon *lspcon = &dig_port->lspcon;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 
 	if (!lspcon->active) {
 		DRM_ERROR("Writing infoframes while LSPCON disabled ?\n");
diff --git a/drivers/gpu/drm/i915/display/intel_lvds.c b/drivers/gpu/drm/i915/display/intel_lvds.c
index b1bc78623647..10696bb99dcf 100644
--- a/drivers/gpu/drm/i915/display/intel_lvds.c
+++ b/drivers/gpu/drm/i915/display/intel_lvds.c
@@ -135,7 +135,7 @@ static void intel_lvds_get_config(struct intel_encoder *encoder,
 	else
 		flags |= DRM_MODE_FLAG_PVSYNC;
 
-	pipe_config->base.adjusted_mode.flags |= flags;
+	pipe_config->hw.adjusted_mode.flags |= flags;
 
 	if (INTEL_GEN(dev_priv) < 5)
 		pipe_config->gmch_pfit.lvds_border_bits =
@@ -148,7 +148,7 @@ static void intel_lvds_get_config(struct intel_encoder *encoder,
 		pipe_config->gmch_pfit.control |= tmp & PANEL_8TO6_DITHER_ENABLE;
 	}
 
-	pipe_config->base.adjusted_mode.crtc_clock = pipe_config->port_clock;
+	pipe_config->hw.adjusted_mode.crtc_clock = pipe_config->port_clock;
 }
 
 static void intel_lvds_pps_get_hw_state(struct drm_i915_private *dev_priv,
@@ -230,8 +230,8 @@ static void intel_pre_enable_lvds(struct intel_encoder *encoder,
 {
 	struct intel_lvds_encoder *lvds_encoder = to_lvds_encoder(&encoder->base);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	enum pipe pipe = crtc->pipe;
 	u32 temp;
 
@@ -392,8 +392,8 @@ static int intel_lvds_compute_config(struct intel_encoder *intel_encoder,
 		to_lvds_encoder(&intel_encoder->base);
 	struct intel_connector *intel_connector =
 		lvds_encoder->attached_connector;
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	unsigned int lvds_bpp;
 
 	/* Should never happen!! */
diff --git a/drivers/gpu/drm/i915/display/intel_opregion.c b/drivers/gpu/drm/i915/display/intel_opregion.c
index 969ade623691..e59b4992ba1b 100644
--- a/drivers/gpu/drm/i915/display/intel_opregion.c
+++ b/drivers/gpu/drm/i915/display/intel_opregion.c
@@ -941,6 +941,13 @@ int intel_opregion_setup(struct drm_i915_private *dev_priv)
 	if (mboxes & MBOX_ACPI) {
 		DRM_DEBUG_DRIVER("Public ACPI methods supported\n");
 		opregion->acpi = base + OPREGION_ACPI_OFFSET;
+		/*
+		 * Indicate we handle monitor hotplug events ourselves so we do
+		 * not need ACPI notifications for them. Disabling these avoids
+		 * triggering the AML code doing the notifation, which may be
+		 * broken as Windows also seems to disable these.
+		 */
+		opregion->acpi->chpd = 1;
 	}
 
 	if (mboxes & MBOX_SWSCI) {
diff --git a/drivers/gpu/drm/i915/display/intel_overlay.c b/drivers/gpu/drm/i915/display/intel_overlay.c
index 848ce07a8ec2..e40c3a0e2cd7 100644
--- a/drivers/gpu/drm/i915/display/intel_overlay.c
+++ b/drivers/gpu/drm/i915/display/intel_overlay.c
@@ -279,12 +279,21 @@ static void intel_overlay_flip_prepare(struct intel_overlay *overlay,
 				       struct i915_vma *vma)
 {
 	enum pipe pipe = overlay->crtc->pipe;
+	struct intel_frontbuffer *from = NULL, *to = NULL;
 
 	WARN_ON(overlay->old_vma);
 
-	intel_frontbuffer_track(overlay->vma ? overlay->vma->obj->frontbuffer : NULL,
-				vma ? vma->obj->frontbuffer : NULL,
-				INTEL_FRONTBUFFER_OVERLAY(pipe));
+	if (overlay->vma)
+		from = intel_frontbuffer_get(overlay->vma->obj);
+	if (vma)
+		to = intel_frontbuffer_get(vma->obj);
+
+	intel_frontbuffer_track(from, to, INTEL_FRONTBUFFER_OVERLAY(pipe));
+
+	if (to)
+		intel_frontbuffer_put(to);
+	if (from)
+		intel_frontbuffer_put(from);
 
 	intel_frontbuffer_flip_prepare(overlay->i915,
 				       INTEL_FRONTBUFFER_OVERLAY(pipe));
@@ -668,8 +677,8 @@ static void update_colorkey(struct intel_overlay *overlay,
 	if (overlay->color_key_enabled)
 		flags |= DST_KEY_ENABLE;
 
-	if (state->base.visible)
-		format = state->base.fb->format->format;
+	if (state->uapi.visible)
+		format = state->hw.fb->format->format;
 
 	switch (format) {
 	case DRM_FORMAT_C8:
@@ -758,15 +767,13 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay,
 
 	atomic_inc(&dev_priv->gpu_error.pending_fb_pin);
 
-	i915_gem_object_lock(new_bo);
 	vma = i915_gem_object_pin_to_display_plane(new_bo,
 						   0, NULL, PIN_MAPPABLE);
-	i915_gem_object_unlock(new_bo);
 	if (IS_ERR(vma)) {
 		ret = PTR_ERR(vma);
 		goto out_pin_section;
 	}
-	intel_frontbuffer_flush(new_bo->frontbuffer, ORIGIN_DIRTYFB);
+	i915_gem_object_flush_frontbuffer(new_bo, ORIGIN_DIRTYFB);
 
 	if (!overlay->active) {
 		u32 oconfig;
@@ -1326,12 +1333,14 @@ err_put_bo:
 void intel_overlay_setup(struct drm_i915_private *dev_priv)
 {
 	struct intel_overlay *overlay;
+	struct intel_engine_cs *engine;
 	int ret;
 
 	if (!HAS_OVERLAY(dev_priv))
 		return;
 
-	if (!HAS_ENGINE(dev_priv, RCS0))
+	engine = dev_priv->engine[RCS0];
+	if (!engine || !engine->kernel_context)
 		return;
 
 	overlay = kzalloc(sizeof(*overlay), GFP_KERNEL);
@@ -1339,7 +1348,7 @@ void intel_overlay_setup(struct drm_i915_private *dev_priv)
 		return;
 
 	overlay->i915 = dev_priv;
-	overlay->context = dev_priv->engine[RCS0]->kernel_context;
+	overlay->context = engine->kernel_context;
 	GEM_BUG_ON(!overlay->context);
 
 	overlay->color_key = 0x0101fe;
diff --git a/drivers/gpu/drm/i915/display/intel_panel.c b/drivers/gpu/drm/i915/display/intel_panel.c
index bc14e9c0285a..7b3ec6eb3382 100644
--- a/drivers/gpu/drm/i915/display/intel_panel.c
+++ b/drivers/gpu/drm/i915/display/intel_panel.c
@@ -178,7 +178,7 @@ intel_pch_panel_fitting(struct intel_crtc *intel_crtc,
 			struct intel_crtc_state *pipe_config,
 			int fitting_mode)
 {
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	int x = 0, y = 0, width = 0, height = 0;
 
 	/* Native modes don't need fitting */
@@ -300,7 +300,7 @@ static inline u32 panel_fitter_scaling(u32 source, u32 target)
 static void i965_scale_aspect(struct intel_crtc_state *pipe_config,
 			      u32 *pfit_control)
 {
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	u32 scaled_width = adjusted_mode->crtc_hdisplay *
 		pipe_config->pipe_src_h;
 	u32 scaled_height = pipe_config->pipe_src_w *
@@ -321,7 +321,7 @@ static void i9xx_scale_aspect(struct intel_crtc_state *pipe_config,
 			      u32 *pfit_control, u32 *pfit_pgm_ratios,
 			      u32 *border)
 {
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	u32 scaled_width = adjusted_mode->crtc_hdisplay *
 		pipe_config->pipe_src_h;
 	u32 scaled_height = pipe_config->pipe_src_w *
@@ -380,7 +380,7 @@ void intel_gmch_panel_fitting(struct intel_crtc *intel_crtc,
 {
 	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
 	u32 pfit_control = 0, pfit_pgm_ratios = 0, border = 0;
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 
 	/* Native modes don't need fitting */
 	if (adjusted_mode->crtc_hdisplay == pipe_config->pipe_src_w &&
@@ -1047,7 +1047,7 @@ static void vlv_enable_backlight(const struct intel_crtc_state *crtc_state,
 	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(crtc_state->uapi.crtc)->pipe;
 	u32 ctl, ctl2;
 
 	ctl2 = I915_READ(VLV_BLC_PWM_CTL2(pipe));
@@ -1077,7 +1077,7 @@ static void bxt_enable_backlight(const struct intel_crtc_state *crtc_state,
 	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(crtc_state->uapi.crtc)->pipe;
 	u32 pwm_ctl, val;
 
 	/* Controller 1 uses the utility pin. */
@@ -1189,7 +1189,7 @@ void intel_panel_enable_backlight(const struct intel_crtc_state *crtc_state,
 	struct intel_connector *connector = to_intel_connector(conn_state->connector);
 	struct drm_i915_private *dev_priv = to_i915(connector->base.dev);
 	struct intel_panel *panel = &connector->panel;
-	enum pipe pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
+	enum pipe pipe = to_intel_crtc(crtc_state->uapi.crtc)->pipe;
 
 	if (!panel->backlight.present)
 		return;
@@ -1840,13 +1840,22 @@ static int pwm_setup_backlight(struct intel_connector *connector,
 			       enum pipe pipe)
 {
 	struct drm_device *dev = connector->base.dev;
+	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_panel *panel = &connector->panel;
+	const char *desc;
 	int retval;
 
-	/* Get the PWM chip for backlight control */
-	panel->backlight.pwm = pwm_get(dev->dev, "pwm_backlight");
+	/* Get the right PWM chip for DSI backlight according to VBT */
+	if (dev_priv->vbt.dsi.config->pwm_blc == PPS_BLC_PMIC) {
+		panel->backlight.pwm = pwm_get(dev->dev, "pwm_pmic_backlight");
+		desc = "PMIC";
+	} else {
+		panel->backlight.pwm = pwm_get(dev->dev, "pwm_soc_backlight");
+		desc = "SoC";
+	}
+
 	if (IS_ERR(panel->backlight.pwm)) {
-		DRM_ERROR("Failed to own the pwm chip\n");
+		DRM_ERROR("Failed to get the %s PWM chip\n", desc);
 		panel->backlight.pwm = NULL;
 		return -ENODEV;
 	}
@@ -1873,6 +1882,7 @@ static int pwm_setup_backlight(struct intel_connector *connector,
 				 CRC_PMIC_PWM_PERIOD_NS);
 	panel->backlight.enabled = panel->backlight.level != 0;
 
+	DRM_INFO("Using %s PWM for LCD backlight control\n", desc);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/i915/display/intel_pipe_crc.c b/drivers/gpu/drm/i915/display/intel_pipe_crc.c
index 6260a2082719..2746512f4466 100644
--- a/drivers/gpu/drm/i915/display/intel_pipe_crc.c
+++ b/drivers/gpu/drm/i915/display/intel_pipe_crc.c
@@ -309,13 +309,13 @@ retry:
 		goto put_state;
 	}
 
-	pipe_config->base.mode_changed = pipe_config->has_psr;
+	pipe_config->uapi.mode_changed = pipe_config->has_psr;
 	pipe_config->crc_enabled = enable;
 
 	if (IS_HASWELL(dev_priv) &&
-	    pipe_config->base.active && crtc->pipe == PIPE_A &&
+	    pipe_config->hw.active && crtc->pipe == PIPE_A &&
 	    pipe_config->cpu_transcoder == TRANSCODER_EDP)
-		pipe_config->base.mode_changed = true;
+		pipe_config->uapi.mode_changed = true;
 
 	ret = drm_atomic_commit(state);
 
diff --git a/drivers/gpu/drm/i915/display/intel_psr.c b/drivers/gpu/drm/i915/display/intel_psr.c
index 6a9f322d3fca..16e9ff47d519 100644
--- a/drivers/gpu/drm/i915/display/intel_psr.c
+++ b/drivers/gpu/drm/i915/display/intel_psr.c
@@ -26,6 +26,7 @@
 #include "display/intel_dp.h"
 
 #include "i915_drv.h"
+#include "intel_atomic.h"
 #include "intel_display_types.h"
 #include "intel_psr.h"
 #include "intel_sprite.h"
@@ -401,7 +402,9 @@ static void intel_psr_enable_sink(struct intel_dp *intel_dp)
 	/* Enable ALPM at sink for psr2 */
 	if (dev_priv->psr.psr2_enabled) {
 		drm_dp_dpcd_writeb(&intel_dp->aux, DP_RECEIVER_ALPM_CONFIG,
-				   DP_ALPM_ENABLE);
+				   DP_ALPM_ENABLE |
+				   DP_ALPM_LOCK_ERROR_IRQ_HPD_ENABLE);
+
 		dpcd_val |= DP_PSR_ENABLE_PSR2 | DP_PSR_IRQ_HPD_WITH_CRC_ERRORS;
 	} else {
 		if (dev_priv->psr.link_standby)
@@ -536,11 +539,11 @@ transcoder_has_psr2(struct drm_i915_private *dev_priv, enum transcoder trans)
 
 static u32 intel_get_frame_time_us(const struct intel_crtc_state *cstate)
 {
-	if (!cstate || !cstate->base.active)
+	if (!cstate || !cstate->hw.active)
 		return 0;
 
 	return DIV_ROUND_UP(1000 * 1000,
-			    drm_mode_vrefresh(&cstate->base.adjusted_mode));
+			    drm_mode_vrefresh(&cstate->hw.adjusted_mode));
 }
 
 static void psr2_program_idle_frames(struct drm_i915_private *dev_priv,
@@ -605,9 +608,9 @@ static bool intel_psr2_config_valid(struct intel_dp *intel_dp,
 				    struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
-	int crtc_hdisplay = crtc_state->base.adjusted_mode.crtc_hdisplay;
-	int crtc_vdisplay = crtc_state->base.adjusted_mode.crtc_vdisplay;
-	int psr_max_h = 0, psr_max_v = 0;
+	int crtc_hdisplay = crtc_state->hw.adjusted_mode.crtc_hdisplay;
+	int crtc_vdisplay = crtc_state->hw.adjusted_mode.crtc_vdisplay;
+	int psr_max_h = 0, psr_max_v = 0, max_bpp = 0;
 
 	if (!dev_priv->psr.sink_psr2_support)
 		return false;
@@ -631,12 +634,15 @@ static bool intel_psr2_config_valid(struct intel_dp *intel_dp,
 	if (INTEL_GEN(dev_priv) >= 12) {
 		psr_max_h = 5120;
 		psr_max_v = 3200;
+		max_bpp = 30;
 	} else if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
 		psr_max_h = 4096;
 		psr_max_v = 2304;
+		max_bpp = 24;
 	} else if (IS_GEN(dev_priv, 9)) {
 		psr_max_h = 3640;
 		psr_max_v = 2304;
+		max_bpp = 24;
 	}
 
 	if (crtc_hdisplay > psr_max_h || crtc_vdisplay > psr_max_v) {
@@ -646,6 +652,12 @@ static bool intel_psr2_config_valid(struct intel_dp *intel_dp,
 		return false;
 	}
 
+	if (crtc_state->pipe_bpp > max_bpp) {
+		DRM_DEBUG_KMS("PSR2 not enabled, pipe bpp %d > max supported %d\n",
+			      crtc_state->pipe_bpp, max_bpp);
+		return false;
+	}
+
 	/*
 	 * HW sends SU blocks of size four scan lines, which means the starting
 	 * X coordinate and Y granularity requirements will always be met. We
@@ -672,7 +684,7 @@ void intel_psr_compute_config(struct intel_dp *intel_dp,
 	struct intel_digital_port *dig_port = dp_to_dig_port(intel_dp);
 	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	int psr_setup_time;
 
 	if (!CAN_PSR(dev_priv))
@@ -792,7 +804,7 @@ static void intel_psr_enable_locked(struct drm_i915_private *dev_priv,
 
 	dev_priv->psr.psr2_enabled = intel_psr2_enabled(dev_priv, crtc_state);
 	dev_priv->psr.busy_frontbuffer_bits = 0;
-	dev_priv->psr.pipe = to_intel_crtc(crtc_state->base.crtc)->pipe;
+	dev_priv->psr.pipe = to_intel_crtc(crtc_state->uapi.crtc)->pipe;
 	dev_priv->psr.dc3co_enabled = !!crtc_state->dc3co_exitline;
 	dev_priv->psr.dc3co_exit_delay = intel_get_frame_time_us(crtc_state);
 	dev_priv->psr.transcoder = crtc_state->cpu_transcoder;
@@ -924,6 +936,9 @@ static void intel_psr_disable_locked(struct intel_dp *intel_dp)
 	/* Disable PSR on Sink */
 	drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_EN_CFG, 0);
 
+	if (dev_priv->psr.psr2_enabled)
+		drm_dp_dpcd_writeb(&intel_dp->aux, DP_RECEIVER_ALPM_CONFIG, 0);
+
 	dev_priv->psr.enabled = false;
 }
 
@@ -1039,7 +1054,7 @@ unlock:
 int intel_psr_wait_for_idle(const struct intel_crtc_state *new_crtc_state,
 			    u32 *out_value)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
 	if (!dev_priv->psr.enabled || !new_crtc_state->has_psr)
@@ -1096,7 +1111,7 @@ static int intel_psr_fastset_force(struct drm_i915_private *dev_priv)
 	struct drm_device *dev = &dev_priv->drm;
 	struct drm_modeset_acquire_ctx ctx;
 	struct drm_atomic_state *state;
-	struct drm_crtc *crtc;
+	struct intel_crtc *crtc;
 	int err;
 
 	state = drm_atomic_state_alloc(dev);
@@ -1107,21 +1122,18 @@ static int intel_psr_fastset_force(struct drm_i915_private *dev_priv)
 	state->acquire_ctx = &ctx;
 
 retry:
-	drm_for_each_crtc(crtc, dev) {
-		struct drm_crtc_state *crtc_state;
-		struct intel_crtc_state *intel_crtc_state;
+	for_each_intel_crtc(dev, crtc) {
+		struct intel_crtc_state *crtc_state =
+			intel_atomic_get_crtc_state(state, crtc);
 
-		crtc_state = drm_atomic_get_crtc_state(state, crtc);
 		if (IS_ERR(crtc_state)) {
 			err = PTR_ERR(crtc_state);
 			goto error;
 		}
 
-		intel_crtc_state = to_intel_crtc_state(crtc_state);
-
-		if (crtc_state->active && intel_crtc_state->has_psr) {
+		if (crtc_state->hw.active && crtc_state->has_psr) {
 			/* Mark mode as changed to trigger a pipe->update() */
-			crtc_state->mode_changed = true;
+			crtc_state->uapi.mode_changed = true;
 			break;
 		}
 	}
@@ -1379,11 +1391,80 @@ void intel_psr_init(struct drm_i915_private *dev_priv)
 	mutex_init(&dev_priv->psr.lock);
 }
 
-void intel_psr_short_pulse(struct intel_dp *intel_dp)
+static int psr_get_status_and_error_status(struct intel_dp *intel_dp,
+					   u8 *status, u8 *error_status)
+{
+	struct drm_dp_aux *aux = &intel_dp->aux;
+	int ret;
+
+	ret = drm_dp_dpcd_readb(aux, DP_PSR_STATUS, status);
+	if (ret != 1)
+		return ret;
+
+	ret = drm_dp_dpcd_readb(aux, DP_PSR_ERROR_STATUS, error_status);
+	if (ret != 1)
+		return ret;
+
+	*status = *status & DP_PSR_SINK_STATE_MASK;
+
+	return 0;
+}
+
+static void psr_alpm_check(struct intel_dp *intel_dp)
+{
+	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
+	struct drm_dp_aux *aux = &intel_dp->aux;
+	struct i915_psr *psr = &dev_priv->psr;
+	u8 val;
+	int r;
+
+	if (!psr->psr2_enabled)
+		return;
+
+	r = drm_dp_dpcd_readb(aux, DP_RECEIVER_ALPM_STATUS, &val);
+	if (r != 1) {
+		DRM_ERROR("Error reading ALPM status\n");
+		return;
+	}
+
+	if (val & DP_ALPM_LOCK_TIMEOUT_ERROR) {
+		intel_psr_disable_locked(intel_dp);
+		psr->sink_not_reliable = true;
+		DRM_DEBUG_KMS("ALPM lock timeout error, disabling PSR\n");
+
+		/* Clearing error */
+		drm_dp_dpcd_writeb(aux, DP_RECEIVER_ALPM_STATUS, val);
+	}
+}
+
+static void psr_capability_changed_check(struct intel_dp *intel_dp)
 {
 	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
 	struct i915_psr *psr = &dev_priv->psr;
 	u8 val;
+	int r;
+
+	r = drm_dp_dpcd_readb(&intel_dp->aux, DP_PSR_ESI, &val);
+	if (r != 1) {
+		DRM_ERROR("Error reading DP_PSR_ESI\n");
+		return;
+	}
+
+	if (val & DP_PSR_CAPS_CHANGE) {
+		intel_psr_disable_locked(intel_dp);
+		psr->sink_not_reliable = true;
+		DRM_DEBUG_KMS("Sink PSR capability changed, disabling PSR\n");
+
+		/* Clearing it */
+		drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_ESI, val);
+	}
+}
+
+void intel_psr_short_pulse(struct intel_dp *intel_dp)
+{
+	struct drm_i915_private *dev_priv = dp_to_i915(intel_dp);
+	struct i915_psr *psr = &dev_priv->psr;
+	u8 status, error_status;
 	const u8 errors = DP_PSR_RFB_STORAGE_ERROR |
 			  DP_PSR_VSC_SDP_UNCORRECTABLE_ERROR |
 			  DP_PSR_LINK_CRC_ERROR;
@@ -1396,38 +1477,34 @@ void intel_psr_short_pulse(struct intel_dp *intel_dp)
 	if (!psr->enabled || psr->dp != intel_dp)
 		goto exit;
 
-	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_PSR_STATUS, &val) != 1) {
-		DRM_ERROR("PSR_STATUS dpcd read failed\n");
+	if (psr_get_status_and_error_status(intel_dp, &status, &error_status)) {
+		DRM_ERROR("Error reading PSR status or error status\n");
 		goto exit;
 	}
 
-	if ((val & DP_PSR_SINK_STATE_MASK) == DP_PSR_SINK_INTERNAL_ERROR) {
-		DRM_DEBUG_KMS("PSR sink internal error, disabling PSR\n");
+	if (status == DP_PSR_SINK_INTERNAL_ERROR || (error_status & errors)) {
 		intel_psr_disable_locked(intel_dp);
 		psr->sink_not_reliable = true;
 	}
 
-	if (drm_dp_dpcd_readb(&intel_dp->aux, DP_PSR_ERROR_STATUS, &val) != 1) {
-		DRM_ERROR("PSR_ERROR_STATUS dpcd read failed\n");
-		goto exit;
-	}
-
-	if (val & DP_PSR_RFB_STORAGE_ERROR)
+	if (status == DP_PSR_SINK_INTERNAL_ERROR && !error_status)
+		DRM_DEBUG_KMS("PSR sink internal error, disabling PSR\n");
+	if (error_status & DP_PSR_RFB_STORAGE_ERROR)
 		DRM_DEBUG_KMS("PSR RFB storage error, disabling PSR\n");
-	if (val & DP_PSR_VSC_SDP_UNCORRECTABLE_ERROR)
+	if (error_status & DP_PSR_VSC_SDP_UNCORRECTABLE_ERROR)
 		DRM_DEBUG_KMS("PSR VSC SDP uncorrectable error, disabling PSR\n");
-	if (val & DP_PSR_LINK_CRC_ERROR)
+	if (error_status & DP_PSR_LINK_CRC_ERROR)
 		DRM_DEBUG_KMS("PSR Link CRC error, disabling PSR\n");
 
-	if (val & ~errors)
+	if (error_status & ~errors)
 		DRM_ERROR("PSR_ERROR_STATUS unhandled errors %x\n",
-			  val & ~errors);
-	if (val & errors) {
-		intel_psr_disable_locked(intel_dp);
-		psr->sink_not_reliable = true;
-	}
+			  error_status & ~errors);
 	/* clear status register */
-	drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_ERROR_STATUS, val);
+	drm_dp_dpcd_writeb(&intel_dp->aux, DP_PSR_ERROR_STATUS, error_status);
+
+	psr_alpm_check(intel_dp);
+	psr_capability_changed_check(intel_dp);
+
 exit:
 	mutex_unlock(&psr->lock);
 }
diff --git a/drivers/gpu/drm/i915/display/intel_sdvo.c b/drivers/gpu/drm/i915/display/intel_sdvo.c
index 5b7f4baf7348..8758ee2a4442 100644
--- a/drivers/gpu/drm/i915/display/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/display/intel_sdvo.c
@@ -1087,7 +1087,7 @@ static bool intel_sdvo_compute_avi_infoframe(struct intel_sdvo *intel_sdvo,
 {
 	struct hdmi_avi_infoframe *frame = &crtc_state->infoframes.avi.avi;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	int ret;
 
 	if (!crtc_state->has_hdmi_sink)
@@ -1276,8 +1276,8 @@ static int intel_sdvo_compute_config(struct intel_encoder *encoder,
 		to_intel_sdvo_connector_state(conn_state);
 	struct intel_sdvo_connector *intel_sdvo_connector =
 		to_intel_sdvo_connector(conn_state->connector);
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
-	struct drm_display_mode *mode = &pipe_config->base.mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
+	struct drm_display_mode *mode = &pipe_config->hw.mode;
 
 	DRM_DEBUG_KMS("forcing bpc to 8 for SDVO\n");
 	pipe_config->pipe_bpp = 8*3;
@@ -1349,9 +1349,9 @@ static int intel_sdvo_compute_config(struct intel_encoder *encoder,
 	if (IS_TV(intel_sdvo_connector))
 		i9xx_adjust_sdvo_tv_clock(pipe_config);
 
-	/* Set user selected PAR to incoming mode's member */
-	if (intel_sdvo_connector->is_hdmi)
-		adjusted_mode->picture_aspect_ratio = conn_state->picture_aspect_ratio;
+	if (conn_state->picture_aspect_ratio)
+		adjusted_mode->picture_aspect_ratio =
+			conn_state->picture_aspect_ratio;
 
 	if (!intel_sdvo_compute_avi_infoframe(intel_sdvo,
 					      pipe_config, conn_state)) {
@@ -1429,13 +1429,13 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder,
 				  const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(intel_encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	const struct drm_display_mode *adjusted_mode = &crtc_state->base.adjusted_mode;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_display_mode *adjusted_mode = &crtc_state->hw.adjusted_mode;
 	const struct intel_sdvo_connector_state *sdvo_state =
 		to_intel_sdvo_connector_state(conn_state);
 	const struct intel_sdvo_connector *intel_sdvo_connector =
 		to_intel_sdvo_connector(conn_state->connector);
-	const struct drm_display_mode *mode = &crtc_state->base.mode;
+	const struct drm_display_mode *mode = &crtc_state->hw.mode;
 	struct intel_sdvo *intel_sdvo = to_sdvo(intel_encoder);
 	u32 sdvox;
 	struct intel_sdvo_in_out_map in_out;
@@ -1629,7 +1629,7 @@ static void intel_sdvo_get_config(struct intel_encoder *encoder,
 			flags |= DRM_MODE_FLAG_NVSYNC;
 	}
 
-	pipe_config->base.adjusted_mode.flags |= flags;
+	pipe_config->hw.adjusted_mode.flags |= flags;
 
 	/*
 	 * pixel multiplier readout is tricky: Only on i915g/gm it is stored in
@@ -1649,7 +1649,7 @@ static void intel_sdvo_get_config(struct intel_encoder *encoder,
 	if (pipe_config->pixel_multiplier)
 		dotclock /= pipe_config->pixel_multiplier;
 
-	pipe_config->base.adjusted_mode.crtc_clock = dotclock;
+	pipe_config->hw.adjusted_mode.crtc_clock = dotclock;
 
 	/* Cross check the port pixel multiplier with the sdvo encoder state. */
 	if (intel_sdvo_get_value(intel_sdvo, SDVO_CMD_GET_CLOCK_RATE_MULT,
@@ -1701,7 +1701,7 @@ static void intel_sdvo_enable_audio(struct intel_sdvo *intel_sdvo,
 				    const struct drm_connector_state *conn_state)
 {
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	struct drm_connector *connector = conn_state->connector;
 	u8 *eld = connector->eld;
 
@@ -1723,7 +1723,7 @@ static void intel_disable_sdvo(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_sdvo *intel_sdvo = to_sdvo(encoder);
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	u32 temp;
 
 	if (old_crtc_state->has_audio)
@@ -1785,7 +1785,7 @@ static void intel_enable_sdvo(struct intel_encoder *encoder,
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_sdvo *intel_sdvo = to_sdvo(encoder);
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	u32 temp;
 	bool input1, input2;
 	int i;
@@ -2654,7 +2654,6 @@ intel_sdvo_add_hdmi_properties(struct intel_sdvo *intel_sdvo,
 		intel_attach_broadcast_rgb_property(&connector->base.base);
 	}
 	intel_attach_aspect_ratio_property(&connector->base.base);
-	connector->base.base.state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
 }
 
 static struct intel_sdvo_connector *intel_sdvo_connector_alloc(void)
diff --git a/drivers/gpu/drm/i915/display/intel_sprite.c b/drivers/gpu/drm/i915/display/intel_sprite.c
index 72fda0430062..3f7b8f2ff671 100644
--- a/drivers/gpu/drm/i915/display/intel_sprite.c
+++ b/drivers/gpu/drm/i915/display/intel_sprite.c
@@ -81,9 +81,9 @@ int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode,
  */
 void intel_pipe_update_start(const struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
-	const struct drm_display_mode *adjusted_mode = &new_crtc_state->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &new_crtc_state->hw.adjusted_mode;
 	long timeout = msecs_to_jiffies_timeout(1);
 	int scanline, min, max, vblank_start;
 	wait_queue_head_t *wq = drm_crtc_vblank_waitqueue(&crtc->base);
@@ -120,7 +120,7 @@ void intel_pipe_update_start(const struct intel_crtc_state *new_crtc_state)
 
 	crtc->debug.min_vbl = min;
 	crtc->debug.max_vbl = max;
-	trace_i915_pipe_update_start(crtc);
+	trace_intel_pipe_update_start(crtc);
 
 	for (;;) {
 		/*
@@ -173,7 +173,7 @@ void intel_pipe_update_start(const struct intel_crtc_state *new_crtc_state)
 	crtc->debug.start_vbl_time = ktime_get();
 	crtc->debug.start_vbl_count = intel_crtc_get_vblank_counter(crtc);
 
-	trace_i915_pipe_update_vblank_evaded(crtc);
+	trace_intel_pipe_update_vblank_evaded(crtc);
 	return;
 
 irq_disable:
@@ -190,27 +190,28 @@ irq_disable:
  */
 void intel_pipe_update_end(struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	enum pipe pipe = crtc->pipe;
 	int scanline_end = intel_get_crtc_scanline(crtc);
 	u32 end_vbl_count = intel_crtc_get_vblank_counter(crtc);
 	ktime_t end_vbl_time = ktime_get();
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 
-	trace_i915_pipe_update_end(crtc, end_vbl_count, scanline_end);
+	trace_intel_pipe_update_end(crtc, end_vbl_count, scanline_end);
 
 	/* We're still in the vblank-evade critical section, this can't race.
 	 * Would be slightly nice to just grab the vblank count and arm the
 	 * event outside of the critical section - the spinlock might spin for a
 	 * while ... */
-	if (new_crtc_state->base.event) {
+	if (new_crtc_state->uapi.event) {
 		WARN_ON(drm_crtc_vblank_get(&crtc->base) != 0);
 
 		spin_lock(&crtc->base.dev->event_lock);
-		drm_crtc_arm_vblank_event(&crtc->base, new_crtc_state->base.event);
+		drm_crtc_arm_vblank_event(&crtc->base,
+				          new_crtc_state->uapi.event);
 		spin_unlock(&crtc->base.dev->event_lock);
 
-		new_crtc_state->base.event = NULL;
+		new_crtc_state->uapi.event = NULL;
 	}
 
 	local_irq_enable();
@@ -239,9 +240,9 @@ void intel_pipe_update_end(struct intel_crtc_state *new_crtc_state)
 
 int intel_plane_check_stride(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	u32 stride, max_stride;
 
 	/*
@@ -251,7 +252,7 @@ int intel_plane_check_stride(const struct intel_plane_state *plane_state)
 	 * kick in due the plane being invisible.
 	 */
 	if (intel_plane_can_remap(plane_state) &&
-	    !plane_state->base.visible)
+	    !plane_state->uapi.visible)
 		return 0;
 
 	/* FIXME other color planes? */
@@ -271,10 +272,10 @@ int intel_plane_check_stride(const struct intel_plane_state *plane_state)
 
 int intel_plane_check_src_coordinates(struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	struct drm_rect *src = &plane_state->base.src;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	struct drm_rect *src = &plane_state->uapi.src;
 	u32 src_x, src_y, src_w, src_h, hsub, vsub;
-	bool rotated = drm_rotation_90_or_270(plane_state->base.rotation);
+	bool rotated = drm_rotation_90_or_270(plane_state->hw.rotation);
 
 	/*
 	 * Hardware doesn't handle subpixel coordinates.
@@ -327,8 +328,8 @@ skl_plane_ratio(const struct intel_crtc_state *crtc_state,
 		const struct intel_plane_state *plane_state,
 		unsigned int *num, unsigned int *den)
 {
-	struct drm_i915_private *dev_priv = to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	struct drm_i915_private *dev_priv = to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 
 	if (fb->format->cpp[0] == 8) {
 		if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) {
@@ -347,7 +348,7 @@ skl_plane_ratio(const struct intel_crtc_state *crtc_state,
 static int skl_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
 			       const struct intel_plane_state *plane_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(plane_state->base.plane->dev);
+	struct drm_i915_private *dev_priv = to_i915(plane_state->uapi.plane->dev);
 	unsigned int pixel_rate = crtc_state->pixel_rate;
 	unsigned int src_w, src_h, dst_w, dst_h;
 	unsigned int num, den;
@@ -358,10 +359,10 @@ static int skl_plane_min_cdclk(const struct intel_crtc_state *crtc_state,
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 		den *= 2;
 
-	src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	src_h = drm_rect_height(&plane_state->base.src) >> 16;
-	dst_w = drm_rect_width(&plane_state->base.dst);
-	dst_h = drm_rect_height(&plane_state->base.dst);
+	src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
+	dst_w = drm_rect_width(&plane_state->uapi.dst);
+	dst_h = drm_rect_height(&plane_state->uapi.dst);
 
 	/* Downscaling limits the maximum pixel rate */
 	dst_w = min(src_w, dst_w);
@@ -395,28 +396,28 @@ skl_program_scaler(struct intel_plane *plane,
 		   const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum pipe pipe = plane->pipe;
 	int scaler_id = plane_state->scaler_id;
 	const struct intel_scaler *scaler =
 		&crtc_state->scaler_state.scalers[scaler_id];
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_y = plane_state->base.dst.y1;
-	u32 crtc_w = drm_rect_width(&plane_state->base.dst);
-	u32 crtc_h = drm_rect_height(&plane_state->base.dst);
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_y = plane_state->uapi.dst.y1;
+	u32 crtc_w = drm_rect_width(&plane_state->uapi.dst);
+	u32 crtc_h = drm_rect_height(&plane_state->uapi.dst);
 	u16 y_hphase, uv_rgb_hphase;
 	u16 y_vphase, uv_rgb_vphase;
 	int hscale, vscale;
 
-	hscale = drm_rect_calc_hscale(&plane_state->base.src,
-				      &plane_state->base.dst,
+	hscale = drm_rect_calc_hscale(&plane_state->uapi.src,
+				      &plane_state->uapi.dst,
 				      0, INT_MAX);
-	vscale = drm_rect_calc_vscale(&plane_state->base.src,
-				      &plane_state->base.dst,
+	vscale = drm_rect_calc_vscale(&plane_state->uapi.src,
+				      &plane_state->uapi.dst,
 				      0, INT_MAX);
 
 	/* TODO: handle sub-pixel coordinates */
-	if (drm_format_info_is_yuv_semiplanar(fb->format) &&
+	if (intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier) &&
 	    !icl_is_hdr_plane(dev_priv, plane->id)) {
 		y_hphase = skl_scaler_calc_phase(1, hscale, false);
 		y_vphase = skl_scaler_calc_phase(1, vscale, false);
@@ -541,10 +542,10 @@ icl_program_input_csc(struct intel_plane *plane,
 	};
 	const u16 *csc;
 
-	if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
-		csc = input_csc_matrix[plane_state->base.color_encoding];
+	if (plane_state->hw.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
+		csc = input_csc_matrix[plane_state->hw.color_encoding];
 	else
-		csc = input_csc_matrix_lr[plane_state->base.color_encoding];
+		csc = input_csc_matrix_lr[plane_state->hw.color_encoding];
 
 	I915_WRITE_FW(PLANE_INPUT_CSC_COEFF(pipe, plane_id, 0), ROFF(csc[0]) |
 		      GOFF(csc[1]));
@@ -558,7 +559,7 @@ icl_program_input_csc(struct intel_plane *plane,
 
 	I915_WRITE_FW(PLANE_INPUT_CSC_PREOFF(pipe, plane_id, 0),
 		      PREOFF_YUV_TO_RGB_HI);
-	if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
+	if (plane_state->hw.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
 		I915_WRITE_FW(PLANE_INPUT_CSC_PREOFF(pipe, plane_id, 1), 0);
 	else
 		I915_WRITE_FW(PLANE_INPUT_CSC_PREOFF(pipe, plane_id, 1),
@@ -574,7 +575,7 @@ static void
 skl_program_plane(struct intel_plane *plane,
 		  const struct intel_crtc_state *crtc_state,
 		  const struct intel_plane_state *plane_state,
-		  int color_plane, bool slave, u32 plane_ctl)
+		  int color_plane)
 {
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	enum plane_id plane_id = plane->id;
@@ -582,19 +583,20 @@ skl_program_plane(struct intel_plane *plane,
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 	u32 surf_addr = plane_state->color_plane[color_plane].offset;
 	u32 stride = skl_plane_stride(plane_state, color_plane);
+	u32 aux_dist = plane_state->color_plane[1].offset - surf_addr;
 	u32 aux_stride = skl_plane_stride(plane_state, 1);
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_y = plane_state->base.dst.y1;
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_y = plane_state->uapi.dst.y1;
 	u32 x = plane_state->color_plane[color_plane].x;
 	u32 y = plane_state->color_plane[color_plane].y;
-	u32 src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	u32 src_h = drm_rect_height(&plane_state->base.src) >> 16;
-	struct intel_plane *linked = plane_state->planar_linked_plane;
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	u8 alpha = plane_state->base.alpha >> 8;
+	u32 src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	u32 src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	u8 alpha = plane_state->hw.alpha >> 8;
 	u32 plane_color_ctl = 0;
 	unsigned long irqflags;
 	u32 keymsk, keymax;
+	u32 plane_ctl = plane_state->ctl;
 
 	plane_ctl |= skl_plane_ctl_crtc(crtc_state);
 
@@ -623,29 +625,13 @@ skl_program_plane(struct intel_plane *plane,
 	I915_WRITE_FW(PLANE_STRIDE(pipe, plane_id), stride);
 	I915_WRITE_FW(PLANE_POS(pipe, plane_id), (crtc_y << 16) | crtc_x);
 	I915_WRITE_FW(PLANE_SIZE(pipe, plane_id), (src_h << 16) | src_w);
-	I915_WRITE_FW(PLANE_AUX_DIST(pipe, plane_id),
-		      (plane_state->color_plane[1].offset - surf_addr) | aux_stride);
 
-	if (icl_is_hdr_plane(dev_priv, plane_id)) {
-		u32 cus_ctl = 0;
-
-		if (linked) {
-			/* Enable and use MPEG-2 chroma siting */
-			cus_ctl = PLANE_CUS_ENABLE |
-				PLANE_CUS_HPHASE_0 |
-				PLANE_CUS_VPHASE_SIGN_NEGATIVE |
-				PLANE_CUS_VPHASE_0_25;
-
-			if (linked->id == PLANE_SPRITE5)
-				cus_ctl |= PLANE_CUS_PLANE_7;
-			else if (linked->id == PLANE_SPRITE4)
-				cus_ctl |= PLANE_CUS_PLANE_6;
-			else
-				MISSING_CASE(linked->id);
-		}
+	if (INTEL_GEN(dev_priv) < 12)
+		aux_dist |= aux_stride;
+	I915_WRITE_FW(PLANE_AUX_DIST(pipe, plane_id), aux_dist);
 
-		I915_WRITE_FW(PLANE_CUS_CTL(pipe, plane_id), cus_ctl);
-	}
+	if (icl_is_hdr_plane(dev_priv, plane_id))
+		I915_WRITE_FW(PLANE_CUS_CTL(pipe, plane_id), plane_state->cus_ctl);
 
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv))
 		I915_WRITE_FW(PLANE_COLOR_CTL(pipe, plane_id), plane_color_ctl);
@@ -675,7 +661,7 @@ skl_program_plane(struct intel_plane *plane,
 	I915_WRITE_FW(PLANE_SURF(pipe, plane_id),
 		      intel_plane_ggtt_offset(plane_state) + surf_addr);
 
-	if (!slave && plane_state->scaler_id >= 0)
+	if (plane_state->scaler_id >= 0)
 		skl_program_scaler(plane, crtc_state, plane_state);
 
 	spin_unlock_irqrestore(&dev_priv->uncore.lock, irqflags);
@@ -688,24 +674,12 @@ skl_update_plane(struct intel_plane *plane,
 {
 	int color_plane = 0;
 
-	if (plane_state->planar_linked_plane) {
-		/* Program the UV plane */
+	if (plane_state->planar_linked_plane && !plane_state->planar_slave)
+		/* Program the UV plane on planar master */
 		color_plane = 1;
-	}
-
-	skl_program_plane(plane, crtc_state, plane_state,
-			  color_plane, false, plane_state->ctl);
-}
 
-static void
-icl_update_slave(struct intel_plane *plane,
-		 const struct intel_crtc_state *crtc_state,
-		 const struct intel_plane_state *plane_state)
-{
-	skl_program_plane(plane, crtc_state, plane_state, 0, true,
-			  plane_state->ctl | PLANE_CTL_YUV420_Y_PLANE);
+	skl_program_plane(plane, crtc_state, plane_state, color_plane);
 }
-
 static void
 skl_disable_plane(struct intel_plane *plane,
 		  const struct intel_crtc_state *crtc_state)
@@ -765,9 +739,9 @@ static void i9xx_plane_linear_gamma(u16 gamma[8])
 static void
 chv_update_csc(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum plane_id plane_id = plane->id;
 	/*
 	 * |r|   | c0 c1 c2 |   |cr|
@@ -793,7 +767,7 @@ chv_update_csc(const struct intel_plane_state *plane_state)
 			    0, 4096,  7601,
 		},
 	};
-	const s16 *csc = csc_matrix[plane_state->base.color_encoding];
+	const s16 *csc = csc_matrix[plane_state->hw.color_encoding];
 
 	/* Seems RGB data bypasses the CSC always */
 	if (!fb->format->is_yuv)
@@ -824,15 +798,15 @@ chv_update_csc(const struct intel_plane_state *plane_state)
 static void
 vlv_update_clrc(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum pipe pipe = plane->pipe;
 	enum plane_id plane_id = plane->id;
 	int contrast, brightness, sh_scale, sh_sin, sh_cos;
 
 	if (fb->format->is_yuv &&
-	    plane_state->base.color_range == DRM_COLOR_YCBCR_LIMITED_RANGE) {
+	    plane_state->hw.color_range == DRM_COLOR_YCBCR_LIMITED_RANGE) {
 		/*
 		 * Expand limited range to full range:
 		 * Contrast is applied first and is used to expand Y range.
@@ -866,7 +840,7 @@ vlv_plane_ratio(const struct intel_crtc_state *crtc_state,
 		unsigned int *num, unsigned int *den)
 {
 	u8 active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int cpp = fb->format->cpp[0];
 
 	/*
@@ -952,8 +926,8 @@ static u32 vlv_sprite_ctl_crtc(const struct intel_crtc_state *crtc_state)
 static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 	u32 sprctl;
 
@@ -972,6 +946,9 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_VYUY:
 		sprctl |= SP_FORMAT_YUV422 | SP_YUV_ORDER_VYUY;
 		break;
+	case DRM_FORMAT_C8:
+		sprctl |= SP_FORMAT_8BPP;
+		break;
 	case DRM_FORMAT_RGB565:
 		sprctl |= SP_FORMAT_BGR565;
 		break;
@@ -987,6 +964,12 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_ABGR2101010:
 		sprctl |= SP_FORMAT_RGBA1010102;
 		break;
+	case DRM_FORMAT_XRGB2101010:
+		sprctl |= SP_FORMAT_BGRX1010102;
+		break;
+	case DRM_FORMAT_ARGB2101010:
+		sprctl |= SP_FORMAT_BGRA1010102;
+		break;
 	case DRM_FORMAT_XBGR8888:
 		sprctl |= SP_FORMAT_RGBX8888;
 		break;
@@ -998,7 +981,7 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 		return 0;
 	}
 
-	if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
+	if (plane_state->hw.color_encoding == DRM_COLOR_YCBCR_BT709)
 		sprctl |= SP_YUV_FORMAT_BT709;
 
 	if (fb->modifier == I915_FORMAT_MOD_X_TILED)
@@ -1018,9 +1001,9 @@ static u32 vlv_sprite_ctl(const struct intel_crtc_state *crtc_state,
 
 static void vlv_update_gamma(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum pipe pipe = plane->pipe;
 	enum plane_id plane_id = plane->id;
 	u16 gamma[8];
@@ -1052,10 +1035,10 @@ vlv_update_plane(struct intel_plane *plane,
 	u32 sprsurf_offset = plane_state->color_plane[0].offset;
 	u32 linear_offset;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_y = plane_state->base.dst.y1;
-	u32 crtc_w = drm_rect_width(&plane_state->base.dst);
-	u32 crtc_h = drm_rect_height(&plane_state->base.dst);
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_y = plane_state->uapi.dst.y1;
+	u32 crtc_w = drm_rect_width(&plane_state->uapi.dst);
+	u32 crtc_h = drm_rect_height(&plane_state->uapi.dst);
 	u32 x = plane_state->color_plane[0].x;
 	u32 y = plane_state->color_plane[0].y;
 	unsigned long irqflags;
@@ -1150,7 +1133,7 @@ static void ivb_plane_ratio(const struct intel_crtc_state *crtc_state,
 			    unsigned int *num, unsigned int *den)
 {
 	u8 active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int cpp = fb->format->cpp[0];
 
 	if (hweight8(active_planes) == 2) {
@@ -1186,7 +1169,7 @@ static void ivb_plane_ratio_scaling(const struct intel_crtc_state *crtc_state,
 				    const struct intel_plane_state *plane_state,
 				    unsigned int *num, unsigned int *den)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int cpp = fb->format->cpp[0];
 
 	switch (cpp) {
@@ -1244,8 +1227,8 @@ static int ivb_sprite_min_cdclk(const struct intel_crtc_state *crtc_state,
 	 */
 	pixel_rate = crtc_state->pixel_rate;
 
-	src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	dst_w = drm_rect_width(&plane_state->base.dst);
+	src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	dst_w = drm_rect_width(&plane_state->uapi.dst);
 
 	if (src_w != dst_w)
 		ivb_plane_ratio_scaling(crtc_state, plane_state, &num, &den);
@@ -1264,7 +1247,7 @@ static void hsw_plane_ratio(const struct intel_crtc_state *crtc_state,
 			    unsigned int *num, unsigned int *den)
 {
 	u8 active_planes = crtc_state->active_planes & ~BIT(PLANE_CURSOR);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int cpp = fb->format->cpp[0];
 
 	if (hweight8(active_planes) == 2) {
@@ -1319,8 +1302,8 @@ static u32 ivb_sprite_ctl_crtc(const struct intel_crtc_state *crtc_state)
 static bool ivb_need_sprite_gamma(const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 
 	return fb->format->cpp[0] == 8 &&
 		(IS_IVYBRIDGE(dev_priv) || IS_HASWELL(dev_priv));
@@ -1330,9 +1313,9 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 	u32 sprctl;
 
@@ -1348,6 +1331,12 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XRGB8888:
 		sprctl |= SPRITE_FORMAT_RGBX888;
 		break;
+	case DRM_FORMAT_XBGR2101010:
+		sprctl |= SPRITE_FORMAT_RGBX101010 | SPRITE_RGB_ORDER_RGBX;
+		break;
+	case DRM_FORMAT_XRGB2101010:
+		sprctl |= SPRITE_FORMAT_RGBX101010;
+		break;
 	case DRM_FORMAT_XBGR16161616F:
 		sprctl |= SPRITE_FORMAT_RGBX161616 | SPRITE_RGB_ORDER_RGBX;
 		break;
@@ -1374,10 +1363,10 @@ static u32 ivb_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	if (!ivb_need_sprite_gamma(plane_state))
 		sprctl |= SPRITE_INT_GAMMA_DISABLE;
 
-	if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
+	if (plane_state->hw.color_encoding == DRM_COLOR_YCBCR_BT709)
 		sprctl |= SPRITE_YUV_TO_RGB_CSC_FORMAT_BT709;
 
-	if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
+	if (plane_state->hw.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
 		sprctl |= SPRITE_YUV_RANGE_CORRECTION_DISABLE;
 
 	if (fb->modifier == I915_FORMAT_MOD_X_TILED)
@@ -1421,7 +1410,7 @@ static void ivb_sprite_linear_gamma(const struct intel_plane_state *plane_state,
 
 static void ivb_update_gamma(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	enum pipe pipe = plane->pipe;
 	u16 gamma[18];
@@ -1460,14 +1449,14 @@ ivb_update_plane(struct intel_plane *plane,
 	u32 sprsurf_offset = plane_state->color_plane[0].offset;
 	u32 linear_offset;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_y = plane_state->base.dst.y1;
-	u32 crtc_w = drm_rect_width(&plane_state->base.dst);
-	u32 crtc_h = drm_rect_height(&plane_state->base.dst);
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_y = plane_state->uapi.dst.y1;
+	u32 crtc_w = drm_rect_width(&plane_state->uapi.dst);
+	u32 crtc_h = drm_rect_height(&plane_state->uapi.dst);
 	u32 x = plane_state->color_plane[0].x;
 	u32 y = plane_state->color_plane[0].y;
-	u32 src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	u32 src_h = drm_rect_height(&plane_state->base.src) >> 16;
+	u32 src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	u32 src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
 	u32 sprctl, sprscale = 0;
 	unsigned long irqflags;
 
@@ -1566,7 +1555,7 @@ ivb_plane_get_hw_state(struct intel_plane *plane,
 static int g4x_sprite_min_cdclk(const struct intel_crtc_state *crtc_state,
 				const struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	unsigned int hscale, pixel_rate;
 	unsigned int limit, decimate;
 
@@ -1580,8 +1569,8 @@ static int g4x_sprite_min_cdclk(const struct intel_crtc_state *crtc_state,
 	pixel_rate = crtc_state->pixel_rate;
 
 	/* Horizontal downscaling limits the maximum pixel rate */
-	hscale = drm_rect_calc_hscale(&plane_state->base.src,
-				      &plane_state->base.dst,
+	hscale = drm_rect_calc_hscale(&plane_state->uapi.src,
+				      &plane_state->uapi.dst,
 				      0, INT_MAX);
 	if (hscale < 0x10000)
 		return pixel_rate;
@@ -1635,9 +1624,9 @@ static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+		to_i915(plane_state->uapi.plane->dev);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 	u32 dvscntr;
 
@@ -1653,6 +1642,12 @@ static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state,
 	case DRM_FORMAT_XRGB8888:
 		dvscntr |= DVS_FORMAT_RGBX888;
 		break;
+	case DRM_FORMAT_XBGR2101010:
+		dvscntr |= DVS_FORMAT_RGBX101010 | DVS_RGB_ORDER_XBGR;
+		break;
+	case DRM_FORMAT_XRGB2101010:
+		dvscntr |= DVS_FORMAT_RGBX101010;
+		break;
 	case DRM_FORMAT_XBGR16161616F:
 		dvscntr |= DVS_FORMAT_RGBX161616 | DVS_RGB_ORDER_XBGR;
 		break;
@@ -1676,10 +1671,10 @@ static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state,
 		return 0;
 	}
 
-	if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
+	if (plane_state->hw.color_encoding == DRM_COLOR_YCBCR_BT709)
 		dvscntr |= DVS_YUV_FORMAT_BT709;
 
-	if (plane_state->base.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
+	if (plane_state->hw.color_range == DRM_COLOR_YCBCR_FULL_RANGE)
 		dvscntr |= DVS_YUV_RANGE_CORRECTION_DISABLE;
 
 	if (fb->modifier == I915_FORMAT_MOD_X_TILED)
@@ -1698,9 +1693,9 @@ static u32 g4x_sprite_ctl(const struct intel_crtc_state *crtc_state,
 
 static void g4x_update_gamma(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum pipe pipe = plane->pipe;
 	u16 gamma[8];
 	int i;
@@ -1730,9 +1725,9 @@ static void ilk_sprite_linear_gamma(u16 gamma[17])
 
 static void ilk_update_gamma(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum pipe pipe = plane->pipe;
 	u16 gamma[17];
 	int i;
@@ -1766,14 +1761,14 @@ g4x_update_plane(struct intel_plane *plane,
 	u32 dvssurf_offset = plane_state->color_plane[0].offset;
 	u32 linear_offset;
 	const struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_y = plane_state->base.dst.y1;
-	u32 crtc_w = drm_rect_width(&plane_state->base.dst);
-	u32 crtc_h = drm_rect_height(&plane_state->base.dst);
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_y = plane_state->uapi.dst.y1;
+	u32 crtc_w = drm_rect_width(&plane_state->uapi.dst);
+	u32 crtc_h = drm_rect_height(&plane_state->uapi.dst);
 	u32 x = plane_state->color_plane[0].x;
 	u32 y = plane_state->color_plane[0].y;
-	u32 src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	u32 src_h = drm_rect_height(&plane_state->base.src) >> 16;
+	u32 src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	u32 src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
 	u32 dvscntr, dvsscale = 0;
 	unsigned long irqflags;
 
@@ -1886,12 +1881,12 @@ static int
 g4x_sprite_check_scaling(struct intel_crtc_state *crtc_state,
 			 struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	const struct drm_rect *src = &plane_state->base.src;
-	const struct drm_rect *dst = &plane_state->base.dst;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	const struct drm_rect *src = &plane_state->uapi.src;
+	const struct drm_rect *dst = &plane_state->uapi.dst;
 	int src_x, src_w, src_h, crtc_w, crtc_h;
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	unsigned int stride = plane_state->color_plane[0].stride;
 	unsigned int cpp = fb->format->cpp[0];
 	unsigned int width_bytes;
@@ -1947,13 +1942,13 @@ static int
 g4x_sprite_check(struct intel_crtc_state *crtc_state,
 		 struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	int min_scale = DRM_PLANE_HELPER_NO_SCALING;
 	int max_scale = DRM_PLANE_HELPER_NO_SCALING;
 	int ret;
 
-	if (intel_fb_scalable(plane_state->base.fb)) {
+	if (intel_fb_scalable(plane_state->hw.fb)) {
 		if (INTEL_GEN(dev_priv) < 7) {
 			min_scale = 1;
 			max_scale = 16 << 16;
@@ -1963,8 +1958,8 @@ g4x_sprite_check(struct intel_crtc_state *crtc_state,
 		}
 	}
 
-	ret = drm_atomic_helper_check_plane_state(&plane_state->base,
-						  &crtc_state->base,
+	ret = drm_atomic_helper_check_plane_state(&plane_state->uapi,
+						  &crtc_state->uapi,
 						  min_scale, max_scale,
 						  true, true);
 	if (ret)
@@ -1974,7 +1969,7 @@ g4x_sprite_check(struct intel_crtc_state *crtc_state,
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	ret = intel_plane_check_src_coordinates(plane_state);
@@ -1995,9 +1990,9 @@ g4x_sprite_check(struct intel_crtc_state *crtc_state,
 
 int chv_plane_check_rotation(const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	unsigned int rotation = plane_state->base.rotation;
+	unsigned int rotation = plane_state->hw.rotation;
 
 	/* CHV ignores the mirror bit when the rotate bit is set :( */
 	if (IS_CHERRYVIEW(dev_priv) &&
@@ -2020,8 +2015,8 @@ vlv_sprite_check(struct intel_crtc_state *crtc_state,
 	if (ret)
 		return ret;
 
-	ret = drm_atomic_helper_check_plane_state(&plane_state->base,
-						  &crtc_state->base,
+	ret = drm_atomic_helper_check_plane_state(&plane_state->uapi,
+						  &crtc_state->uapi,
 						  DRM_PLANE_HELPER_NO_SCALING,
 						  DRM_PLANE_HELPER_NO_SCALING,
 						  true, true);
@@ -2032,7 +2027,7 @@ vlv_sprite_check(struct intel_crtc_state *crtc_state,
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	ret = intel_plane_check_src_coordinates(plane_state);
@@ -2047,10 +2042,10 @@ vlv_sprite_check(struct intel_crtc_state *crtc_state,
 static int skl_plane_check_fb(const struct intel_crtc_state *crtc_state,
 			      const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
 	struct drm_format_name_buf format_name;
 
 	if (!fb)
@@ -2105,12 +2100,13 @@ static int skl_plane_check_fb(const struct intel_crtc_state *crtc_state,
 	}
 
 	/* Y-tiling is not supported in IF-ID Interlace mode */
-	if (crtc_state->base.enable &&
-	    crtc_state->base.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE &&
+	if (crtc_state->hw.enable &&
+	    crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE &&
 	    (fb->modifier == I915_FORMAT_MOD_Y_TILED ||
 	     fb->modifier == I915_FORMAT_MOD_Yf_TILED ||
 	     fb->modifier == I915_FORMAT_MOD_Y_TILED_CCS ||
-	     fb->modifier == I915_FORMAT_MOD_Yf_TILED_CCS)) {
+	     fb->modifier == I915_FORMAT_MOD_Yf_TILED_CCS ||
+	     fb->modifier == I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS)) {
 		DRM_DEBUG_KMS("Y/Yf tiling not supported in IF-ID mode\n");
 		return -EINVAL;
 	}
@@ -2122,9 +2118,9 @@ static int skl_plane_check_dst_coordinates(const struct intel_crtc_state *crtc_s
 					   const struct intel_plane_state *plane_state)
 {
 	struct drm_i915_private *dev_priv =
-		to_i915(plane_state->base.plane->dev);
-	int crtc_x = plane_state->base.dst.x1;
-	int crtc_w = drm_rect_width(&plane_state->base.dst);
+		to_i915(plane_state->uapi.plane->dev);
+	int crtc_x = plane_state->uapi.dst.x1;
+	int crtc_w = drm_rect_width(&plane_state->uapi.dst);
 	int pipe_src_w = crtc_state->pipe_src_w;
 
 	/*
@@ -2150,12 +2146,13 @@ static int skl_plane_check_dst_coordinates(const struct intel_crtc_state *crtc_s
 
 static int skl_plane_check_nv12_rotation(const struct intel_plane_state *plane_state)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
-	unsigned int rotation = plane_state->base.rotation;
-	int src_w = drm_rect_width(&plane_state->base.src) >> 16;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	unsigned int rotation = plane_state->hw.rotation;
+	int src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
 
 	/* Display WA #1106 */
-	if (drm_format_info_is_yuv_semiplanar(fb->format) && src_w & 3 &&
+	if (intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier) &&
+	    src_w & 3 &&
 	    (rotation == DRM_MODE_ROTATE_270 ||
 	     rotation == (DRM_MODE_REFLECT_X | DRM_MODE_ROTATE_90))) {
 		DRM_DEBUG_KMS("src width must be multiple of 4 for rotated planar YUV\n");
@@ -2175,7 +2172,7 @@ static int skl_plane_max_scale(struct drm_i915_private *dev_priv,
 	 * FIXME need to properly check this later.
 	 */
 	if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv) ||
-	    !drm_format_info_is_yuv_semiplanar(fb->format))
+	    !intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier))
 		return 0x30000 - 1;
 	else
 		return 0x20000 - 1;
@@ -2184,9 +2181,9 @@ static int skl_plane_max_scale(struct drm_i915_private *dev_priv,
 static int skl_plane_check(struct intel_crtc_state *crtc_state,
 			   struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int min_scale = DRM_PLANE_HELPER_NO_SCALING;
 	int max_scale = DRM_PLANE_HELPER_NO_SCALING;
 	int ret;
@@ -2201,8 +2198,8 @@ static int skl_plane_check(struct intel_crtc_state *crtc_state,
 		max_scale = skl_plane_max_scale(dev_priv, fb);
 	}
 
-	ret = drm_atomic_helper_check_plane_state(&plane_state->base,
-						  &crtc_state->base,
+	ret = drm_atomic_helper_check_plane_state(&plane_state->uapi,
+						  &crtc_state->uapi,
 						  min_scale, max_scale,
 						  true, true);
 	if (ret)
@@ -2212,7 +2209,7 @@ static int skl_plane_check(struct intel_crtc_state *crtc_state,
 	if (ret)
 		return ret;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	ret = skl_plane_check_dst_coordinates(crtc_state, plane_state);
@@ -2228,8 +2225,8 @@ static int skl_plane_check(struct intel_crtc_state *crtc_state,
 		return ret;
 
 	/* HW only has 8 bits pixel precision, disable plane if invisible */
-	if (!(plane_state->base.alpha >> 8))
-		plane_state->base.visible = false;
+	if (!(plane_state->hw.alpha >> 8))
+		plane_state->uapi.visible = false;
 
 	plane_state->ctl = skl_plane_ctl(crtc_state, plane_state);
 
@@ -2237,6 +2234,15 @@ static int skl_plane_check(struct intel_crtc_state *crtc_state,
 		plane_state->color_ctl = glk_plane_color_ctl(crtc_state,
 							     plane_state);
 
+	if (intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier) &&
+	    icl_is_hdr_plane(dev_priv, plane->id))
+		/* Enable and use MPEG-2 chroma siting */
+		plane_state->cus_ctl = PLANE_CUS_ENABLE |
+			PLANE_CUS_HPHASE_0 |
+			PLANE_CUS_VPHASE_SIGN_NEGATIVE | PLANE_CUS_VPHASE_0_25;
+	else
+		plane_state->cus_ctl = 0;
+
 	return 0;
 }
 
@@ -2248,7 +2254,7 @@ static bool has_dst_key_in_primary_plane(struct drm_i915_private *dev_priv)
 static void intel_plane_set_ckey(struct intel_plane_state *plane_state,
 				 const struct drm_intel_sprite_colorkey *set)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	struct drm_intel_sprite_colorkey *key = &plane_state->ckey;
 
@@ -2375,6 +2381,8 @@ static const u64 i9xx_plane_format_modifiers[] = {
 static const u32 snb_plane_formats[] = {
 	DRM_FORMAT_XRGB8888,
 	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_XBGR2101010,
 	DRM_FORMAT_XRGB16161616F,
 	DRM_FORMAT_XBGR16161616F,
 	DRM_FORMAT_YUYV,
@@ -2384,12 +2392,30 @@ static const u32 snb_plane_formats[] = {
 };
 
 static const u32 vlv_plane_formats[] = {
+	DRM_FORMAT_C8,
 	DRM_FORMAT_RGB565,
-	DRM_FORMAT_ABGR8888,
-	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB8888,
 	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_YVYU,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_VYUY,
+};
+
+static const u32 chv_pipe_b_sprite_formats[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
 	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ARGB2101010,
 	DRM_FORMAT_ABGR2101010,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
@@ -2462,6 +2488,8 @@ static const u32 icl_sdr_y_plane_formats[] = {
 	DRM_FORMAT_ABGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ARGB2101010,
+	DRM_FORMAT_ABGR2101010,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
 	DRM_FORMAT_UYVY,
@@ -2483,6 +2511,8 @@ static const u32 icl_sdr_uv_plane_formats[] = {
 	DRM_FORMAT_ABGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ARGB2101010,
+	DRM_FORMAT_ABGR2101010,
 	DRM_FORMAT_YUYV,
 	DRM_FORMAT_YVYU,
 	DRM_FORMAT_UYVY,
@@ -2508,6 +2538,8 @@ static const u32 icl_hdr_plane_formats[] = {
 	DRM_FORMAT_ABGR8888,
 	DRM_FORMAT_XRGB2101010,
 	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ARGB2101010,
+	DRM_FORMAT_ABGR2101010,
 	DRM_FORMAT_XRGB16161616F,
 	DRM_FORMAT_XBGR16161616F,
 	DRM_FORMAT_ARGB16161616F,
@@ -2546,7 +2578,8 @@ static const u64 skl_plane_format_modifiers_ccs[] = {
 	DRM_FORMAT_MOD_INVALID
 };
 
-static const u64 gen12_plane_format_modifiers_noccs[] = {
+static const u64 gen12_plane_format_modifiers_ccs[] = {
+	I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS,
 	I915_FORMAT_MOD_Y_TILED,
 	I915_FORMAT_MOD_X_TILED,
 	DRM_FORMAT_MOD_LINEAR,
@@ -2593,6 +2626,8 @@ static bool snb_sprite_format_mod_supported(struct drm_plane *_plane,
 	switch (format) {
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR8888:
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_XBGR2101010:
 	case DRM_FORMAT_XRGB16161616F:
 	case DRM_FORMAT_XBGR16161616F:
 	case DRM_FORMAT_YUYV:
@@ -2620,6 +2655,7 @@ static bool vlv_sprite_format_mod_supported(struct drm_plane *_plane,
 	}
 
 	switch (format) {
+	case DRM_FORMAT_C8:
 	case DRM_FORMAT_RGB565:
 	case DRM_FORMAT_ABGR8888:
 	case DRM_FORMAT_ARGB8888:
@@ -2627,6 +2663,8 @@ static bool vlv_sprite_format_mod_supported(struct drm_plane *_plane,
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR2101010:
 	case DRM_FORMAT_ABGR2101010:
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_ARGB2101010:
 	case DRM_FORMAT_YUYV:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
@@ -2671,6 +2709,8 @@ static bool skl_plane_format_mod_supported(struct drm_plane *_plane,
 	case DRM_FORMAT_RGB565:
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_ABGR2101010:
 	case DRM_FORMAT_YUYV:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
@@ -2710,6 +2750,7 @@ static bool gen12_plane_format_mod_supported(struct drm_plane *_plane,
 	case DRM_FORMAT_MOD_LINEAR:
 	case I915_FORMAT_MOD_X_TILED:
 	case I915_FORMAT_MOD_Y_TILED:
+	case I915_FORMAT_MOD_Y_TILED_GEN12_RC_CCS:
 		break;
 	default:
 		return false;
@@ -2720,9 +2761,14 @@ static bool gen12_plane_format_mod_supported(struct drm_plane *_plane,
 	case DRM_FORMAT_XBGR8888:
 	case DRM_FORMAT_ARGB8888:
 	case DRM_FORMAT_ABGR8888:
+		if (is_ccs_modifier(modifier))
+			return true;
+		/* fall through */
 	case DRM_FORMAT_RGB565:
 	case DRM_FORMAT_XRGB2101010:
 	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ARGB2101010:
+	case DRM_FORMAT_ABGR2101010:
 	case DRM_FORMAT_YUYV:
 	case DRM_FORMAT_YVYU:
 	case DRM_FORMAT_UYVY:
@@ -2916,8 +2962,6 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 	plane->get_hw_state = skl_plane_get_hw_state;
 	plane->check_plane = skl_plane_check;
 	plane->min_cdclk = skl_plane_min_cdclk;
-	if (icl_is_nv12_y_plane(plane_id))
-		plane->update_slave = icl_update_slave;
 
 	if (INTEL_GEN(dev_priv) >= 11)
 		formats = icl_get_plane_formats(dev_priv, pipe,
@@ -2929,13 +2973,11 @@ skl_universal_plane_create(struct drm_i915_private *dev_priv,
 		formats = skl_get_plane_formats(dev_priv, pipe,
 						plane_id, &num_formats);
 
+	plane->has_ccs = skl_plane_has_ccs(dev_priv, pipe, plane_id);
 	if (INTEL_GEN(dev_priv) >= 12) {
-		/* TODO: Implement support for gen-12 CCS modifiers */
-		plane->has_ccs = false;
-		modifiers = gen12_plane_format_modifiers_noccs;
+		modifiers = gen12_plane_format_modifiers_ccs;
 		plane_funcs = &gen12_plane_funcs;
 	} else {
-		plane->has_ccs = skl_plane_has_ccs(dev_priv, pipe, plane_id);
 		if (plane->has_ccs)
 			modifiers = skl_plane_format_modifiers_ccs;
 		else
@@ -3025,8 +3067,13 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		plane->check_plane = vlv_sprite_check;
 		plane->min_cdclk = vlv_plane_min_cdclk;
 
-		formats = vlv_plane_formats;
-		num_formats = ARRAY_SIZE(vlv_plane_formats);
+		if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B) {
+			formats = chv_pipe_b_sprite_formats;
+			num_formats = ARRAY_SIZE(chv_pipe_b_sprite_formats);
+		} else {
+			formats = vlv_plane_formats;
+			num_formats = ARRAY_SIZE(vlv_plane_formats);
+		}
 		modifiers = i9xx_plane_format_modifiers;
 
 		plane_funcs = &vlv_sprite_funcs;
diff --git a/drivers/gpu/drm/i915/display/intel_tv.c b/drivers/gpu/drm/i915/display/intel_tv.c
index 9983fadf6c28..50703536436c 100644
--- a/drivers/gpu/drm/i915/display/intel_tv.c
+++ b/drivers/gpu/drm/i915/display/intel_tv.c
@@ -924,7 +924,7 @@ intel_enable_tv(struct intel_encoder *encoder,
 
 	/* Prevents vblank waits from timing out in intel_tv_detect_type() */
 	intel_wait_for_vblank(dev_priv,
-			      to_intel_crtc(pipe_config->base.crtc)->pipe);
+			      to_intel_crtc(pipe_config->uapi.crtc)->pipe);
 
 	I915_WRITE(TV_CTL, I915_READ(TV_CTL) | TV_ENC_ENABLE);
 }
@@ -1085,7 +1085,7 @@ intel_tv_get_config(struct intel_encoder *encoder,
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct drm_display_mode *adjusted_mode =
-		&pipe_config->base.adjusted_mode;
+		&pipe_config->hw.adjusted_mode;
 	struct drm_display_mode mode = {};
 	u32 tv_ctl, hctl1, hctl3, vctl1, vctl2, tmp;
 	struct tv_mode tv_mode = {};
@@ -1188,7 +1188,7 @@ intel_tv_compute_config(struct intel_encoder *encoder,
 		to_intel_tv_connector_state(conn_state);
 	const struct tv_mode *tv_mode = intel_tv_mode_find(conn_state);
 	struct drm_display_mode *adjusted_mode =
-		&pipe_config->base.adjusted_mode;
+		&pipe_config->hw.adjusted_mode;
 	int hdisplay = adjusted_mode->crtc_hdisplay;
 	int vdisplay = adjusted_mode->crtc_vdisplay;
 
@@ -1417,7 +1417,7 @@ static void intel_tv_pre_enable(struct intel_encoder *encoder,
 				const struct drm_connector_state *conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_tv *intel_tv = enc_to_tv(encoder);
 	const struct intel_tv_connector_state *tv_conn_state =
 		to_intel_tv_connector_state(conn_state);
diff --git a/drivers/gpu/drm/i915/display/intel_vbt_defs.h b/drivers/gpu/drm/i915/display/intel_vbt_defs.h
index 69a7cb1fa121..4d0c23b29248 100644
--- a/drivers/gpu/drm/i915/display/intel_vbt_defs.h
+++ b/drivers/gpu/drm/i915/display/intel_vbt_defs.h
@@ -115,6 +115,7 @@ enum bdb_block_id {
 	BDB_MIPI_CONFIG			= 52,
 	BDB_MIPI_SEQUENCE		= 53,
 	BDB_COMPRESSION_PARAMETERS	= 56,
+	BDB_GENERIC_DTD			= 58,
 	BDB_SKIP			= 254, /* VBIOS private block, ignore */
 };
 
@@ -368,7 +369,7 @@ struct child_device_config {
 			u16 dtd_buf_ptr;			/* 161 */
 			u8 edidless_efp:1;			/* 161 */
 			u8 compression_enable:1;		/* 198 */
-			u8 compression_method:1;		/* 198 */
+			u8 compression_method_cps:1;		/* 198 */
 			u8 ganged_edp:1;			/* 202 */
 			u8 reserved0:4;
 			u8 compression_structure_index:4;	/* 198 */
@@ -793,6 +794,35 @@ struct bdb_lfp_backlight_data {
 } __packed;
 
 /*
+ * Block 44 - LFP Power Conservation Features Block
+ */
+
+struct als_data_entry {
+	u16 backlight_adjust;
+	u16 lux;
+} __packed;
+
+struct agressiveness_profile_entry {
+	u8 dpst_agressiveness : 4;
+	u8 lace_agressiveness : 4;
+} __packed;
+
+struct bdb_lfp_power {
+	u8 lfp_feature_bits;
+	struct als_data_entry als[5];
+	u8 lace_aggressiveness_profile;
+	u16 dpst;
+	u16 psr;
+	u16 drrs;
+	u16 lace_support;
+	u16 adt;
+	u16 dmrrs;
+	u16 adb;
+	u16 lace_enabled_status;
+	struct agressiveness_profile_entry aggressivenes[16];
+} __packed;
+
+/*
  * Block 52 - MIPI Configuration Block
  */
 
@@ -863,4 +893,34 @@ struct bdb_compression_parameters {
 	struct dsc_compression_parameters_entry data[16];
 } __packed;
 
+/*
+ * Block 58 - Generic DTD Block
+ */
+
+struct generic_dtd_entry {
+	u32 pixel_clock;
+	u16 hactive;
+	u16 hblank;
+	u16 hfront_porch;
+	u16 hsync;
+	u16 vactive;
+	u16 vblank;
+	u16 vfront_porch;
+	u16 vsync;
+	u16 width_mm;
+	u16 height_mm;
+
+	/* Flags */
+	u8 rsvd_flags:6;
+	u8 vsync_positive_polarity:1;
+	u8 hsync_positive_polarity:1;
+
+	u8 rsvd[3];
+} __packed;
+
+struct bdb_generic_dtd {
+	u16 gdtd_size;
+	struct generic_dtd_entry dtd[];	/* up to 24 DTD's */
+} __packed;
+
 #endif /* _INTEL_VBT_DEFS_H_ */
diff --git a/drivers/gpu/drm/i915/display/intel_vdsc.c b/drivers/gpu/drm/i915/display/intel_vdsc.c
index 896b0c334f5e..6bab08db5d75 100644
--- a/drivers/gpu/drm/i915/display/intel_vdsc.c
+++ b/drivers/gpu/drm/i915/display/intel_vdsc.c
@@ -10,6 +10,7 @@
 
 #include "i915_drv.h"
 #include "intel_display_types.h"
+#include "intel_dsi.h"
 #include "intel_vdsc.h"
 
 enum ROW_INDEX_BPP {
@@ -30,10 +31,8 @@ enum COLUMN_INDEX_BPC {
 	MAX_COLUMN_INDEX
 };
 
-#define DSC_SUPPORTED_VERSION_MIN		1
-
 /* From DSC_v1.11 spec, rc_parameter_Set syntax element typically constant */
-static u16 rc_buf_thresh[] = {
+static const u16 rc_buf_thresh[] = {
 	896, 1792, 2688, 3584, 4480, 5376, 6272, 6720, 7168, 7616,
 	7744, 7872, 8000, 8064
 };
@@ -53,7 +52,7 @@ struct rc_parameters {
  * Selected Rate Control Related Parameter Recommended Values
  * from DSC_v1.11 spec & C Model release: DSC_model_20161212
  */
-static struct rc_parameters rc_params[][MAX_COLUMN_INDEX] = {
+static const struct rc_parameters rc_parameters[][MAX_COLUMN_INDEX] = {
 {
 	/* 6BPP/8BPC */
 	{ 768, 15, 6144, 3, 13, 11, 11, {
@@ -319,63 +318,84 @@ static int get_column_index_for_rc_params(u8 bits_per_component)
 	}
 }
 
-int intel_dp_compute_dsc_params(struct intel_dp *intel_dp,
-				struct intel_crtc_state *pipe_config)
+static const struct rc_parameters *get_rc_params(u16 compressed_bpp,
+						 u8 bits_per_component)
+{
+	int row_index, column_index;
+
+	row_index = get_row_index_for_rc_params(compressed_bpp);
+	if (row_index < 0)
+		return NULL;
+
+	column_index = get_column_index_for_rc_params(bits_per_component);
+	if (column_index < 0)
+		return NULL;
+
+	return &rc_parameters[row_index][column_index];
+}
+
+bool intel_dsc_source_support(struct intel_encoder *encoder,
+			      const struct intel_crtc_state *crtc_state)
+{
+	const struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *i915 = to_i915(encoder->base.dev);
+	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
+	enum pipe pipe = crtc->pipe;
+
+	if (!INTEL_INFO(i915)->display.has_dsc)
+		return false;
+
+	/* On TGL, DSC is supported on all Pipes */
+	if (INTEL_GEN(i915) >= 12)
+		return true;
+
+	if (INTEL_GEN(i915) >= 10 &&
+	    (pipe != PIPE_A ||
+	     (cpu_transcoder == TRANSCODER_EDP ||
+	      cpu_transcoder == TRANSCODER_DSI_0 ||
+	      cpu_transcoder == TRANSCODER_DSI_1)))
+		return true;
+
+	return false;
+}
+
+static bool is_pipe_dsc(const struct intel_crtc_state *crtc_state)
+{
+	const struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	const struct drm_i915_private *i915 = to_i915(crtc->base.dev);
+	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
+
+	if (INTEL_GEN(i915) >= 12)
+		return true;
+
+	if (cpu_transcoder == TRANSCODER_EDP ||
+	    cpu_transcoder == TRANSCODER_DSI_0 ||
+	    cpu_transcoder == TRANSCODER_DSI_1)
+		return false;
+
+	/* There's no pipe A DSC engine on ICL */
+	WARN_ON(crtc->pipe == PIPE_A);
+
+	return true;
+}
+
+int intel_dsc_compute_params(struct intel_encoder *encoder,
+			     struct intel_crtc_state *pipe_config)
 {
 	struct drm_dsc_config *vdsc_cfg = &pipe_config->dsc.config;
 	u16 compressed_bpp = pipe_config->dsc.compressed_bpp;
+	const struct rc_parameters *rc_params;
 	u8 i = 0;
-	int row_index = 0;
-	int column_index = 0;
-	u8 line_buf_depth = 0;
 
-	vdsc_cfg->pic_width = pipe_config->base.adjusted_mode.crtc_hdisplay;
-	vdsc_cfg->pic_height = pipe_config->base.adjusted_mode.crtc_vdisplay;
+	vdsc_cfg->pic_width = pipe_config->hw.adjusted_mode.crtc_hdisplay;
+	vdsc_cfg->pic_height = pipe_config->hw.adjusted_mode.crtc_vdisplay;
 	vdsc_cfg->slice_width = DIV_ROUND_UP(vdsc_cfg->pic_width,
 					     pipe_config->dsc.slice_count);
-	/*
-	 * Slice Height of 8 works for all currently available panels. So start
-	 * with that if pic_height is an integral multiple of 8.
-	 * Eventually add logic to try multiple slice heights.
-	 */
-	if (vdsc_cfg->pic_height % 8 == 0)
-		vdsc_cfg->slice_height = 8;
-	else if (vdsc_cfg->pic_height % 4 == 0)
-		vdsc_cfg->slice_height = 4;
-	else
-		vdsc_cfg->slice_height = 2;
-
-	/* Values filled from DSC Sink DPCD */
-	vdsc_cfg->dsc_version_major =
-		(intel_dp->dsc_dpcd[DP_DSC_REV - DP_DSC_SUPPORT] &
-		 DP_DSC_MAJOR_MASK) >> DP_DSC_MAJOR_SHIFT;
-	vdsc_cfg->dsc_version_minor =
-		min(DSC_SUPPORTED_VERSION_MIN,
-		    (intel_dp->dsc_dpcd[DP_DSC_REV - DP_DSC_SUPPORT] &
-		     DP_DSC_MINOR_MASK) >> DP_DSC_MINOR_SHIFT);
-
-	vdsc_cfg->convert_rgb = intel_dp->dsc_dpcd[DP_DSC_DEC_COLOR_FORMAT_CAP - DP_DSC_SUPPORT] &
-		DP_DSC_RGB;
-
-	line_buf_depth = drm_dp_dsc_sink_line_buf_depth(intel_dp->dsc_dpcd);
-	if (!line_buf_depth) {
-		DRM_DEBUG_KMS("DSC Sink Line Buffer Depth invalid\n");
-		return -EINVAL;
-	}
-	if (vdsc_cfg->dsc_version_minor == 2)
-		vdsc_cfg->line_buf_depth = (line_buf_depth == DSC_1_2_MAX_LINEBUF_DEPTH_BITS) ?
-			DSC_1_2_MAX_LINEBUF_DEPTH_VAL : line_buf_depth;
-	else
-		vdsc_cfg->line_buf_depth = (line_buf_depth > DSC_1_1_MAX_LINEBUF_DEPTH_BITS) ?
-			DSC_1_1_MAX_LINEBUF_DEPTH_BITS : line_buf_depth;
 
 	/* Gen 11 does not support YCbCr */
 	vdsc_cfg->simple_422 = false;
 	/* Gen 11 does not support VBR */
 	vdsc_cfg->vbr_enable = false;
-	vdsc_cfg->block_pred_enable =
-			intel_dp->dsc_dpcd[DP_DSC_BLK_PREDICTION_SUPPORT - DP_DSC_SUPPORT] &
-		DP_DSC_BLK_PREDICTION_IS_SUPPORTED;
 
 	/* Gen 11 only supports integral values of bpp */
 	vdsc_cfg->bits_per_pixel = compressed_bpp << 4;
@@ -399,39 +419,29 @@ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp,
 		vdsc_cfg->rc_buf_thresh[13] = 0x7D;
 	}
 
-	row_index = get_row_index_for_rc_params(compressed_bpp);
-	column_index =
-		get_column_index_for_rc_params(vdsc_cfg->bits_per_component);
-
-	if (row_index < 0 || column_index < 0)
+	rc_params = get_rc_params(compressed_bpp, vdsc_cfg->bits_per_component);
+	if (!rc_params)
 		return -EINVAL;
 
-	vdsc_cfg->first_line_bpg_offset =
-		rc_params[row_index][column_index].first_line_bpg_offset;
-	vdsc_cfg->initial_xmit_delay =
-		rc_params[row_index][column_index].initial_xmit_delay;
-	vdsc_cfg->initial_offset =
-		rc_params[row_index][column_index].initial_offset;
-	vdsc_cfg->flatness_min_qp =
-		rc_params[row_index][column_index].flatness_min_qp;
-	vdsc_cfg->flatness_max_qp =
-		rc_params[row_index][column_index].flatness_max_qp;
-	vdsc_cfg->rc_quant_incr_limit0 =
-		rc_params[row_index][column_index].rc_quant_incr_limit0;
-	vdsc_cfg->rc_quant_incr_limit1 =
-		rc_params[row_index][column_index].rc_quant_incr_limit1;
+	vdsc_cfg->first_line_bpg_offset = rc_params->first_line_bpg_offset;
+	vdsc_cfg->initial_xmit_delay = rc_params->initial_xmit_delay;
+	vdsc_cfg->initial_offset = rc_params->initial_offset;
+	vdsc_cfg->flatness_min_qp = rc_params->flatness_min_qp;
+	vdsc_cfg->flatness_max_qp = rc_params->flatness_max_qp;
+	vdsc_cfg->rc_quant_incr_limit0 = rc_params->rc_quant_incr_limit0;
+	vdsc_cfg->rc_quant_incr_limit1 = rc_params->rc_quant_incr_limit1;
 
 	for (i = 0; i < DSC_NUM_BUF_RANGES; i++) {
 		vdsc_cfg->rc_range_params[i].range_min_qp =
-			rc_params[row_index][column_index].rc_range_params[i].range_min_qp;
+			rc_params->rc_range_params[i].range_min_qp;
 		vdsc_cfg->rc_range_params[i].range_max_qp =
-			rc_params[row_index][column_index].rc_range_params[i].range_max_qp;
+			rc_params->rc_range_params[i].range_max_qp;
 		/*
 		 * Range BPG Offset uses 2's complement and is only a 6 bits. So
 		 * mask it to get only 6 bits.
 		 */
 		vdsc_cfg->rc_range_params[i].range_bpg_offset =
-			rc_params[row_index][column_index].rc_range_params[i].range_bpg_offset &
+			rc_params->rc_range_params[i].range_bpg_offset &
 			DSC_RANGE_BPG_OFFSET_MASK;
 	}
 
@@ -453,41 +463,42 @@ int intel_dp_compute_dsc_params(struct intel_dp *intel_dp,
 	vdsc_cfg->initial_scale_value = (vdsc_cfg->rc_model_size << 3) /
 		(vdsc_cfg->rc_model_size - vdsc_cfg->initial_offset);
 
-	return drm_dsc_compute_rc_parameters(vdsc_cfg);
+	return 0;
 }
 
 enum intel_display_power_domain
 intel_dsc_power_domain(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *i915 = to_i915(crtc_state->base.crtc->dev);
-	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	struct drm_i915_private *i915 = to_i915(crtc->base.dev);
+	enum pipe pipe = crtc->pipe;
 
 	/*
-	 * On ICL VDSC/joining for eDP transcoder uses a separate power well,
-	 * PW2. This requires POWER_DOMAIN_TRANSCODER_VDSC_PW2 power domain.
-	 * For any other transcoder, VDSC/joining uses the power well associated
-	 * with the pipe/transcoder in use. Hence another reference on the
-	 * transcoder power domain will suffice.
+	 * VDSC/joining uses a separate power well, PW2, and requires
+	 * POWER_DOMAIN_TRANSCODER_VDSC_PW2 power domain in two cases:
+	 *
+	 *  - ICL eDP/DSI transcoder
+	 *  - TGL pipe A
 	 *
-	 * On TGL we have the same mapping, but for transcoder A (the special
-	 * TRANSCODER_EDP is gone).
+	 * For any other pipe, VDSC/joining uses the power well associated with
+	 * the pipe in use. Hence another reference on the pipe power domain
+	 * will suffice. (Except no VDSC/joining on ICL pipe A.)
 	 */
-	if (INTEL_GEN(i915) >= 12 && cpu_transcoder == TRANSCODER_A)
-		return POWER_DOMAIN_TRANSCODER_VDSC_PW2;
-	else if (cpu_transcoder == TRANSCODER_EDP)
+	if (INTEL_GEN(i915) >= 12 && pipe == PIPE_A)
 		return POWER_DOMAIN_TRANSCODER_VDSC_PW2;
+	else if (is_pipe_dsc(crtc_state))
+		return POWER_DOMAIN_PIPE(pipe);
 	else
-		return POWER_DOMAIN_TRANSCODER(cpu_transcoder);
+		return POWER_DOMAIN_TRANSCODER_VDSC_PW2;
 }
 
-static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
-						const struct intel_crtc_state *crtc_state)
+static void intel_dsc_pps_configure(struct intel_encoder *encoder,
+				    const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	const struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
 	enum pipe pipe = crtc->pipe;
-	enum transcoder cpu_transcoder = crtc_state->cpu_transcoder;
 	u32 pps_val = 0;
 	u32 rc_buf_thresh_dword[4];
 	u32 rc_range_params_dword[8];
@@ -508,7 +519,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	if (vdsc_cfg->vbr_enable)
 		pps_val |= DSC_VBR_ENABLE;
 	DRM_INFO("PPS0 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_0, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -527,7 +538,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val = 0;
 	pps_val |= DSC_BPP(vdsc_cfg->bits_per_pixel);
 	DRM_INFO("PPS1 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_1, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -547,7 +558,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_PIC_HEIGHT(vdsc_cfg->pic_height) |
 		DSC_PIC_WIDTH(vdsc_cfg->pic_width / num_vdsc_instances);
 	DRM_INFO("PPS2 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_2, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -567,7 +578,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_SLICE_HEIGHT(vdsc_cfg->slice_height) |
 		DSC_SLICE_WIDTH(vdsc_cfg->slice_width);
 	DRM_INFO("PPS3 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_3, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -587,7 +598,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_INITIAL_XMIT_DELAY(vdsc_cfg->initial_xmit_delay) |
 		DSC_INITIAL_DEC_DELAY(vdsc_cfg->initial_dec_delay);
 	DRM_INFO("PPS4 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_4, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -607,7 +618,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_SCALE_INC_INT(vdsc_cfg->scale_increment_interval) |
 		DSC_SCALE_DEC_INT(vdsc_cfg->scale_decrement_interval);
 	DRM_INFO("PPS5 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_5, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -629,7 +640,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		DSC_FLATNESS_MIN_QP(vdsc_cfg->flatness_min_qp) |
 		DSC_FLATNESS_MAX_QP(vdsc_cfg->flatness_max_qp);
 	DRM_INFO("PPS6 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_6, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -649,7 +660,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_SLICE_BPG_OFFSET(vdsc_cfg->slice_bpg_offset) |
 		DSC_NFL_BPG_OFFSET(vdsc_cfg->nfl_bpg_offset);
 	DRM_INFO("PPS7 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_7, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -669,7 +680,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_FINAL_OFFSET(vdsc_cfg->final_offset) |
 		DSC_INITIAL_OFFSET(vdsc_cfg->initial_offset);
 	DRM_INFO("PPS8 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_8, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -689,7 +700,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	pps_val |= DSC_RC_MODEL_SIZE(DSC_RC_MODEL_SIZE_CONST) |
 		DSC_RC_EDGE_FACTOR(DSC_RC_EDGE_FACTOR_CONST);
 	DRM_INFO("PPS9 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_9, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -711,7 +722,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		DSC_RC_TARGET_OFF_HIGH(DSC_RC_TGT_OFFSET_HI_CONST) |
 		DSC_RC_TARGET_OFF_LOW(DSC_RC_TGT_OFFSET_LO_CONST);
 	DRM_INFO("PPS10 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_10, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -734,7 +745,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		DSC_SLICE_ROW_PER_FRAME(vdsc_cfg->pic_height /
 					vdsc_cfg->slice_height);
 	DRM_INFO("PPS16 = 0x%08x\n", pps_val);
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_PICTURE_PARAMETER_SET_16, pps_val);
 		/*
 		 * If 2 VDSC instances are needed, configure PPS for second
@@ -758,7 +769,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		DRM_INFO(" RC_BUF_THRESH%d = 0x%08x\n", i,
 			 rc_buf_thresh_dword[i / 4]);
 	}
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_RC_BUF_THRESH_0, rc_buf_thresh_dword[0]);
 		I915_WRITE(DSCA_RC_BUF_THRESH_0_UDW, rc_buf_thresh_dword[1]);
 		I915_WRITE(DSCA_RC_BUF_THRESH_1, rc_buf_thresh_dword[2]);
@@ -807,7 +818,7 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 		DRM_INFO(" RC_RANGE_PARAM_%d = 0x%08x\n", i,
 			 rc_range_params_dword[i / 2]);
 	}
-	if (cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		I915_WRITE(DSCA_RC_RANGE_PARAMETERS_0,
 			   rc_range_params_dword[0]);
 		I915_WRITE(DSCA_RC_RANGE_PARAMETERS_0_UDW,
@@ -880,8 +891,75 @@ static void intel_configure_pps_for_dsc_encoder(struct intel_encoder *encoder,
 	}
 }
 
-static void intel_dp_write_dsc_pps_sdp(struct intel_encoder *encoder,
-				       const struct intel_crtc_state *crtc_state)
+void intel_dsc_get_config(struct intel_encoder *encoder,
+			  struct intel_crtc_state *crtc_state)
+{
+	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
+	struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
+	enum pipe pipe = crtc->pipe;
+	enum intel_display_power_domain power_domain;
+	intel_wakeref_t wakeref;
+	u32 dss_ctl1, dss_ctl2, val;
+
+	if (!intel_dsc_source_support(encoder, crtc_state))
+		return;
+
+	power_domain = intel_dsc_power_domain(crtc_state);
+
+	wakeref = intel_display_power_get_if_enabled(dev_priv, power_domain);
+	if (!wakeref)
+		return;
+
+	if (!is_pipe_dsc(crtc_state)) {
+		dss_ctl1 = I915_READ(DSS_CTL1);
+		dss_ctl2 = I915_READ(DSS_CTL2);
+	} else {
+		dss_ctl1 = I915_READ(ICL_PIPE_DSS_CTL1(pipe));
+		dss_ctl2 = I915_READ(ICL_PIPE_DSS_CTL2(pipe));
+	}
+
+	crtc_state->dsc.compression_enable = dss_ctl2 & LEFT_BRANCH_VDSC_ENABLE;
+	if (!crtc_state->dsc.compression_enable)
+		goto out;
+
+	crtc_state->dsc.dsc_split = (dss_ctl2 & RIGHT_BRANCH_VDSC_ENABLE) &&
+		(dss_ctl1 & JOINER_ENABLE);
+
+	/* FIXME: add more state readout as needed */
+
+	/* PPS1 */
+	if (!is_pipe_dsc(crtc_state))
+		val = I915_READ(DSCA_PICTURE_PARAMETER_SET_1);
+	else
+		val = I915_READ(ICL_DSC0_PICTURE_PARAMETER_SET_1(pipe));
+	vdsc_cfg->bits_per_pixel = val;
+	crtc_state->dsc.compressed_bpp = vdsc_cfg->bits_per_pixel >> 4;
+out:
+	intel_display_power_put(dev_priv, power_domain, wakeref);
+}
+
+static void intel_dsc_dsi_pps_write(struct intel_encoder *encoder,
+				    const struct intel_crtc_state *crtc_state)
+{
+	const struct drm_dsc_config *vdsc_cfg = &crtc_state->dsc.config;
+	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
+	struct mipi_dsi_device *dsi;
+	struct drm_dsc_picture_parameter_set pps;
+	enum port port;
+
+	drm_dsc_pps_payload_pack(&pps, vdsc_cfg);
+
+	for_each_dsi_port(port, intel_dsi->ports) {
+		dsi = intel_dsi->dsi_hosts[port]->device;
+
+		mipi_dsi_picture_parameter_set(dsi, &pps);
+		mipi_dsi_compression_mode(dsi, true);
+	}
+}
+
+static void intel_dsc_dp_pps_write(struct intel_encoder *encoder,
+				   const struct intel_crtc_state *crtc_state)
 {
 	struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base);
 	struct intel_digital_port *intel_dig_port = dp_to_dig_port(intel_dp);
@@ -902,7 +980,7 @@ static void intel_dp_write_dsc_pps_sdp(struct intel_encoder *encoder,
 void intel_dsc_enable(struct intel_encoder *encoder,
 		      const struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	enum pipe pipe = crtc->pipe;
 	i915_reg_t dss_ctl1_reg, dss_ctl2_reg;
@@ -916,11 +994,14 @@ void intel_dsc_enable(struct intel_encoder *encoder,
 	intel_display_power_get(dev_priv,
 				intel_dsc_power_domain(crtc_state));
 
-	intel_configure_pps_for_dsc_encoder(encoder, crtc_state);
+	intel_dsc_pps_configure(encoder, crtc_state);
 
-	intel_dp_write_dsc_pps_sdp(encoder, crtc_state);
+	if (encoder->type == INTEL_OUTPUT_DSI)
+		intel_dsc_dsi_pps_write(encoder, crtc_state);
+	else
+		intel_dsc_dp_pps_write(encoder, crtc_state);
 
-	if (crtc_state->cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(crtc_state)) {
 		dss_ctl1_reg = DSS_CTL1;
 		dss_ctl2_reg = DSS_CTL2;
 	} else {
@@ -938,7 +1019,7 @@ void intel_dsc_enable(struct intel_encoder *encoder,
 
 void intel_dsc_disable(const struct intel_crtc_state *old_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(old_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	enum pipe pipe = crtc->pipe;
 	i915_reg_t dss_ctl1_reg, dss_ctl2_reg;
@@ -947,7 +1028,7 @@ void intel_dsc_disable(const struct intel_crtc_state *old_crtc_state)
 	if (!old_crtc_state->dsc.compression_enable)
 		return;
 
-	if (old_crtc_state->cpu_transcoder == TRANSCODER_EDP) {
+	if (!is_pipe_dsc(old_crtc_state)) {
 		dss_ctl1_reg = DSS_CTL1;
 		dss_ctl2_reg = DSS_CTL2;
 	} else {
diff --git a/drivers/gpu/drm/i915/display/intel_vdsc.h b/drivers/gpu/drm/i915/display/intel_vdsc.h
index 90d3f6017fcb..e56a3254c214 100644
--- a/drivers/gpu/drm/i915/display/intel_vdsc.h
+++ b/drivers/gpu/drm/i915/display/intel_vdsc.h
@@ -6,15 +6,20 @@
 #ifndef __INTEL_VDSC_H__
 #define __INTEL_VDSC_H__
 
+#include <linux/types.h>
+
 struct intel_encoder;
 struct intel_crtc_state;
-struct intel_dp;
 
+bool intel_dsc_source_support(struct intel_encoder *encoder,
+			      const struct intel_crtc_state *crtc_state);
 void intel_dsc_enable(struct intel_encoder *encoder,
 		      const struct intel_crtc_state *crtc_state);
 void intel_dsc_disable(const struct intel_crtc_state *crtc_state);
-int intel_dp_compute_dsc_params(struct intel_dp *intel_dp,
-				struct intel_crtc_state *pipe_config);
+int intel_dsc_compute_params(struct intel_encoder *encoder,
+			     struct intel_crtc_state *pipe_config);
+void intel_dsc_get_config(struct intel_encoder *encoder,
+			  struct intel_crtc_state *crtc_state);
 enum intel_display_power_domain
 intel_dsc_power_domain(const struct intel_crtc_state *crtc_state);
 
diff --git a/drivers/gpu/drm/i915/display/vlv_dsi.c b/drivers/gpu/drm/i915/display/vlv_dsi.c
index 0ca49b1604c6..21e820299107 100644
--- a/drivers/gpu/drm/i915/display/vlv_dsi.c
+++ b/drivers/gpu/drm/i915/display/vlv_dsi.c
@@ -261,9 +261,9 @@ static int intel_dsi_compute_config(struct intel_encoder *encoder,
 	struct intel_dsi *intel_dsi = container_of(encoder, struct intel_dsi,
 						   base);
 	struct intel_connector *intel_connector = intel_dsi->attached_connector;
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	const struct drm_display_mode *fixed_mode = intel_connector->panel.fixed_mode;
-	struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	int ret;
 
 	DRM_DEBUG_KMS("\n");
@@ -624,7 +624,7 @@ static void intel_dsi_port_enable(struct intel_encoder *encoder,
 				  const struct intel_crtc_state *crtc_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	enum port port;
 
@@ -746,7 +746,7 @@ static void intel_dsi_pre_enable(struct intel_encoder *encoder,
 				 const struct drm_connector_state *conn_state)
 {
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
-	struct drm_crtc *crtc = pipe_config->base.crtc;
+	struct drm_crtc *crtc = pipe_config->uapi.crtc;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	enum pipe pipe = intel_crtc->pipe;
@@ -882,8 +882,8 @@ static void intel_dsi_clear_device_ready(struct intel_encoder *encoder)
 }
 
 static void intel_dsi_post_disable(struct intel_encoder *encoder,
-				   const struct intel_crtc_state *pipe_config,
-				   const struct drm_connector_state *conn_state)
+				   const struct intel_crtc_state *old_crtc_state,
+				   const struct drm_connector_state *old_conn_state)
 {
 	struct drm_i915_private *dev_priv = to_i915(encoder->base.dev);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
@@ -892,6 +892,12 @@ static void intel_dsi_post_disable(struct intel_encoder *encoder,
 
 	DRM_DEBUG_KMS("\n");
 
+	if (IS_GEN9_LP(dev_priv)) {
+		intel_crtc_vblank_off(old_crtc_state);
+
+		skylake_scaler_disable(old_crtc_state);
+	}
+
 	if (is_vid_mode(intel_dsi)) {
 		for_each_dsi_port(port, intel_dsi->ports)
 			vlv_dsi_wait_for_fifo_empty(intel_dsi, port);
@@ -1032,9 +1038,9 @@ static void bxt_dsi_get_pipe_config(struct intel_encoder *encoder,
 	struct drm_device *dev = encoder->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct drm_display_mode *adjusted_mode =
-					&pipe_config->base.adjusted_mode;
+					&pipe_config->hw.adjusted_mode;
 	struct drm_display_mode *adjusted_mode_sw;
-	struct intel_crtc *crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(&encoder->base);
 	unsigned int lane_count = intel_dsi->lane_count;
 	unsigned int bpp, fmt;
@@ -1045,7 +1051,7 @@ static void bxt_dsi_get_pipe_config(struct intel_encoder *encoder,
 				crtc_hblank_start_sw, crtc_hblank_end_sw;
 
 	/* FIXME: hw readout should not depend on SW state */
-	adjusted_mode_sw = &crtc->config->base.adjusted_mode;
+	adjusted_mode_sw = &crtc->config->hw.adjusted_mode;
 
 	/*
 	 * Atleast one port is active as encoder->get_config called only if
@@ -1204,7 +1210,7 @@ static void intel_dsi_get_config(struct intel_encoder *encoder,
 	}
 
 	if (pclk) {
-		pipe_config->base.adjusted_mode.crtc_clock = pclk;
+		pipe_config->hw.adjusted_mode.crtc_clock = pclk;
 		pipe_config->port_clock = pclk;
 	}
 }
@@ -1315,9 +1321,9 @@ static void intel_dsi_prepare(struct intel_encoder *intel_encoder,
 	struct drm_encoder *encoder = &intel_encoder->base;
 	struct drm_device *dev = encoder->dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
-	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->uapi.crtc);
 	struct intel_dsi *intel_dsi = enc_to_intel_dsi(encoder);
-	const struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode;
+	const struct drm_display_mode *adjusted_mode = &pipe_config->hw.adjusted_mode;
 	enum port port;
 	unsigned int bpp = mipi_dsi_pixel_format_to_bpp(intel_dsi->pixel_format);
 	u32 val, tmp;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_clflush.c b/drivers/gpu/drm/i915/gem/i915_gem_clflush.c
index b9f504ba3b32..34be4c0ee7c5 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_clflush.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_clflush.c
@@ -20,33 +20,31 @@ static void __do_clflush(struct drm_i915_gem_object *obj)
 {
 	GEM_BUG_ON(!i915_gem_object_has_pages(obj));
 	drm_clflush_sg(obj->mm.pages);
-	intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU);
+
+	i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
 }
 
 static int clflush_work(struct dma_fence_work *base)
 {
 	struct clflush *clflush = container_of(base, typeof(*clflush), base);
-	struct drm_i915_gem_object *obj = fetch_and_zero(&clflush->obj);
+	struct drm_i915_gem_object *obj = clflush->obj;
 	int err;
 
 	err = i915_gem_object_pin_pages(obj);
 	if (err)
-		goto put;
+		return err;
 
 	__do_clflush(obj);
 	i915_gem_object_unpin_pages(obj);
 
-put:
-	i915_gem_object_put(obj);
-	return err;
+	return 0;
 }
 
 static void clflush_release(struct dma_fence_work *base)
 {
 	struct clflush *clflush = container_of(base, typeof(*clflush), base);
 
-	if (clflush->obj)
-		i915_gem_object_put(clflush->obj);
+	i915_gem_object_put(clflush->obj);
 }
 
 static const struct dma_fence_work_ops clflush_ops = {
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c
index 337ba17b1e0e..dc90b044a217 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c
@@ -69,7 +69,9 @@
 
 #include <drm/i915_drm.h>
 
+#include "gt/intel_context.h"
 #include "gt/intel_engine_heartbeat.h"
+#include "gt/intel_engine_pm.h"
 #include "gt/intel_engine_user.h"
 #include "gt/intel_lrc_reg.h"
 #include "gt/intel_ring.h"
@@ -169,12 +171,80 @@ lookup_user_engine(struct i915_gem_context *ctx,
 	return i915_gem_context_get_engine(ctx, idx);
 }
 
+static struct i915_address_space *
+context_get_vm_rcu(struct i915_gem_context *ctx)
+{
+	GEM_BUG_ON(!rcu_access_pointer(ctx->vm));
+
+	do {
+		struct i915_address_space *vm;
+
+		/*
+		 * We do not allow downgrading from full-ppgtt [to a shared
+		 * global gtt], so ctx->vm cannot become NULL.
+		 */
+		vm = rcu_dereference(ctx->vm);
+		if (!kref_get_unless_zero(&vm->ref))
+			continue;
+
+		/*
+		 * This ppgtt may have be reallocated between
+		 * the read and the kref, and reassigned to a third
+		 * context. In order to avoid inadvertent sharing
+		 * of this ppgtt with that third context (and not
+		 * src), we have to confirm that we have the same
+		 * ppgtt after passing through the strong memory
+		 * barrier implied by a successful
+		 * kref_get_unless_zero().
+		 *
+		 * Once we have acquired the current ppgtt of ctx,
+		 * we no longer care if it is released from ctx, as
+		 * it cannot be reallocated elsewhere.
+		 */
+
+		if (vm == rcu_access_pointer(ctx->vm))
+			return rcu_pointer_handoff(vm);
+
+		i915_vm_put(vm);
+	} while (1);
+}
+
+static void intel_context_set_gem(struct intel_context *ce,
+				  struct i915_gem_context *ctx)
+{
+	GEM_BUG_ON(rcu_access_pointer(ce->gem_context));
+	RCU_INIT_POINTER(ce->gem_context, ctx);
+
+	if (!test_bit(CONTEXT_ALLOC_BIT, &ce->flags))
+		ce->ring = __intel_context_ring_size(SZ_16K);
+
+	if (rcu_access_pointer(ctx->vm)) {
+		struct i915_address_space *vm;
+
+		rcu_read_lock();
+		vm = context_get_vm_rcu(ctx); /* hmm */
+		rcu_read_unlock();
+
+		i915_vm_put(ce->vm);
+		ce->vm = vm;
+	}
+
+	GEM_BUG_ON(ce->timeline);
+	if (ctx->timeline)
+		ce->timeline = intel_timeline_get(ctx->timeline);
+
+	if (ctx->sched.priority >= I915_PRIORITY_NORMAL &&
+	    intel_engine_has_semaphores(ce->engine))
+		__set_bit(CONTEXT_USE_SEMAPHORES, &ce->flags);
+}
+
 static void __free_engines(struct i915_gem_engines *e, unsigned int count)
 {
 	while (count--) {
 		if (!e->engines[count])
 			continue;
 
+		RCU_INIT_POINTER(e->engines[count]->gem_context, NULL);
 		intel_context_put(e->engines[count]);
 	}
 	kfree(e);
@@ -211,12 +281,14 @@ static struct i915_gem_engines *default_engines(struct i915_gem_context *ctx)
 		GEM_BUG_ON(engine->legacy_idx >= I915_NUM_ENGINES);
 		GEM_BUG_ON(e->engines[engine->legacy_idx]);
 
-		ce = intel_context_create(ctx, engine);
+		ce = intel_context_create(engine);
 		if (IS_ERR(ce)) {
 			__free_engines(e, e->num_engines + 1);
 			return ERR_CAST(ce);
 		}
 
+		intel_context_set_gem(ce, ctx);
+
 		e->engines[engine->legacy_idx] = ce;
 		e->num_engines = max(e->num_engines, engine->legacy_idx);
 	}
@@ -236,14 +308,10 @@ static void i915_gem_context_free(struct i915_gem_context *ctx)
 	free_engines(rcu_access_pointer(ctx->engines));
 	mutex_destroy(&ctx->engines_mutex);
 
-	kfree(ctx->jump_whitelist);
-
 	if (ctx->timeline)
 		intel_timeline_put(ctx->timeline);
 
-	kfree(ctx->name);
 	put_pid(ctx->pid);
-
 	mutex_destroy(&ctx->mutex);
 
 	kfree_rcu(ctx, rcu);
@@ -389,15 +457,6 @@ static void kill_context(struct i915_gem_context *ctx)
 	struct intel_context *ce;
 
 	/*
-	 * If we are already banned, it was due to a guilty request causing
-	 * a reset and the entire context being evicted from the GPU.
-	 */
-	if (i915_gem_context_is_banned(ctx))
-		return;
-
-	i915_gem_context_set_banned(ctx);
-
-	/*
 	 * Map the user's engine back to the actual engines; one virtual
 	 * engine will be mapped to multiple engines, and using ctx->engine[]
 	 * the same engine may be have multiple instances in the user's map.
@@ -407,6 +466,9 @@ static void kill_context(struct i915_gem_context *ctx)
 	for_each_gem_engine(ce, __context_engines_static(ctx), it) {
 		struct intel_engine_cs *engine;
 
+		if (intel_context_set_banned(ce))
+			continue;
+
 		/*
 		 * Check the current active state of this context; if we
 		 * are currently executing on the GPU we need to evict
@@ -427,11 +489,29 @@ static void kill_context(struct i915_gem_context *ctx)
 	}
 }
 
+static void set_closed_name(struct i915_gem_context *ctx)
+{
+	char *s;
+
+	/* Replace '[]' with '<>' to indicate closed in debug prints */
+
+	s = strrchr(ctx->name, '[');
+	if (!s)
+		return;
+
+	*s = '<';
+
+	s = strchr(s + 1, ']');
+	if (s)
+		*s = '>';
+}
+
 static void context_close(struct i915_gem_context *ctx)
 {
 	struct i915_address_space *vm;
 
 	i915_gem_context_set_closed(ctx);
+	set_closed_name(ctx);
 
 	mutex_lock(&ctx->mutex);
 
@@ -529,9 +609,6 @@ __create_context(struct drm_i915_private *i915)
 	for (i = 0; i < ARRAY_SIZE(ctx->hang_timestamp); i++)
 		ctx->hang_timestamp[i] = jiffies - CONTEXT_FAST_HANG_JIFFIES;
 
-	ctx->jump_whitelist = NULL;
-	ctx->jump_whitelist_cmds = 0;
-
 	spin_lock(&i915->gem.contexts.lock);
 	list_add_tail(&ctx->link, &i915->gem.contexts.list);
 	spin_unlock(&i915->gem.contexts.lock);
@@ -661,37 +738,6 @@ i915_gem_create_context(struct drm_i915_private *i915, unsigned int flags)
 	return ctx;
 }
 
-static void
-destroy_kernel_context(struct i915_gem_context **ctxp)
-{
-	struct i915_gem_context *ctx;
-
-	/* Keep the context ref so that we can free it immediately ourselves */
-	ctx = i915_gem_context_get(fetch_and_zero(ctxp));
-	GEM_BUG_ON(!i915_gem_context_is_kernel(ctx));
-
-	context_close(ctx);
-	i915_gem_context_free(ctx);
-}
-
-struct i915_gem_context *
-i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio)
-{
-	struct i915_gem_context *ctx;
-
-	ctx = i915_gem_create_context(i915, 0);
-	if (IS_ERR(ctx))
-		return ctx;
-
-	i915_gem_context_clear_bannable(ctx);
-	i915_gem_context_set_persistence(ctx);
-	ctx->sched.priority = I915_USER_PRIORITY(prio);
-
-	GEM_BUG_ON(!i915_gem_context_is_kernel(ctx));
-
-	return ctx;
-}
-
 static void init_contexts(struct i915_gem_contexts *gc)
 {
 	spin_lock_init(&gc->lock);
@@ -701,32 +747,16 @@ static void init_contexts(struct i915_gem_contexts *gc)
 	init_llist_head(&gc->free_list);
 }
 
-int i915_gem_init_contexts(struct drm_i915_private *i915)
+void i915_gem_init__contexts(struct drm_i915_private *i915)
 {
-	struct i915_gem_context *ctx;
-
-	/* Reassure ourselves we are only called once */
-	GEM_BUG_ON(i915->kernel_context);
-
 	init_contexts(&i915->gem.contexts);
-
-	/* lowest priority; idle task */
-	ctx = i915_gem_context_create_kernel(i915, I915_PRIORITY_MIN);
-	if (IS_ERR(ctx)) {
-		DRM_ERROR("Failed to create default global context\n");
-		return PTR_ERR(ctx);
-	}
-	i915->kernel_context = ctx;
-
 	DRM_DEBUG_DRIVER("%s context support initialized\n",
 			 DRIVER_CAPS(i915)->has_logical_contexts ?
 			 "logical" : "fake");
-	return 0;
 }
 
 void i915_gem_driver_release__contexts(struct drm_i915_private *i915)
 {
-	destroy_kernel_context(&i915->kernel_context);
 	flush_work(&i915->gem.contexts.free_work);
 }
 
@@ -757,12 +787,8 @@ static int gem_context_register(struct i915_gem_context *ctx,
 	mutex_unlock(&ctx->mutex);
 
 	ctx->pid = get_task_pid(current, PIDTYPE_PID);
-	ctx->name = kasprintf(GFP_KERNEL, "%s[%d]",
-			      current->comm, pid_nr(ctx->pid));
-	if (!ctx->name) {
-		ret = -ENOMEM;
-		goto err_pid;
-	}
+	snprintf(ctx->name, sizeof(ctx->name), "%s[%d]",
+		 current->comm, pid_nr(ctx->pid));
 
 	/* And finally expose ourselves to userspace via the idr */
 	mutex_lock(&fpriv->context_idr_lock);
@@ -771,8 +797,6 @@ static int gem_context_register(struct i915_gem_context *ctx,
 	if (ret >= 0)
 		goto out;
 
-	kfree(fetch_and_zero(&ctx->name));
-err_pid:
 	put_pid(fetch_and_zero(&ctx->pid));
 out:
 	return ret;
@@ -801,7 +825,6 @@ int i915_gem_context_open(struct drm_i915_private *i915,
 	if (err < 0)
 		goto err_ctx;
 
-	GEM_BUG_ON(i915_gem_context_is_kernel(ctx));
 	GEM_BUG_ON(err > 0);
 
 	return 0;
@@ -1012,7 +1035,7 @@ static int get_ppgtt(struct drm_i915_file_private *file_priv,
 		return -ENODEV;
 
 	rcu_read_lock();
-	vm = i915_vm_get(ctx->vm);
+	vm = context_get_vm_rcu(ctx);
 	rcu_read_unlock();
 
 	ret = mutex_lock_interruptible(&file_priv->vm_idr_lock);
@@ -1049,7 +1072,7 @@ static void set_ppgtt_barrier(void *data)
 
 static int emit_ppgtt_update(struct i915_request *rq, void *data)
 {
-	struct i915_address_space *vm = rq->hw_context->vm;
+	struct i915_address_space *vm = rq->context->vm;
 	struct intel_engine_cs *engine = rq->engine;
 	u32 base = engine->mmio_base;
 	u32 *cs;
@@ -1096,9 +1119,6 @@ static int emit_ppgtt_update(struct i915_request *rq, void *data)
 		}
 		*cs++ = MI_NOOP;
 		intel_ring_advance(rq, cs);
-	} else {
-		/* ppGTT is not part of the legacy context image */
-		gen6_ppgtt_pin(i915_vm_to_ppgtt(vm));
 	}
 
 	return 0;
@@ -1106,10 +1126,20 @@ static int emit_ppgtt_update(struct i915_request *rq, void *data)
 
 static bool skip_ppgtt_update(struct intel_context *ce, void *data)
 {
+	if (!test_bit(CONTEXT_ALLOC_BIT, &ce->flags))
+		return true;
+
 	if (HAS_LOGICAL_RING_CONTEXTS(ce->engine->i915))
-		return !ce->state;
-	else
-		return !atomic_read(&ce->pin_count);
+		return false;
+
+	if (!atomic_read(&ce->pin_count))
+		return true;
+
+	/* ppGTT is not part of the legacy context image */
+	if (gen6_ppgtt_pin(i915_vm_to_ppgtt(ce->vm)))
+		return true;
+
+	return false;
 }
 
 static int set_ppgtt(struct drm_i915_file_private *file_priv,
@@ -1217,7 +1247,7 @@ gen8_modify_rpcs(struct intel_context *ce, struct intel_sseu sseu)
 	if (!intel_context_is_pinned(ce))
 		return 0;
 
-	rq = i915_request_create(ce->engine->kernel_context);
+	rq = intel_engine_create_kernel_request(ce->engine);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
@@ -1485,12 +1515,14 @@ set_engines__load_balance(struct i915_user_extension __user *base, void *data)
 		}
 	}
 
-	ce = intel_execlists_create_virtual(set->ctx, siblings, n);
+	ce = intel_execlists_create_virtual(siblings, n);
 	if (IS_ERR(ce)) {
 		err = PTR_ERR(ce);
 		goto out_siblings;
 	}
 
+	intel_context_set_gem(ce, set->ctx);
+
 	if (cmpxchg(&set->engines->engines[idx], NULL, ce)) {
 		intel_context_put(ce);
 		err = -EEXIST;
@@ -1660,12 +1692,14 @@ set_engines(struct i915_gem_context *ctx,
 			return -ENOENT;
 		}
 
-		ce = intel_context_create(ctx, engine);
+		ce = intel_context_create(engine);
 		if (IS_ERR(ce)) {
 			__free_engines(set.engines, n);
 			return PTR_ERR(ce);
 		}
 
+		intel_context_set_gem(ce, ctx);
+
 		set.engines->engines[n] = ce;
 	}
 	set.engines->num_engines = num_engines;
@@ -1806,6 +1840,44 @@ set_persistence(struct i915_gem_context *ctx,
 	return __context_set_persistence(ctx, args->value);
 }
 
+static void __apply_priority(struct intel_context *ce, void *arg)
+{
+	struct i915_gem_context *ctx = arg;
+
+	if (!intel_engine_has_semaphores(ce->engine))
+		return;
+
+	if (ctx->sched.priority >= I915_PRIORITY_NORMAL)
+		intel_context_set_use_semaphores(ce);
+	else
+		intel_context_clear_use_semaphores(ce);
+}
+
+static int set_priority(struct i915_gem_context *ctx,
+			const struct drm_i915_gem_context_param *args)
+{
+	s64 priority = args->value;
+
+	if (args->size)
+		return -EINVAL;
+
+	if (!(ctx->i915->caps.scheduler & I915_SCHEDULER_CAP_PRIORITY))
+		return -ENODEV;
+
+	if (priority > I915_CONTEXT_MAX_USER_PRIORITY ||
+	    priority < I915_CONTEXT_MIN_USER_PRIORITY)
+		return -EINVAL;
+
+	if (priority > I915_CONTEXT_DEFAULT_PRIORITY &&
+	    !capable(CAP_SYS_NICE))
+		return -EPERM;
+
+	ctx->sched.priority = I915_USER_PRIORITY(priority);
+	context_apply_all(ctx, __apply_priority, ctx);
+
+	return 0;
+}
+
 static int ctx_setparam(struct drm_i915_file_private *fpriv,
 			struct i915_gem_context *ctx,
 			struct drm_i915_gem_context_param *args)
@@ -1852,23 +1924,7 @@ static int ctx_setparam(struct drm_i915_file_private *fpriv,
 		break;
 
 	case I915_CONTEXT_PARAM_PRIORITY:
-		{
-			s64 priority = args->value;
-
-			if (args->size)
-				ret = -EINVAL;
-			else if (!(ctx->i915->caps.scheduler & I915_SCHEDULER_CAP_PRIORITY))
-				ret = -ENODEV;
-			else if (priority > I915_CONTEXT_MAX_USER_PRIORITY ||
-				 priority < I915_CONTEXT_MIN_USER_PRIORITY)
-				ret = -EINVAL;
-			else if (priority > I915_CONTEXT_DEFAULT_PRIORITY &&
-				 !capable(CAP_SYS_NICE))
-				ret = -EPERM;
-			else
-				ctx->sched.priority =
-					I915_USER_PRIORITY(priority);
-		}
+		ret = set_priority(ctx, args);
 		break;
 
 	case I915_CONTEXT_PARAM_SSEU:
@@ -1948,20 +2004,23 @@ static int clone_engines(struct i915_gem_context *dst,
 		 */
 		if (intel_engine_is_virtual(engine))
 			clone->engines[n] =
-				intel_execlists_clone_virtual(dst, engine);
+				intel_execlists_clone_virtual(engine);
 		else
-			clone->engines[n] = intel_context_create(dst, engine);
+			clone->engines[n] = intel_context_create(engine);
 		if (IS_ERR_OR_NULL(clone->engines[n])) {
 			__free_engines(clone, n);
 			goto err_unlock;
 		}
+
+		intel_context_set_gem(clone->engines[n], dst);
 	}
 	clone->num_engines = n;
 
 	user_engines = i915_gem_context_user_engines(src);
 	i915_gem_context_unlock_engines(src);
 
-	free_engines(dst->engines);
+	/* Serialised by constructor */
+	free_engines(__context_engines_static(dst));
 	RCU_INIT_POINTER(dst->engines, clone);
 	if (user_engines)
 		i915_gem_context_set_user_engines(dst);
@@ -1996,7 +2055,8 @@ static int clone_sseu(struct i915_gem_context *dst,
 	unsigned long n;
 	int err;
 
-	clone = dst->engines; /* no locking required; sole access */
+	/* no locking required; sole access under constructor*/
+	clone = __context_engines_static(dst);
 	if (e->num_engines != clone->num_engines) {
 		err = -EINVAL;
 		goto unlock;
@@ -2041,47 +2101,21 @@ static int clone_vm(struct i915_gem_context *dst,
 	struct i915_address_space *vm;
 	int err = 0;
 
-	rcu_read_lock();
-	do {
-		vm = rcu_dereference(src->vm);
-		if (!vm)
-			break;
-
-		if (!kref_get_unless_zero(&vm->ref))
-			continue;
-
-		/*
-		 * This ppgtt may have be reallocated between
-		 * the read and the kref, and reassigned to a third
-		 * context. In order to avoid inadvertent sharing
-		 * of this ppgtt with that third context (and not
-		 * src), we have to confirm that we have the same
-		 * ppgtt after passing through the strong memory
-		 * barrier implied by a successful
-		 * kref_get_unless_zero().
-		 *
-		 * Once we have acquired the current ppgtt of src,
-		 * we no longer care if it is released from src, as
-		 * it cannot be reallocated elsewhere.
-		 */
-
-		if (vm == rcu_access_pointer(src->vm))
-			break;
+	if (!rcu_access_pointer(src->vm))
+		return 0;
 
-		i915_vm_put(vm);
-	} while (1);
+	rcu_read_lock();
+	vm = context_get_vm_rcu(src);
 	rcu_read_unlock();
 
-	if (vm) {
-		if (!mutex_lock_interruptible(&dst->mutex)) {
-			__assign_ppgtt(dst, vm);
-			mutex_unlock(&dst->mutex);
-		} else {
-			err = -EINTR;
-		}
-		i915_vm_put(vm);
+	if (!mutex_lock_interruptible(&dst->mutex)) {
+		__assign_ppgtt(dst, vm);
+		mutex_unlock(&dst->mutex);
+	} else {
+		err = -EINTR;
 	}
 
+	i915_vm_put(vm);
 	return err;
 }
 
@@ -2167,8 +2201,7 @@ int i915_gem_context_create_ioctl(struct drm_device *dev, void *data,
 	ext_data.fpriv = file->driver_priv;
 	if (client_is_banned(ext_data.fpriv)) {
 		DRM_DEBUG("client %s[%d] banned from creating ctx\n",
-			  current->comm,
-			  pid_nr(get_task_pid(current, PIDTYPE_PID)));
+			  current->comm, task_pid_nr(current));
 		return -EIO;
 	}
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.h b/drivers/gpu/drm/i915/gem/i915_gem_context.h
index 18e50a769a6e..14f3cc1b7583 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context.h
@@ -91,26 +91,6 @@ static inline void i915_gem_context_clear_persistence(struct i915_gem_context *c
 	clear_bit(UCONTEXT_PERSISTENCE, &ctx->user_flags);
 }
 
-static inline bool i915_gem_context_is_banned(const struct i915_gem_context *ctx)
-{
-	return test_bit(CONTEXT_BANNED, &ctx->flags);
-}
-
-static inline void i915_gem_context_set_banned(struct i915_gem_context *ctx)
-{
-	set_bit(CONTEXT_BANNED, &ctx->flags);
-}
-
-static inline bool i915_gem_context_force_single_submission(const struct i915_gem_context *ctx)
-{
-	return test_bit(CONTEXT_FORCE_SINGLE_SUBMISSION, &ctx->flags);
-}
-
-static inline void i915_gem_context_set_force_single_submission(struct i915_gem_context *ctx)
-{
-	__set_bit(CONTEXT_FORCE_SINGLE_SUBMISSION, &ctx->flags);
-}
-
 static inline bool
 i915_gem_context_user_engines(const struct i915_gem_context *ctx)
 {
@@ -129,31 +109,8 @@ i915_gem_context_clear_user_engines(struct i915_gem_context *ctx)
 	clear_bit(CONTEXT_USER_ENGINES, &ctx->flags);
 }
 
-static inline bool
-i915_gem_context_nopreempt(const struct i915_gem_context *ctx)
-{
-	return test_bit(CONTEXT_NOPREEMPT, &ctx->flags);
-}
-
-static inline void
-i915_gem_context_set_nopreempt(struct i915_gem_context *ctx)
-{
-	set_bit(CONTEXT_NOPREEMPT, &ctx->flags);
-}
-
-static inline void
-i915_gem_context_clear_nopreempt(struct i915_gem_context *ctx)
-{
-	clear_bit(CONTEXT_NOPREEMPT, &ctx->flags);
-}
-
-static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx)
-{
-	return !ctx->file_priv;
-}
-
 /* i915_gem_context.c */
-int __must_check i915_gem_init_contexts(struct drm_i915_private *i915);
+void i915_gem_init__contexts(struct drm_i915_private *i915);
 void i915_gem_driver_release__contexts(struct drm_i915_private *i915);
 
 int i915_gem_context_open(struct drm_i915_private *i915,
@@ -178,9 +135,6 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data,
 int i915_gem_context_reset_stats_ioctl(struct drm_device *dev, void *data,
 				       struct drm_file *file);
 
-struct i915_gem_context *
-i915_gem_context_create_kernel(struct drm_i915_private *i915, int prio);
-
 static inline struct i915_gem_context *
 i915_gem_context_get(struct i915_gem_context *ctx)
 {
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h
index 3870dd5daaa0..017ca803ab47 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_context_types.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_context_types.h
@@ -100,15 +100,6 @@ struct i915_gem_context {
 	 */
 	struct pid *pid;
 
-	/**
-	 * @name: arbitrary name
-	 *
-	 * A name is constructed for the context from the creator's process
-	 * name, pid and user handle in order to uniquely identify the
-	 * context in messages.
-	 */
-	const char *name;
-
 	/** link: place with &drm_i915_private.context_list */
 	struct list_head link;
 	struct llist_node free_link;
@@ -143,11 +134,8 @@ struct i915_gem_context {
 	 * @flags: small set of booleans
 	 */
 	unsigned long flags;
-#define CONTEXT_BANNED			0
-#define CONTEXT_CLOSED			1
-#define CONTEXT_FORCE_SINGLE_SUBMISSION	2
-#define CONTEXT_USER_ENGINES		3
-#define CONTEXT_NOPREEMPT		4
+#define CONTEXT_CLOSED			0
+#define CONTEXT_USER_ENGINES		1
 
 	struct mutex mutex;
 
@@ -177,12 +165,14 @@ struct i915_gem_context {
 	 */
 	struct radix_tree_root handles_vma;
 
-	/** jump_whitelist: Bit array for tracking cmds during cmdparsing
-	 *  Guarded by struct_mutex
+	/**
+	 * @name: arbitrary name, used for user debug
+	 *
+	 * A name is constructed for the context from the creator's process
+	 * name, pid and user handle in order to uniquely identify the
+	 * context in messages.
 	 */
-	unsigned long *jump_whitelist;
-	/** jump_whitelist_cmds: No of cmd slots available */
-	u32 jump_whitelist_cmds;
+	char name[TASK_COMM_LEN + 8];
 };
 
 #endif /* __I915_GEM_CONTEXT_TYPES_H__ */
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_domain.c b/drivers/gpu/drm/i915/gem/i915_gem_domain.c
index 9937b4c341f1..0cc40e77bbd2 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_domain.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_domain.c
@@ -12,6 +12,8 @@
 #include "i915_gem_ioctls.h"
 #include "i915_gem_object.h"
 #include "i915_vma.h"
+#include "i915_gem_lmem.h"
+#include "i915_gem_mman.h"
 
 static void __i915_gem_object_flush_for_display(struct drm_i915_gem_object *obj)
 {
@@ -148,9 +150,17 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 	GEM_BUG_ON((obj->write_domain & ~I915_GEM_DOMAIN_GTT) != 0);
 	obj->read_domains |= I915_GEM_DOMAIN_GTT;
 	if (write) {
+		struct i915_vma *vma;
+
 		obj->read_domains = I915_GEM_DOMAIN_GTT;
 		obj->write_domain = I915_GEM_DOMAIN_GTT;
 		obj->mm.dirty = true;
+
+		spin_lock(&obj->vma.lock);
+		for_each_ggtt_vma(vma, obj)
+			if (i915_vma_is_bound(vma, I915_VMA_GLOBAL_BIND))
+				i915_vma_set_ggtt_write(vma);
+		spin_unlock(&obj->vma.lock);
 	}
 
 	i915_gem_object_unpin_pages(obj);
@@ -175,138 +185,34 @@ i915_gem_object_set_to_gtt_domain(struct drm_i915_gem_object *obj, bool write)
 int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj,
 				    enum i915_cache_level cache_level)
 {
-	struct i915_vma *vma;
 	int ret;
 
-	assert_object_held(obj);
-
 	if (obj->cache_level == cache_level)
 		return 0;
 
-	/* Inspect the list of currently bound VMA and unbind any that would
-	 * be invalid given the new cache-level. This is principally to
-	 * catch the issue of the CS prefetch crossing page boundaries and
-	 * reading an invalid PTE on older architectures.
-	 */
-restart:
-	list_for_each_entry(vma, &obj->vma.list, obj_link) {
-		if (!drm_mm_node_allocated(&vma->node))
-			continue;
-
-		if (i915_vma_is_pinned(vma)) {
-			DRM_DEBUG("can not change the cache level of pinned objects\n");
-			return -EBUSY;
-		}
-
-		if (!i915_vma_is_closed(vma) &&
-		    i915_gem_valid_gtt_space(vma, cache_level))
-			continue;
-
-		ret = i915_vma_unbind(vma);
-		if (ret)
-			return ret;
-
-		/* As unbinding may affect other elements in the
-		 * obj->vma_list (due to side-effects from retiring
-		 * an active vma), play safe and restart the iterator.
-		 */
-		goto restart;
-	}
+	ret = i915_gem_object_wait(obj,
+				   I915_WAIT_INTERRUPTIBLE |
+				   I915_WAIT_ALL,
+				   MAX_SCHEDULE_TIMEOUT);
+	if (ret)
+		return ret;
 
-	/* We can reuse the existing drm_mm nodes but need to change the
-	 * cache-level on the PTE. We could simply unbind them all and
-	 * rebind with the correct cache-level on next use. However since
-	 * we already have a valid slot, dma mapping, pages etc, we may as
-	 * rewrite the PTE in the belief that doing so tramples upon less
-	 * state and so involves less work.
-	 */
-	if (atomic_read(&obj->bind_count)) {
-		struct drm_i915_private *i915 = to_i915(obj->base.dev);
+	ret = i915_gem_object_lock_interruptible(obj);
+	if (ret)
+		return ret;
 
-		/* Before we change the PTE, the GPU must not be accessing it.
-		 * If we wait upon the object, we know that all the bound
-		 * VMA are no longer active.
-		 */
-		ret = i915_gem_object_wait(obj,
-					   I915_WAIT_INTERRUPTIBLE |
-					   I915_WAIT_ALL,
-					   MAX_SCHEDULE_TIMEOUT);
-		if (ret)
-			return ret;
-
-		if (!HAS_LLC(i915) && cache_level != I915_CACHE_NONE) {
-			intel_wakeref_t wakeref =
-				intel_runtime_pm_get(&i915->runtime_pm);
-
-			/*
-			 * Access to snoopable pages through the GTT is
-			 * incoherent and on some machines causes a hard
-			 * lockup. Relinquish the CPU mmaping to force
-			 * userspace to refault in the pages and we can
-			 * then double check if the GTT mapping is still
-			 * valid for that pointer access.
-			 */
-			ret = mutex_lock_interruptible(&i915->ggtt.vm.mutex);
-			if (ret) {
-				intel_runtime_pm_put(&i915->runtime_pm,
-						     wakeref);
-				return ret;
-			}
-
-			if (obj->userfault_count)
-				__i915_gem_object_release_mmap(obj);
-
-			/*
-			 * As we no longer need a fence for GTT access,
-			 * we can relinquish it now (and so prevent having
-			 * to steal a fence from someone else on the next
-			 * fence request). Note GPU activity would have
-			 * dropped the fence as all snoopable access is
-			 * supposed to be linear.
-			 */
-			for_each_ggtt_vma(vma, obj) {
-				ret = i915_vma_revoke_fence(vma);
-				if (ret)
-					break;
-			}
-			mutex_unlock(&i915->ggtt.vm.mutex);
-			intel_runtime_pm_put(&i915->runtime_pm, wakeref);
-			if (ret)
-				return ret;
-		} else {
-			/*
-			 * We either have incoherent backing store and
-			 * so no GTT access or the architecture is fully
-			 * coherent. In such cases, existing GTT mmaps
-			 * ignore the cache bit in the PTE and we can
-			 * rewrite it without confusing the GPU or having
-			 * to force userspace to fault back in its mmaps.
-			 */
-		}
-
-		list_for_each_entry(vma, &obj->vma.list, obj_link) {
-			if (!drm_mm_node_allocated(&vma->node))
-				continue;
-
-			/* Wait for an earlier async bind, need to rewrite it */
-			ret = i915_vma_sync(vma);
-			if (ret)
-				return ret;
-
-			ret = i915_vma_bind(vma, cache_level, PIN_UPDATE, NULL);
-			if (ret)
-				return ret;
-		}
+	/* Always invalidate stale cachelines */
+	if (obj->cache_level != cache_level) {
+		i915_gem_object_set_cache_coherency(obj, cache_level);
+		obj->cache_dirty = true;
 	}
 
-	list_for_each_entry(vma, &obj->vma.list, obj_link) {
-		if (i915_vm_has_cache_coloring(vma->vm))
-			vma->node.color = cache_level;
-	}
-	i915_gem_object_set_cache_coherency(obj, cache_level);
-	obj->cache_dirty = true; /* Always invalidate stale cachelines */
+	i915_gem_object_unlock(obj);
 
-	return 0;
+	/* The cache-level will be applied when each vma is rebound. */
+	return i915_gem_object_unbind(obj,
+				      I915_GEM_OBJECT_UNBIND_ACTIVE |
+				      I915_GEM_OBJECT_UNBIND_BARRIER);
 }
 
 int i915_gem_get_caching_ioctl(struct drm_device *dev, void *data,
@@ -387,20 +293,7 @@ int i915_gem_set_caching_ioctl(struct drm_device *dev, void *data,
 		goto out;
 	}
 
-	if (obj->cache_level == level)
-		goto out;
-
-	ret = i915_gem_object_wait(obj,
-				   I915_WAIT_INTERRUPTIBLE,
-				   MAX_SCHEDULE_TIMEOUT);
-	if (ret)
-		goto out;
-
-	ret = i915_gem_object_lock_interruptible(obj);
-	if (ret == 0) {
-		ret = i915_gem_object_set_cache_level(obj, level);
-		i915_gem_object_unlock(obj);
-	}
+	ret = i915_gem_object_set_cache_level(obj, level);
 
 out:
 	i915_gem_object_put(obj);
@@ -419,10 +312,13 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 				     const struct i915_ggtt_view *view,
 				     unsigned int flags)
 {
+	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	struct i915_vma *vma;
 	int ret;
 
-	assert_object_held(obj);
+	/* Frame buffer must be in LMEM (no migration yet) */
+	if (HAS_LMEM(i915) && !i915_gem_object_is_lmem(obj))
+		return ERR_PTR(-EINVAL);
 
 	/*
 	 * The display engine is not coherent with the LLC cache on gen6.  As
@@ -435,7 +331,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 	 * with that bit in the PTE to main memory with just one PIPE_CONTROL.
 	 */
 	ret = i915_gem_object_set_cache_level(obj,
-					      HAS_WT(to_i915(obj->base.dev)) ?
+					      HAS_WT(i915) ?
 					      I915_CACHE_WT : I915_CACHE_NONE);
 	if (ret)
 		return ERR_PTR(ret);
@@ -462,13 +358,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj,
 
 	vma->display_alignment = max_t(u64, vma->display_alignment, alignment);
 
-	__i915_gem_object_flush_for_display(obj);
-
-	/*
-	 * It should now be out of any other write domains, and we can update
-	 * the domain values for our changes.
-	 */
-	obj->read_domains |= I915_GEM_DOMAIN_GTT;
+	i915_gem_object_flush_if_display(obj);
 
 	return vma;
 }
@@ -479,8 +369,11 @@ static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
 	struct i915_vma *vma;
 
 	GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
+	if (!atomic_read(&obj->bind_count))
+		return;
 
 	mutex_lock(&i915->ggtt.vm.mutex);
+	spin_lock(&obj->vma.lock);
 	for_each_ggtt_vma(vma, obj) {
 		if (!drm_mm_node_allocated(&vma->node))
 			continue;
@@ -488,6 +381,7 @@ static void i915_gem_object_bump_inactive_ggtt(struct drm_i915_gem_object *obj)
 		GEM_BUG_ON(vma->vm != &i915->ggtt.vm);
 		list_move_tail(&vma->vm_link, &vma->vm->bound_list);
 	}
+	spin_unlock(&obj->vma.lock);
 	mutex_unlock(&i915->ggtt.vm.mutex);
 
 	if (i915_gem_object_is_shrinkable(obj)) {
@@ -664,7 +558,7 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data,
 	i915_gem_object_unlock(obj);
 
 	if (write_domain)
-		intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU);
+		i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
 
 out_unpin:
 	i915_gem_object_unpin_pages(obj);
@@ -784,7 +678,7 @@ int i915_gem_object_prepare_write(struct drm_i915_gem_object *obj,
 	}
 
 out:
-	intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU);
+	i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
 	obj->mm.dirty = true;
 	/* return with the pages pinned */
 	return 0;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
index f0998f1225af..cbd2bcade3c8 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c
@@ -25,6 +25,7 @@
 #include "i915_gem_clflush.h"
 #include "i915_gem_context.h"
 #include "i915_gem_ioctls.h"
+#include "i915_sw_fence_work.h"
 #include "i915_trace.h"
 
 enum {
@@ -228,6 +229,7 @@ struct i915_execbuffer {
 
 	struct i915_request *request; /** our request to build */
 	struct i915_vma *batch; /** identity of the batch obj/vma */
+	struct i915_vma *trampoline; /** trampoline used for chaining */
 
 	/** actual size of execobj[] as we may extend it for the cmdparser */
 	unsigned int buffer_count;
@@ -253,7 +255,6 @@ struct i915_execbuffer {
 		bool has_fence : 1;
 		bool needs_unfenced : 1;
 
-		struct intel_context *ce;
 		struct i915_request *rq;
 		u32 *rq_cmd;
 		unsigned int rq_size;
@@ -277,25 +278,6 @@ struct i915_execbuffer {
 
 #define exec_entry(EB, VMA) (&(EB)->exec[(VMA)->exec_flags - (EB)->flags])
 
-/*
- * Used to convert any address to canonical form.
- * Starting from gen8, some commands (e.g. STATE_BASE_ADDRESS,
- * MI_LOAD_REGISTER_MEM and others, see Broadwell PRM Vol2a) require the
- * addresses to be in a canonical form:
- * "GraphicsAddress[63:48] are ignored by the HW and assumed to be in correct
- * canonical form [63:48] == [47]."
- */
-#define GEN8_HIGH_ADDRESS_BIT 47
-static inline u64 gen8_canonical_addr(u64 address)
-{
-	return sign_extend64(address, GEN8_HIGH_ADDRESS_BIT);
-}
-
-static inline u64 gen8_noncanonical_addr(u64 address)
-{
-	return address & GENMASK_ULL(GEN8_HIGH_ADDRESS_BIT, 0);
-}
-
 static inline bool eb_use_cmdparser(const struct i915_execbuffer *eb)
 {
 	return intel_engine_requires_cmd_parser(eb->engine) ||
@@ -748,9 +730,6 @@ static int eb_lookup_vmas(struct i915_execbuffer *eb)
 	unsigned int i, batch;
 	int err;
 
-	if (unlikely(i915_gem_context_is_banned(eb->gem_context)))
-		return -EIO;
-
 	INIT_LIST_HEAD(&eb->relocs);
 	INIT_LIST_HEAD(&eb->unbound);
 
@@ -886,9 +865,6 @@ static void eb_destroy(const struct i915_execbuffer *eb)
 {
 	GEM_BUG_ON(eb->reloc_cache.rq);
 
-	if (eb->reloc_cache.ce)
-		intel_context_put(eb->reloc_cache.ce);
-
 	if (eb->lut_size > 0)
 		kfree(eb->buckets);
 }
@@ -912,7 +888,6 @@ static void reloc_cache_init(struct reloc_cache *cache,
 	cache->has_fence = cache->gen < 4;
 	cache->needs_unfenced = INTEL_INFO(i915)->unfenced_needs_alignment;
 	cache->node.flags = 0;
-	cache->ce = NULL;
 	cache->rq = NULL;
 	cache->rq_size = 0;
 }
@@ -1182,7 +1157,7 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb,
 	if (err)
 		goto err_unmap;
 
-	rq = intel_context_create_request(cache->ce);
+	rq = i915_request_create(eb->context);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto err_unpin;
@@ -1246,36 +1221,9 @@ static u32 *reloc_gpu(struct i915_execbuffer *eb,
 	if (unlikely(!cache->rq)) {
 		int err;
 
-		/* If we need to copy for the cmdparser, we will stall anyway */
-		if (eb_use_cmdparser(eb))
-			return ERR_PTR(-EWOULDBLOCK);
-
 		if (!intel_engine_can_store_dword(eb->engine))
 			return ERR_PTR(-ENODEV);
 
-		if (!cache->ce) {
-			struct intel_context *ce;
-
-			/*
-			 * The CS pre-parser can pre-fetch commands across
-			 * memory sync points and starting gen12 it is able to
-			 * pre-fetch across BB_START and BB_END boundaries
-			 * (within the same context). We therefore use a
-			 * separate context gen12+ to guarantee that the reloc
-			 * writes land before the parser gets to the target
-			 * memory location.
-			 */
-			if (cache->gen >= 12)
-				ce = intel_context_create(eb->context->gem_context,
-							  eb->engine);
-			else
-				ce = intel_context_get(eb->context);
-			if (IS_ERR(ce))
-				return ERR_CAST(ce);
-
-			cache->ce = ce;
-		}
-
 		err = __reloc_gpu_alloc(eb, vma, len);
 		if (unlikely(err))
 			return ERR_PTR(err);
@@ -1943,15 +1891,15 @@ err_skip:
 	return err;
 }
 
-static bool i915_gem_check_execbuffer(struct drm_i915_gem_execbuffer2 *exec)
+static int i915_gem_check_execbuffer(struct drm_i915_gem_execbuffer2 *exec)
 {
 	if (exec->flags & __I915_EXEC_ILLEGAL_FLAGS)
-		return false;
+		return -EINVAL;
 
 	/* Kernel clipping was a DRI1 misfeature */
 	if (!(exec->flags & I915_EXEC_FENCE_ARRAY)) {
 		if (exec->num_cliprects || exec->cliprects_ptr)
-			return false;
+			return -EINVAL;
 	}
 
 	if (exec->DR4 == 0xffffffff) {
@@ -1959,12 +1907,12 @@ static bool i915_gem_check_execbuffer(struct drm_i915_gem_execbuffer2 *exec)
 		exec->DR4 = 0;
 	}
 	if (exec->DR1 || exec->DR4)
-		return false;
+		return -EINVAL;
 
 	if ((exec->batch_start_offset | exec->batch_len) & 0x7)
-		return false;
+		return -EINVAL;
 
-	return true;
+	return 0;
 }
 
 static int i915_reset_gen7_sol_offsets(struct i915_request *rq)
@@ -1993,99 +1941,179 @@ static int i915_reset_gen7_sol_offsets(struct i915_request *rq)
 }
 
 static struct i915_vma *
-shadow_batch_pin(struct i915_execbuffer *eb, struct drm_i915_gem_object *obj)
+shadow_batch_pin(struct drm_i915_gem_object *obj,
+		 struct i915_address_space *vm,
+		 unsigned int flags)
 {
-	struct drm_i915_private *dev_priv = eb->i915;
-	struct i915_vma * const vma = *eb->vma;
-	struct i915_address_space *vm;
-	u64 flags;
+	struct i915_vma *vma;
+	int err;
 
-	/*
-	 * PPGTT backed shadow buffers must be mapped RO, to prevent
-	 * post-scan tampering
-	 */
-	if (CMDPARSER_USES_GGTT(dev_priv)) {
-		flags = PIN_GLOBAL;
-		vm = &dev_priv->ggtt.vm;
-	} else if (vma->vm->has_read_only) {
-		flags = PIN_USER;
-		vm = vma->vm;
-		i915_gem_object_set_readonly(obj);
-	} else {
-		DRM_DEBUG("Cannot prevent post-scan tampering without RO capable vm\n");
-		return ERR_PTR(-EINVAL);
-	}
+	vma = i915_vma_instance(obj, vm, NULL);
+	if (IS_ERR(vma))
+		return vma;
+
+	err = i915_vma_pin(vma, 0, 0, flags);
+	if (err)
+		return ERR_PTR(err);
 
-	return i915_gem_object_pin(obj, vm, NULL, 0, 0, flags);
+	return vma;
 }
 
-static struct i915_vma *eb_parse(struct i915_execbuffer *eb)
+struct eb_parse_work {
+	struct dma_fence_work base;
+	struct intel_engine_cs *engine;
+	struct i915_vma *batch;
+	struct i915_vma *shadow;
+	struct i915_vma *trampoline;
+	unsigned int batch_offset;
+	unsigned int batch_length;
+};
+
+static int __eb_parse(struct dma_fence_work *work)
 {
-	struct intel_engine_pool_node *pool;
-	struct i915_vma *vma;
-	u64 batch_start;
-	u64 shadow_batch_start;
+	struct eb_parse_work *pw = container_of(work, typeof(*pw), base);
+
+	return intel_engine_cmd_parser(pw->engine,
+				       pw->batch,
+				       pw->batch_offset,
+				       pw->batch_length,
+				       pw->shadow,
+				       pw->trampoline);
+}
+
+static const struct dma_fence_work_ops eb_parse_ops = {
+	.name = "eb_parse",
+	.work = __eb_parse,
+};
+
+static int eb_parse_pipeline(struct i915_execbuffer *eb,
+			     struct i915_vma *shadow,
+			     struct i915_vma *trampoline)
+{
+	struct eb_parse_work *pw;
 	int err;
 
-	pool = intel_engine_get_pool(eb->engine, eb->batch_len);
-	if (IS_ERR(pool))
-		return ERR_CAST(pool);
+	pw = kzalloc(sizeof(*pw), GFP_KERNEL);
+	if (!pw)
+		return -ENOMEM;
 
-	vma = shadow_batch_pin(eb, pool->obj);
-	if (IS_ERR(vma))
-		goto err;
+	dma_fence_work_init(&pw->base, &eb_parse_ops);
 
-	batch_start = gen8_canonical_addr(eb->batch->node.start) +
-		      eb->batch_start_offset;
+	pw->engine = eb->engine;
+	pw->batch = eb->batch;
+	pw->batch_offset = eb->batch_start_offset;
+	pw->batch_length = eb->batch_len;
+	pw->shadow = shadow;
+	pw->trampoline = trampoline;
 
-	shadow_batch_start = gen8_canonical_addr(vma->node.start);
+	dma_resv_lock(pw->batch->resv, NULL);
 
-	err = intel_engine_cmd_parser(eb->gem_context,
-				      eb->engine,
-				      eb->batch->obj,
-				      batch_start,
-				      eb->batch_start_offset,
-				      eb->batch_len,
-				      pool->obj,
-				      shadow_batch_start);
+	err = dma_resv_reserve_shared(pw->batch->resv, 1);
+	if (err)
+		goto err_batch_unlock;
 
-	if (err) {
-		i915_vma_unpin(vma);
+	/* Wait for all writes (and relocs) into the batch to complete */
+	err = i915_sw_fence_await_reservation(&pw->base.chain,
+					      pw->batch->resv, NULL, false,
+					      0, I915_FENCE_GFP);
+	if (err < 0)
+		goto err_batch_unlock;
+
+	/* Keep the batch alive and unwritten as we parse */
+	dma_resv_add_shared_fence(pw->batch->resv, &pw->base.dma);
+
+	dma_resv_unlock(pw->batch->resv);
 
+	/* Force execution to wait for completion of the parser */
+	dma_resv_lock(shadow->resv, NULL);
+	dma_resv_add_excl_fence(shadow->resv, &pw->base.dma);
+	dma_resv_unlock(shadow->resv);
+
+	dma_fence_work_commit(&pw->base);
+	return 0;
+
+err_batch_unlock:
+	dma_resv_unlock(pw->batch->resv);
+	kfree(pw);
+	return err;
+}
+
+static int eb_parse(struct i915_execbuffer *eb)
+{
+	struct intel_engine_pool_node *pool;
+	struct i915_vma *shadow, *trampoline;
+	unsigned int len;
+	int err;
+
+	if (!eb_use_cmdparser(eb))
+		return 0;
+
+	len = eb->batch_len;
+	if (!CMDPARSER_USES_GGTT(eb->i915)) {
 		/*
-		 * Unsafe GGTT-backed buffers can still be submitted safely
-		 * as non-secure.
-		 * For PPGTT backing however, we have no choice but to forcibly
-		 * reject unsafe buffers
+		 * ppGTT backed shadow buffers must be mapped RO, to prevent
+		 * post-scan tampering
 		 */
-		if (CMDPARSER_USES_GGTT(eb->i915) && (err == -EACCES))
-			/* Execute original buffer non-secure */
-			vma = NULL;
-		else
-			vma = ERR_PTR(err);
+		if (!eb->context->vm->has_read_only) {
+			DRM_DEBUG("Cannot prevent post-scan tampering without RO capable vm\n");
+			return -EINVAL;
+		}
+	} else {
+		len += I915_CMD_PARSER_TRAMPOLINE_SIZE;
+	}
+
+	pool = intel_engine_get_pool(eb->engine, len);
+	if (IS_ERR(pool))
+		return PTR_ERR(pool);
+
+	shadow = shadow_batch_pin(pool->obj, eb->context->vm, PIN_USER);
+	if (IS_ERR(shadow)) {
+		err = PTR_ERR(shadow);
 		goto err;
 	}
+	i915_gem_object_set_readonly(shadow->obj);
+
+	trampoline = NULL;
+	if (CMDPARSER_USES_GGTT(eb->i915)) {
+		trampoline = shadow;
+
+		shadow = shadow_batch_pin(pool->obj,
+					  &eb->engine->gt->ggtt->vm,
+					  PIN_GLOBAL);
+		if (IS_ERR(shadow)) {
+			err = PTR_ERR(shadow);
+			shadow = trampoline;
+			goto err_shadow;
+		}
+
+		eb->batch_flags |= I915_DISPATCH_SECURE;
+	}
+
+	err = eb_parse_pipeline(eb, shadow, trampoline);
+	if (err)
+		goto err_trampoline;
 
-	eb->vma[eb->buffer_count] = i915_vma_get(vma);
+	eb->vma[eb->buffer_count] = i915_vma_get(shadow);
 	eb->flags[eb->buffer_count] =
 		__EXEC_OBJECT_HAS_PIN | __EXEC_OBJECT_HAS_REF;
-	vma->exec_flags = &eb->flags[eb->buffer_count];
+	shadow->exec_flags = &eb->flags[eb->buffer_count];
 	eb->buffer_count++;
 
+	eb->trampoline = trampoline;
 	eb->batch_start_offset = 0;
-	eb->batch = vma;
+	eb->batch = shadow;
 
-	if (CMDPARSER_USES_GGTT(eb->i915))
-		eb->batch_flags |= I915_DISPATCH_SECURE;
-
-	/* eb->batch_len unchanged */
-
-	vma->private = pool;
-	return vma;
+	shadow->private = pool;
+	return 0;
 
+err_trampoline:
+	if (trampoline)
+		i915_vma_unpin(trampoline);
+err_shadow:
+	i915_vma_unpin(shadow);
 err:
 	intel_engine_pool_put(pool);
-	return vma;
+	return err;
 }
 
 static void
@@ -2134,7 +2162,17 @@ static int eb_submit(struct i915_execbuffer *eb)
 	if (err)
 		return err;
 
-	if (i915_gem_context_nopreempt(eb->gem_context))
+	if (eb->trampoline) {
+		GEM_BUG_ON(eb->batch_start_offset);
+		err = eb->engine->emit_bb_start(eb->request,
+						eb->trampoline->node.start +
+						eb->batch_len,
+						0, 0);
+		if (err)
+			return err;
+	}
+
+	if (intel_context_nopreempt(eb->context))
 		eb->request->flags |= I915_REQUEST_NOPREEMPT;
 
 	return 0;
@@ -2220,6 +2258,9 @@ static int __eb_pin_engine(struct i915_execbuffer *eb, struct intel_context *ce)
 	if (err)
 		return err;
 
+	if (unlikely(intel_context_is_banned(ce)))
+		return -EIO;
+
 	/*
 	 * Pinning the contexts may generate requests in order to acquire
 	 * GGTT space, so do this first before we reserve a seqno for
@@ -2515,6 +2556,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	eb.buffer_count = args->buffer_count;
 	eb.batch_start_offset = args->batch_start_offset;
 	eb.batch_len = args->batch_len;
+	eb.trampoline = NULL;
 
 	eb.batch_flags = 0;
 	if (args->flags & I915_EXEC_SECURE) {
@@ -2606,15 +2648,9 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	if (eb.batch_len == 0)
 		eb.batch_len = eb.batch->size - eb.batch_start_offset;
 
-	if (eb_use_cmdparser(&eb)) {
-		struct i915_vma *vma;
-
-		vma = eb_parse(&eb);
-		if (IS_ERR(vma)) {
-			err = PTR_ERR(vma);
-			goto err_vma;
-		}
-	}
+	err = eb_parse(&eb);
+	if (err)
+		goto err_vma;
 
 	/*
 	 * snb/ivb/vlv conflate the "batch in ppgtt" bit with the "non-secure
@@ -2694,6 +2730,7 @@ i915_gem_do_execbuffer(struct drm_device *dev,
 	err = eb_submit(&eb);
 err_request:
 	add_to_client(eb.request, file);
+	i915_request_get(eb.request);
 	i915_request_add(eb.request);
 
 	if (fences)
@@ -2709,6 +2746,7 @@ err_request:
 			fput(out_fence->file);
 		}
 	}
+	i915_request_put(eb.request);
 
 err_batch_unpin:
 	if (eb.batch_flags & I915_DISPATCH_SECURE)
@@ -2718,6 +2756,8 @@ err_batch_unpin:
 err_vma:
 	if (eb.exec)
 		eb_release_vmas(&eb);
+	if (eb.trampoline)
+		i915_vma_unpin(eb.trampoline);
 	mutex_unlock(&dev->struct_mutex);
 err_engine:
 	eb_unpin_engine(&eb);
@@ -2787,8 +2827,9 @@ i915_gem_execbuffer_ioctl(struct drm_device *dev, void *data,
 	exec2.flags = I915_EXEC_RENDER;
 	i915_execbuffer2_set_context_id(exec2, 0);
 
-	if (!i915_gem_check_execbuffer(&exec2))
-		return -EINVAL;
+	err = i915_gem_check_execbuffer(&exec2);
+	if (err)
+		return err;
 
 	/* Copy in the exec list from userland */
 	exec_list = kvmalloc_array(count, sizeof(*exec_list),
@@ -2865,8 +2906,9 @@ i915_gem_execbuffer2_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 	}
 
-	if (!i915_gem_check_execbuffer(args))
-		return -EINVAL;
+	err = i915_gem_check_execbuffer(args);
+	if (err)
+		return err;
 
 	/* Allocate an extra slot for use by the command parser */
 	exec2_list = kvmalloc_array(count + 1, eb_element_size(),
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_ioctls.h b/drivers/gpu/drm/i915/gem/i915_gem_ioctls.h
index ddc7f2a52b3e..87d8b27f426d 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_ioctls.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_ioctls.h
@@ -28,8 +28,8 @@ int i915_gem_madvise_ioctl(struct drm_device *dev, void *data,
 			   struct drm_file *file);
 int i915_gem_mmap_ioctl(struct drm_device *dev, void *data,
 			struct drm_file *file);
-int i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data,
-			    struct drm_file *file);
+int i915_gem_mmap_offset_ioctl(struct drm_device *dev, void *data,
+			       struct drm_file *file);
 int i915_gem_pread_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file);
 int i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_lmem.c b/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
index 0e2bf6b7e143..520cc9cac471 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_lmem.c
@@ -79,9 +79,6 @@ __i915_gem_lmem_object_create(struct intel_memory_region *mem,
 	struct drm_i915_private *i915 = mem->i915;
 	struct drm_i915_gem_object *obj;
 
-	if (size > BIT(mem->mm.max_order) * mem->mm.chunk_size)
-		return ERR_PTR(-E2BIG);
-
 	obj = i915_gem_object_alloc();
 	if (!obj)
 		return ERR_PTR(-ENOMEM);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_mman.c b/drivers/gpu/drm/i915/gem/i915_gem_mman.c
index e3002849844b..879fff8adc48 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_mman.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_mman.c
@@ -5,6 +5,7 @@
  */
 
 #include <linux/mman.h>
+#include <linux/pfn_t.h>
 #include <linux/sizes.h>
 
 #include "gt/intel_gt.h"
@@ -14,7 +15,9 @@
 #include "i915_gem_gtt.h"
 #include "i915_gem_ioctls.h"
 #include "i915_gem_object.h"
+#include "i915_gem_mman.h"
 #include "i915_trace.h"
+#include "i915_user_extensions.h"
 #include "i915_vma.h"
 
 static inline bool
@@ -144,6 +147,9 @@ static unsigned int tile_row_pages(const struct drm_i915_gem_object *obj)
  * 3 - Remove implicit set-domain(GTT) and synchronisation on initial
  *     pagefault; swapin remains transparent.
  *
+ * 4 - Support multiple fault handlers per object depending on object's
+ *     backing storage (a.k.a. MMAP_OFFSET).
+ *
  * Restrictions:
  *
  *  * snoopable objects cannot be accessed via the GTT. It can cause machine
@@ -171,7 +177,7 @@ static unsigned int tile_row_pages(const struct drm_i915_gem_object *obj)
  */
 int i915_gem_mmap_gtt_version(void)
 {
-	return 3;
+	return 4;
 }
 
 static inline struct i915_ggtt_view
@@ -197,29 +203,83 @@ compute_partial_view(const struct drm_i915_gem_object *obj,
 	return view;
 }
 
-/**
- * i915_gem_fault - fault a page into the GTT
- * @vmf: fault info
- *
- * The fault handler is set up by drm_gem_mmap() when a object is GTT mapped
- * from userspace.  The fault handler takes care of binding the object to
- * the GTT (if needed), allocating and programming a fence register (again,
- * only if needed based on whether the old reg is still valid or the object
- * is tiled) and inserting a new PTE into the faulting process.
- *
- * Note that the faulting process may involve evicting existing objects
- * from the GTT and/or fence registers to make room.  So performance may
- * suffer if the GTT working set is large or there are few fence registers
- * left.
- *
- * The current feature set supported by i915_gem_fault() and thus GTT mmaps
- * is exposed via I915_PARAM_MMAP_GTT_VERSION (see i915_gem_mmap_gtt_version).
- */
-vm_fault_t i915_gem_fault(struct vm_fault *vmf)
+static vm_fault_t i915_error_to_vmf_fault(int err)
+{
+	switch (err) {
+	default:
+		WARN_ONCE(err, "unhandled error in %s: %i\n", __func__, err);
+		/* fallthrough */
+	case -EIO: /* shmemfs failure from swap device */
+	case -EFAULT: /* purged object */
+	case -ENODEV: /* bad object, how did you get here! */
+		return VM_FAULT_SIGBUS;
+
+	case -ENOSPC: /* shmemfs allocation failure */
+	case -ENOMEM: /* our allocation failure */
+		return VM_FAULT_OOM;
+
+	case 0:
+	case -EAGAIN:
+	case -ERESTARTSYS:
+	case -EINTR:
+	case -EBUSY:
+		/*
+		 * EBUSY is ok: this just means that another thread
+		 * already did the job.
+		 */
+		return VM_FAULT_NOPAGE;
+	}
+}
+
+static vm_fault_t vm_fault_cpu(struct vm_fault *vmf)
+{
+	struct vm_area_struct *area = vmf->vma;
+	struct i915_mmap_offset *mmo = area->vm_private_data;
+	struct drm_i915_gem_object *obj = mmo->obj;
+	unsigned long i, size = area->vm_end - area->vm_start;
+	bool write = area->vm_flags & VM_WRITE;
+	vm_fault_t ret = VM_FAULT_SIGBUS;
+	int err;
+
+	if (!i915_gem_object_has_struct_page(obj))
+		return ret;
+
+	/* Sanity check that we allow writing into this object */
+	if (i915_gem_object_is_readonly(obj) && write)
+		return ret;
+
+	err = i915_gem_object_pin_pages(obj);
+	if (err)
+		return i915_error_to_vmf_fault(err);
+
+	/* PTEs are revoked in obj->ops->put_pages() */
+	for (i = 0; i < size >> PAGE_SHIFT; i++) {
+		struct page *page = i915_gem_object_get_page(obj, i);
+
+		ret = vmf_insert_pfn(area,
+				     (unsigned long)area->vm_start + i * PAGE_SIZE,
+				     page_to_pfn(page));
+		if (ret != VM_FAULT_NOPAGE)
+			break;
+	}
+
+	if (write) {
+		GEM_BUG_ON(!i915_gem_object_has_pinned_pages(obj));
+		obj->cache_dirty = true; /* XXX flush after PAT update? */
+		obj->mm.dirty = true;
+	}
+
+	i915_gem_object_unpin_pages(obj);
+
+	return ret;
+}
+
+static vm_fault_t vm_fault_gtt(struct vm_fault *vmf)
 {
 #define MIN_CHUNK_PAGES (SZ_1M >> PAGE_SHIFT)
 	struct vm_area_struct *area = vmf->vma;
-	struct drm_i915_gem_object *obj = to_intel_bo(area->vm_private_data);
+	struct i915_mmap_offset *mmo = area->vm_private_data;
+	struct drm_i915_gem_object *obj = mmo->obj;
 	struct drm_device *dev = obj->base.dev;
 	struct drm_i915_private *i915 = to_i915(dev);
 	struct intel_runtime_pm *rpm = &i915->runtime_pm;
@@ -312,6 +372,9 @@ vm_fault_t i915_gem_fault(struct vm_fault *vmf)
 		list_add(&obj->userfault_link, &i915->ggtt.userfault_list);
 	mutex_unlock(&i915->ggtt.vm.mutex);
 
+	/* Track the mmo associated with the fenced vma */
+	vma->mmo = mmo;
+
 	if (IS_ACTIVE(CONFIG_DRM_I915_USERFAULT_AUTOSUSPEND))
 		intel_wakeref_auto(&i915->ggtt.userfault_wakeref,
 				   msecs_to_jiffies_timeout(CONFIG_DRM_I915_USERFAULT_AUTOSUSPEND));
@@ -332,67 +395,36 @@ err_rpm:
 	intel_runtime_pm_put(rpm, wakeref);
 	i915_gem_object_unpin_pages(obj);
 err:
-	switch (ret) {
-	default:
-		WARN_ONCE(ret, "unhandled error in %s: %i\n", __func__, ret);
-		/* fallthrough */
-	case -EIO: /* shmemfs failure from swap device */
-	case -EFAULT: /* purged object */
-	case -ENODEV: /* bad object, how did you get here! */
-		return VM_FAULT_SIGBUS;
-
-	case -ENOSPC: /* shmemfs allocation failure */
-	case -ENOMEM: /* our allocation failure */
-		return VM_FAULT_OOM;
-
-	case 0:
-	case -EAGAIN:
-	case -ERESTARTSYS:
-	case -EINTR:
-	case -EBUSY:
-		/*
-		 * EBUSY is ok: this just means that another thread
-		 * already did the job.
-		 */
-		return VM_FAULT_NOPAGE;
-	}
+	return i915_error_to_vmf_fault(ret);
 }
 
-void __i915_gem_object_release_mmap(struct drm_i915_gem_object *obj)
+void __i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj)
 {
 	struct i915_vma *vma;
 
 	GEM_BUG_ON(!obj->userfault_count);
 
-	obj->userfault_count = 0;
-	list_del(&obj->userfault_link);
-	drm_vma_node_unmap(&obj->base.vma_node,
-			   obj->base.dev->anon_inode->i_mapping);
-
 	for_each_ggtt_vma(vma, obj)
-		i915_vma_unset_userfault(vma);
+		i915_vma_revoke_mmap(vma);
+
+	GEM_BUG_ON(obj->userfault_count);
 }
 
-/**
- * i915_gem_object_release_mmap - remove physical page mappings
- * @obj: obj in question
- *
- * Preserve the reservation of the mmapping with the DRM core code, but
- * relinquish ownership of the pages back to the system.
- *
+/*
  * It is vital that we remove the page mapping if we have mapped a tiled
  * object through the GTT and then lose the fence register due to
  * resource pressure. Similarly if the object has been moved out of the
  * aperture, than pages mapped into userspace must be revoked. Removing the
  * mapping will then trigger a page fault on the next user access, allowing
- * fixup by i915_gem_fault().
+ * fixup by vm_fault_gtt().
  */
-void i915_gem_object_release_mmap(struct drm_i915_gem_object *obj)
+static void i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj)
 {
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	intel_wakeref_t wakeref;
 
-	/* Serialisation between user GTT access and our code depends upon
+	/*
+	 * Serialisation between user GTT access and our code depends upon
 	 * revoking the CPU's PTE whilst the mutex is held. The next user
 	 * pagefault then has to wait until we release the mutex.
 	 *
@@ -406,9 +438,10 @@ void i915_gem_object_release_mmap(struct drm_i915_gem_object *obj)
 	if (!obj->userfault_count)
 		goto out;
 
-	__i915_gem_object_release_mmap(obj);
+	__i915_gem_object_release_mmap_gtt(obj);
 
-	/* Ensure that the CPU's PTE are revoked and there are not outstanding
+	/*
+	 * Ensure that the CPU's PTE are revoked and there are not outstanding
 	 * memory transactions from userspace before we return. The TLB
 	 * flushing implied above by changing the PTE above *should* be
 	 * sufficient, an extra barrier here just provides us with a bit
@@ -422,54 +455,149 @@ out:
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
 }
 
-static int create_mmap_offset(struct drm_i915_gem_object *obj)
+void i915_gem_object_release_mmap_offset(struct drm_i915_gem_object *obj)
+{
+	struct i915_mmap_offset *mmo;
+
+	spin_lock(&obj->mmo.lock);
+	list_for_each_entry(mmo, &obj->mmo.offsets, offset) {
+		/*
+		 * vma_node_unmap for GTT mmaps handled already in
+		 * __i915_gem_object_release_mmap_gtt
+		 */
+		if (mmo->mmap_type == I915_MMAP_TYPE_GTT)
+			continue;
+
+		spin_unlock(&obj->mmo.lock);
+		drm_vma_node_unmap(&mmo->vma_node,
+				   obj->base.dev->anon_inode->i_mapping);
+		spin_lock(&obj->mmo.lock);
+	}
+	spin_unlock(&obj->mmo.lock);
+}
+
+/**
+ * i915_gem_object_release_mmap - remove physical page mappings
+ * @obj: obj in question
+ *
+ * Preserve the reservation of the mmapping with the DRM core code, but
+ * relinquish ownership of the pages back to the system.
+ */
+void i915_gem_object_release_mmap(struct drm_i915_gem_object *obj)
+{
+	i915_gem_object_release_mmap_gtt(obj);
+	i915_gem_object_release_mmap_offset(obj);
+}
+
+static struct i915_mmap_offset *
+mmap_offset_attach(struct drm_i915_gem_object *obj,
+		   enum i915_mmap_type mmap_type,
+		   struct drm_file *file)
 {
 	struct drm_i915_private *i915 = to_i915(obj->base.dev);
-	struct intel_gt *gt = &i915->gt;
+	struct i915_mmap_offset *mmo;
 	int err;
 
-	err = drm_gem_create_mmap_offset(&obj->base);
+	mmo = kmalloc(sizeof(*mmo), GFP_KERNEL);
+	if (!mmo)
+		return ERR_PTR(-ENOMEM);
+
+	mmo->obj = obj;
+	mmo->dev = obj->base.dev;
+	mmo->file = file;
+	mmo->mmap_type = mmap_type;
+	drm_vma_node_reset(&mmo->vma_node);
+
+	err = drm_vma_offset_add(mmo->dev->vma_offset_manager, &mmo->vma_node,
+				 obj->base.size / PAGE_SIZE);
 	if (likely(!err))
-		return 0;
+		goto out;
 
 	/* Attempt to reap some mmap space from dead objects */
-	err = intel_gt_retire_requests_timeout(gt, MAX_SCHEDULE_TIMEOUT);
+	err = intel_gt_retire_requests_timeout(&i915->gt, MAX_SCHEDULE_TIMEOUT);
 	if (err)
-		return err;
+		goto err;
 
 	i915_gem_drain_freed_objects(i915);
-	return drm_gem_create_mmap_offset(&obj->base);
+	err = drm_vma_offset_add(mmo->dev->vma_offset_manager, &mmo->vma_node,
+				 obj->base.size / PAGE_SIZE);
+	if (err)
+		goto err;
+
+out:
+	if (file)
+		drm_vma_node_allow(&mmo->vma_node, file);
+
+	spin_lock(&obj->mmo.lock);
+	list_add(&mmo->offset, &obj->mmo.offsets);
+	spin_unlock(&obj->mmo.lock);
+
+	return mmo;
+
+err:
+	kfree(mmo);
+	return ERR_PTR(err);
 }
 
-int
-i915_gem_mmap_gtt(struct drm_file *file,
-		  struct drm_device *dev,
-		  u32 handle,
-		  u64 *offset)
+static int
+__assign_mmap_offset(struct drm_file *file,
+		     u32 handle,
+		     enum i915_mmap_type mmap_type,
+		     u64 *offset)
 {
 	struct drm_i915_gem_object *obj;
-	int ret;
+	struct i915_mmap_offset *mmo;
+	int err;
 
 	obj = i915_gem_object_lookup(file, handle);
 	if (!obj)
 		return -ENOENT;
 
-	if (i915_gem_object_never_bind_ggtt(obj)) {
-		ret = -ENODEV;
+	if (mmap_type == I915_MMAP_TYPE_GTT &&
+	    i915_gem_object_never_bind_ggtt(obj)) {
+		err = -ENODEV;
+		goto out;
+	}
+
+	if (mmap_type != I915_MMAP_TYPE_GTT &&
+	    !i915_gem_object_has_struct_page(obj)) {
+		err = -ENODEV;
 		goto out;
 	}
 
-	ret = create_mmap_offset(obj);
-	if (ret == 0)
-		*offset = drm_vma_node_offset_addr(&obj->base.vma_node);
+	mmo = mmap_offset_attach(obj, mmap_type, file);
+	if (IS_ERR(mmo)) {
+		err = PTR_ERR(mmo);
+		goto out;
+	}
 
+	*offset = drm_vma_node_offset_addr(&mmo->vma_node);
+	err = 0;
 out:
 	i915_gem_object_put(obj);
-	return ret;
+	return err;
+}
+
+int
+i915_gem_dumb_mmap_offset(struct drm_file *file,
+			  struct drm_device *dev,
+			  u32 handle,
+			  u64 *offset)
+{
+	enum i915_mmap_type mmap_type;
+
+	if (boot_cpu_has(X86_FEATURE_PAT))
+		mmap_type = I915_MMAP_TYPE_WC;
+	else if (!i915_ggtt_has_aperture(&to_i915(dev)->ggtt))
+		return -ENODEV;
+	else
+		mmap_type = I915_MMAP_TYPE_GTT;
+
+	return __assign_mmap_offset(file, handle, mmap_type, offset);
 }
 
 /**
- * i915_gem_mmap_gtt_ioctl - prepare an object for GTT mmap'ing
+ * i915_gem_mmap_offset_ioctl - prepare an object for GTT mmap'ing
  * @dev: DRM device
  * @data: GTT mapping ioctl data
  * @file: GEM object info
@@ -484,12 +612,179 @@ out:
  * userspace.
  */
 int
-i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data,
-			struct drm_file *file)
+i915_gem_mmap_offset_ioctl(struct drm_device *dev, void *data,
+			   struct drm_file *file)
 {
-	struct drm_i915_gem_mmap_gtt *args = data;
+	struct drm_i915_private *i915 = to_i915(dev);
+	struct drm_i915_gem_mmap_offset *args = data;
+	enum i915_mmap_type type;
+	int err;
 
-	return i915_gem_mmap_gtt(file, dev, args->handle, &args->offset);
+	/*
+	 * Historically we failed to check args.pad and args.offset
+	 * and so we cannot use those fields for user input and we cannot
+	 * add -EINVAL for them as the ABI is fixed, i.e. old userspace
+	 * may be feeding in garbage in those fields.
+	 *
+	 * if (args->pad) return -EINVAL; is verbotten!
+	 */
+
+	err = i915_user_extensions(u64_to_user_ptr(args->extensions),
+				   NULL, 0, NULL);
+	if (err)
+		return err;
+
+	switch (args->flags) {
+	case I915_MMAP_OFFSET_GTT:
+		if (!i915_ggtt_has_aperture(&i915->ggtt))
+			return -ENODEV;
+		type = I915_MMAP_TYPE_GTT;
+		break;
+
+	case I915_MMAP_OFFSET_WC:
+		if (!boot_cpu_has(X86_FEATURE_PAT))
+			return -ENODEV;
+		type = I915_MMAP_TYPE_WC;
+		break;
+
+	case I915_MMAP_OFFSET_WB:
+		type = I915_MMAP_TYPE_WB;
+		break;
+
+	case I915_MMAP_OFFSET_UC:
+		if (!boot_cpu_has(X86_FEATURE_PAT))
+			return -ENODEV;
+		type = I915_MMAP_TYPE_UC;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return __assign_mmap_offset(file, args->handle, type, &args->offset);
+}
+
+static void vm_open(struct vm_area_struct *vma)
+{
+	struct i915_mmap_offset *mmo = vma->vm_private_data;
+	struct drm_i915_gem_object *obj = mmo->obj;
+
+	GEM_BUG_ON(!obj);
+	i915_gem_object_get(obj);
+}
+
+static void vm_close(struct vm_area_struct *vma)
+{
+	struct i915_mmap_offset *mmo = vma->vm_private_data;
+	struct drm_i915_gem_object *obj = mmo->obj;
+
+	GEM_BUG_ON(!obj);
+	i915_gem_object_put(obj);
+}
+
+static const struct vm_operations_struct vm_ops_gtt = {
+	.fault = vm_fault_gtt,
+	.open = vm_open,
+	.close = vm_close,
+};
+
+static const struct vm_operations_struct vm_ops_cpu = {
+	.fault = vm_fault_cpu,
+	.open = vm_open,
+	.close = vm_close,
+};
+
+/*
+ * This overcomes the limitation in drm_gem_mmap's assignment of a
+ * drm_gem_object as the vma->vm_private_data. Since we need to
+ * be able to resolve multiple mmap offsets which could be tied
+ * to a single gem object.
+ */
+int i915_gem_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	struct drm_vma_offset_node *node;
+	struct drm_file *priv = filp->private_data;
+	struct drm_device *dev = priv->minor->dev;
+	struct i915_mmap_offset *mmo = NULL;
+	struct drm_gem_object *obj = NULL;
+
+	if (drm_dev_is_unplugged(dev))
+		return -ENODEV;
+
+	drm_vma_offset_lock_lookup(dev->vma_offset_manager);
+	node = drm_vma_offset_exact_lookup_locked(dev->vma_offset_manager,
+						  vma->vm_pgoff,
+						  vma_pages(vma));
+	if (likely(node)) {
+		mmo = container_of(node, struct i915_mmap_offset,
+				   vma_node);
+		/*
+		 * In our dependency chain, the drm_vma_offset_node
+		 * depends on the validity of the mmo, which depends on
+		 * the gem object. However the only reference we have
+		 * at this point is the mmo (as the parent of the node).
+		 * Try to check if the gem object was at least cleared.
+		 */
+		if (!mmo || !mmo->obj) {
+			drm_vma_offset_unlock_lookup(dev->vma_offset_manager);
+			return -EINVAL;
+		}
+		/*
+		 * Skip 0-refcnted objects as it is in the process of being
+		 * destroyed and will be invalid when the vma manager lock
+		 * is released.
+		 */
+		obj = &mmo->obj->base;
+		if (!kref_get_unless_zero(&obj->refcount))
+			obj = NULL;
+	}
+	drm_vma_offset_unlock_lookup(dev->vma_offset_manager);
+	if (!obj)
+		return -EINVAL;
+
+	if (!drm_vma_node_is_allowed(node, priv)) {
+		drm_gem_object_put_unlocked(obj);
+		return -EACCES;
+	}
+
+	if (i915_gem_object_is_readonly(to_intel_bo(obj))) {
+		if (vma->vm_flags & VM_WRITE) {
+			drm_gem_object_put_unlocked(obj);
+			return -EINVAL;
+		}
+		vma->vm_flags &= ~VM_MAYWRITE;
+	}
+
+	vma->vm_flags |= VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP;
+	vma->vm_private_data = mmo;
+
+	switch (mmo->mmap_type) {
+	case I915_MMAP_TYPE_WC:
+		vma->vm_page_prot =
+			pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+		vma->vm_ops = &vm_ops_cpu;
+		break;
+
+	case I915_MMAP_TYPE_WB:
+		vma->vm_page_prot = vm_get_page_prot(vma->vm_flags);
+		vma->vm_ops = &vm_ops_cpu;
+		break;
+
+	case I915_MMAP_TYPE_UC:
+		vma->vm_page_prot =
+			pgprot_noncached(vm_get_page_prot(vma->vm_flags));
+		vma->vm_ops = &vm_ops_cpu;
+		break;
+
+	case I915_MMAP_TYPE_GTT:
+		vma->vm_page_prot =
+			pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+		vma->vm_ops = &vm_ops_gtt;
+		break;
+	}
+	vma->vm_page_prot = pgprot_decrypted(vma->vm_page_prot);
+
+	return 0;
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_mman.h b/drivers/gpu/drm/i915/gem/i915_gem_mman.h
new file mode 100644
index 000000000000..862e01b7cb69
--- /dev/null
+++ b/drivers/gpu/drm/i915/gem/i915_gem_mman.h
@@ -0,0 +1,31 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef __I915_GEM_MMAN_H__
+#define __I915_GEM_MMAN_H__
+
+#include <linux/mm_types.h>
+#include <linux/types.h>
+
+struct drm_device;
+struct drm_file;
+struct drm_i915_gem_object;
+struct file;
+struct i915_mmap_offset;
+struct mutex;
+
+int i915_gem_mmap_gtt_version(void);
+int i915_gem_mmap(struct file *filp, struct vm_area_struct *vma);
+
+int i915_gem_dumb_mmap_offset(struct drm_file *file_priv,
+			      struct drm_device *dev,
+			      u32 handle, u64 *offset);
+
+void __i915_gem_object_release_mmap_gtt(struct drm_i915_gem_object *obj);
+void i915_gem_object_release_mmap(struct drm_i915_gem_object *obj);
+void i915_gem_object_release_mmap_offset(struct drm_i915_gem_object *obj);
+
+#endif
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c
index a50296cce0d8..46bacc82ddc4 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c
@@ -22,11 +22,14 @@
  *
  */
 
+#include <linux/sched/mm.h>
+
 #include "display/intel_frontbuffer.h"
 #include "gt/intel_gt.h"
 #include "i915_drv.h"
 #include "i915_gem_clflush.h"
 #include "i915_gem_context.h"
+#include "i915_gem_mman.h"
 #include "i915_gem_object.h"
 #include "i915_globals.h"
 #include "i915_trace.h"
@@ -59,6 +62,9 @@ void i915_gem_object_init(struct drm_i915_gem_object *obj,
 
 	INIT_LIST_HEAD(&obj->lut_list);
 
+	spin_lock_init(&obj->mmo.lock);
+	INIT_LIST_HEAD(&obj->mmo.offsets);
+
 	init_rcu_head(&obj->rcu);
 
 	obj->ops = ops;
@@ -95,6 +101,7 @@ void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file)
 	struct drm_i915_gem_object *obj = to_intel_bo(gem);
 	struct drm_i915_file_private *fpriv = file->driver_priv;
 	struct i915_lut_handle *lut, *ln;
+	struct i915_mmap_offset *mmo;
 	LIST_HEAD(close);
 
 	i915_gem_object_lock(obj);
@@ -109,6 +116,17 @@ void i915_gem_close_object(struct drm_gem_object *gem, struct drm_file *file)
 	}
 	i915_gem_object_unlock(obj);
 
+	spin_lock(&obj->mmo.lock);
+	list_for_each_entry(mmo, &obj->mmo.offsets, offset) {
+		if (mmo->file != file)
+			continue;
+
+		spin_unlock(&obj->mmo.lock);
+		drm_vma_node_revoke(&mmo->vma_node, file);
+		spin_lock(&obj->mmo.lock);
+	}
+	spin_unlock(&obj->mmo.lock);
+
 	list_for_each_entry_safe(lut, ln, &close, obj_link) {
 		struct i915_gem_context *ctx = lut->ctx;
 		struct i915_vma *vma;
@@ -156,6 +174,8 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 
 	wakeref = intel_runtime_pm_get(&i915->runtime_pm);
 	llist_for_each_entry_safe(obj, on, freed, freed) {
+		struct i915_mmap_offset *mmo, *mn;
+
 		trace_i915_gem_object_destroy(obj);
 
 		if (!list_empty(&obj->vma.list)) {
@@ -174,19 +194,28 @@ static void __i915_gem_free_objects(struct drm_i915_private *i915,
 				GEM_BUG_ON(vma->obj != obj);
 				spin_unlock(&obj->vma.lock);
 
-				i915_vma_destroy(vma);
+				__i915_vma_put(vma);
 
 				spin_lock(&obj->vma.lock);
 			}
 			spin_unlock(&obj->vma.lock);
 		}
 
+		i915_gem_object_release_mmap(obj);
+
+		list_for_each_entry_safe(mmo, mn, &obj->mmo.offsets, offset) {
+			drm_vma_offset_remove(obj->base.dev->vma_offset_manager,
+					      &mmo->vma_node);
+			kfree(mmo);
+		}
+		INIT_LIST_HEAD(&obj->mmo.offsets);
+
 		GEM_BUG_ON(atomic_read(&obj->bind_count));
 		GEM_BUG_ON(obj->userfault_count);
 		GEM_BUG_ON(!list_empty(&obj->lut_list));
 
 		atomic_set(&obj->mm.pages_pin_count, 0);
-		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+		__i915_gem_object_put_pages(obj);
 		GEM_BUG_ON(i915_gem_object_has_pages(obj));
 		bitmap_free(obj->bit_17);
 
@@ -277,18 +306,14 @@ i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj,
 
 	switch (obj->write_domain) {
 	case I915_GEM_DOMAIN_GTT:
-		for_each_ggtt_vma(vma, obj)
-			intel_gt_flush_ggtt_writes(vma->vm->gt);
-
-		intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU);
-
+		spin_lock(&obj->vma.lock);
 		for_each_ggtt_vma(vma, obj) {
-			if (vma->iomap)
-				continue;
-
-			i915_vma_unset_ggtt_write(vma);
+			if (i915_vma_unset_ggtt_write(vma))
+				intel_gt_flush_ggtt_writes(vma->vm->gt);
 		}
+		spin_unlock(&obj->vma.lock);
 
+		i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
 		break;
 
 	case I915_GEM_DOMAIN_WC:
@@ -308,6 +333,30 @@ i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj,
 	obj->write_domain = 0;
 }
 
+void __i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj,
+					 enum fb_op_origin origin)
+{
+	struct intel_frontbuffer *front;
+
+	front = __intel_frontbuffer_get(obj);
+	if (front) {
+		intel_frontbuffer_flush(front, origin);
+		intel_frontbuffer_put(front);
+	}
+}
+
+void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj,
+					      enum fb_op_origin origin)
+{
+	struct intel_frontbuffer *front;
+
+	front = __intel_frontbuffer_get(obj);
+	if (front) {
+		intel_frontbuffer_invalidate(front, origin);
+		intel_frontbuffer_put(front);
+	}
+}
+
 void i915_gem_init__objects(struct drm_i915_private *i915)
 {
 	INIT_WORK(&i915->mm.free_work, __i915_gem_free_work);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h
index 458cd51331f1..858f8bf49a04 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h
@@ -13,8 +13,8 @@
 
 #include <drm/i915_drm.h>
 
+#include "display/intel_frontbuffer.h"
 #include "i915_gem_object_types.h"
-
 #include "i915_gem_gtt.h"
 
 void i915_gem_init__objects(struct drm_i915_private *i915);
@@ -132,13 +132,13 @@ void i915_gem_object_unlock_fence(struct drm_i915_gem_object *obj,
 static inline void
 i915_gem_object_set_readonly(struct drm_i915_gem_object *obj)
 {
-	obj->base.vma_node.readonly = true;
+	obj->flags |= I915_BO_READONLY;
 }
 
 static inline bool
 i915_gem_object_is_readonly(const struct drm_i915_gem_object *obj)
 {
-	return obj->base.vma_node.readonly;
+	return obj->flags & I915_BO_READONLY;
 }
 
 static inline bool
@@ -271,10 +271,27 @@ void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 int ____i915_gem_object_get_pages(struct drm_i915_gem_object *obj);
 int __i915_gem_object_get_pages(struct drm_i915_gem_object *obj);
 
+enum i915_mm_subclass { /* lockdep subclass for obj->mm.lock/struct_mutex */
+	I915_MM_NORMAL = 0,
+	/*
+	 * Only used by struct_mutex, when called "recursively" from
+	 * direct-reclaim-esque. Safe because there is only every one
+	 * struct_mutex in the entire system.
+	 */
+	I915_MM_SHRINKER = 1,
+	/*
+	 * Used for obj->mm.lock when allocating pages. Safe because the object
+	 * isn't yet on any LRU, and therefore the shrinker can't deadlock on
+	 * it. As soon as the object has pages, obj->mm.lock nests within
+	 * fs_reclaim.
+	 */
+	I915_MM_GET_PAGES = 1,
+};
+
 static inline int __must_check
 i915_gem_object_pin_pages(struct drm_i915_gem_object *obj)
 {
-	might_lock(&obj->mm.lock);
+	might_lock_nested(&obj->mm.lock, I915_MM_GET_PAGES);
 
 	if (atomic_inc_not_zero(&obj->mm.pages_pin_count))
 		return 0;
@@ -317,13 +334,7 @@ i915_gem_object_unpin_pages(struct drm_i915_gem_object *obj)
 	__i915_gem_object_unpin_pages(obj);
 }
 
-enum i915_mm_subclass { /* lockdep subclass for obj->mm.lock/struct_mutex */
-	I915_MM_NORMAL = 0,
-	I915_MM_SHRINKER /* called "recursively" from direct-reclaim-esque */
-};
-
-int __i915_gem_object_put_pages(struct drm_i915_gem_object *obj,
-				enum i915_mm_subclass subclass);
+int __i915_gem_object_put_pages(struct drm_i915_gem_object *obj);
 void i915_gem_object_truncate(struct drm_i915_gem_object *obj);
 void i915_gem_object_writeback(struct drm_i915_gem_object *obj);
 
@@ -376,9 +387,6 @@ static inline void i915_gem_object_unpin_map(struct drm_i915_gem_object *obj)
 	i915_gem_object_unpin_pages(obj);
 }
 
-void __i915_gem_object_release_mmap(struct drm_i915_gem_object *obj);
-void i915_gem_object_release_mmap(struct drm_i915_gem_object *obj);
-
 void
 i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj,
 				   unsigned int flush_domains);
@@ -463,4 +471,25 @@ int i915_gem_object_wait_priority(struct drm_i915_gem_object *obj,
 				  unsigned int flags,
 				  const struct i915_sched_attr *attr);
 
+void __i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj,
+					 enum fb_op_origin origin);
+void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj,
+					      enum fb_op_origin origin);
+
+static inline void
+i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj,
+				  enum fb_op_origin origin)
+{
+	if (unlikely(rcu_access_pointer(obj->frontbuffer)))
+		__i915_gem_object_flush_frontbuffer(obj, origin);
+}
+
+static inline void
+i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj,
+				       enum fb_op_origin origin)
+{
+	if (unlikely(rcu_access_pointer(obj->frontbuffer)))
+		__i915_gem_object_invalidate_frontbuffer(obj, origin);
+}
+
 #endif
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
index 96008374a412..88e268633fdc 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
@@ -63,6 +63,23 @@ struct drm_i915_gem_object_ops {
 	void (*release)(struct drm_i915_gem_object *obj);
 };
 
+enum i915_mmap_type {
+	I915_MMAP_TYPE_GTT = 0,
+	I915_MMAP_TYPE_WC,
+	I915_MMAP_TYPE_WB,
+	I915_MMAP_TYPE_UC,
+};
+
+struct i915_mmap_offset {
+	struct drm_device *dev;
+	struct drm_vma_offset_node vma_node;
+	struct drm_i915_gem_object *obj;
+	struct drm_file *file;
+	enum i915_mmap_type mmap_type;
+
+	struct list_head offset;
+};
+
 struct drm_i915_gem_object {
 	struct drm_gem_object base;
 
@@ -118,12 +135,18 @@ struct drm_i915_gem_object {
 	unsigned int userfault_count;
 	struct list_head userfault_link;
 
+	struct {
+		spinlock_t lock; /* Protects access to mmo offsets */
+		struct list_head offsets;
+	} mmo;
+
 	I915_SELFTEST_DECLARE(struct list_head st_link);
 
 	unsigned long flags;
 #define I915_BO_ALLOC_CONTIGUOUS BIT(0)
 #define I915_BO_ALLOC_VOLATILE   BIT(1)
 #define I915_BO_ALLOC_FLAGS (I915_BO_ALLOC_CONTIGUOUS | I915_BO_ALLOC_VOLATILE)
+#define I915_BO_READONLY         BIT(2)
 
 	/*
 	 * Is the object to be mapped as read-only to the GPU
@@ -150,7 +173,7 @@ struct drm_i915_gem_object {
 	 */
 	u16 write_domain;
 
-	struct intel_frontbuffer *frontbuffer;
+	struct intel_frontbuffer __rcu *frontbuffer;
 
 	/** Current tiling stride for the object, if it's tiled. */
 	unsigned int tiling_and_stride;
@@ -162,7 +185,11 @@ struct drm_i915_gem_object {
 	atomic_t bind_count;
 
 	struct {
-		struct mutex lock; /* protects the pages and their use */
+		/*
+		 * Protects the pages and their use. Do not use directly, but
+		 * instead go through the pin/unpin interfaces.
+		 */
+		struct mutex lock;
 		atomic_t pages_pin_count;
 		atomic_t shrink_pin;
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pages.c b/drivers/gpu/drm/i915/gem/i915_gem_pages.c
index 29f4c2850745..75197ca696a8 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pages.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pages.c
@@ -8,6 +8,7 @@
 #include "i915_gem_object.h"
 #include "i915_scatterlist.h"
 #include "i915_gem_lmem.h"
+#include "i915_gem_mman.h"
 
 void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 				 struct sg_table *pages,
@@ -106,7 +107,7 @@ int __i915_gem_object_get_pages(struct drm_i915_gem_object *obj)
 {
 	int err;
 
-	err = mutex_lock_interruptible(&obj->mm.lock);
+	err = mutex_lock_interruptible_nested(&obj->mm.lock, I915_MM_GET_PAGES);
 	if (err)
 		return err;
 
@@ -190,8 +191,7 @@ __i915_gem_object_unset_pages(struct drm_i915_gem_object *obj)
 	return pages;
 }
 
-int __i915_gem_object_put_pages(struct drm_i915_gem_object *obj,
-				enum i915_mm_subclass subclass)
+int __i915_gem_object_put_pages(struct drm_i915_gem_object *obj)
 {
 	struct sg_table *pages;
 	int err;
@@ -202,12 +202,14 @@ int __i915_gem_object_put_pages(struct drm_i915_gem_object *obj,
 	GEM_BUG_ON(atomic_read(&obj->bind_count));
 
 	/* May be called by shrinker from within get_pages() (on another bo) */
-	mutex_lock_nested(&obj->mm.lock, subclass);
+	mutex_lock(&obj->mm.lock);
 	if (unlikely(atomic_read(&obj->mm.pages_pin_count))) {
 		err = -EBUSY;
 		goto unlock;
 	}
 
+	i915_gem_object_release_mmap_offset(obj);
+
 	/*
 	 * ->put_pages might need to allocate memory for the bit17 swizzle
 	 * array, hence protect them from being reaped by removing them from gtt
@@ -308,7 +310,7 @@ void *i915_gem_object_pin_map(struct drm_i915_gem_object *obj,
 	if (!i915_gem_object_type_has(obj, flags))
 		return ERR_PTR(-ENXIO);
 
-	err = mutex_lock_interruptible(&obj->mm.lock);
+	err = mutex_lock_interruptible_nested(&obj->mm.lock, I915_MM_GET_PAGES);
 	if (err)
 		return ERR_PTR(err);
 
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_phys.c b/drivers/gpu/drm/i915/gem/i915_gem_phys.c
index 8043ff63d73f..b1b7c1b3038a 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_phys.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_phys.c
@@ -164,7 +164,7 @@ int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align)
 	if (err)
 		return err;
 
-	mutex_lock(&obj->mm.lock);
+	mutex_lock_nested(&obj->mm.lock, I915_MM_GET_PAGES);
 
 	if (obj->mm.madv != I915_MADV_WILLNEED) {
 		err = -EFAULT;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pm.c b/drivers/gpu/drm/i915/gem/i915_gem_pm.c
index f88ee1317bb4..c8264eb036bf 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pm.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pm.c
@@ -13,7 +13,7 @@
 
 void i915_gem_suspend(struct drm_i915_private *i915)
 {
-	GEM_TRACE("\n");
+	GEM_TRACE("%s\n", dev_name(i915->drm.dev));
 
 	intel_wakeref_auto(&i915->ggtt.userfault_wakeref, 0);
 	flush_workqueue(i915->wq);
@@ -99,30 +99,12 @@ void i915_gem_suspend_late(struct drm_i915_private *i915)
 
 void i915_gem_resume(struct drm_i915_private *i915)
 {
-	GEM_TRACE("\n");
-
-	intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
-
-	if (intel_gt_init_hw(&i915->gt))
-		goto err_wedged;
+	GEM_TRACE("%s\n", dev_name(i915->drm.dev));
 
 	/*
 	 * As we didn't flush the kernel context before suspend, we cannot
 	 * guarantee that the context image is complete. So let's just reset
 	 * it and start again.
 	 */
-	if (intel_gt_resume(&i915->gt))
-		goto err_wedged;
-
-out_unlock:
-	intel_uncore_forcewake_put(&i915->uncore, FORCEWAKE_ALL);
-	return;
-
-err_wedged:
-	if (!intel_gt_is_wedged(&i915->gt)) {
-		dev_err(i915->drm.dev,
-			"Failed to re-initialize GPU, declaring it wedged!\n");
-		intel_gt_set_wedged(&i915->gt);
-	}
-	goto out_unlock;
+	intel_gt_resume(&i915->gt);
 }
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_region.c b/drivers/gpu/drm/i915/gem/i915_gem_region.c
index 2f7bcfb9c964..d50adac12249 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_region.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_region.c
@@ -85,7 +85,7 @@ i915_gem_object_get_pages_buddy(struct drm_i915_gem_object *obj)
 		}
 
 		prev_end = offset + block_size;
-	};
+	}
 
 	sg_page_sizes |= sg->length;
 	sg_mark_end(sg);
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
index f2418a1cfe68..f7e4b39c734f 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_shrinker.c
@@ -57,7 +57,7 @@ static bool unsafe_drop_pages(struct drm_i915_gem_object *obj,
 		flags = I915_GEM_OBJECT_UNBIND_ACTIVE;
 
 	if (i915_gem_object_unbind(obj, flags) == 0)
-		__i915_gem_object_put_pages(obj, I915_MM_SHRINKER);
+		__i915_gem_object_put_pages(obj);
 
 	return !i915_gem_object_has_pages(obj);
 }
@@ -209,8 +209,7 @@ i915_gem_shrink(struct drm_i915_private *i915,
 
 			if (unsafe_drop_pages(obj, shrink)) {
 				/* May arrive from get_pages on another bo */
-				mutex_lock_nested(&obj->mm.lock,
-						  I915_MM_SHRINKER);
+				mutex_lock(&obj->mm.lock);
 				if (!i915_gem_object_has_pages(obj)) {
 					try_to_writeback(obj, shrink);
 					count += obj->base.size >> PAGE_SHIFT;
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_stolen.c b/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
index a2d49c04e6a4..afb08a1704a2 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_stolen.c
@@ -26,48 +26,49 @@
  * for is a boon.
  */
 
-int i915_gem_stolen_insert_node_in_range(struct drm_i915_private *dev_priv,
+int i915_gem_stolen_insert_node_in_range(struct drm_i915_private *i915,
 					 struct drm_mm_node *node, u64 size,
 					 unsigned alignment, u64 start, u64 end)
 {
 	int ret;
 
-	if (!drm_mm_initialized(&dev_priv->mm.stolen))
+	if (!drm_mm_initialized(&i915->mm.stolen))
 		return -ENODEV;
 
 	/* WaSkipStolenMemoryFirstPage:bdw+ */
-	if (INTEL_GEN(dev_priv) >= 8 && start < 4096)
+	if (INTEL_GEN(i915) >= 8 && start < 4096)
 		start = 4096;
 
-	mutex_lock(&dev_priv->mm.stolen_lock);
-	ret = drm_mm_insert_node_in_range(&dev_priv->mm.stolen, node,
+	mutex_lock(&i915->mm.stolen_lock);
+	ret = drm_mm_insert_node_in_range(&i915->mm.stolen, node,
 					  size, alignment, 0,
 					  start, end, DRM_MM_INSERT_BEST);
-	mutex_unlock(&dev_priv->mm.stolen_lock);
+	mutex_unlock(&i915->mm.stolen_lock);
 
 	return ret;
 }
 
-int i915_gem_stolen_insert_node(struct drm_i915_private *dev_priv,
+int i915_gem_stolen_insert_node(struct drm_i915_private *i915,
 				struct drm_mm_node *node, u64 size,
 				unsigned alignment)
 {
-	return i915_gem_stolen_insert_node_in_range(dev_priv, node, size,
+	return i915_gem_stolen_insert_node_in_range(i915, node, size,
 						    alignment, 0, U64_MAX);
 }
 
-void i915_gem_stolen_remove_node(struct drm_i915_private *dev_priv,
+void i915_gem_stolen_remove_node(struct drm_i915_private *i915,
 				 struct drm_mm_node *node)
 {
-	mutex_lock(&dev_priv->mm.stolen_lock);
+	mutex_lock(&i915->mm.stolen_lock);
 	drm_mm_remove_node(node);
-	mutex_unlock(&dev_priv->mm.stolen_lock);
+	mutex_unlock(&i915->mm.stolen_lock);
 }
 
-static int i915_adjust_stolen(struct drm_i915_private *dev_priv,
+static int i915_adjust_stolen(struct drm_i915_private *i915,
 			      struct resource *dsm)
 {
-	struct i915_ggtt *ggtt = &dev_priv->ggtt;
+	struct i915_ggtt *ggtt = &i915->ggtt;
+	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
 	struct resource *r;
 
 	if (dsm->start == 0 || dsm->end <= dsm->start)
@@ -79,14 +80,14 @@ static int i915_adjust_stolen(struct drm_i915_private *dev_priv,
 	 */
 
 	/* Make sure we don't clobber the GTT if it's within stolen memory */
-	if (INTEL_GEN(dev_priv) <= 4 &&
-	    !IS_G33(dev_priv) && !IS_PINEVIEW(dev_priv) && !IS_G4X(dev_priv)) {
+	if (INTEL_GEN(i915) <= 4 &&
+	    !IS_G33(i915) && !IS_PINEVIEW(i915) && !IS_G4X(i915)) {
 		struct resource stolen[2] = {*dsm, *dsm};
 		struct resource ggtt_res;
 		resource_size_t ggtt_start;
 
-		ggtt_start = I915_READ(PGTBL_CTL);
-		if (IS_GEN(dev_priv, 4))
+		ggtt_start = intel_uncore_read(uncore, PGTBL_CTL);
+		if (IS_GEN(i915, 4))
 			ggtt_start = (ggtt_start & PGTBL_ADDRESS_LO_MASK) |
 				     (ggtt_start & PGTBL_ADDRESS_HI_MASK) << 28;
 		else
@@ -120,7 +121,7 @@ static int i915_adjust_stolen(struct drm_i915_private *dev_priv,
 	 * kernel. So if the region is already marked as busy, something
 	 * is seriously wrong.
 	 */
-	r = devm_request_mem_region(dev_priv->drm.dev, dsm->start,
+	r = devm_request_mem_region(i915->drm.dev, dsm->start,
 				    resource_size(dsm),
 				    "Graphics Stolen Memory");
 	if (r == NULL) {
@@ -133,14 +134,14 @@ static int i915_adjust_stolen(struct drm_i915_private *dev_priv,
 		 * reservation starting from 1 instead of 0.
 		 * There's also BIOS with off-by-one on the other end.
 		 */
-		r = devm_request_mem_region(dev_priv->drm.dev, dsm->start + 1,
+		r = devm_request_mem_region(i915->drm.dev, dsm->start + 1,
 					    resource_size(dsm) - 2,
 					    "Graphics Stolen Memory");
 		/*
 		 * GEN3 firmware likes to smash pci bridges into the stolen
 		 * range. Apparently this works.
 		 */
-		if (r == NULL && !IS_GEN(dev_priv, 3)) {
+		if (!r && !IS_GEN(i915, 3)) {
 			DRM_ERROR("conflict detected with stolen region: %pR\n",
 				  dsm);
 
@@ -151,25 +152,27 @@ static int i915_adjust_stolen(struct drm_i915_private *dev_priv,
 	return 0;
 }
 
-static void i915_gem_cleanup_stolen(struct drm_i915_private *dev_priv)
+static void i915_gem_cleanup_stolen(struct drm_i915_private *i915)
 {
-	if (!drm_mm_initialized(&dev_priv->mm.stolen))
+	if (!drm_mm_initialized(&i915->mm.stolen))
 		return;
 
-	drm_mm_takedown(&dev_priv->mm.stolen);
+	drm_mm_takedown(&i915->mm.stolen);
 }
 
-static void g4x_get_stolen_reserved(struct drm_i915_private *dev_priv,
+static void g4x_get_stolen_reserved(struct drm_i915_private *i915,
+				    struct intel_uncore *uncore,
 				    resource_size_t *base,
 				    resource_size_t *size)
 {
-	u32 reg_val = I915_READ(IS_GM45(dev_priv) ?
-				CTG_STOLEN_RESERVED :
-				ELK_STOLEN_RESERVED);
-	resource_size_t stolen_top = dev_priv->dsm.end + 1;
+	u32 reg_val = intel_uncore_read(uncore,
+					IS_GM45(i915) ?
+					CTG_STOLEN_RESERVED :
+					ELK_STOLEN_RESERVED);
+	resource_size_t stolen_top = i915->dsm.end + 1;
 
 	DRM_DEBUG_DRIVER("%s_STOLEN_RESERVED = %08x\n",
-			 IS_GM45(dev_priv) ? "CTG" : "ELK", reg_val);
+			 IS_GM45(i915) ? "CTG" : "ELK", reg_val);
 
 	if ((reg_val & G4X_STOLEN_RESERVED_ENABLE) == 0)
 		return;
@@ -178,7 +181,7 @@ static void g4x_get_stolen_reserved(struct drm_i915_private *dev_priv,
 	 * Whether ILK really reuses the ELK register for this is unclear.
 	 * Let's see if we catch anyone with this supposedly enabled on ILK.
 	 */
-	WARN(IS_GEN(dev_priv, 5), "ILK stolen reserved found? 0x%08x\n",
+	WARN(IS_GEN(i915, 5), "ILK stolen reserved found? 0x%08x\n",
 	     reg_val);
 
 	if (!(reg_val & G4X_STOLEN_RESERVED_ADDR2_MASK))
@@ -190,11 +193,12 @@ static void g4x_get_stolen_reserved(struct drm_i915_private *dev_priv,
 	*size = stolen_top - *base;
 }
 
-static void gen6_get_stolen_reserved(struct drm_i915_private *dev_priv,
+static void gen6_get_stolen_reserved(struct drm_i915_private *i915,
+				     struct intel_uncore *uncore,
 				     resource_size_t *base,
 				     resource_size_t *size)
 {
-	u32 reg_val = I915_READ(GEN6_STOLEN_RESERVED);
+	u32 reg_val = intel_uncore_read(uncore, GEN6_STOLEN_RESERVED);
 
 	DRM_DEBUG_DRIVER("GEN6_STOLEN_RESERVED = %08x\n", reg_val);
 
@@ -222,12 +226,13 @@ static void gen6_get_stolen_reserved(struct drm_i915_private *dev_priv,
 	}
 }
 
-static void vlv_get_stolen_reserved(struct drm_i915_private *dev_priv,
+static void vlv_get_stolen_reserved(struct drm_i915_private *i915,
+				    struct intel_uncore *uncore,
 				    resource_size_t *base,
 				    resource_size_t *size)
 {
-	u32 reg_val = I915_READ(GEN6_STOLEN_RESERVED);
-	resource_size_t stolen_top = dev_priv->dsm.end + 1;
+	u32 reg_val = intel_uncore_read(uncore, GEN6_STOLEN_RESERVED);
+	resource_size_t stolen_top = i915->dsm.end + 1;
 
 	DRM_DEBUG_DRIVER("GEN6_STOLEN_RESERVED = %08x\n", reg_val);
 
@@ -250,11 +255,12 @@ static void vlv_get_stolen_reserved(struct drm_i915_private *dev_priv,
 	*base = stolen_top - *size;
 }
 
-static void gen7_get_stolen_reserved(struct drm_i915_private *dev_priv,
+static void gen7_get_stolen_reserved(struct drm_i915_private *i915,
+				     struct intel_uncore *uncore,
 				     resource_size_t *base,
 				     resource_size_t *size)
 {
-	u32 reg_val = I915_READ(GEN6_STOLEN_RESERVED);
+	u32 reg_val = intel_uncore_read(uncore, GEN6_STOLEN_RESERVED);
 
 	DRM_DEBUG_DRIVER("GEN6_STOLEN_RESERVED = %08x\n", reg_val);
 
@@ -276,11 +282,12 @@ static void gen7_get_stolen_reserved(struct drm_i915_private *dev_priv,
 	}
 }
 
-static void chv_get_stolen_reserved(struct drm_i915_private *dev_priv,
+static void chv_get_stolen_reserved(struct drm_i915_private *i915,
+				    struct intel_uncore *uncore,
 				    resource_size_t *base,
 				    resource_size_t *size)
 {
-	u32 reg_val = I915_READ(GEN6_STOLEN_RESERVED);
+	u32 reg_val = intel_uncore_read(uncore, GEN6_STOLEN_RESERVED);
 
 	DRM_DEBUG_DRIVER("GEN6_STOLEN_RESERVED = %08x\n", reg_val);
 
@@ -308,12 +315,13 @@ static void chv_get_stolen_reserved(struct drm_i915_private *dev_priv,
 	}
 }
 
-static void bdw_get_stolen_reserved(struct drm_i915_private *dev_priv,
+static void bdw_get_stolen_reserved(struct drm_i915_private *i915,
+				    struct intel_uncore *uncore,
 				    resource_size_t *base,
 				    resource_size_t *size)
 {
-	u32 reg_val = I915_READ(GEN6_STOLEN_RESERVED);
-	resource_size_t stolen_top = dev_priv->dsm.end + 1;
+	u32 reg_val = intel_uncore_read(uncore, GEN6_STOLEN_RESERVED);
+	resource_size_t stolen_top = i915->dsm.end + 1;
 
 	DRM_DEBUG_DRIVER("GEN6_STOLEN_RESERVED = %08x\n", reg_val);
 
@@ -328,10 +336,11 @@ static void bdw_get_stolen_reserved(struct drm_i915_private *dev_priv,
 }
 
 static void icl_get_stolen_reserved(struct drm_i915_private *i915,
+				    struct intel_uncore *uncore,
 				    resource_size_t *base,
 				    resource_size_t *size)
 {
-	u64 reg_val = intel_uncore_read64(&i915->uncore, GEN6_STOLEN_RESERVED);
+	u64 reg_val = intel_uncore_read64(uncore, GEN6_STOLEN_RESERVED);
 
 	DRM_DEBUG_DRIVER("GEN6_STOLEN_RESERVED = 0x%016llx\n", reg_val);
 
@@ -356,22 +365,23 @@ static void icl_get_stolen_reserved(struct drm_i915_private *i915,
 	}
 }
 
-static int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
+static int i915_gem_init_stolen(struct drm_i915_private *i915)
 {
+	struct intel_uncore *uncore = &i915->uncore;
 	resource_size_t reserved_base, stolen_top;
 	resource_size_t reserved_total, reserved_size;
 
-	mutex_init(&dev_priv->mm.stolen_lock);
+	mutex_init(&i915->mm.stolen_lock);
 
-	if (intel_vgpu_active(dev_priv)) {
-		dev_notice(dev_priv->drm.dev,
+	if (intel_vgpu_active(i915)) {
+		dev_notice(i915->drm.dev,
 			   "%s, disabling use of stolen memory\n",
 			   "iGVT-g active");
 		return 0;
 	}
 
-	if (intel_vtd_active() && INTEL_GEN(dev_priv) < 8) {
-		dev_notice(dev_priv->drm.dev,
+	if (intel_vtd_active() && INTEL_GEN(i915) < 8) {
+		dev_notice(i915->drm.dev,
 			   "%s, disabling use of stolen memory\n",
 			   "DMAR active");
 		return 0;
@@ -380,58 +390,59 @@ static int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 	if (resource_size(&intel_graphics_stolen_res) == 0)
 		return 0;
 
-	dev_priv->dsm = intel_graphics_stolen_res;
+	i915->dsm = intel_graphics_stolen_res;
 
-	if (i915_adjust_stolen(dev_priv, &dev_priv->dsm))
+	if (i915_adjust_stolen(i915, &i915->dsm))
 		return 0;
 
-	GEM_BUG_ON(dev_priv->dsm.start == 0);
-	GEM_BUG_ON(dev_priv->dsm.end <= dev_priv->dsm.start);
+	GEM_BUG_ON(i915->dsm.start == 0);
+	GEM_BUG_ON(i915->dsm.end <= i915->dsm.start);
 
-	stolen_top = dev_priv->dsm.end + 1;
+	stolen_top = i915->dsm.end + 1;
 	reserved_base = stolen_top;
 	reserved_size = 0;
 
-	switch (INTEL_GEN(dev_priv)) {
+	switch (INTEL_GEN(i915)) {
 	case 2:
 	case 3:
 		break;
 	case 4:
-		if (!IS_G4X(dev_priv))
+		if (!IS_G4X(i915))
 			break;
 		/* fall through */
 	case 5:
-		g4x_get_stolen_reserved(dev_priv,
+		g4x_get_stolen_reserved(i915, uncore,
 					&reserved_base, &reserved_size);
 		break;
 	case 6:
-		gen6_get_stolen_reserved(dev_priv,
+		gen6_get_stolen_reserved(i915, uncore,
 					 &reserved_base, &reserved_size);
 		break;
 	case 7:
-		if (IS_VALLEYVIEW(dev_priv))
-			vlv_get_stolen_reserved(dev_priv,
+		if (IS_VALLEYVIEW(i915))
+			vlv_get_stolen_reserved(i915, uncore,
 						&reserved_base, &reserved_size);
 		else
-			gen7_get_stolen_reserved(dev_priv,
+			gen7_get_stolen_reserved(i915, uncore,
 						 &reserved_base, &reserved_size);
 		break;
 	case 8:
 	case 9:
 	case 10:
-		if (IS_LP(dev_priv))
-			chv_get_stolen_reserved(dev_priv,
+		if (IS_LP(i915))
+			chv_get_stolen_reserved(i915, uncore,
 						&reserved_base, &reserved_size);
 		else
-			bdw_get_stolen_reserved(dev_priv,
+			bdw_get_stolen_reserved(i915, uncore,
 						&reserved_base, &reserved_size);
 		break;
 	default:
-		MISSING_CASE(INTEL_GEN(dev_priv));
+		MISSING_CASE(INTEL_GEN(i915));
 		/* fall-through */
 	case 11:
 	case 12:
-		icl_get_stolen_reserved(dev_priv, &reserved_base,
+		icl_get_stolen_reserved(i915, uncore,
+					&reserved_base,
 					&reserved_size);
 		break;
 	}
@@ -448,12 +459,12 @@ static int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 		reserved_size = 0;
 	}
 
-	dev_priv->dsm_reserved =
-		(struct resource) DEFINE_RES_MEM(reserved_base, reserved_size);
+	i915->dsm_reserved =
+		(struct resource)DEFINE_RES_MEM(reserved_base, reserved_size);
 
-	if (!resource_contains(&dev_priv->dsm, &dev_priv->dsm_reserved)) {
+	if (!resource_contains(&i915->dsm, &i915->dsm_reserved)) {
 		DRM_ERROR("Stolen reserved area %pR outside stolen memory %pR\n",
-			  &dev_priv->dsm_reserved, &dev_priv->dsm);
+			  &i915->dsm_reserved, &i915->dsm);
 		return 0;
 	}
 
@@ -462,14 +473,14 @@ static int i915_gem_init_stolen(struct drm_i915_private *dev_priv)
 	reserved_total = stolen_top - reserved_base;
 
 	DRM_DEBUG_DRIVER("Memory reserved for graphics device: %lluK, usable: %lluK\n",
-			 (u64)resource_size(&dev_priv->dsm) >> 10,
-			 ((u64)resource_size(&dev_priv->dsm) - reserved_total) >> 10);
+			 (u64)resource_size(&i915->dsm) >> 10,
+			 ((u64)resource_size(&i915->dsm) - reserved_total) >> 10);
 
-	dev_priv->stolen_usable_size =
-		resource_size(&dev_priv->dsm) - reserved_total;
+	i915->stolen_usable_size =
+		resource_size(&i915->dsm) - reserved_total;
 
 	/* Basic memrange allocator for stolen space. */
-	drm_mm_init(&dev_priv->mm.stolen, 0, dev_priv->stolen_usable_size);
+	drm_mm_init(&i915->mm.stolen, 0, i915->stolen_usable_size);
 
 	return 0;
 }
@@ -478,11 +489,11 @@ static struct sg_table *
 i915_pages_create_for_stolen(struct drm_device *dev,
 			     resource_size_t offset, resource_size_t size)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
+	struct drm_i915_private *i915 = to_i915(dev);
 	struct sg_table *st;
 	struct scatterlist *sg;
 
-	GEM_BUG_ON(range_overflows(offset, size, resource_size(&dev_priv->dsm)));
+	GEM_BUG_ON(range_overflows(offset, size, resource_size(&i915->dsm)));
 
 	/* We hide that we have no struct page backing our stolen object
 	 * by wrapping the contiguous physical allocation with a fake
@@ -502,7 +513,7 @@ i915_pages_create_for_stolen(struct drm_device *dev,
 	sg->offset = 0;
 	sg->length = size;
 
-	sg_dma_address(sg) = (dma_addr_t)dev_priv->dsm.start + offset;
+	sg_dma_address(sg) = (dma_addr_t)i915->dsm.start + offset;
 	sg_dma_len(sg) = size;
 
 	return st;
@@ -533,16 +544,15 @@ static void i915_gem_object_put_pages_stolen(struct drm_i915_gem_object *obj,
 static void
 i915_gem_object_release_stolen(struct drm_i915_gem_object *obj)
 {
-	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+	struct drm_i915_private *i915 = to_i915(obj->base.dev);
 	struct drm_mm_node *stolen = fetch_and_zero(&obj->stolen);
 
 	GEM_BUG_ON(!stolen);
 
-	i915_gem_stolen_remove_node(dev_priv, stolen);
-	kfree(stolen);
+	i915_gem_object_release_memory_region(obj);
 
-	if (obj->mm.region)
-		i915_gem_object_release_memory_region(obj);
+	i915_gem_stolen_remove_node(i915, stolen);
+	kfree(stolen);
 }
 
 static const struct drm_i915_gem_object_ops i915_gem_object_stolen_ops = {
@@ -552,9 +562,8 @@ static const struct drm_i915_gem_object_ops i915_gem_object_stolen_ops = {
 };
 
 static struct drm_i915_gem_object *
-__i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
-				struct drm_mm_node *stolen,
-				struct intel_memory_region *mem)
+__i915_gem_object_create_stolen(struct intel_memory_region *mem,
+				struct drm_mm_node *stolen)
 {
 	static struct lock_class_key lock_class;
 	struct drm_i915_gem_object *obj;
@@ -565,20 +574,19 @@ __i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
 	if (!obj)
 		goto err;
 
-	drm_gem_private_object_init(&dev_priv->drm, &obj->base, stolen->size);
+	drm_gem_private_object_init(&mem->i915->drm, &obj->base, stolen->size);
 	i915_gem_object_init(obj, &i915_gem_object_stolen_ops, &lock_class);
 
 	obj->stolen = stolen;
 	obj->read_domains = I915_GEM_DOMAIN_CPU | I915_GEM_DOMAIN_GTT;
-	cache_level = HAS_LLC(dev_priv) ? I915_CACHE_LLC : I915_CACHE_NONE;
+	cache_level = HAS_LLC(mem->i915) ? I915_CACHE_LLC : I915_CACHE_NONE;
 	i915_gem_object_set_cache_coherency(obj, cache_level);
 
 	err = i915_gem_object_pin_pages(obj);
 	if (err)
 		goto cleanup;
 
-	if (mem)
-		i915_gem_object_init_memory_region(obj, mem, 0);
+	i915_gem_object_init_memory_region(obj, mem, 0);
 
 	return obj;
 
@@ -593,12 +601,12 @@ _i915_gem_object_create_stolen(struct intel_memory_region *mem,
 			       resource_size_t size,
 			       unsigned int flags)
 {
-	struct drm_i915_private *dev_priv = mem->i915;
+	struct drm_i915_private *i915 = mem->i915;
 	struct drm_i915_gem_object *obj;
 	struct drm_mm_node *stolen;
 	int ret;
 
-	if (!drm_mm_initialized(&dev_priv->mm.stolen))
+	if (!drm_mm_initialized(&i915->mm.stolen))
 		return ERR_PTR(-ENODEV);
 
 	if (size == 0)
@@ -608,30 +616,30 @@ _i915_gem_object_create_stolen(struct intel_memory_region *mem,
 	if (!stolen)
 		return ERR_PTR(-ENOMEM);
 
-	ret = i915_gem_stolen_insert_node(dev_priv, stolen, size, 4096);
+	ret = i915_gem_stolen_insert_node(i915, stolen, size, 4096);
 	if (ret) {
 		obj = ERR_PTR(ret);
 		goto err_free;
 	}
 
-	obj = __i915_gem_object_create_stolen(dev_priv, stolen, mem);
+	obj = __i915_gem_object_create_stolen(mem, stolen);
 	if (IS_ERR(obj))
 		goto err_remove;
 
 	return obj;
 
 err_remove:
-	i915_gem_stolen_remove_node(dev_priv, stolen);
+	i915_gem_stolen_remove_node(i915, stolen);
 err_free:
 	kfree(stolen);
 	return obj;
 }
 
 struct drm_i915_gem_object *
-i915_gem_object_create_stolen(struct drm_i915_private *dev_priv,
+i915_gem_object_create_stolen(struct drm_i915_private *i915,
 			      resource_size_t size)
 {
-	return i915_gem_object_create_region(dev_priv->mm.regions[INTEL_REGION_STOLEN],
+	return i915_gem_object_create_region(i915->mm.regions[INTEL_REGION_STOLEN],
 					     size, I915_BO_ALLOC_CONTIGUOUS);
 }
 
@@ -665,18 +673,19 @@ struct intel_memory_region *i915_gem_stolen_setup(struct drm_i915_private *i915)
 }
 
 struct drm_i915_gem_object *
-i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv,
+i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *i915,
 					       resource_size_t stolen_offset,
 					       resource_size_t gtt_offset,
 					       resource_size_t size)
 {
-	struct i915_ggtt *ggtt = &dev_priv->ggtt;
+	struct intel_memory_region *mem = i915->mm.regions[INTEL_REGION_STOLEN];
+	struct i915_ggtt *ggtt = &i915->ggtt;
 	struct drm_i915_gem_object *obj;
 	struct drm_mm_node *stolen;
 	struct i915_vma *vma;
 	int ret;
 
-	if (!drm_mm_initialized(&dev_priv->mm.stolen))
+	if (!drm_mm_initialized(&i915->mm.stolen))
 		return ERR_PTR(-ENODEV);
 
 	DRM_DEBUG_DRIVER("creating preallocated stolen object: stolen_offset=%pa, gtt_offset=%pa, size=%pa\n",
@@ -694,19 +703,19 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_i915_private *dev_priv
 
 	stolen->start = stolen_offset;
 	stolen->size = size;
-	mutex_lock(&dev_priv->mm.stolen_lock);
-	ret = drm_mm_reserve_node(&dev_priv->mm.stolen, stolen);
-	mutex_unlock(&dev_priv->mm.stolen_lock);
+	mutex_lock(&i915->mm.stolen_lock);
+	ret = drm_mm_reserve_node(&i915->mm.stolen, stolen);
+	mutex_unlock(&i915->mm.stolen_lock);
 	if (ret) {
 		DRM_DEBUG_DRIVER("failed to allocate stolen space\n");
 		kfree(stolen);
 		return ERR_PTR(ret);
 	}
 
-	obj = __i915_gem_object_create_stolen(dev_priv, stolen, NULL);
+	obj = __i915_gem_object_create_stolen(mem, stolen);
 	if (IS_ERR(obj)) {
 		DRM_DEBUG_DRIVER("failed to allocate stolen object\n");
-		i915_gem_stolen_remove_node(dev_priv, stolen);
+		i915_gem_stolen_remove_node(i915, stolen);
 		kfree(stolen);
 		return obj;
 	}
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_tiling.c b/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
index 1fa592d82af5..6c7825a2dc2a 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_tiling.c
@@ -11,6 +11,7 @@
 #include "i915_drv.h"
 #include "i915_gem.h"
 #include "i915_gem_ioctls.h"
+#include "i915_gem_mman.h"
 #include "i915_gem_object.h"
 
 /**
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_userptr.c b/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
index 4c72d74d6576..e5558af111e2 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_userptr.c
@@ -129,9 +129,10 @@ userptr_mn_invalidate_range_start(struct mmu_notifier *_mn,
 		spin_unlock(&mn->lock);
 
 		ret = i915_gem_object_unbind(obj,
-					     I915_GEM_OBJECT_UNBIND_ACTIVE);
+					     I915_GEM_OBJECT_UNBIND_ACTIVE |
+					     I915_GEM_OBJECT_UNBIND_BARRIER);
 		if (ret == 0)
-			ret = __i915_gem_object_put_pages(obj, I915_MM_SHRINKER);
+			ret = __i915_gem_object_put_pages(obj);
 		i915_gem_object_put(obj);
 		if (ret)
 			return ret;
@@ -459,31 +460,36 @@ __i915_gem_userptr_get_pages_worker(struct work_struct *_work)
 	if (pvec != NULL) {
 		struct mm_struct *mm = obj->userptr.mm->mm;
 		unsigned int flags = 0;
+		int locked = 0;
 
 		if (!i915_gem_object_is_readonly(obj))
 			flags |= FOLL_WRITE;
 
 		ret = -EFAULT;
 		if (mmget_not_zero(mm)) {
-			down_read(&mm->mmap_sem);
 			while (pinned < npages) {
+				if (!locked) {
+					down_read(&mm->mmap_sem);
+					locked = 1;
+				}
 				ret = get_user_pages_remote
 					(work->task, mm,
 					 obj->userptr.ptr + pinned * PAGE_SIZE,
 					 npages - pinned,
 					 flags,
-					 pvec + pinned, NULL, NULL);
+					 pvec + pinned, NULL, &locked);
 				if (ret < 0)
 					break;
 
 				pinned += ret;
 			}
-			up_read(&mm->mmap_sem);
+			if (locked)
+				up_read(&mm->mmap_sem);
 			mmput(mm);
 		}
 	}
 
-	mutex_lock(&obj->mm.lock);
+	mutex_lock_nested(&obj->mm.lock, I915_MM_GET_PAGES);
 	if (obj->userptr.work == &work->work) {
 		struct sg_table *pages = ERR_PTR(ret);
 
@@ -773,15 +779,11 @@ i915_gem_userptr_ioctl(struct drm_device *dev,
 		return -EFAULT;
 
 	if (args->flags & I915_USERPTR_READ_ONLY) {
-		struct i915_address_space *vm;
-
 		/*
 		 * On almost all of the older hw, we cannot tell the GPU that
 		 * a page is readonly.
 		 */
-		vm = rcu_dereference_protected(dev_priv->kernel_context->vm,
-					       true); /* static vm */
-		if (!vm || !vm->has_read_only)
+		if (!dev_priv->gt.vm->has_read_only)
 			return -ENODEV;
 	}
 
diff --git a/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c b/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c
index 892d12db6c49..fa16f2c3f3ac 100644
--- a/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c
+++ b/drivers/gpu/drm/i915/gem/selftests/huge_gem_object.c
@@ -12,10 +12,14 @@ static void huge_free_pages(struct drm_i915_gem_object *obj,
 			    struct sg_table *pages)
 {
 	unsigned long nreal = obj->scratch / PAGE_SIZE;
-	struct scatterlist *sg;
+	struct sgt_iter sgt_iter;
+	struct page *page;
 
-	for (sg = pages->sgl; sg && nreal--; sg = __sg_next(sg))
-		__free_page(sg_page(sg));
+	for_each_sgt_page(page, sgt_iter, pages) {
+		__free_page(page);
+		if (!--nreal)
+			break;
+	}
 
 	sg_free_table(pages);
 	kfree(pages);
@@ -70,7 +74,6 @@ static int huge_get_pages(struct drm_i915_gem_object *obj)
 
 err:
 	huge_free_pages(obj, pages);
-
 	return -ENOMEM;
 #undef GFP
 }
diff --git a/drivers/gpu/drm/i915/gem/selftests/huge_pages.c b/drivers/gpu/drm/i915/gem/selftests/huge_pages.c
index 688c49a24f32..2479395c1873 100644
--- a/drivers/gpu/drm/i915/gem/selftests/huge_pages.c
+++ b/drivers/gpu/drm/i915/gem/selftests/huge_pages.c
@@ -517,7 +517,7 @@ static int igt_mock_memory_region_huge_pages(void *arg)
 			i915_vma_unpin(vma);
 			i915_vma_close(vma);
 
-			__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+			__i915_gem_object_put_pages(obj);
 			i915_gem_object_put(obj);
 		}
 	}
@@ -650,7 +650,7 @@ static int igt_mock_ppgtt_misaligned_dma(void *arg)
 		i915_vma_close(vma);
 
 		i915_gem_object_unpin_pages(obj);
-		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+		__i915_gem_object_put_pages(obj);
 		i915_gem_object_put(obj);
 	}
 
@@ -678,7 +678,7 @@ static void close_object_list(struct list_head *objects,
 
 		list_del(&obj->st_link);
 		i915_gem_object_unpin_pages(obj);
-		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+		__i915_gem_object_put_pages(obj);
 		i915_gem_object_put(obj);
 	}
 }
@@ -948,7 +948,7 @@ static int igt_mock_ppgtt_64K(void *arg)
 			i915_vma_close(vma);
 
 			i915_gem_object_unpin_pages(obj);
-			__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+			__i915_gem_object_put_pages(obj);
 			i915_gem_object_put(obj);
 		}
 	}
@@ -1110,8 +1110,7 @@ static int __igt_write_huge(struct intel_context *ce,
 out_vma_unpin:
 	i915_vma_unpin(vma);
 out_vma_close:
-	i915_vma_destroy(vma);
-
+	__i915_vma_put(vma);
 	return err;
 }
 
@@ -1301,7 +1300,7 @@ static int igt_ppgtt_exhaust_huge(void *arg)
 			}
 
 			i915_gem_object_unpin_pages(obj);
-			__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+			__i915_gem_object_put_pages(obj);
 			i915_gem_object_put(obj);
 		}
 	}
@@ -1420,7 +1419,7 @@ try_again:
 
 		err = i915_gem_object_pin_pages(obj);
 		if (err) {
-			if (err == -ENXIO) {
+			if (err == -ENXIO || err == -E2BIG) {
 				i915_gem_object_put(obj);
 				size >>= 1;
 				goto try_again;
@@ -1442,7 +1441,7 @@ try_again:
 		}
 out_unpin:
 		i915_gem_object_unpin_pages(obj);
-		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+		__i915_gem_object_put_pages(obj);
 out_put:
 		i915_gem_object_put(obj);
 
@@ -1530,7 +1529,7 @@ static int igt_ppgtt_sanity_check(void *arg)
 			err = igt_write_huge(ctx, obj);
 
 			i915_gem_object_unpin_pages(obj);
-			__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+			__i915_gem_object_put_pages(obj);
 			i915_gem_object_put(obj);
 
 			if (err) {
@@ -1912,9 +1911,9 @@ int i915_gem_huge_page_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(igt_ppgtt_smoke_huge),
 		SUBTEST(igt_ppgtt_sanity_check),
 	};
-	struct drm_file *file;
 	struct i915_gem_context *ctx;
 	struct i915_address_space *vm;
+	struct file *file;
 	int err;
 
 	if (!HAS_PPGTT(i915)) {
@@ -1944,6 +1943,6 @@ int i915_gem_huge_page_live_selftests(struct drm_i915_private *i915)
 	err = i915_subtests(tests, ctx);
 
 out_file:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c
index da8edee4fe0a..b972be165e85 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_client_blt.c
@@ -24,6 +24,7 @@ static int __igt_client_fill(struct intel_engine_cs *engine)
 
 	prandom_seed_state(&prng, i915_selftest.random_seed);
 
+	intel_engine_pm_get(engine);
 	do {
 		const u32 max_block_size = S16_MAX * PAGE_SIZE;
 		u32 sz = min_t(u64, ce->vm->total >> 4, prandom_u32_state(&prng));
@@ -99,6 +100,7 @@ err_put:
 err_flush:
 	if (err == -ENOMEM)
 		err = 0;
+	intel_engine_pm_put(engine);
 
 	return err;
 }
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c
index 2b29f6b4e1dd..49edc51111d5 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_coherency.c
@@ -6,6 +6,7 @@
 
 #include <linux/prime_numbers.h>
 
+#include "gt/intel_engine_pm.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
 #include "gt/intel_ring.h"
@@ -200,7 +201,7 @@ static int gpu_set(struct context *ctx, unsigned long offset, u32 v)
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
-	rq = i915_request_create(ctx->engine->kernel_context);
+	rq = intel_engine_create_kernel_request(ctx->engine);
 	if (IS_ERR(rq)) {
 		i915_vma_unpin(vma);
 		return PTR_ERR(rq);
@@ -326,6 +327,7 @@ static int igt_gem_coherency(void *arg)
 	ctx.engine = random_engine(i915, &prng);
 	GEM_BUG_ON(!ctx.engine);
 	pr_info("%s: using %s\n", __func__, ctx.engine->name);
+	intel_engine_pm_get(ctx.engine);
 
 	for (over = igt_coherency_mode; over->name; over++) {
 		if (!over->set)
@@ -404,6 +406,7 @@ static int igt_gem_coherency(void *arg)
 		}
 	}
 free:
+	intel_engine_pm_put(ctx.engine);
 	kfree(offsets);
 	return err;
 
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
index 62fabc023a83..7fc46861a54d 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
@@ -7,6 +7,7 @@
 #include <linux/prime_numbers.h>
 
 #include "gem/i915_gem_pm.h"
+#include "gt/intel_engine_pm.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_requests.h"
 #include "gt/intel_reset.h"
@@ -26,6 +27,12 @@
 
 #define DW_PER_PAGE (PAGE_SIZE / sizeof(u32))
 
+static inline struct i915_address_space *ctx_vm(struct i915_gem_context *ctx)
+{
+	/* single threaded, private ctx */
+	return rcu_dereference_protected(ctx->vm, true);
+}
+
 static int live_nop_switch(void *arg)
 {
 	const unsigned int nctx = 1024;
@@ -33,7 +40,7 @@ static int live_nop_switch(void *arg)
 	struct intel_engine_cs *engine;
 	struct i915_gem_context **ctx;
 	struct igt_live_test t;
-	struct drm_file *file;
+	struct file *file;
 	unsigned long n;
 	int err = -ENODEV;
 
@@ -67,25 +74,34 @@ static int live_nop_switch(void *arg)
 	}
 
 	for_each_uabi_engine(engine, i915) {
-		struct i915_request *rq;
+		struct i915_request *rq = NULL;
 		unsigned long end_time, prime;
 		ktime_t times[2] = {};
 
 		times[0] = ktime_get_raw();
 		for (n = 0; n < nctx; n++) {
-			rq = igt_request_alloc(ctx[n], engine);
-			if (IS_ERR(rq)) {
-				err = PTR_ERR(rq);
+			struct i915_request *this;
+
+			this = igt_request_alloc(ctx[n], engine);
+			if (IS_ERR(this)) {
+				err = PTR_ERR(this);
 				goto out_file;
 			}
-			i915_request_add(rq);
+			if (rq) {
+				i915_request_await_dma_fence(this, &rq->fence);
+				i915_request_put(rq);
+			}
+			rq = i915_request_get(this);
+			i915_request_add(this);
 		}
 		if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 			pr_err("Failed to populated %d contexts\n", nctx);
 			intel_gt_set_wedged(&i915->gt);
+			i915_request_put(rq);
 			err = -EIO;
 			goto out_file;
 		}
+		i915_request_put(rq);
 
 		times[1] = ktime_get_raw();
 
@@ -100,13 +116,21 @@ static int live_nop_switch(void *arg)
 		for_each_prime_number_from(prime, 2, 8192) {
 			times[1] = ktime_get_raw();
 
+			rq = NULL;
 			for (n = 0; n < prime; n++) {
-				rq = igt_request_alloc(ctx[n % nctx], engine);
-				if (IS_ERR(rq)) {
-					err = PTR_ERR(rq);
+				struct i915_request *this;
+
+				this = igt_request_alloc(ctx[n % nctx], engine);
+				if (IS_ERR(this)) {
+					err = PTR_ERR(this);
 					goto out_file;
 				}
 
+				if (rq) { /* Force submission order */
+					i915_request_await_dma_fence(this, &rq->fence);
+					i915_request_put(rq);
+				}
+
 				/*
 				 * This space is left intentionally blank.
 				 *
@@ -121,14 +145,18 @@ static int live_nop_switch(void *arg)
 				 * for latency.
 				 */
 
-				i915_request_add(rq);
+				rq = i915_request_get(this);
+				i915_request_add(this);
 			}
+			GEM_BUG_ON(!rq);
 			if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 				pr_err("Switching between %ld contexts timed out\n",
 				       prime);
 				intel_gt_set_wedged(&i915->gt);
+				i915_request_put(rq);
 				break;
 			}
+			i915_request_put(rq);
 
 			times[1] = ktime_sub(ktime_get_raw(), times[1]);
 			if (prime == 2)
@@ -149,7 +177,7 @@ static int live_nop_switch(void *arg)
 	}
 
 out_file:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
@@ -255,7 +283,7 @@ static int live_parallel_switch(void *arg)
 	int (* const *fn)(void *arg);
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
-	struct drm_file *file;
+	struct file *file;
 	int n, m, count;
 	int err = 0;
 
@@ -309,7 +337,7 @@ static int live_parallel_switch(void *arg)
 			if (!data[m].ce[0])
 				continue;
 
-			ce = intel_context_create(ctx, data[m].ce[0]->engine);
+			ce = intel_context_create(data[m].ce[0]->engine);
 			if (IS_ERR(ce))
 				goto out;
 
@@ -377,7 +405,7 @@ out:
 	}
 	kfree(data);
 out_file:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
@@ -502,17 +530,17 @@ out_unmap:
 	return err;
 }
 
-static int file_add_object(struct drm_file *file,
-			    struct drm_i915_gem_object *obj)
+static int file_add_object(struct file *file, struct drm_i915_gem_object *obj)
 {
 	int err;
 
 	GEM_BUG_ON(obj->base.handle_count);
 
 	/* tie the object to the drm_file for easy reaping */
-	err = idr_alloc(&file->object_idr, &obj->base, 1, 0, GFP_KERNEL);
+	err = idr_alloc(&to_drm_file(file)->object_idr,
+			&obj->base, 1, 0, GFP_KERNEL);
 	if (err < 0)
-		return  err;
+		return err;
 
 	i915_gem_object_get(obj);
 	obj->base.handle_count++;
@@ -521,7 +549,7 @@ static int file_add_object(struct drm_file *file,
 
 static struct drm_i915_gem_object *
 create_test_object(struct i915_address_space *vm,
-		   struct drm_file *file,
+		   struct file *file,
 		   struct list_head *objects)
 {
 	struct drm_i915_gem_object *obj;
@@ -621,9 +649,9 @@ static int igt_ctx_exec(void *arg)
 		unsigned long ncontexts, ndwords, dw;
 		struct i915_request *tq[5] = {};
 		struct igt_live_test t;
-		struct drm_file *file;
 		IGT_TIMEOUT(end_time);
 		LIST_HEAD(objects);
+		struct file *file;
 
 		if (!intel_engine_can_store_dword(engine))
 			continue;
@@ -716,7 +744,7 @@ out_file:
 		if (igt_live_test_end(&t))
 			err = -EIO;
 
-		mock_file_free(i915, file);
+		fput(file);
 		if (err)
 			return err;
 
@@ -733,7 +761,7 @@ static int igt_shared_ctx_exec(void *arg)
 	struct i915_gem_context *parent;
 	struct intel_engine_cs *engine;
 	struct igt_live_test t;
-	struct drm_file *file;
+	struct file *file;
 	int err = 0;
 
 	/*
@@ -786,14 +814,15 @@ static int igt_shared_ctx_exec(void *arg)
 			}
 
 			mutex_lock(&ctx->mutex);
-			__assign_ppgtt(ctx, parent->vm);
+			__assign_ppgtt(ctx, ctx_vm(parent));
 			mutex_unlock(&ctx->mutex);
 
 			ce = i915_gem_context_get_engine(ctx, engine->legacy_idx);
 			GEM_BUG_ON(IS_ERR(ce));
 
 			if (!obj) {
-				obj = create_test_object(parent->vm, file, &objects);
+				obj = create_test_object(ctx_vm(parent),
+							 file, &objects);
 				if (IS_ERR(obj)) {
 					err = PTR_ERR(obj);
 					intel_context_put(ce);
@@ -854,7 +883,7 @@ out_test:
 	if (igt_live_test_end(&t))
 		err = -EIO;
 out_file:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
@@ -1140,8 +1169,7 @@ out:
 		igt_spinner_end(spin);
 
 	if ((flags & TEST_IDLE) && ret == 0) {
-		ret = intel_gt_wait_for_idle(ce->engine->gt,
-					     MAX_SCHEDULE_TIMEOUT);
+		ret = igt_flush_test(ce->engine->i915);
 		if (ret)
 			return ret;
 
@@ -1163,9 +1191,11 @@ __sseu_test(const char *name,
 	struct igt_spinner *spin = NULL;
 	int ret;
 
+	intel_engine_pm_get(ce->engine);
+
 	ret = __sseu_prepare(name, flags, ce, &spin);
 	if (ret)
-		return ret;
+		goto out_pm;
 
 	ret = intel_context_reconfigure_sseu(ce, sseu);
 	if (ret)
@@ -1180,6 +1210,8 @@ out_spin:
 		igt_spinner_fini(spin);
 		kfree(spin);
 	}
+out_pm:
+	intel_engine_pm_put(ce->engine);
 	return ret;
 }
 
@@ -1232,8 +1264,7 @@ __igt_ctx_sseu(struct drm_i915_private *i915,
 			hweight32(engine->sseu.slice_mask),
 			hweight32(pg_sseu.slice_mask));
 
-		ce = intel_context_create(engine->kernel_context->gem_context,
-					  engine);
+		ce = intel_context_create(engine);
 		if (IS_ERR(ce)) {
 			ret = PTR_ERR(ce);
 			goto out_put;
@@ -1311,16 +1342,18 @@ static int igt_ctx_sseu(void *arg)
 static int igt_ctx_readonly(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
+	unsigned long idx, ndwords, dw, num_engines;
 	struct drm_i915_gem_object *obj = NULL;
 	struct i915_request *tq[5] = {};
+	struct i915_gem_engines_iter it;
 	struct i915_address_space *vm;
 	struct i915_gem_context *ctx;
-	unsigned long idx, ndwords, dw;
+	struct intel_context *ce;
 	struct igt_live_test t;
-	struct drm_file *file;
 	I915_RND_STATE(prng);
 	IGT_TIMEOUT(end_time);
 	LIST_HEAD(objects);
+	struct file *file;
 	int err = -ENODEV;
 
 	/*
@@ -1343,21 +1376,21 @@ static int igt_ctx_readonly(void *arg)
 		goto out_file;
 	}
 
-	rcu_read_lock();
-	vm = rcu_dereference(ctx->vm) ?: &i915->ggtt.alias->vm;
+	vm = ctx_vm(ctx) ?: &i915->ggtt.alias->vm;
 	if (!vm || !vm->has_read_only) {
-		rcu_read_unlock();
 		err = 0;
 		goto out_file;
 	}
-	rcu_read_unlock();
+
+	num_engines = 0;
+	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it)
+		if (intel_engine_can_store_dword(ce->engine))
+			num_engines++;
+	i915_gem_context_unlock_engines(ctx);
 
 	ndwords = 0;
 	dw = 0;
 	while (!time_after(jiffies, end_time)) {
-		struct i915_gem_engines_iter it;
-		struct intel_context *ce;
-
 		for_each_gem_engine(ce,
 				    i915_gem_context_lock_engines(ctx), it) {
 			if (!intel_engine_can_store_dword(ce->engine))
@@ -1380,7 +1413,7 @@ static int igt_ctx_readonly(void *arg)
 				pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
 				       ndwords, dw, max_dwords(obj),
 				       ce->engine->name,
-				       yesno(!!rcu_access_pointer(ctx->vm)),
+				       yesno(!!ctx_vm(ctx)),
 				       err);
 				i915_gem_context_unlock_engines(ctx);
 				goto out_file;
@@ -1400,8 +1433,8 @@ static int igt_ctx_readonly(void *arg)
 		}
 		i915_gem_context_unlock_engines(ctx);
 	}
-	pr_info("Submitted %lu dwords (across %u engines)\n",
-		ndwords, RUNTIME_INFO(i915)->num_engines);
+	pr_info("Submitted %lu dwords (across %lu engines)\n",
+		ndwords, num_engines);
 
 	dw = 0;
 	idx = 0;
@@ -1426,7 +1459,7 @@ out_file:
 	if (igt_live_test_end(&t))
 		err = -EIO;
 
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
@@ -1466,7 +1499,7 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 	cmd = i915_gem_object_pin_map(obj, I915_MAP_WB);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
-		goto err;
+		goto out;
 	}
 
 	*cmd++ = MI_STORE_DWORD_IMM_GEN4;
@@ -1488,12 +1521,12 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 	vma = i915_vma_instance(obj, vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
-		goto err_vm;
+		goto out_vm;
 	}
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_OFFSET_FIXED);
 	if (err)
-		goto err_vm;
+		goto out_vm;
 
 	err = check_scratch(vm, offset);
 	if (err)
@@ -1517,22 +1550,20 @@ static int write_to_scratch(struct i915_gem_context *ctx,
 	if (err)
 		goto skip_request;
 
-	i915_vma_unpin_and_release(&vma, 0);
+	i915_vma_unpin(vma);
 
 	i915_request_add(rq);
 
-	i915_vm_put(vm);
-	return 0;
-
+	goto out_vm;
 skip_request:
 	i915_request_skip(rq, err);
 err_request:
 	i915_request_add(rq);
 err_unpin:
 	i915_vma_unpin(vma);
-err_vm:
+out_vm:
 	i915_vm_put(vm);
-err:
+out:
 	i915_gem_object_put(obj);
 	return err;
 }
@@ -1560,7 +1591,7 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 	cmd = i915_gem_object_pin_map(obj, I915_MAP_WB);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
-		goto err;
+		goto out;
 	}
 
 	memset(cmd, POISON_INUSE, PAGE_SIZE);
@@ -1592,12 +1623,12 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 	vma = i915_vma_instance(obj, vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
-		goto err_vm;
+		goto out_vm;
 	}
 
 	err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_OFFSET_FIXED);
 	if (err)
-		goto err_vm;
+		goto out_vm;
 
 	err = check_scratch(vm, offset);
 	if (err)
@@ -1630,29 +1661,27 @@ static int read_from_scratch(struct i915_gem_context *ctx,
 	err = i915_gem_object_set_to_cpu_domain(obj, false);
 	i915_gem_object_unlock(obj);
 	if (err)
-		goto err_vm;
+		goto out_vm;
 
 	cmd = i915_gem_object_pin_map(obj, I915_MAP_WB);
 	if (IS_ERR(cmd)) {
 		err = PTR_ERR(cmd);
-		goto err_vm;
+		goto out_vm;
 	}
 
 	*value = cmd[result / sizeof(*cmd)];
 	i915_gem_object_unpin_map(obj);
-	i915_gem_object_put(obj);
-
-	return 0;
 
+	goto out_vm;
 skip_request:
 	i915_request_skip(rq, err);
 err_request:
 	i915_request_add(rq);
 err_unpin:
 	i915_vma_unpin(vma);
-err_vm:
+out_vm:
 	i915_vm_put(vm);
-err:
+out:
 	i915_gem_object_put(obj);
 	return err;
 }
@@ -1661,11 +1690,11 @@ static int igt_vm_isolation(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct i915_gem_context *ctx_a, *ctx_b;
+	unsigned long num_engines, count;
 	struct intel_engine_cs *engine;
 	struct igt_live_test t;
-	struct drm_file *file;
 	I915_RND_STATE(prng);
-	unsigned long count;
+	struct file *file;
 	u64 vm_total;
 	int err;
 
@@ -1698,14 +1727,15 @@ static int igt_vm_isolation(void *arg)
 	}
 
 	/* We can only test vm isolation, if the vm are distinct */
-	if (ctx_a->vm == ctx_b->vm)
+	if (ctx_vm(ctx_a) == ctx_vm(ctx_b))
 		goto out_file;
 
-	vm_total = ctx_a->vm->total;
-	GEM_BUG_ON(ctx_b->vm->total != vm_total);
+	vm_total = ctx_vm(ctx_a)->total;
+	GEM_BUG_ON(ctx_vm(ctx_b)->total != vm_total);
 	vm_total -= I915_GTT_PAGE_SIZE;
 
 	count = 0;
+	num_engines = 0;
 	for_each_uabi_engine(engine, i915) {
 		IGT_TIMEOUT(end_time);
 		unsigned long this = 0;
@@ -1743,14 +1773,15 @@ static int igt_vm_isolation(void *arg)
 			this++;
 		}
 		count += this;
+		num_engines++;
 	}
-	pr_info("Checked %lu scratch offsets across %d engines\n",
-		count, RUNTIME_INFO(i915)->num_engines);
+	pr_info("Checked %lu scratch offsets across %lu engines\n",
+		count, num_engines);
 
 out_file:
 	if (igt_live_test_end(&t))
 		err = -EIO;
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c
index 29b2077b73d2..cbf796da64e3 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_mman.c
@@ -6,12 +6,14 @@
 
 #include <linux/prime_numbers.h>
 
+#include "gt/intel_engine_pm.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
 #include "huge_gem_object.h"
 #include "i915_selftest.h"
 #include "selftests/i915_random.h"
 #include "selftests/igt_flush_test.h"
+#include "selftests/igt_mmap.h"
 
 struct tile {
 	unsigned int width;
@@ -161,7 +163,7 @@ static int check_partial_mapping(struct drm_i915_gem_object *obj,
 	kunmap(p);
 
 out:
-	i915_vma_destroy(vma);
+	__i915_vma_put(vma);
 	return err;
 }
 
@@ -255,7 +257,7 @@ static int check_partial_mappings(struct drm_i915_gem_object *obj,
 		if (err)
 			return err;
 
-		i915_vma_destroy(vma);
+		__i915_vma_put(vma);
 
 		if (igt_timeout(end_time,
 				"%s: timed out after tiling=%d stride=%d\n",
@@ -535,7 +537,7 @@ static int make_obj_busy(struct drm_i915_gem_object *obj)
 		if (err)
 			return err;
 
-		rq = i915_request_create(engine->kernel_context);
+		rq = intel_engine_create_kernel_request(engine);
 		if (IS_ERR(rq)) {
 			i915_vma_unpin(vma);
 			return PTR_ERR(rq);
@@ -563,16 +565,16 @@ static bool assert_mmap_offset(struct drm_i915_private *i915,
 			       int expected)
 {
 	struct drm_i915_gem_object *obj;
-	int err;
+	struct i915_mmap_offset *mmo;
 
 	obj = i915_gem_object_create_internal(i915, size);
 	if (IS_ERR(obj))
 		return PTR_ERR(obj);
 
-	err = create_mmap_offset(obj);
+	mmo = mmap_offset_attach(obj, I915_MMAP_OFFSET_GTT, NULL);
 	i915_gem_object_put(obj);
 
-	return err == expected;
+	return PTR_ERR_OR_ZERO(mmo) == expected;
 }
 
 static void disable_retire_worker(struct drm_i915_private *i915)
@@ -606,28 +608,50 @@ static int igt_mmap_offset_exhaustion(void *arg)
 	struct drm_i915_private *i915 = arg;
 	struct drm_mm *mm = &i915->drm.vma_offset_manager->vm_addr_space_mm;
 	struct drm_i915_gem_object *obj;
-	struct drm_mm_node resv, *hole;
-	u64 hole_start, hole_end;
-	int loop, err;
+	struct drm_mm_node *hole, *next;
+	struct i915_mmap_offset *mmo;
+	int loop, err = 0;
 
 	/* Disable background reaper */
 	disable_retire_worker(i915);
 	GEM_BUG_ON(!i915->gt.awake);
+	intel_gt_retire_requests(&i915->gt);
+	i915_gem_drain_freed_objects(i915);
 
 	/* Trim the device mmap space to only a page */
-	memset(&resv, 0, sizeof(resv));
-	drm_mm_for_each_hole(hole, mm, hole_start, hole_end) {
-		resv.start = hole_start;
-		resv.size = hole_end - hole_start - 1; /* PAGE_SIZE units */
-		mmap_offset_lock(i915);
-		err = drm_mm_reserve_node(mm, &resv);
-		mmap_offset_unlock(i915);
+	mmap_offset_lock(i915);
+	loop = 1; /* PAGE_SIZE units */
+	list_for_each_entry_safe(hole, next, &mm->hole_stack, hole_stack) {
+		struct drm_mm_node *resv;
+
+		resv = kzalloc(sizeof(*resv), GFP_NOWAIT);
+		if (!resv) {
+			err = -ENOMEM;
+			goto out_park;
+		}
+
+		resv->start = drm_mm_hole_node_start(hole) + loop;
+		resv->size = hole->hole_size - loop;
+		resv->color = -1ul;
+		loop = 0;
+
+		if (!resv->size) {
+			kfree(resv);
+			continue;
+		}
+
+		pr_debug("Reserving hole [%llx + %llx]\n",
+			 resv->start, resv->size);
+
+		err = drm_mm_reserve_node(mm, resv);
 		if (err) {
 			pr_err("Failed to trim VMA manager, err=%d\n", err);
+			kfree(resv);
 			goto out_park;
 		}
-		break;
 	}
+	GEM_BUG_ON(!list_is_singular(&mm->hole_stack));
+	mmap_offset_unlock(i915);
 
 	/* Just fits! */
 	if (!assert_mmap_offset(i915, PAGE_SIZE, 0)) {
@@ -650,9 +674,10 @@ static int igt_mmap_offset_exhaustion(void *arg)
 		goto out;
 	}
 
-	err = create_mmap_offset(obj);
-	if (err) {
+	mmo = mmap_offset_attach(obj, I915_MMAP_OFFSET_GTT, NULL);
+	if (IS_ERR(mmo)) {
 		pr_err("Unable to insert object into reclaimed hole\n");
+		err = PTR_ERR(mmo);
 		goto err_obj;
 	}
 
@@ -684,9 +709,15 @@ static int igt_mmap_offset_exhaustion(void *arg)
 
 out:
 	mmap_offset_lock(i915);
-	drm_mm_remove_node(&resv);
-	mmap_offset_unlock(i915);
 out_park:
+	drm_mm_for_each_node_safe(hole, next, mm) {
+		if (hole->color != -1ul)
+			continue;
+
+		drm_mm_remove_node(hole);
+		kfree(hole);
+	}
+	mmap_offset_unlock(i915);
 	restore_retire_worker(i915);
 	return err;
 err_obj:
@@ -694,12 +725,258 @@ err_obj:
 	goto out;
 }
 
+#define expand32(x) (((x) << 0) | ((x) << 8) | ((x) << 16) | ((x) << 24))
+static int igt_mmap(void *arg, enum i915_mmap_type type)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	struct i915_mmap_offset *mmo;
+	struct vm_area_struct *area;
+	unsigned long addr;
+	void *vaddr;
+	int err = 0, i;
+
+	if (!i915_ggtt_has_aperture(&i915->ggtt))
+		return 0;
+
+	obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	vaddr = i915_gem_object_pin_map(obj, I915_MAP_WB);
+	if (IS_ERR(vaddr)) {
+		err = PTR_ERR(vaddr);
+		goto out;
+	}
+	memset(vaddr, POISON_INUSE, PAGE_SIZE);
+	i915_gem_object_flush_map(obj);
+	i915_gem_object_unpin_map(obj);
+
+	mmo = mmap_offset_attach(obj, type, NULL);
+	if (IS_ERR(mmo)) {
+		err = PTR_ERR(mmo);
+		goto out;
+	}
+
+	addr = igt_mmap_node(i915, &mmo->vma_node, 0, PROT_WRITE, MAP_SHARED);
+	if (IS_ERR_VALUE(addr)) {
+		err = addr;
+		goto out;
+	}
+
+	pr_debug("igt_mmap() @ %lx\n", addr);
+
+	area = find_vma(current->mm, addr);
+	if (!area) {
+		pr_err("Did not create a vm_area_struct for the mmap\n");
+		err = -EINVAL;
+		goto out_unmap;
+	}
+
+	if (area->vm_private_data != mmo) {
+		pr_err("vm_area_struct did not point back to our mmap_offset object!\n");
+		err = -EINVAL;
+		goto out_unmap;
+	}
+
+	for (i = 0; i < PAGE_SIZE / sizeof(u32); i++) {
+		u32 __user *ux = u64_to_user_ptr((u64)(addr + i * sizeof(*ux)));
+		u32 x;
+
+		if (get_user(x, ux)) {
+			pr_err("Unable to read from mmap, offset:%zd\n",
+			       i * sizeof(x));
+			err = -EFAULT;
+			break;
+		}
+
+		if (x != expand32(POISON_INUSE)) {
+			pr_err("Read incorrect value from mmap, offset:%zd, found:%x, expected:%x\n",
+			       i * sizeof(x), x, expand32(POISON_INUSE));
+			err = -EINVAL;
+			break;
+		}
+
+		x = expand32(POISON_FREE);
+		if (put_user(x, ux)) {
+			pr_err("Unable to write to mmap, offset:%zd\n",
+			       i * sizeof(x));
+			err = -EFAULT;
+			break;
+		}
+	}
+
+out_unmap:
+	vm_munmap(addr, PAGE_SIZE);
+
+	vaddr = i915_gem_object_pin_map(obj, I915_MAP_FORCE_WC);
+	if (IS_ERR(vaddr)) {
+		err = PTR_ERR(vaddr);
+		goto out;
+	}
+	if (err == 0 && memchr_inv(vaddr, POISON_FREE, PAGE_SIZE)) {
+		pr_err("Write via mmap did not land in backing store\n");
+		err = -EINVAL;
+	}
+	i915_gem_object_unpin_map(obj);
+
+out:
+	i915_gem_object_put(obj);
+	return err;
+}
+
+static int igt_mmap_gtt(void *arg)
+{
+	return igt_mmap(arg, I915_MMAP_TYPE_GTT);
+}
+
+static int igt_mmap_cpu(void *arg)
+{
+	return igt_mmap(arg, I915_MMAP_TYPE_WC);
+}
+
+static int check_present_pte(pte_t *pte, unsigned long addr, void *data)
+{
+	if (!pte_present(*pte) || pte_none(*pte)) {
+		pr_err("missing PTE:%lx\n",
+		       (addr - (unsigned long)data) >> PAGE_SHIFT);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int check_absent_pte(pte_t *pte, unsigned long addr, void *data)
+{
+	if (pte_present(*pte) && !pte_none(*pte)) {
+		pr_err("present PTE:%lx; expected to be revoked\n",
+		       (addr - (unsigned long)data) >> PAGE_SHIFT);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int check_present(unsigned long addr, unsigned long len)
+{
+	return apply_to_page_range(current->mm, addr, len,
+				   check_present_pte, (void *)addr);
+}
+
+static int check_absent(unsigned long addr, unsigned long len)
+{
+	return apply_to_page_range(current->mm, addr, len,
+				   check_absent_pte, (void *)addr);
+}
+
+static int prefault_range(u64 start, u64 len)
+{
+	const char __user *addr, *end;
+	char __maybe_unused c;
+	int err;
+
+	addr = u64_to_user_ptr(start);
+	end = addr + len;
+
+	for (; addr < end; addr += PAGE_SIZE) {
+		err = __get_user(c, addr);
+		if (err)
+			return err;
+	}
+
+	return __get_user(c, end - 1);
+}
+
+static int igt_mmap_revoke(void *arg, enum i915_mmap_type type)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	struct i915_mmap_offset *mmo;
+	unsigned long addr;
+	int err;
+
+	if (!i915_ggtt_has_aperture(&i915->ggtt))
+		return 0;
+
+	obj = i915_gem_object_create_internal(i915, SZ_4M);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
+
+	mmo = mmap_offset_attach(obj, type, NULL);
+	if (IS_ERR(mmo)) {
+		err = PTR_ERR(mmo);
+		goto out;
+	}
+
+	addr = igt_mmap_node(i915, &mmo->vma_node, 0, PROT_WRITE, MAP_SHARED);
+	if (IS_ERR_VALUE(addr)) {
+		err = addr;
+		goto out;
+	}
+
+	err = prefault_range(addr, obj->base.size);
+	if (err)
+		goto out_unmap;
+
+	GEM_BUG_ON(mmo->mmap_type == I915_MMAP_TYPE_GTT &&
+		   !atomic_read(&obj->bind_count));
+
+	err = check_present(addr, obj->base.size);
+	if (err)
+		goto out_unmap;
+
+	/*
+	 * After unbinding the object from the GGTT, its address may be reused
+	 * for other objects. Ergo we have to revoke the previous mmap PTE
+	 * access as it no longer points to the same object.
+	 */
+	err = i915_gem_object_unbind(obj, I915_GEM_OBJECT_UNBIND_ACTIVE);
+	if (err) {
+		pr_err("Failed to unbind object!\n");
+		goto out_unmap;
+	}
+	GEM_BUG_ON(atomic_read(&obj->bind_count));
+
+	if (type != I915_MMAP_TYPE_GTT) {
+		__i915_gem_object_put_pages(obj);
+		if (i915_gem_object_has_pages(obj)) {
+			pr_err("Failed to put-pages object!\n");
+			err = -EINVAL;
+			goto out_unmap;
+		}
+	}
+
+	err = check_absent(addr, obj->base.size);
+	if (err)
+		goto out_unmap;
+
+out_unmap:
+	vm_munmap(addr, obj->base.size);
+out:
+	i915_gem_object_put(obj);
+	return err;
+}
+
+static int igt_mmap_gtt_revoke(void *arg)
+{
+	return igt_mmap_revoke(arg, I915_MMAP_TYPE_GTT);
+}
+
+static int igt_mmap_cpu_revoke(void *arg)
+{
+	return igt_mmap_revoke(arg, I915_MMAP_TYPE_WC);
+}
+
 int i915_gem_mman_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
 		SUBTEST(igt_partial_tiling),
 		SUBTEST(igt_smoke_tiling),
 		SUBTEST(igt_mmap_offset_exhaustion),
+		SUBTEST(igt_mmap_gtt),
+		SUBTEST(igt_mmap_cpu),
+		SUBTEST(igt_mmap_gtt_revoke),
+		SUBTEST(igt_mmap_cpu_revoke),
 	};
 
 	return i915_subtests(tests, i915);
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c
index e8132aca0bb6..62077fe46715 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_object_blt.c
@@ -41,6 +41,7 @@ static int __perf_fill_blt(struct drm_i915_gem_object *obj)
 		if (!engine)
 			return 0;
 
+		intel_engine_pm_get(engine);
 		for (pass = 0; pass < ARRAY_SIZE(t); pass++) {
 			struct intel_context *ce = engine->kernel_context;
 			ktime_t t0, t1;
@@ -49,17 +50,20 @@ static int __perf_fill_blt(struct drm_i915_gem_object *obj)
 
 			err = i915_gem_object_fill_blt(obj, ce, 0);
 			if (err)
-				return err;
+				break;
 
 			err = i915_gem_object_wait(obj,
 						   I915_WAIT_ALL,
 						   MAX_SCHEDULE_TIMEOUT);
 			if (err)
-				return err;
+				break;
 
 			t1 = ktime_get();
 			t[pass] = ktime_sub(t1, t0);
 		}
+		intel_engine_pm_put(engine);
+		if (err)
+			return err;
 
 		sort(t, ARRAY_SIZE(t), sizeof(*t), wrap_ktime_compare, NULL);
 		pr_info("%s: blt %zd KiB fill: %lld MiB/s\n",
@@ -109,6 +113,7 @@ static int __perf_copy_blt(struct drm_i915_gem_object *src,
 		struct intel_engine_cs *engine;
 		ktime_t t[5];
 		int pass;
+		int err = 0;
 
 		engine = intel_engine_lookup_user(i915,
 						  I915_ENGINE_CLASS_COPY,
@@ -116,26 +121,29 @@ static int __perf_copy_blt(struct drm_i915_gem_object *src,
 		if (!engine)
 			return 0;
 
+		intel_engine_pm_get(engine);
 		for (pass = 0; pass < ARRAY_SIZE(t); pass++) {
 			struct intel_context *ce = engine->kernel_context;
 			ktime_t t0, t1;
-			int err;
 
 			t0 = ktime_get();
 
 			err = i915_gem_object_copy_blt(src, dst, ce);
 			if (err)
-				return err;
+				break;
 
 			err = i915_gem_object_wait(dst,
 						   I915_WAIT_ALL,
 						   MAX_SCHEDULE_TIMEOUT);
 			if (err)
-				return err;
+				break;
 
 			t1 = ktime_get();
 			t[pass] = ktime_sub(t1, t0);
 		}
+		intel_engine_pm_put(engine);
+		if (err)
+			return err;
 
 		sort(t, ARRAY_SIZE(t), sizeof(*t), wrap_ktime_compare, NULL);
 		pr_info("%s: blt %zd KiB copy: %lld MiB/s\n",
@@ -186,6 +194,8 @@ err_src:
 
 struct igt_thread_arg {
 	struct drm_i915_private *i915;
+	struct i915_gem_context *ctx;
+	struct file *file;
 	struct rnd_state prng;
 	unsigned int n_cpus;
 };
@@ -198,24 +208,20 @@ static int igt_fill_blt_thread(void *arg)
 	struct drm_i915_gem_object *obj;
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
-	struct drm_file *file;
 	unsigned int prio;
 	IGT_TIMEOUT(end);
 	int err;
 
-	file = mock_file(i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
+	ctx = thread->ctx;
+	if (!ctx) {
+		ctx = live_context(i915, thread->file);
+		if (IS_ERR(ctx))
+			return PTR_ERR(ctx);
 
-	ctx = live_context(i915, file);
-	if (IS_ERR(ctx)) {
-		err = PTR_ERR(ctx);
-		goto out_file;
+		prio = i915_prandom_u32_max_state(I915_PRIORITY_MAX, prng);
+		ctx->sched.priority = I915_USER_PRIORITY(prio);
 	}
 
-	prio = i915_prandom_u32_max_state(I915_PRIORITY_MAX, prng);
-	ctx->sched.priority = I915_USER_PRIORITY(prio);
-
 	ce = i915_gem_context_get_engine(ctx, BCS0);
 	GEM_BUG_ON(IS_ERR(ce));
 
@@ -300,8 +306,6 @@ err_flush:
 		err = 0;
 
 	intel_context_put(ce);
-out_file:
-	mock_file_free(i915, file);
 	return err;
 }
 
@@ -313,24 +317,20 @@ static int igt_copy_blt_thread(void *arg)
 	struct drm_i915_gem_object *src, *dst;
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
-	struct drm_file *file;
 	unsigned int prio;
 	IGT_TIMEOUT(end);
 	int err;
 
-	file = mock_file(i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
+	ctx = thread->ctx;
+	if (!ctx) {
+		ctx = live_context(i915, thread->file);
+		if (IS_ERR(ctx))
+			return PTR_ERR(ctx);
 
-	ctx = live_context(i915, file);
-	if (IS_ERR(ctx)) {
-		err = PTR_ERR(ctx);
-		goto out_file;
+		prio = i915_prandom_u32_max_state(I915_PRIORITY_MAX, prng);
+		ctx->sched.priority = I915_USER_PRIORITY(prio);
 	}
 
-	prio = i915_prandom_u32_max_state(I915_PRIORITY_MAX, prng);
-	ctx->sched.priority = I915_USER_PRIORITY(prio);
-
 	ce = i915_gem_context_get_engine(ctx, BCS0);
 	GEM_BUG_ON(IS_ERR(ce));
 
@@ -431,19 +431,18 @@ err_flush:
 		err = 0;
 
 	intel_context_put(ce);
-out_file:
-	mock_file_free(i915, file);
 	return err;
 }
 
 static int igt_threaded_blt(struct drm_i915_private *i915,
-			    int (*blt_fn)(void *arg))
+			    int (*blt_fn)(void *arg),
+			    unsigned int flags)
+#define SINGLE_CTX BIT(0)
 {
 	struct igt_thread_arg *thread;
 	struct task_struct **tsk;
+	unsigned int n_cpus, i;
 	I915_RND_STATE(prng);
-	unsigned int n_cpus;
-	unsigned int i;
 	int err = 0;
 
 	n_cpus = num_online_cpus() + 1;
@@ -453,13 +452,27 @@ static int igt_threaded_blt(struct drm_i915_private *i915,
 		return 0;
 
 	thread = kcalloc(n_cpus, sizeof(struct igt_thread_arg), GFP_KERNEL);
-	if (!thread) {
-		kfree(tsk);
-		return 0;
+	if (!thread)
+		goto out_tsk;
+
+	thread[0].file = mock_file(i915);
+	if (IS_ERR(thread[0].file)) {
+		err = PTR_ERR(thread[0].file);
+		goto out_thread;
+	}
+
+	if (flags & SINGLE_CTX) {
+		thread[0].ctx = live_context(i915, thread[0].file);
+		if (IS_ERR(thread[0].ctx)) {
+			err = PTR_ERR(thread[0].ctx);
+			goto out_file;
+		}
 	}
 
 	for (i = 0; i < n_cpus; ++i) {
 		thread[i].i915 = i915;
+		thread[i].file = thread[0].file;
+		thread[i].ctx = thread[0].ctx;
 		thread[i].n_cpus = n_cpus;
 		thread[i].prng =
 			I915_RND_STATE_INITIALIZER(prandom_u32_state(&prng));
@@ -488,29 +501,42 @@ static int igt_threaded_blt(struct drm_i915_private *i915,
 		put_task_struct(tsk[i]);
 	}
 
-	kfree(tsk);
+out_file:
+	fput(thread[0].file);
+out_thread:
 	kfree(thread);
-
+out_tsk:
+	kfree(tsk);
 	return err;
 }
 
 static int igt_fill_blt(void *arg)
 {
-	return igt_threaded_blt(arg, igt_fill_blt_thread);
+	return igt_threaded_blt(arg, igt_fill_blt_thread, 0);
+}
+
+static int igt_fill_blt_ctx0(void *arg)
+{
+	return igt_threaded_blt(arg, igt_fill_blt_thread, SINGLE_CTX);
 }
 
 static int igt_copy_blt(void *arg)
 {
-	return igt_threaded_blt(arg, igt_copy_blt_thread);
+	return igt_threaded_blt(arg, igt_copy_blt_thread, 0);
+}
+
+static int igt_copy_blt_ctx0(void *arg)
+{
+	return igt_threaded_blt(arg, igt_copy_blt_thread, SINGLE_CTX);
 }
 
 int i915_gem_object_blt_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
-		SUBTEST(perf_fill_blt),
-		SUBTEST(perf_copy_blt),
 		SUBTEST(igt_fill_blt),
+		SUBTEST(igt_fill_blt_ctx0),
 		SUBTEST(igt_copy_blt),
+		SUBTEST(igt_copy_blt_ctx0),
 	};
 
 	if (intel_gt_is_wedged(&i915->gt))
@@ -521,3 +547,16 @@ int i915_gem_object_blt_live_selftests(struct drm_i915_private *i915)
 
 	return i915_live_subtests(tests, i915);
 }
+
+int i915_gem_object_blt_perf_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(perf_fill_blt),
+		SUBTEST(perf_copy_blt),
+	};
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	return i915_live_subtests(tests, i915);
+}
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_context.c b/drivers/gpu/drm/i915/gem/selftests/mock_context.c
index 29b8984f0e47..7d7e13dc2fdf 100644
--- a/drivers/gpu/drm/i915/gem/selftests/mock_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/mock_context.c
@@ -5,6 +5,7 @@
  */
 
 #include "mock_context.h"
+#include "selftests/mock_drm.h"
 #include "selftests/mock_gtt.h"
 
 struct i915_gem_context *
@@ -36,9 +37,7 @@ mock_context(struct drm_i915_private *i915,
 	if (name) {
 		struct i915_ppgtt *ppgtt;
 
-		ctx->name = kstrdup(name, GFP_KERNEL);
-		if (!ctx->name)
-			goto err_put;
+		strncpy(ctx->name, name, sizeof(ctx->name));
 
 		ppgtt = mock_ppgtt(i915, name);
 		if (!ppgtt)
@@ -74,7 +73,7 @@ void mock_init_contexts(struct drm_i915_private *i915)
 }
 
 struct i915_gem_context *
-live_context(struct drm_i915_private *i915, struct drm_file *file)
+live_context(struct drm_i915_private *i915, struct file *file)
 {
 	struct i915_gem_context *ctx;
 	int err;
@@ -83,7 +82,7 @@ live_context(struct drm_i915_private *i915, struct drm_file *file)
 	if (IS_ERR(ctx))
 		return ctx;
 
-	err = gem_context_register(ctx, file->driver_priv);
+	err = gem_context_register(ctx, to_drm_file(file)->driver_priv);
 	if (err < 0)
 		goto err_ctx;
 
@@ -97,7 +96,16 @@ err_ctx:
 struct i915_gem_context *
 kernel_context(struct drm_i915_private *i915)
 {
-	return i915_gem_context_create_kernel(i915, I915_PRIORITY_NORMAL);
+	struct i915_gem_context *ctx;
+
+	ctx = i915_gem_create_context(i915, 0);
+	if (IS_ERR(ctx))
+		return ctx;
+
+	i915_gem_context_clear_bannable(ctx);
+	i915_gem_context_set_persistence(ctx);
+
+	return ctx;
 }
 
 void kernel_context_close(struct i915_gem_context *ctx)
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_context.h b/drivers/gpu/drm/i915/gem/selftests/mock_context.h
index 0b926653914f..fb83d2f09212 100644
--- a/drivers/gpu/drm/i915/gem/selftests/mock_context.h
+++ b/drivers/gpu/drm/i915/gem/selftests/mock_context.h
@@ -7,6 +7,9 @@
 #ifndef __MOCK_CONTEXT_H
 #define __MOCK_CONTEXT_H
 
+struct file;
+struct drm_i915_private;
+
 void mock_init_contexts(struct drm_i915_private *i915);
 
 struct i915_gem_context *
@@ -16,7 +19,7 @@ mock_context(struct drm_i915_private *i915,
 void mock_context_close(struct i915_gem_context *ctx);
 
 struct i915_gem_context *
-live_context(struct drm_i915_private *i915, struct drm_file *file);
+live_context(struct drm_i915_private *i915, struct file *file);
 
 struct i915_gem_context *kernel_context(struct drm_i915_private *i915);
 void kernel_context_close(struct i915_gem_context *ctx);
diff --git a/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.h b/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.h
index f0f8bbd82dfc..22818bbb139d 100644
--- a/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.h
+++ b/drivers/gpu/drm/i915/gem/selftests/mock_dmabuf.h
@@ -14,7 +14,7 @@ struct mock_dmabuf {
 	struct page *pages[];
 };
 
-static struct mock_dmabuf *to_mock(struct dma_buf *buf)
+static inline struct mock_dmabuf *to_mock(struct dma_buf *buf)
 {
 	return buf->priv;
 }
diff --git a/drivers/gpu/drm/i915/gt/debugfs_engines.c b/drivers/gpu/drm/i915/gt/debugfs_engines.c
new file mode 100644
index 000000000000..6a5e9ab20b94
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/debugfs_engines.c
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: MIT
+
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <drm/drm_print.h>
+
+#include "debugfs_engines.h"
+#include "debugfs_gt.h"
+#include "i915_drv.h" /* for_each_engine! */
+#include "intel_engine.h"
+
+static int engines_show(struct seq_file *m, void *data)
+{
+	struct intel_gt *gt = m->private;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	struct drm_printer p;
+
+	p = drm_seq_file_printer(m);
+	for_each_engine(engine, gt, id)
+		intel_engine_dump(engine, &p, "%s\n", engine->name);
+
+	return 0;
+}
+DEFINE_GT_DEBUGFS_ATTRIBUTE(engines);
+
+void debugfs_engines_register(struct intel_gt *gt, struct dentry *root)
+{
+	static const struct debugfs_gt_file files[] = {
+		{ "engines", &engines_fops },
+	};
+
+	debugfs_gt_register_files(gt, root, files, ARRAY_SIZE(files));
+}
diff --git a/drivers/gpu/drm/i915/gt/debugfs_engines.h b/drivers/gpu/drm/i915/gt/debugfs_engines.h
new file mode 100644
index 000000000000..f69257eaa1cc
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/debugfs_engines.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef DEBUGFS_ENGINES_H
+#define DEBUGFS_ENGINES_H
+
+struct intel_gt;
+struct dentry;
+
+void debugfs_engines_register(struct intel_gt *gt, struct dentry *root);
+
+#endif /* DEBUGFS_ENGINES_H */
diff --git a/drivers/gpu/drm/i915/gt/debugfs_gt.c b/drivers/gpu/drm/i915/gt/debugfs_gt.c
new file mode 100644
index 000000000000..75255aaacaed
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/debugfs_gt.c
@@ -0,0 +1,42 @@
+// SPDX-License-Identifier: MIT
+
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/debugfs.h>
+
+#include "debugfs_engines.h"
+#include "debugfs_gt.h"
+#include "debugfs_gt_pm.h"
+#include "i915_drv.h"
+
+void debugfs_gt_register(struct intel_gt *gt)
+{
+	struct dentry *root;
+
+	if (!gt->i915->drm.primary->debugfs_root)
+		return;
+
+	root = debugfs_create_dir("gt", gt->i915->drm.primary->debugfs_root);
+	if (IS_ERR(root))
+		return;
+
+	debugfs_engines_register(gt, root);
+	debugfs_gt_pm_register(gt, root);
+}
+
+void debugfs_gt_register_files(struct intel_gt *gt,
+			       struct dentry *root,
+			       const struct debugfs_gt_file *files,
+			       unsigned long count)
+{
+	while (count--) {
+		if (!files->eval || files->eval(gt))
+			debugfs_create_file(files->name,
+					    0444, root, gt,
+					    files->fops);
+
+		files++;
+	}
+}
diff --git a/drivers/gpu/drm/i915/gt/debugfs_gt.h b/drivers/gpu/drm/i915/gt/debugfs_gt.h
new file mode 100644
index 000000000000..4ea0f06cda8f
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/debugfs_gt.h
@@ -0,0 +1,39 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef DEBUGFS_GT_H
+#define DEBUGFS_GT_H
+
+#include <linux/file.h>
+
+struct intel_gt;
+
+#define DEFINE_GT_DEBUGFS_ATTRIBUTE(__name)				\
+	static int __name ## _open(struct inode *inode, struct file *file) \
+{									\
+	return single_open(file, __name ## _show, inode->i_private);	\
+}									\
+static const struct file_operations __name ## _fops = {			\
+	.owner = THIS_MODULE,						\
+	.open = __name ## _open,					\
+	.read = seq_read,						\
+	.llseek = seq_lseek,						\
+	.release = single_release,					\
+}
+
+void debugfs_gt_register(struct intel_gt *gt);
+
+struct debugfs_gt_file {
+	const char *name;
+	const struct file_operations *fops;
+	bool (*eval)(const struct intel_gt *gt);
+};
+
+void debugfs_gt_register_files(struct intel_gt *gt,
+			       struct dentry *root,
+			       const struct debugfs_gt_file *files,
+			       unsigned long count);
+
+#endif /* DEBUGFS_GT_H */
diff --git a/drivers/gpu/drm/i915/gt/debugfs_gt_pm.c b/drivers/gpu/drm/i915/gt/debugfs_gt_pm.c
new file mode 100644
index 000000000000..059c9e5c002e
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/debugfs_gt_pm.c
@@ -0,0 +1,601 @@
+// SPDX-License-Identifier: MIT
+
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <linux/seq_file.h>
+
+#include "debugfs_gt.h"
+#include "debugfs_gt_pm.h"
+#include "i915_drv.h"
+#include "intel_gt.h"
+#include "intel_llc.h"
+#include "intel_rc6.h"
+#include "intel_rps.h"
+#include "intel_runtime_pm.h"
+#include "intel_sideband.h"
+#include "intel_uncore.h"
+
+static int fw_domains_show(struct seq_file *m, void *data)
+{
+	struct intel_gt *gt = m->private;
+	struct intel_uncore *uncore = gt->uncore;
+	struct intel_uncore_forcewake_domain *fw_domain;
+	unsigned int tmp;
+
+	seq_printf(m, "user.bypass_count = %u\n",
+		   uncore->user_forcewake_count);
+
+	for_each_fw_domain(fw_domain, uncore, tmp)
+		seq_printf(m, "%s.wake_count = %u\n",
+			   intel_uncore_forcewake_domain_to_str(fw_domain->id),
+			   READ_ONCE(fw_domain->wake_count));
+
+	return 0;
+}
+DEFINE_GT_DEBUGFS_ATTRIBUTE(fw_domains);
+
+static void print_rc6_res(struct seq_file *m,
+			  const char *title,
+			  const i915_reg_t reg)
+{
+	struct intel_gt *gt = m->private;
+	intel_wakeref_t wakeref;
+
+	with_intel_runtime_pm(gt->uncore->rpm, wakeref)
+		seq_printf(m, "%s %u (%llu us)\n", title,
+			   intel_uncore_read(gt->uncore, reg),
+			   intel_rc6_residency_us(&gt->rc6, reg));
+}
+
+static int vlv_drpc(struct seq_file *m)
+{
+	struct intel_gt *gt = m->private;
+	struct intel_uncore *uncore = gt->uncore;
+	u32 rcctl1, pw_status;
+
+	pw_status = intel_uncore_read(uncore, VLV_GTLC_PW_STATUS);
+	rcctl1 = intel_uncore_read(uncore, GEN6_RC_CONTROL);
+
+	seq_printf(m, "RC6 Enabled: %s\n",
+		   yesno(rcctl1 & (GEN7_RC_CTL_TO_MODE |
+					GEN6_RC_CTL_EI_MODE(1))));
+	seq_printf(m, "Render Power Well: %s\n",
+		   (pw_status & VLV_GTLC_PW_RENDER_STATUS_MASK) ? "Up" : "Down");
+	seq_printf(m, "Media Power Well: %s\n",
+		   (pw_status & VLV_GTLC_PW_MEDIA_STATUS_MASK) ? "Up" : "Down");
+
+	print_rc6_res(m, "Render RC6 residency since boot:", VLV_GT_RENDER_RC6);
+	print_rc6_res(m, "Media RC6 residency since boot:", VLV_GT_MEDIA_RC6);
+
+	return fw_domains_show(m, NULL);
+}
+
+static int gen6_drpc(struct seq_file *m)
+{
+	struct intel_gt *gt = m->private;
+	struct drm_i915_private *i915 = gt->i915;
+	struct intel_uncore *uncore = gt->uncore;
+	u32 gt_core_status, rcctl1, rc6vids = 0;
+	u32 gen9_powergate_enable = 0, gen9_powergate_status = 0;
+
+	gt_core_status = intel_uncore_read_fw(uncore, GEN6_GT_CORE_STATUS);
+
+	rcctl1 = intel_uncore_read(uncore, GEN6_RC_CONTROL);
+	if (INTEL_GEN(i915) >= 9) {
+		gen9_powergate_enable =
+			intel_uncore_read(uncore, GEN9_PG_ENABLE);
+		gen9_powergate_status =
+			intel_uncore_read(uncore, GEN9_PWRGT_DOMAIN_STATUS);
+	}
+
+	if (INTEL_GEN(i915) <= 7)
+		sandybridge_pcode_read(i915, GEN6_PCODE_READ_RC6VIDS,
+				       &rc6vids, NULL);
+
+	seq_printf(m, "RC1e Enabled: %s\n",
+		   yesno(rcctl1 & GEN6_RC_CTL_RC1e_ENABLE));
+	seq_printf(m, "RC6 Enabled: %s\n",
+		   yesno(rcctl1 & GEN6_RC_CTL_RC6_ENABLE));
+	if (INTEL_GEN(i915) >= 9) {
+		seq_printf(m, "Render Well Gating Enabled: %s\n",
+			   yesno(gen9_powergate_enable & GEN9_RENDER_PG_ENABLE));
+		seq_printf(m, "Media Well Gating Enabled: %s\n",
+			   yesno(gen9_powergate_enable & GEN9_MEDIA_PG_ENABLE));
+	}
+	seq_printf(m, "Deep RC6 Enabled: %s\n",
+		   yesno(rcctl1 & GEN6_RC_CTL_RC6p_ENABLE));
+	seq_printf(m, "Deepest RC6 Enabled: %s\n",
+		   yesno(rcctl1 & GEN6_RC_CTL_RC6pp_ENABLE));
+	seq_puts(m, "Current RC state: ");
+	switch (gt_core_status & GEN6_RCn_MASK) {
+	case GEN6_RC0:
+		if (gt_core_status & GEN6_CORE_CPD_STATE_MASK)
+			seq_puts(m, "Core Power Down\n");
+		else
+			seq_puts(m, "on\n");
+		break;
+	case GEN6_RC3:
+		seq_puts(m, "RC3\n");
+		break;
+	case GEN6_RC6:
+		seq_puts(m, "RC6\n");
+		break;
+	case GEN6_RC7:
+		seq_puts(m, "RC7\n");
+		break;
+	default:
+		seq_puts(m, "Unknown\n");
+		break;
+	}
+
+	seq_printf(m, "Core Power Down: %s\n",
+		   yesno(gt_core_status & GEN6_CORE_CPD_STATE_MASK));
+	if (INTEL_GEN(i915) >= 9) {
+		seq_printf(m, "Render Power Well: %s\n",
+			   (gen9_powergate_status &
+			    GEN9_PWRGT_RENDER_STATUS_MASK) ? "Up" : "Down");
+		seq_printf(m, "Media Power Well: %s\n",
+			   (gen9_powergate_status &
+			    GEN9_PWRGT_MEDIA_STATUS_MASK) ? "Up" : "Down");
+	}
+
+	/* Not exactly sure what this is */
+	print_rc6_res(m, "RC6 \"Locked to RPn\" residency since boot:",
+		      GEN6_GT_GFX_RC6_LOCKED);
+	print_rc6_res(m, "RC6 residency since boot:", GEN6_GT_GFX_RC6);
+	print_rc6_res(m, "RC6+ residency since boot:", GEN6_GT_GFX_RC6p);
+	print_rc6_res(m, "RC6++ residency since boot:", GEN6_GT_GFX_RC6pp);
+
+	if (INTEL_GEN(i915) <= 7) {
+		seq_printf(m, "RC6   voltage: %dmV\n",
+			   GEN6_DECODE_RC6_VID(((rc6vids >> 0) & 0xff)));
+		seq_printf(m, "RC6+  voltage: %dmV\n",
+			   GEN6_DECODE_RC6_VID(((rc6vids >> 8) & 0xff)));
+		seq_printf(m, "RC6++ voltage: %dmV\n",
+			   GEN6_DECODE_RC6_VID(((rc6vids >> 16) & 0xff)));
+	}
+
+	return fw_domains_show(m, NULL);
+}
+
+static int ilk_drpc(struct seq_file *m)
+{
+	struct intel_gt *gt = m->private;
+	struct intel_uncore *uncore = gt->uncore;
+	u32 rgvmodectl, rstdbyctl;
+	u16 crstandvid;
+
+	rgvmodectl = intel_uncore_read(uncore, MEMMODECTL);
+	rstdbyctl = intel_uncore_read(uncore, RSTDBYCTL);
+	crstandvid = intel_uncore_read16(uncore, CRSTANDVID);
+
+	seq_printf(m, "HD boost: %s\n", yesno(rgvmodectl & MEMMODE_BOOST_EN));
+	seq_printf(m, "Boost freq: %d\n",
+		   (rgvmodectl & MEMMODE_BOOST_FREQ_MASK) >>
+		   MEMMODE_BOOST_FREQ_SHIFT);
+	seq_printf(m, "HW control enabled: %s\n",
+		   yesno(rgvmodectl & MEMMODE_HWIDLE_EN));
+	seq_printf(m, "SW control enabled: %s\n",
+		   yesno(rgvmodectl & MEMMODE_SWMODE_EN));
+	seq_printf(m, "Gated voltage change: %s\n",
+		   yesno(rgvmodectl & MEMMODE_RCLK_GATE));
+	seq_printf(m, "Starting frequency: P%d\n",
+		   (rgvmodectl & MEMMODE_FSTART_MASK) >> MEMMODE_FSTART_SHIFT);
+	seq_printf(m, "Max P-state: P%d\n",
+		   (rgvmodectl & MEMMODE_FMAX_MASK) >> MEMMODE_FMAX_SHIFT);
+	seq_printf(m, "Min P-state: P%d\n", (rgvmodectl & MEMMODE_FMIN_MASK));
+	seq_printf(m, "RS1 VID: %d\n", (crstandvid & 0x3f));
+	seq_printf(m, "RS2 VID: %d\n", ((crstandvid >> 8) & 0x3f));
+	seq_printf(m, "Render standby enabled: %s\n",
+		   yesno(!(rstdbyctl & RCX_SW_EXIT)));
+	seq_puts(m, "Current RS state: ");
+	switch (rstdbyctl & RSX_STATUS_MASK) {
+	case RSX_STATUS_ON:
+		seq_puts(m, "on\n");
+		break;
+	case RSX_STATUS_RC1:
+		seq_puts(m, "RC1\n");
+		break;
+	case RSX_STATUS_RC1E:
+		seq_puts(m, "RC1E\n");
+		break;
+	case RSX_STATUS_RS1:
+		seq_puts(m, "RS1\n");
+		break;
+	case RSX_STATUS_RS2:
+		seq_puts(m, "RS2 (RC6)\n");
+		break;
+	case RSX_STATUS_RS3:
+		seq_puts(m, "RC3 (RC6+)\n");
+		break;
+	default:
+		seq_puts(m, "unknown\n");
+		break;
+	}
+
+	return 0;
+}
+
+static int drpc_show(struct seq_file *m, void *unused)
+{
+	struct intel_gt *gt = m->private;
+	struct drm_i915_private *i915 = gt->i915;
+	intel_wakeref_t wakeref;
+	int err = -ENODEV;
+
+	with_intel_runtime_pm(gt->uncore->rpm, wakeref) {
+		if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915))
+			err = vlv_drpc(m);
+		else if (INTEL_GEN(i915) >= 6)
+			err = gen6_drpc(m);
+		else
+			err = ilk_drpc(m);
+	}
+
+	return err;
+}
+DEFINE_GT_DEBUGFS_ATTRIBUTE(drpc);
+
+static int frequency_show(struct seq_file *m, void *unused)
+{
+	struct intel_gt *gt = m->private;
+	struct drm_i915_private *i915 = gt->i915;
+	struct intel_uncore *uncore = gt->uncore;
+	struct intel_rps *rps = &gt->rps;
+	intel_wakeref_t wakeref;
+
+	wakeref = intel_runtime_pm_get(uncore->rpm);
+
+	if (IS_GEN(i915, 5)) {
+		u16 rgvswctl = intel_uncore_read16(uncore, MEMSWCTL);
+		u16 rgvstat = intel_uncore_read16(uncore, MEMSTAT_ILK);
+
+		seq_printf(m, "Requested P-state: %d\n", (rgvswctl >> 8) & 0xf);
+		seq_printf(m, "Requested VID: %d\n", rgvswctl & 0x3f);
+		seq_printf(m, "Current VID: %d\n", (rgvstat & MEMSTAT_VID_MASK) >>
+			   MEMSTAT_VID_SHIFT);
+		seq_printf(m, "Current P-state: %d\n",
+			   (rgvstat & MEMSTAT_PSTATE_MASK) >> MEMSTAT_PSTATE_SHIFT);
+	} else if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)) {
+		u32 rpmodectl, freq_sts;
+
+		rpmodectl = intel_uncore_read(uncore, GEN6_RP_CONTROL);
+		seq_printf(m, "Video Turbo Mode: %s\n",
+			   yesno(rpmodectl & GEN6_RP_MEDIA_TURBO));
+		seq_printf(m, "HW control enabled: %s\n",
+			   yesno(rpmodectl & GEN6_RP_ENABLE));
+		seq_printf(m, "SW control enabled: %s\n",
+			   yesno((rpmodectl & GEN6_RP_MEDIA_MODE_MASK) ==
+				  GEN6_RP_MEDIA_SW_MODE));
+
+		vlv_punit_get(i915);
+		freq_sts = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+		vlv_punit_put(i915);
+
+		seq_printf(m, "PUNIT_REG_GPU_FREQ_STS: 0x%08x\n", freq_sts);
+		seq_printf(m, "DDR freq: %d MHz\n", i915->mem_freq);
+
+		seq_printf(m, "actual GPU freq: %d MHz\n",
+			   intel_gpu_freq(rps, (freq_sts >> 8) & 0xff));
+
+		seq_printf(m, "current GPU freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->cur_freq));
+
+		seq_printf(m, "max GPU freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->max_freq));
+
+		seq_printf(m, "min GPU freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->min_freq));
+
+		seq_printf(m, "idle GPU freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->idle_freq));
+
+		seq_printf(m, "efficient (RPe) frequency: %d MHz\n",
+			   intel_gpu_freq(rps, rps->efficient_freq));
+	} else if (INTEL_GEN(i915) >= 6) {
+		u32 rp_state_limits;
+		u32 gt_perf_status;
+		u32 rp_state_cap;
+		u32 rpmodectl, rpinclimit, rpdeclimit;
+		u32 rpstat, cagf, reqf;
+		u32 rpupei, rpcurup, rpprevup;
+		u32 rpdownei, rpcurdown, rpprevdown;
+		u32 pm_ier, pm_imr, pm_isr, pm_iir, pm_mask;
+		int max_freq;
+
+		rp_state_limits = intel_uncore_read(uncore, GEN6_RP_STATE_LIMITS);
+		if (IS_GEN9_LP(i915)) {
+			rp_state_cap = intel_uncore_read(uncore, BXT_RP_STATE_CAP);
+			gt_perf_status = intel_uncore_read(uncore, BXT_GT_PERF_STATUS);
+		} else {
+			rp_state_cap = intel_uncore_read(uncore, GEN6_RP_STATE_CAP);
+			gt_perf_status = intel_uncore_read(uncore, GEN6_GT_PERF_STATUS);
+		}
+
+		/* RPSTAT1 is in the GT power well */
+		intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+
+		reqf = intel_uncore_read(uncore, GEN6_RPNSWREQ);
+		if (INTEL_GEN(i915) >= 9) {
+			reqf >>= 23;
+		} else {
+			reqf &= ~GEN6_TURBO_DISABLE;
+			if (IS_HASWELL(i915) || IS_BROADWELL(i915))
+				reqf >>= 24;
+			else
+				reqf >>= 25;
+		}
+		reqf = intel_gpu_freq(rps, reqf);
+
+		rpmodectl = intel_uncore_read(uncore, GEN6_RP_CONTROL);
+		rpinclimit = intel_uncore_read(uncore, GEN6_RP_UP_THRESHOLD);
+		rpdeclimit = intel_uncore_read(uncore, GEN6_RP_DOWN_THRESHOLD);
+
+		rpstat = intel_uncore_read(uncore, GEN6_RPSTAT1);
+		rpupei = intel_uncore_read(uncore, GEN6_RP_CUR_UP_EI) & GEN6_CURICONT_MASK;
+		rpcurup = intel_uncore_read(uncore, GEN6_RP_CUR_UP) & GEN6_CURBSYTAVG_MASK;
+		rpprevup = intel_uncore_read(uncore, GEN6_RP_PREV_UP) & GEN6_CURBSYTAVG_MASK;
+		rpdownei = intel_uncore_read(uncore, GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK;
+		rpcurdown = intel_uncore_read(uncore, GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK;
+		rpprevdown = intel_uncore_read(uncore, GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK;
+		cagf = intel_rps_read_actual_frequency(rps);
+
+		intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
+
+		if (INTEL_GEN(i915) >= 11) {
+			pm_ier = intel_uncore_read(uncore, GEN11_GPM_WGBOXPERF_INTR_ENABLE);
+			pm_imr = intel_uncore_read(uncore, GEN11_GPM_WGBOXPERF_INTR_MASK);
+			/*
+			 * The equivalent to the PM ISR & IIR cannot be read
+			 * without affecting the current state of the system
+			 */
+			pm_isr = 0;
+			pm_iir = 0;
+		} else if (INTEL_GEN(i915) >= 8) {
+			pm_ier = intel_uncore_read(uncore, GEN8_GT_IER(2));
+			pm_imr = intel_uncore_read(uncore, GEN8_GT_IMR(2));
+			pm_isr = intel_uncore_read(uncore, GEN8_GT_ISR(2));
+			pm_iir = intel_uncore_read(uncore, GEN8_GT_IIR(2));
+		} else {
+			pm_ier = intel_uncore_read(uncore, GEN6_PMIER);
+			pm_imr = intel_uncore_read(uncore, GEN6_PMIMR);
+			pm_isr = intel_uncore_read(uncore, GEN6_PMISR);
+			pm_iir = intel_uncore_read(uncore, GEN6_PMIIR);
+		}
+		pm_mask = intel_uncore_read(uncore, GEN6_PMINTRMSK);
+
+		seq_printf(m, "Video Turbo Mode: %s\n",
+			   yesno(rpmodectl & GEN6_RP_MEDIA_TURBO));
+		seq_printf(m, "HW control enabled: %s\n",
+			   yesno(rpmodectl & GEN6_RP_ENABLE));
+		seq_printf(m, "SW control enabled: %s\n",
+			   yesno((rpmodectl & GEN6_RP_MEDIA_MODE_MASK) ==
+				  GEN6_RP_MEDIA_SW_MODE));
+
+		seq_printf(m, "PM IER=0x%08x IMR=0x%08x, MASK=0x%08x\n",
+			   pm_ier, pm_imr, pm_mask);
+		if (INTEL_GEN(i915) <= 10)
+			seq_printf(m, "PM ISR=0x%08x IIR=0x%08x\n",
+				   pm_isr, pm_iir);
+		seq_printf(m, "pm_intrmsk_mbz: 0x%08x\n",
+			   rps->pm_intrmsk_mbz);
+		seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status);
+		seq_printf(m, "Render p-state ratio: %d\n",
+			   (gt_perf_status & (INTEL_GEN(i915) >= 9 ? 0x1ff00 : 0xff00)) >> 8);
+		seq_printf(m, "Render p-state VID: %d\n",
+			   gt_perf_status & 0xff);
+		seq_printf(m, "Render p-state limit: %d\n",
+			   rp_state_limits & 0xff);
+		seq_printf(m, "RPSTAT1: 0x%08x\n", rpstat);
+		seq_printf(m, "RPMODECTL: 0x%08x\n", rpmodectl);
+		seq_printf(m, "RPINCLIMIT: 0x%08x\n", rpinclimit);
+		seq_printf(m, "RPDECLIMIT: 0x%08x\n", rpdeclimit);
+		seq_printf(m, "RPNSWREQ: %dMHz\n", reqf);
+		seq_printf(m, "CAGF: %dMHz\n", cagf);
+		seq_printf(m, "RP CUR UP EI: %d (%dus)\n",
+			   rpupei, GT_PM_INTERVAL_TO_US(i915, rpupei));
+		seq_printf(m, "RP CUR UP: %d (%dus)\n",
+			   rpcurup, GT_PM_INTERVAL_TO_US(i915, rpcurup));
+		seq_printf(m, "RP PREV UP: %d (%dus)\n",
+			   rpprevup, GT_PM_INTERVAL_TO_US(i915, rpprevup));
+		seq_printf(m, "Up threshold: %d%%\n",
+			   rps->power.up_threshold);
+
+		seq_printf(m, "RP CUR DOWN EI: %d (%dus)\n",
+			   rpdownei, GT_PM_INTERVAL_TO_US(i915, rpdownei));
+		seq_printf(m, "RP CUR DOWN: %d (%dus)\n",
+			   rpcurdown, GT_PM_INTERVAL_TO_US(i915, rpcurdown));
+		seq_printf(m, "RP PREV DOWN: %d (%dus)\n",
+			   rpprevdown, GT_PM_INTERVAL_TO_US(i915, rpprevdown));
+		seq_printf(m, "Down threshold: %d%%\n",
+			   rps->power.down_threshold);
+
+		max_freq = (IS_GEN9_LP(i915) ? rp_state_cap >> 0 :
+			    rp_state_cap >> 16) & 0xff;
+		max_freq *= (IS_GEN9_BC(i915) ||
+			     INTEL_GEN(i915) >= 10 ? GEN9_FREQ_SCALER : 1);
+		seq_printf(m, "Lowest (RPN) frequency: %dMHz\n",
+			   intel_gpu_freq(rps, max_freq));
+
+		max_freq = (rp_state_cap & 0xff00) >> 8;
+		max_freq *= (IS_GEN9_BC(i915) ||
+			     INTEL_GEN(i915) >= 10 ? GEN9_FREQ_SCALER : 1);
+		seq_printf(m, "Nominal (RP1) frequency: %dMHz\n",
+			   intel_gpu_freq(rps, max_freq));
+
+		max_freq = (IS_GEN9_LP(i915) ? rp_state_cap >> 16 :
+			    rp_state_cap >> 0) & 0xff;
+		max_freq *= (IS_GEN9_BC(i915) ||
+			     INTEL_GEN(i915) >= 10 ? GEN9_FREQ_SCALER : 1);
+		seq_printf(m, "Max non-overclocked (RP0) frequency: %dMHz\n",
+			   intel_gpu_freq(rps, max_freq));
+		seq_printf(m, "Max overclocked frequency: %dMHz\n",
+			   intel_gpu_freq(rps, rps->max_freq));
+
+		seq_printf(m, "Current freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->cur_freq));
+		seq_printf(m, "Actual freq: %d MHz\n", cagf);
+		seq_printf(m, "Idle freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->idle_freq));
+		seq_printf(m, "Min freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->min_freq));
+		seq_printf(m, "Boost freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->boost_freq));
+		seq_printf(m, "Max freq: %d MHz\n",
+			   intel_gpu_freq(rps, rps->max_freq));
+		seq_printf(m,
+			   "efficient (RPe) frequency: %d MHz\n",
+			   intel_gpu_freq(rps, rps->efficient_freq));
+	} else {
+		seq_puts(m, "no P-state info available\n");
+	}
+
+	seq_printf(m, "Current CD clock frequency: %d kHz\n", i915->cdclk.hw.cdclk);
+	seq_printf(m, "Max CD clock frequency: %d kHz\n", i915->max_cdclk_freq);
+	seq_printf(m, "Max pixel clock frequency: %d kHz\n", i915->max_dotclk_freq);
+
+	intel_runtime_pm_put(uncore->rpm, wakeref);
+
+	return 0;
+}
+DEFINE_GT_DEBUGFS_ATTRIBUTE(frequency);
+
+static int llc_show(struct seq_file *m, void *data)
+{
+	struct intel_gt *gt = m->private;
+	struct drm_i915_private *i915 = gt->i915;
+	const bool edram = INTEL_GEN(i915) > 8;
+	struct intel_rps *rps = &gt->rps;
+	unsigned int max_gpu_freq, min_gpu_freq;
+	intel_wakeref_t wakeref;
+	int gpu_freq, ia_freq;
+
+	seq_printf(m, "LLC: %s\n", yesno(HAS_LLC(i915)));
+	seq_printf(m, "%s: %uMB\n", edram ? "eDRAM" : "eLLC",
+		   i915->edram_size_mb);
+
+	min_gpu_freq = rps->min_freq;
+	max_gpu_freq = rps->max_freq;
+	if (IS_GEN9_BC(i915) || INTEL_GEN(i915) >= 10) {
+		/* Convert GT frequency to 50 HZ units */
+		min_gpu_freq /= GEN9_FREQ_SCALER;
+		max_gpu_freq /= GEN9_FREQ_SCALER;
+	}
+
+	seq_puts(m, "GPU freq (MHz)\tEffective CPU freq (MHz)\tEffective Ring freq (MHz)\n");
+
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
+	for (gpu_freq = min_gpu_freq; gpu_freq <= max_gpu_freq; gpu_freq++) {
+		ia_freq = gpu_freq;
+		sandybridge_pcode_read(i915,
+				       GEN6_PCODE_READ_MIN_FREQ_TABLE,
+				       &ia_freq, NULL);
+		seq_printf(m, "%d\t\t%d\t\t\t\t%d\n",
+			   intel_gpu_freq(rps,
+					  (gpu_freq *
+					   (IS_GEN9_BC(i915) ||
+					    INTEL_GEN(i915) >= 10 ?
+					    GEN9_FREQ_SCALER : 1))),
+			   ((ia_freq >> 0) & 0xff) * 100,
+			   ((ia_freq >> 8) & 0xff) * 100);
+	}
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
+
+	return 0;
+}
+
+static bool llc_eval(const struct intel_gt *gt)
+{
+	return HAS_LLC(gt->i915);
+}
+
+DEFINE_GT_DEBUGFS_ATTRIBUTE(llc);
+
+static const char *rps_power_to_str(unsigned int power)
+{
+	static const char * const strings[] = {
+		[LOW_POWER] = "low power",
+		[BETWEEN] = "mixed",
+		[HIGH_POWER] = "high power",
+	};
+
+	if (power >= ARRAY_SIZE(strings) || !strings[power])
+		return "unknown";
+
+	return strings[power];
+}
+
+static int rps_boost_show(struct seq_file *m, void *data)
+{
+	struct intel_gt *gt = m->private;
+	struct drm_i915_private *i915 = gt->i915;
+	struct intel_rps *rps = &gt->rps;
+
+	seq_printf(m, "RPS enabled? %d\n", rps->enabled);
+	seq_printf(m, "GPU busy? %s\n", yesno(gt->awake));
+	seq_printf(m, "Boosts outstanding? %d\n",
+		   atomic_read(&rps->num_waiters));
+	seq_printf(m, "Interactive? %d\n", READ_ONCE(rps->power.interactive));
+	seq_printf(m, "Frequency requested %d, actual %d\n",
+		   intel_gpu_freq(rps, rps->cur_freq),
+		   intel_rps_read_actual_frequency(rps));
+	seq_printf(m, "  min hard:%d, soft:%d; max soft:%d, hard:%d\n",
+		   intel_gpu_freq(rps, rps->min_freq),
+		   intel_gpu_freq(rps, rps->min_freq_softlimit),
+		   intel_gpu_freq(rps, rps->max_freq_softlimit),
+		   intel_gpu_freq(rps, rps->max_freq));
+	seq_printf(m, "  idle:%d, efficient:%d, boost:%d\n",
+		   intel_gpu_freq(rps, rps->idle_freq),
+		   intel_gpu_freq(rps, rps->efficient_freq),
+		   intel_gpu_freq(rps, rps->boost_freq));
+
+	seq_printf(m, "Wait boosts: %d\n", atomic_read(&rps->boosts));
+
+	if (INTEL_GEN(i915) >= 6 && rps->enabled && gt->awake) {
+		struct intel_uncore *uncore = gt->uncore;
+		u32 rpup, rpupei;
+		u32 rpdown, rpdownei;
+
+		intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+		rpup = intel_uncore_read_fw(uncore, GEN6_RP_CUR_UP) & GEN6_RP_EI_MASK;
+		rpupei = intel_uncore_read_fw(uncore, GEN6_RP_CUR_UP_EI) & GEN6_RP_EI_MASK;
+		rpdown = intel_uncore_read_fw(uncore, GEN6_RP_CUR_DOWN) & GEN6_RP_EI_MASK;
+		rpdownei = intel_uncore_read_fw(uncore, GEN6_RP_CUR_DOWN_EI) & GEN6_RP_EI_MASK;
+		intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
+
+		seq_printf(m, "\nRPS Autotuning (current \"%s\" window):\n",
+			   rps_power_to_str(rps->power.mode));
+		seq_printf(m, "  Avg. up: %d%% [above threshold? %d%%]\n",
+			   rpup && rpupei ? 100 * rpup / rpupei : 0,
+			   rps->power.up_threshold);
+		seq_printf(m, "  Avg. down: %d%% [below threshold? %d%%]\n",
+			   rpdown && rpdownei ? 100 * rpdown / rpdownei : 0,
+			   rps->power.down_threshold);
+	} else {
+		seq_puts(m, "\nRPS Autotuning inactive\n");
+	}
+
+	return 0;
+}
+
+static bool rps_eval(const struct intel_gt *gt)
+{
+	return HAS_RPS(gt->i915);
+}
+
+DEFINE_GT_DEBUGFS_ATTRIBUTE(rps_boost);
+
+void debugfs_gt_pm_register(struct intel_gt *gt, struct dentry *root)
+{
+	static const struct debugfs_gt_file files[] = {
+		{ "drpc", &drpc_fops, NULL },
+		{ "frequency", &frequency_fops, NULL },
+		{ "forcewake", &fw_domains_fops, NULL },
+		{ "llc", &llc_fops, llc_eval },
+		{ "rps_boost", &rps_boost_fops, rps_eval },
+	};
+
+	debugfs_gt_register_files(gt, root, files, ARRAY_SIZE(files));
+}
diff --git a/drivers/gpu/drm/i915/gt/debugfs_gt_pm.h b/drivers/gpu/drm/i915/gt/debugfs_gt_pm.h
new file mode 100644
index 000000000000..4cf5f5c9da7d
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/debugfs_gt_pm.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef DEBUGFS_GT_PM_H
+#define DEBUGFS_GT_PM_H
+
+struct intel_gt;
+struct dentry;
+
+void debugfs_gt_pm_register(struct intel_gt *gt, struct dentry *root);
+
+#endif /* DEBUGFS_GT_PM_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c b/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
index 55317081d48b..0ba524a414c6 100644
--- a/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
+++ b/drivers/gpu/drm/i915/gt/intel_breadcrumbs.c
@@ -28,6 +28,8 @@
 
 #include "i915_drv.h"
 #include "i915_trace.h"
+#include "intel_gt_pm.h"
+#include "intel_gt_requests.h"
 
 static void irq_enable(struct intel_engine_cs *engine)
 {
@@ -53,15 +55,17 @@ static void irq_disable(struct intel_engine_cs *engine)
 
 static void __intel_breadcrumbs_disarm_irq(struct intel_breadcrumbs *b)
 {
+	struct intel_engine_cs *engine =
+		container_of(b, struct intel_engine_cs, breadcrumbs);
+
 	lockdep_assert_held(&b->irq_lock);
 
 	GEM_BUG_ON(!b->irq_enabled);
 	if (!--b->irq_enabled)
-		irq_disable(container_of(b,
-					 struct intel_engine_cs,
-					 breadcrumbs));
+		irq_disable(engine);
 
 	b->irq_armed = false;
+	intel_gt_pm_put_async(engine->gt);
 }
 
 void intel_engine_disarm_breadcrumbs(struct intel_engine_cs *engine)
@@ -127,16 +131,23 @@ __dma_fence_signal__notify(struct dma_fence *fence,
 	}
 }
 
-void intel_engine_breadcrumbs_irq(struct intel_engine_cs *engine)
+static void add_retire(struct intel_breadcrumbs *b, struct intel_timeline *tl)
 {
-	struct intel_breadcrumbs *b = &engine->breadcrumbs;
+	struct intel_engine_cs *engine =
+		container_of(b, struct intel_engine_cs, breadcrumbs);
+
+	intel_engine_add_retire(engine, tl);
+}
+
+static void signal_irq_work(struct irq_work *work)
+{
+	struct intel_breadcrumbs *b = container_of(work, typeof(*b), irq_work);
 	const ktime_t timestamp = ktime_get();
 	struct intel_context *ce, *cn;
 	struct list_head *pos, *next;
-	unsigned long flags;
 	LIST_HEAD(signal);
 
-	spin_lock_irqsave(&b->irq_lock, flags);
+	spin_lock(&b->irq_lock);
 
 	if (b->irq_armed && list_empty(&b->signalers))
 		__intel_breadcrumbs_disarm_irq(b);
@@ -177,44 +188,41 @@ void intel_engine_breadcrumbs_irq(struct intel_engine_cs *engine)
 		if (!list_is_first(pos, &ce->signals)) {
 			/* Advance the list to the first incomplete request */
 			__list_del_many(&ce->signals, pos);
-			if (&ce->signals == pos) /* now empty */
+			if (&ce->signals == pos) { /* now empty */
 				list_del_init(&ce->signal_link);
+				add_retire(b, ce->timeline);
+			}
 		}
 	}
 
-	spin_unlock_irqrestore(&b->irq_lock, flags);
+	spin_unlock(&b->irq_lock);
 
 	list_for_each_safe(pos, next, &signal) {
 		struct i915_request *rq =
 			list_entry(pos, typeof(*rq), signal_link);
 		struct list_head cb_list;
 
-		spin_lock_irqsave(&rq->lock, flags);
+		spin_lock(&rq->lock);
 		list_replace(&rq->fence.cb_list, &cb_list);
 		__dma_fence_signal__timestamp(&rq->fence, timestamp);
 		__dma_fence_signal__notify(&rq->fence, &cb_list);
-		spin_unlock_irqrestore(&rq->lock, flags);
+		spin_unlock(&rq->lock);
 
 		i915_request_put(rq);
 	}
 }
 
-static void signal_irq_work(struct irq_work *work)
-{
-	struct intel_engine_cs *engine =
-		container_of(work, typeof(*engine), breadcrumbs.irq_work);
-
-	intel_engine_breadcrumbs_irq(engine);
-}
-
-static void __intel_breadcrumbs_arm_irq(struct intel_breadcrumbs *b)
+static bool __intel_breadcrumbs_arm_irq(struct intel_breadcrumbs *b)
 {
 	struct intel_engine_cs *engine =
 		container_of(b, struct intel_engine_cs, breadcrumbs);
 
 	lockdep_assert_held(&b->irq_lock);
 	if (b->irq_armed)
-		return;
+		return true;
+
+	if (!intel_gt_pm_get_if_awake(engine->gt))
+		return false;
 
 	/*
 	 * The breadcrumb irq will be disarmed on the interrupt after the
@@ -234,6 +242,8 @@ static void __intel_breadcrumbs_arm_irq(struct intel_breadcrumbs *b)
 
 	if (!b->irq_enabled++)
 		irq_enable(engine);
+
+	return true;
 }
 
 void intel_engine_init_breadcrumbs(struct intel_engine_cs *engine)
@@ -271,19 +281,20 @@ bool i915_request_enable_breadcrumb(struct i915_request *rq)
 
 	if (test_bit(I915_FENCE_FLAG_ACTIVE, &rq->fence.flags)) {
 		struct intel_breadcrumbs *b = &rq->engine->breadcrumbs;
-		struct intel_context *ce = rq->hw_context;
+		struct intel_context *ce = rq->context;
 		struct list_head *pos;
 
 		spin_lock(&b->irq_lock);
 		GEM_BUG_ON(test_bit(I915_FENCE_FLAG_SIGNAL, &rq->fence.flags));
 
-		__intel_breadcrumbs_arm_irq(b);
+		if (!__intel_breadcrumbs_arm_irq(b))
+			goto unlock;
 
 		/*
 		 * We keep the seqno in retirement order, so we can break
-		 * inside intel_engine_breadcrumbs_irq as soon as we've passed
-		 * the last completed request (or seen a request that hasn't
-		 * event started). We could iterate the timeline->requests list,
+		 * inside intel_engine_signal_breadcrumbs as soon as we've
+		 * passed the last completed request (or seen a request that
+		 * hasn't event started). We could walk the timeline->requests,
 		 * but keeping a separate signalers_list has the advantage of
 		 * hopefully being much smaller than the full list and so
 		 * provides faster iteration and detection when there are no
@@ -306,6 +317,7 @@ bool i915_request_enable_breadcrumb(struct i915_request *rq)
 		GEM_BUG_ON(!check_signal_order(ce, rq));
 
 		set_bit(I915_FENCE_FLAG_SIGNAL, &rq->fence.flags);
+unlock:
 		spin_unlock(&b->irq_lock);
 	}
 
@@ -326,7 +338,7 @@ void i915_request_cancel_breadcrumb(struct i915_request *rq)
 	 */
 	spin_lock(&b->irq_lock);
 	if (test_bit(I915_FENCE_FLAG_SIGNAL, &rq->fence.flags)) {
-		struct intel_context *ce = rq->hw_context;
+		struct intel_context *ce = rq->context;
 
 		list_del(&rq->signal_link);
 		if (list_empty(&ce->signals))
diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c
index ef7bc41ffffa..fbaa9df6f436 100644
--- a/drivers/gpu/drm/i915/gt/intel_context.c
+++ b/drivers/gpu/drm/i915/gt/intel_context.c
@@ -31,8 +31,7 @@ void intel_context_free(struct intel_context *ce)
 }
 
 struct intel_context *
-intel_context_create(struct i915_gem_context *ctx,
-		     struct intel_engine_cs *engine)
+intel_context_create(struct intel_engine_cs *engine)
 {
 	struct intel_context *ce;
 
@@ -40,7 +39,7 @@ intel_context_create(struct i915_gem_context *ctx,
 	if (!ce)
 		return ERR_PTR(-ENOMEM);
 
-	intel_context_init(ce, ctx, engine);
+	intel_context_init(ce, engine);
 	return ce;
 }
 
@@ -68,11 +67,8 @@ int __intel_context_do_pin(struct intel_context *ce)
 		if (err)
 			goto err;
 
-		GEM_TRACE("%s context:%llx pin ring:{head:%04x, tail:%04x}\n",
-			  ce->engine->name, ce->timeline->fence_context,
-			  ce->ring->head, ce->ring->tail);
-
-		i915_gem_context_get(ce->gem_context); /* for ctx->ppgtt */
+		CE_TRACE(ce, "pin ring:{head:%04x, tail:%04x}\n",
+			 ce->ring->head, ce->ring->tail);
 
 		smp_mb__before_atomic(); /* flush pin before it is visible */
 	}
@@ -98,12 +94,10 @@ void intel_context_unpin(struct intel_context *ce)
 	mutex_lock_nested(&ce->pin_mutex, SINGLE_DEPTH_NESTING);
 
 	if (likely(atomic_dec_and_test(&ce->pin_count))) {
-		GEM_TRACE("%s context:%llx retire\n",
-			  ce->engine->name, ce->timeline->fence_context);
+		CE_TRACE(ce, "retire\n");
 
 		ce->ops->unpin(ce);
 
-		i915_gem_context_put(ce->gem_context);
 		intel_context_active_release(ce);
 	}
 
@@ -113,13 +107,10 @@ void intel_context_unpin(struct intel_context *ce)
 
 static int __context_pin_state(struct i915_vma *vma)
 {
-	u64 flags;
+	unsigned int bias = i915_ggtt_pin_bias(vma) | PIN_OFFSET_BIAS;
 	int err;
 
-	flags = i915_ggtt_pin_bias(vma) | PIN_OFFSET_BIAS;
-	flags |= PIN_HIGH | PIN_GLOBAL;
-
-	err = i915_vma_pin(vma, 0, 0, flags);
+	err = i915_ggtt_pin(vma, 0, bias | PIN_HIGH);
 	if (err)
 		return err;
 
@@ -144,9 +135,9 @@ static void __intel_context_retire(struct i915_active *active)
 {
 	struct intel_context *ce = container_of(active, typeof(*ce), active);
 
-	GEM_TRACE("%s context:%llx retire\n",
-		  ce->engine->name, ce->timeline->fence_context);
+	CE_TRACE(ce, "retire\n");
 
+	set_bit(CONTEXT_VALID_BIT, &ce->flags);
 	if (ce->state)
 		__context_unpin_state(ce->state);
 
@@ -198,7 +189,7 @@ int intel_context_active_acquire(struct intel_context *ce)
 		return err;
 
 	/* Preallocate tracking nodes */
-	if (!i915_gem_context_is_kernel(ce->gem_context)) {
+	if (!intel_context_is_barrier(ce)) {
 		err = i915_active_acquire_preallocate_barrier(&ce->active,
 							      ce->engine);
 		if (err) {
@@ -219,30 +210,19 @@ void intel_context_active_release(struct intel_context *ce)
 
 void
 intel_context_init(struct intel_context *ce,
-		   struct i915_gem_context *ctx,
 		   struct intel_engine_cs *engine)
 {
-	struct i915_address_space *vm;
-
 	GEM_BUG_ON(!engine->cops);
+	GEM_BUG_ON(!engine->gt->vm);
 
 	kref_init(&ce->ref);
 
-	ce->gem_context = ctx;
-	rcu_read_lock();
-	vm = rcu_dereference(ctx->vm);
-	if (vm)
-		ce->vm = i915_vm_get(vm);
-	else
-		ce->vm = i915_vm_get(&engine->gt->ggtt->vm);
-	rcu_read_unlock();
-	if (ctx->timeline)
-		ce->timeline = intel_timeline_get(ctx->timeline);
-
 	ce->engine = engine;
 	ce->ops = engine->cops;
 	ce->sseu = engine->sseu;
-	ce->ring = __intel_context_ring_size(SZ_16K);
+	ce->ring = __intel_context_ring_size(SZ_4K);
+
+	ce->vm = i915_vm_get(engine->gt->vm);
 
 	INIT_LIST_HEAD(&ce->signal_link);
 	INIT_LIST_HEAD(&ce->signals);
@@ -307,30 +287,11 @@ int intel_context_prepare_remote_request(struct intel_context *ce,
 	int err;
 
 	/* Only suitable for use in remotely modifying this context */
-	GEM_BUG_ON(rq->hw_context == ce);
+	GEM_BUG_ON(rq->context == ce);
 
 	if (rcu_access_pointer(rq->timeline) != tl) { /* timeline sharing! */
-		/*
-		 * Ideally, we just want to insert our foreign fence as
-		 * a barrier into the remove context, such that this operation
-		 * occurs after all current operations in that context, and
-		 * all future operations must occur after this.
-		 *
-		 * Currently, the timeline->last_request tracking is guarded
-		 * by its mutex and so we must obtain that to atomically
-		 * insert our barrier. However, since we already hold our
-		 * timeline->mutex, we must be careful against potential
-		 * inversion if we are the kernel_context as the remote context
-		 * will itself poke at the kernel_context when it needs to
-		 * unpin. Ergo, if already locked, we drop both locks and
-		 * try again (through the magic of userspace repeating EAGAIN).
-		 */
-		if (!mutex_trylock(&tl->mutex))
-			return -EAGAIN;
-
 		/* Queue this switch after current activity by this context. */
 		err = i915_active_fence_set(&tl->last_request, rq);
-		mutex_unlock(&tl->mutex);
 		if (err)
 			return err;
 	}
diff --git a/drivers/gpu/drm/i915/gt/intel_context.h b/drivers/gpu/drm/i915/gt/intel_context.h
index 68b3d317d959..1d4a1b1357cf 100644
--- a/drivers/gpu/drm/i915/gt/intel_context.h
+++ b/drivers/gpu/drm/i915/gt/intel_context.h
@@ -7,7 +7,9 @@
 #ifndef __INTEL_CONTEXT_H__
 #define __INTEL_CONTEXT_H__
 
+#include <linux/bitops.h>
 #include <linux/lockdep.h>
+#include <linux/types.h>
 
 #include "i915_active.h"
 #include "intel_context_types.h"
@@ -15,14 +17,19 @@
 #include "intel_ring_types.h"
 #include "intel_timeline_types.h"
 
+#define CE_TRACE(ce, fmt, ...) do {					\
+	const struct intel_context *ce__ = (ce);			\
+	ENGINE_TRACE(ce__->engine, "context:%llx" fmt,			\
+		     ce__->timeline->fence_context,			\
+		     ##__VA_ARGS__);					\
+} while (0)
+
 void intel_context_init(struct intel_context *ce,
-			struct i915_gem_context *ctx,
 			struct intel_engine_cs *engine);
 void intel_context_fini(struct intel_context *ce);
 
 struct intel_context *
-intel_context_create(struct i915_gem_context *ctx,
-		     struct intel_engine_cs *engine);
+intel_context_create(struct intel_engine_cs *engine);
 
 void intel_context_free(struct intel_context *ce);
 
@@ -153,4 +160,64 @@ static inline struct intel_ring *__intel_context_ring_size(u64 sz)
 	return u64_to_ptr(struct intel_ring, sz);
 }
 
+static inline bool intel_context_is_barrier(const struct intel_context *ce)
+{
+	return test_bit(CONTEXT_BARRIER_BIT, &ce->flags);
+}
+
+static inline bool intel_context_use_semaphores(const struct intel_context *ce)
+{
+	return test_bit(CONTEXT_USE_SEMAPHORES, &ce->flags);
+}
+
+static inline void intel_context_set_use_semaphores(struct intel_context *ce)
+{
+	set_bit(CONTEXT_USE_SEMAPHORES, &ce->flags);
+}
+
+static inline void intel_context_clear_use_semaphores(struct intel_context *ce)
+{
+	clear_bit(CONTEXT_USE_SEMAPHORES, &ce->flags);
+}
+
+static inline bool intel_context_is_banned(const struct intel_context *ce)
+{
+	return test_bit(CONTEXT_BANNED, &ce->flags);
+}
+
+static inline bool intel_context_set_banned(struct intel_context *ce)
+{
+	return test_and_set_bit(CONTEXT_BANNED, &ce->flags);
+}
+
+static inline bool
+intel_context_force_single_submission(const struct intel_context *ce)
+{
+	return test_bit(CONTEXT_FORCE_SINGLE_SUBMISSION, &ce->flags);
+}
+
+static inline void
+intel_context_set_single_submission(struct intel_context *ce)
+{
+	__set_bit(CONTEXT_FORCE_SINGLE_SUBMISSION, &ce->flags);
+}
+
+static inline bool
+intel_context_nopreempt(const struct intel_context *ce)
+{
+	return test_bit(CONTEXT_NOPREEMPT, &ce->flags);
+}
+
+static inline void
+intel_context_set_nopreempt(struct intel_context *ce)
+{
+	set_bit(CONTEXT_NOPREEMPT, &ce->flags);
+}
+
+static inline void
+intel_context_clear_nopreempt(struct intel_context *ce)
+{
+	clear_bit(CONTEXT_NOPREEMPT, &ce->flags);
+}
+
 #endif /* __INTEL_CONTEXT_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_context_types.h b/drivers/gpu/drm/i915/gt/intel_context_types.h
index 6959b05ae5f8..9527a659546c 100644
--- a/drivers/gpu/drm/i915/gt/intel_context_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_context_types.h
@@ -44,7 +44,7 @@ struct intel_context {
 #define intel_context_inflight_count(ce) ptr_unmask_bits((ce)->inflight, 2)
 
 	struct i915_address_space *vm;
-	struct i915_gem_context *gem_context;
+	struct i915_gem_context __rcu *gem_context;
 
 	struct list_head signal_link;
 	struct list_head signals;
@@ -54,7 +54,13 @@ struct intel_context {
 	struct intel_timeline *timeline;
 
 	unsigned long flags;
-#define CONTEXT_ALLOC_BIT 0
+#define CONTEXT_BARRIER_BIT		0
+#define CONTEXT_ALLOC_BIT		1
+#define CONTEXT_VALID_BIT		2
+#define CONTEXT_USE_SEMAPHORES		3
+#define CONTEXT_BANNED			4
+#define CONTEXT_FORCE_SINGLE_SUBMISSION	5
+#define CONTEXT_NOPREEMPT		6
 
 	u32 *lrc_reg_state;
 	u64 lrc_desc;
diff --git a/drivers/gpu/drm/i915/gt/intel_engine.h b/drivers/gpu/drm/i915/gt/intel_engine.h
index 01765a7ec18f..79ecac5ac0ab 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine.h
@@ -29,6 +29,13 @@ struct intel_gt;
 #define CACHELINE_BYTES 64
 #define CACHELINE_DWORDS (CACHELINE_BYTES / sizeof(u32))
 
+#define ENGINE_TRACE(e, fmt, ...) do {					\
+	const struct intel_engine_cs *e__ __maybe_unused = (e);		\
+	GEM_TRACE("%s %s: " fmt,					\
+		  dev_name(e__->i915->drm.dev), e__->name,		\
+		  ##__VA_ARGS__);					\
+} while (0)
+
 /*
  * The register defines to be used with the following macros need to accept a
  * base param, e.g:
@@ -177,15 +184,15 @@ void intel_engine_stop(struct intel_engine_cs *engine);
 void intel_engine_cleanup(struct intel_engine_cs *engine);
 
 int intel_engines_init_mmio(struct intel_gt *gt);
-int intel_engines_setup(struct intel_gt *gt);
 int intel_engines_init(struct intel_gt *gt);
-void intel_engines_cleanup(struct intel_gt *gt);
+
+void intel_engines_release(struct intel_gt *gt);
+void intel_engines_free(struct intel_gt *gt);
 
 int intel_engine_init_common(struct intel_engine_cs *engine);
 void intel_engine_cleanup_common(struct intel_engine_cs *engine);
 
 int intel_ring_submission_setup(struct intel_engine_cs *engine);
-int intel_ring_submission_init(struct intel_engine_cs *engine);
 
 int intel_engine_stop_cs(struct intel_engine_cs *engine);
 void intel_engine_cancel_stop_cs(struct intel_engine_cs *engine);
@@ -206,13 +213,11 @@ void intel_engine_fini_breadcrumbs(struct intel_engine_cs *engine);
 void intel_engine_disarm_breadcrumbs(struct intel_engine_cs *engine);
 
 static inline void
-intel_engine_queue_breadcrumbs(struct intel_engine_cs *engine)
+intel_engine_signal_breadcrumbs(struct intel_engine_cs *engine)
 {
 	irq_work_queue(&engine->breadcrumbs.irq_work);
 }
 
-void intel_engine_breadcrumbs_irq(struct intel_engine_cs *engine);
-
 void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine);
 void intel_engine_fini_breadcrumbs(struct intel_engine_cs *engine);
 
@@ -270,14 +275,14 @@ gen8_emit_ggtt_write(u32 *cs, u32 value, u32 gtt_offset, u32 flags)
 static inline void __intel_engine_reset(struct intel_engine_cs *engine,
 					bool stalled)
 {
-	if (engine->reset.reset)
-		engine->reset.reset(engine, stalled);
+	if (engine->reset.rewind)
+		engine->reset.rewind(engine, stalled);
 	engine->serial++; /* contexts lost */
 }
 
 bool intel_engines_are_idle(struct intel_gt *gt);
 bool intel_engine_is_idle(struct intel_engine_cs *engine);
-void intel_engine_flush_submission(struct intel_engine_cs *engine);
+bool intel_engine_flush_submission(struct intel_engine_cs *engine);
 
 void intel_engines_reset_default_submission(struct intel_gt *gt);
 
@@ -296,7 +301,7 @@ ktime_t intel_engine_get_busy_time(struct intel_engine_cs *engine);
 struct i915_request *
 intel_engine_find_active_request(struct intel_engine_cs *engine);
 
-u32 intel_engine_context_size(struct drm_i915_private *i915, u8 class);
+u32 intel_engine_context_size(struct intel_gt *gt, u8 class);
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
index 813bd3a610d2..ddf9543b1261 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c
@@ -141,7 +141,7 @@ static const struct engine_info intel_engines[] = {
 
 /**
  * intel_engine_context_size() - return the size of the context for an engine
- * @dev_priv: i915 device private
+ * @gt: the gt
  * @class: engine class
  *
  * Each engine class may require a different amount of space for a context
@@ -153,17 +153,18 @@ static const struct engine_info intel_engines[] = {
  * in LRC mode, but does not include the "shared data page" used with
  * GuC submission. The caller should account for this if using the GuC.
  */
-u32 intel_engine_context_size(struct drm_i915_private *dev_priv, u8 class)
+u32 intel_engine_context_size(struct intel_gt *gt, u8 class)
 {
+	struct intel_uncore *uncore = gt->uncore;
 	u32 cxt_size;
 
 	BUILD_BUG_ON(I915_GTT_PAGE_SIZE != PAGE_SIZE);
 
 	switch (class) {
 	case RENDER_CLASS:
-		switch (INTEL_GEN(dev_priv)) {
+		switch (INTEL_GEN(gt->i915)) {
 		default:
-			MISSING_CASE(INTEL_GEN(dev_priv));
+			MISSING_CASE(INTEL_GEN(gt->i915));
 			return DEFAULT_LR_CONTEXT_RENDER_SIZE;
 		case 12:
 		case 11:
@@ -175,14 +176,14 @@ u32 intel_engine_context_size(struct drm_i915_private *dev_priv, u8 class)
 		case 8:
 			return GEN8_LR_CONTEXT_RENDER_SIZE;
 		case 7:
-			if (IS_HASWELL(dev_priv))
+			if (IS_HASWELL(gt->i915))
 				return HSW_CXT_TOTAL_SIZE;
 
-			cxt_size = I915_READ(GEN7_CXT_SIZE);
+			cxt_size = intel_uncore_read(uncore, GEN7_CXT_SIZE);
 			return round_up(GEN7_CXT_TOTAL_SIZE(cxt_size) * 64,
 					PAGE_SIZE);
 		case 6:
-			cxt_size = I915_READ(CXT_SIZE);
+			cxt_size = intel_uncore_read(uncore, CXT_SIZE);
 			return round_up(GEN6_CXT_TOTAL_SIZE(cxt_size) * 64,
 					PAGE_SIZE);
 		case 5:
@@ -197,9 +198,9 @@ u32 intel_engine_context_size(struct drm_i915_private *dev_priv, u8 class)
 			 * minimum allocation anyway so it should all come
 			 * out in the wash.
 			 */
-			cxt_size = I915_READ(CXT_SIZE) + 1;
+			cxt_size = intel_uncore_read(uncore, CXT_SIZE) + 1;
 			DRM_DEBUG_DRIVER("gen%d CXT_SIZE = %d bytes [0x%08x]\n",
-					 INTEL_GEN(dev_priv),
+					 INTEL_GEN(gt->i915),
 					 cxt_size * 64,
 					 cxt_size - 1);
 			return round_up(cxt_size * 64, PAGE_SIZE);
@@ -216,7 +217,7 @@ u32 intel_engine_context_size(struct drm_i915_private *dev_priv, u8 class)
 	case VIDEO_DECODE_CLASS:
 	case VIDEO_ENHANCEMENT_CLASS:
 	case COPY_ENGINE_CLASS:
-		if (INTEL_GEN(dev_priv) < 8)
+		if (INTEL_GEN(gt->i915) < 8)
 			return 0;
 		return GEN8_LR_CONTEXT_OTHER_SIZE;
 	}
@@ -318,14 +319,7 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	engine->props.timeslice_duration_ms =
 		CONFIG_DRM_I915_TIMESLICE_DURATION;
 
-	/*
-	 * To be overridden by the backend on setup. However to facilitate
-	 * cleanup on error during setup, we always provide the destroy vfunc.
-	 */
-	engine->destroy = (typeof(engine->destroy))kfree;
-
-	engine->context_size = intel_engine_context_size(gt->i915,
-							 engine->class);
+	engine->context_size = intel_engine_context_size(gt, engine->class);
 	if (WARN_ON(engine->context_size > BIT(20)))
 		engine->context_size = 0;
 	if (engine->context_size)
@@ -334,6 +328,7 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	/* Nothing to do here, execute in order of dependencies */
 	engine->schedule = NULL;
 
+	ewma__engine_latency_init(&engine->latency);
 	seqlock_init(&engine->stats.lock);
 
 	ATOMIC_INIT_NOTIFIER_HEAD(&engine->context_status_notifier);
@@ -344,7 +339,6 @@ static int intel_engine_setup(struct intel_gt *gt, enum intel_engine_id id)
 	gt->engine_class[info->class][info->instance] = engine;
 	gt->engine[id] = engine;
 
-	intel_engine_add_user(engine);
 	gt->i915->engine[id] = engine;
 
 	return 0;
@@ -390,21 +384,39 @@ static void intel_setup_engine_capabilities(struct intel_gt *gt)
 }
 
 /**
- * intel_engines_cleanup() - free the resources allocated for Command Streamers
+ * intel_engines_release() - free the resources allocated for Command Streamers
  * @gt: pointer to struct intel_gt
  */
-void intel_engines_cleanup(struct intel_gt *gt)
+void intel_engines_release(struct intel_gt *gt)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 
+	/* Decouple the backend; but keep the layout for late GPU resets */
 	for_each_engine(engine, gt, id) {
-		engine->destroy(engine);
-		gt->engine[id] = NULL;
+		if (!engine->release)
+			continue;
+
+		engine->release(engine);
+		engine->release = NULL;
+
+		memset(&engine->reset, 0, sizeof(engine->reset));
+
 		gt->i915->engine[id] = NULL;
 	}
 }
 
+void intel_engines_free(struct intel_gt *gt)
+{
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, gt, id) {
+		kfree(engine);
+		gt->engine[id] = NULL;
+	}
+}
+
 /**
  * intel_engines_init_mmio() - allocate and prepare the Engine Command Streamers
  * @gt: pointer to struct intel_gt
@@ -455,38 +467,7 @@ int intel_engines_init_mmio(struct intel_gt *gt)
 	return 0;
 
 cleanup:
-	intel_engines_cleanup(gt);
-	return err;
-}
-
-/**
- * intel_engines_init() - init the Engine Command Streamers
- * @gt: pointer to struct intel_gt
- *
- * Return: non-zero if the initialization failed.
- */
-int intel_engines_init(struct intel_gt *gt)
-{
-	int (*init)(struct intel_engine_cs *engine);
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	int err;
-
-	if (HAS_EXECLISTS(gt->i915))
-		init = intel_execlists_submission_init;
-	else
-		init = intel_ring_submission_init;
-
-	for_each_engine(engine, gt, id) {
-		err = init(engine);
-		if (err)
-			goto cleanup;
-	}
-
-	return 0;
-
-cleanup:
-	intel_engines_cleanup(gt);
+	intel_engines_free(gt);
 	return err;
 }
 
@@ -601,7 +582,7 @@ err:
 	return ret;
 }
 
-static int intel_engine_setup_common(struct intel_engine_cs *engine)
+static int engine_setup_common(struct intel_engine_cs *engine)
 {
 	int err;
 
@@ -631,49 +612,6 @@ static int intel_engine_setup_common(struct intel_engine_cs *engine)
 	return 0;
 }
 
-/**
- * intel_engines_setup- setup engine state not requiring hw access
- * @gt: pointer to struct intel_gt
- *
- * Initializes engine structure members shared between legacy and execlists
- * submission modes which do not require hardware access.
- *
- * Typically done early in the submission mode specific engine setup stage.
- */
-int intel_engines_setup(struct intel_gt *gt)
-{
-	int (*setup)(struct intel_engine_cs *engine);
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	int err;
-
-	if (HAS_EXECLISTS(gt->i915))
-		setup = intel_execlists_submission_setup;
-	else
-		setup = intel_ring_submission_setup;
-
-	for_each_engine(engine, gt, id) {
-		err = intel_engine_setup_common(engine);
-		if (err)
-			goto cleanup;
-
-		err = setup(engine);
-		if (err)
-			goto cleanup;
-
-		/* We expect the backend to take control over its state */
-		GEM_BUG_ON(engine->destroy == (typeof(engine->destroy))kfree);
-
-		GEM_BUG_ON(!engine->cops);
-	}
-
-	return 0;
-
-cleanup:
-	intel_engines_cleanup(gt);
-	return err;
-}
-
 struct measure_breadcrumb {
 	struct i915_request rq;
 	struct intel_timeline timeline;
@@ -757,13 +695,13 @@ create_kernel_context(struct intel_engine_cs *engine)
 	struct intel_context *ce;
 	int err;
 
-	ce = intel_context_create(engine->i915->kernel_context, engine);
+	ce = intel_context_create(engine);
 	if (IS_ERR(ce))
 		return ce;
 
-	ce->ring = __intel_context_ring_size(SZ_4K);
+	__set_bit(CONTEXT_BARRIER_BIT, &ce->flags);
 
-	err = intel_context_pin(ce);
+	err = intel_context_pin(ce); /* perma-pin so it is always available */
 	if (err) {
 		intel_context_put(ce);
 		return ERR_PTR(err);
@@ -791,13 +729,19 @@ create_kernel_context(struct intel_engine_cs *engine)
  *
  * Returns zero on success or an error code on failure.
  */
-int intel_engine_init_common(struct intel_engine_cs *engine)
+static int engine_init_common(struct intel_engine_cs *engine)
 {
 	struct intel_context *ce;
 	int ret;
 
 	engine->set_default_submission(engine);
 
+	ret = measure_breadcrumb_dw(engine);
+	if (ret < 0)
+		return ret;
+
+	engine->emit_fini_breadcrumb_dw = ret;
+
 	/*
 	 * We may need to do things with the shrinker which
 	 * require us to immediately switch back to the default
@@ -812,18 +756,38 @@ int intel_engine_init_common(struct intel_engine_cs *engine)
 
 	engine->kernel_context = ce;
 
-	ret = measure_breadcrumb_dw(engine);
-	if (ret < 0)
-		goto err_unpin;
+	return 0;
+}
 
-	engine->emit_fini_breadcrumb_dw = ret;
+int intel_engines_init(struct intel_gt *gt)
+{
+	int (*setup)(struct intel_engine_cs *engine);
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err;
 
-	return 0;
+	if (HAS_EXECLISTS(gt->i915))
+		setup = intel_execlists_submission_setup;
+	else
+		setup = intel_ring_submission_setup;
 
-err_unpin:
-	intel_context_unpin(ce);
-	intel_context_put(ce);
-	return ret;
+	for_each_engine(engine, gt, id) {
+		err = engine_setup_common(engine);
+		if (err)
+			return err;
+
+		err = setup(engine);
+		if (err)
+			return err;
+
+		err = engine_init_common(engine);
+		if (err)
+			return err;
+
+		intel_engine_add_user(engine);
+	}
+
+	return 0;
 }
 
 /**
@@ -836,6 +800,7 @@ err_unpin:
 void intel_engine_cleanup_common(struct intel_engine_cs *engine)
 {
 	GEM_BUG_ON(!list_empty(&engine->active.requests));
+	tasklet_kill(&engine->execlists.tasklet); /* flush the callback */
 
 	cleanup_status_page(engine);
 
@@ -911,7 +876,7 @@ int intel_engine_stop_cs(struct intel_engine_cs *engine)
 	if (INTEL_GEN(engine->i915) < 3)
 		return -ENODEV;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	intel_uncore_write_fw(uncore, mode, _MASKED_BIT_ENABLE(STOP_RING));
 
@@ -920,7 +885,7 @@ int intel_engine_stop_cs(struct intel_engine_cs *engine)
 					 mode, MODE_IDLE, MODE_IDLE,
 					 1000, stop_timeout(engine),
 					 NULL)) {
-		GEM_TRACE("%s: timed out on STOP_RING -> IDLE\n", engine->name);
+		ENGINE_TRACE(engine, "timed out on STOP_RING -> IDLE\n");
 		err = -ETIMEDOUT;
 	}
 
@@ -932,7 +897,7 @@ int intel_engine_stop_cs(struct intel_engine_cs *engine)
 
 void intel_engine_cancel_stop_cs(struct intel_engine_cs *engine)
 {
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	ENGINE_WRITE_FW(engine, RING_MI_MODE, _MASKED_BIT_DISABLE(STOP_RING));
 }
@@ -1082,9 +1047,10 @@ static bool ring_is_idle(struct intel_engine_cs *engine)
 	return idle;
 }
 
-void intel_engine_flush_submission(struct intel_engine_cs *engine)
+bool intel_engine_flush_submission(struct intel_engine_cs *engine)
 {
 	struct tasklet_struct *t = &engine->execlists.tasklet;
+	bool active = tasklet_is_locked(t);
 
 	if (__tasklet_is_scheduled(t)) {
 		local_bh_disable();
@@ -1095,10 +1061,13 @@ void intel_engine_flush_submission(struct intel_engine_cs *engine)
 			tasklet_unlock(t);
 		}
 		local_bh_enable();
+		active = true;
 	}
 
 	/* Otherwise flush the tasklet if it was running on another cpu */
 	tasklet_unlock_wait(t);
+
+	return active;
 }
 
 /**
@@ -1478,6 +1447,10 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 		drm_printf(m, "*** WEDGED ***\n");
 
 	drm_printf(m, "\tAwake? %d\n", atomic_read(&engine->wakeref.count));
+	drm_printf(m, "\tBarriers?: %s\n",
+		   yesno(!llist_empty(&engine->barrier_tasks)));
+	drm_printf(m, "\tLatency: %luus\n",
+		   ewma__engine_latency_read(&engine->latency));
 
 	rcu_read_lock();
 	rq = READ_ONCE(engine->heartbeat.systole);
@@ -1517,9 +1490,9 @@ void intel_engine_dump(struct intel_engine_cs *engine,
 
 		print_request_ring(m, rq);
 
-		if (rq->hw_context->lrc_reg_state) {
+		if (rq->context->lrc_reg_state) {
 			drm_printf(m, "Logical Ring Context:\n");
-			hexdump(m, rq->hw_context->lrc_reg_state, PAGE_SIZE);
+			hexdump(m, rq->context->lrc_reg_state, PAGE_SIZE);
 		}
 	}
 	spin_unlock_irqrestore(&engine->active.lock, flags);
@@ -1580,7 +1553,7 @@ int intel_enable_engine_stats(struct intel_engine_cs *engine)
 
 		for (port = execlists->pending; (rq = *port); port++) {
 			/* Exclude any contexts already counted in active */
-			if (!intel_context_inflight_count(rq->hw_context))
+			if (!intel_context_inflight_count(rq->context))
 				engine->stats.active++;
 		}
 
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
index 06aa14c7aa8c..742628e40201 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_heartbeat.c
@@ -63,15 +63,15 @@ static void heartbeat(struct work_struct *wrk)
 	struct intel_context *ce = engine->kernel_context;
 	struct i915_request *rq;
 
-	if (!intel_engine_pm_get_if_awake(engine))
-		return;
-
 	rq = engine->heartbeat.systole;
 	if (rq && i915_request_completed(rq)) {
 		i915_request_put(rq);
 		engine->heartbeat.systole = NULL;
 	}
 
+	if (!intel_engine_pm_get_if_awake(engine))
+		return;
+
 	if (intel_gt_is_wedged(engine->gt))
 		goto out;
 
@@ -215,18 +215,26 @@ out_rpm:
 int intel_engine_flush_barriers(struct intel_engine_cs *engine)
 {
 	struct i915_request *rq;
+	int err = 0;
 
 	if (llist_empty(&engine->barrier_tasks))
 		return 0;
 
+	if (!intel_engine_pm_get_if_awake(engine))
+		return 0;
+
 	rq = i915_request_create(engine->kernel_context);
-	if (IS_ERR(rq))
-		return PTR_ERR(rq);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		goto out_rpm;
+	}
 
 	idle_pulse(engine, rq);
 	i915_request_add(rq);
 
-	return 0;
+out_rpm:
+	intel_engine_pm_put(engine);
+	return err;
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_pm.c b/drivers/gpu/drm/i915/gt/intel_engine_pm.c
index c1dd0cd3efc7..010620b78202 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_pm.c
+++ b/drivers/gpu/drm/i915/gt/intel_engine_pm.c
@@ -6,6 +6,7 @@
 
 #include "i915_drv.h"
 
+#include "intel_context.h"
 #include "intel_engine.h"
 #include "intel_engine_heartbeat.h"
 #include "intel_engine_pm.h"
@@ -21,7 +22,7 @@ static int __engine_unpark(struct intel_wakeref *wf)
 		container_of(wf, typeof(*engine), wakeref);
 	void *map;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	intel_gt_pm_get(engine->gt);
 
@@ -73,6 +74,15 @@ static inline void __timeline_mark_unlock(struct intel_context *ce,
 
 #endif /* !IS_ENABLED(CONFIG_LOCKDEP) */
 
+static void duration(struct dma_fence *fence, struct dma_fence_cb *cb)
+{
+	struct i915_request *rq = to_request(fence);
+
+	ewma__engine_latency_add(&rq->engine->latency,
+				 ktime_us_delta(rq->fence.timestamp,
+						rq->duration.emitted));
+}
+
 static void
 __queue_and_release_pm(struct i915_request *rq,
 		       struct intel_timeline *tl,
@@ -80,7 +90,7 @@ __queue_and_release_pm(struct i915_request *rq,
 {
 	struct intel_gt_timelines *timelines = &engine->gt->timelines;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	/*
 	 * We have to serialise all potential retirement paths with our
@@ -113,6 +123,8 @@ static bool switch_to_kernel_context(struct intel_engine_cs *engine)
 	unsigned long flags;
 	bool result = true;
 
+	GEM_BUG_ON(!intel_context_is_barrier(ce));
+
 	/* Already inside the kernel context, safe to power down. */
 	if (engine->wakeref_serial == engine->serial)
 		return true;
@@ -163,7 +175,18 @@ static bool switch_to_kernel_context(struct intel_engine_cs *engine)
 
 	/* Install ourselves as a preemption barrier */
 	rq->sched.attr.priority = I915_PRIORITY_BARRIER;
-	__i915_request_commit(rq);
+	if (likely(!__i915_request_commit(rq))) { /* engine should be idle! */
+		/*
+		 * Use an interrupt for precise measurement of duration,
+		 * otherwise we rely on someone else retiring all the requests
+		 * which may delay the signaling (i.e. we will likely wait
+		 * until the background request retirement running every
+		 * second or two).
+		 */
+		BUILD_BUG_ON(sizeof(rq->duration) > sizeof(rq->submitq));
+		dma_fence_add_callback(&rq->fence, &rq->duration.cb, duration);
+		rq->duration.emitted = ktime_get();
+	}
 
 	/* Expose ourselves to the world */
 	__queue_and_release_pm(rq, ce->timeline, engine);
@@ -183,7 +206,7 @@ static void call_idle_barriers(struct intel_engine_cs *engine)
 			container_of((struct list_head *)node,
 				     typeof(*cb), node);
 
-		cb->func(NULL, cb);
+		cb->func(ERR_PTR(-EAGAIN), cb);
 	}
 }
 
@@ -204,7 +227,7 @@ static int __engine_park(struct intel_wakeref *wf)
 	if (!switch_to_kernel_context(engine))
 		return -EBUSY;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	call_idle_barriers(engine); /* cleanup after wedging */
 
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_pm.h b/drivers/gpu/drm/i915/gt/intel_engine_pm.h
index 24e20344dc22..e52c2b0cb245 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_pm.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_pm.h
@@ -7,6 +7,7 @@
 #ifndef INTEL_ENGINE_PM_H
 #define INTEL_ENGINE_PM_H
 
+#include "i915_request.h"
 #include "intel_engine_types.h"
 #include "intel_wakeref.h"
 
@@ -41,6 +42,26 @@ static inline void intel_engine_pm_flush(struct intel_engine_cs *engine)
 	intel_wakeref_unlock_wait(&engine->wakeref);
 }
 
+static inline struct i915_request *
+intel_engine_create_kernel_request(struct intel_engine_cs *engine)
+{
+	struct i915_request *rq;
+
+	/*
+	 * The engine->kernel_context is special as it is used inside
+	 * the engine-pm barrier (see __engine_park()), circumventing
+	 * the usual mutexes and relying on the engine-pm barrier
+	 * instead. So whenever we use the engine->kernel_context
+	 * outside of the barrier, we must manually handle the
+	 * engine wakeref to serialise with the use inside.
+	 */
+	intel_engine_pm_get(engine);
+	rq = i915_request_create(engine->kernel_context);
+	intel_engine_pm_put(engine);
+
+	return rq;
+}
+
 void intel_engine_init__pm(struct intel_engine_cs *engine);
 
 #endif /* INTEL_ENGINE_PM_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h
index 17f1f1441efc..00287515e7af 100644
--- a/drivers/gpu/drm/i915/gt/intel_engine_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h
@@ -7,6 +7,7 @@
 #ifndef __INTEL_ENGINE_TYPES__
 #define __INTEL_ENGINE_TYPES__
 
+#include <linux/average.h>
 #include <linux/hashtable.h>
 #include <linux/irq_work.h>
 #include <linux/kref.h>
@@ -119,6 +120,9 @@ enum intel_engine_id {
 #define INVALID_ENGINE ((enum intel_engine_id)-1)
 };
 
+/* A simple estimator for the round-trip latency of an engine */
+DECLARE_EWMA(_engine_latency, 6, 4)
+
 struct st_preempt_hang {
 	struct completion completion;
 	unsigned int count;
@@ -316,6 +320,13 @@ struct intel_engine_cs {
 		struct intel_timeline *timeline;
 	} legacy;
 
+	/*
+	 * We track the average duration of the idle pulse on parking the
+	 * engine to keep an estimate of the how the fast the engine is
+	 * under ideal conditions.
+	 */
+	struct ewma__engine_latency latency;
+
 	/* Rather than have every client wait upon all user interrupts,
 	 * with the herd waking after every interrupt and each doing the
 	 * heavyweight seqno dance, we delegate the task (of being the
@@ -389,7 +400,10 @@ struct intel_engine_cs {
 
 	struct {
 		void (*prepare)(struct intel_engine_cs *engine);
-		void (*reset)(struct intel_engine_cs *engine, bool stalled);
+
+		void (*rewind)(struct intel_engine_cs *engine, bool stalled);
+		void (*cancel)(struct intel_engine_cs *engine);
+
 		void (*finish)(struct intel_engine_cs *engine);
 	} reset;
 
@@ -439,15 +453,7 @@ struct intel_engine_cs {
 	void		(*schedule)(struct i915_request *request,
 				    const struct i915_sched_attr *attr);
 
-	/*
-	 * Cancel all requests on the hardware, or queued for execution.
-	 * This should only cancel the ready requests that have been
-	 * submitted to the engine (via the engine->submit_request callback).
-	 * This is called when marking the device as wedged.
-	 */
-	void		(*cancel_requests)(struct intel_engine_cs *engine);
-
-	void		(*destroy)(struct intel_engine_cs *engine);
+	void		(*release)(struct intel_engine_cs *engine);
 
 	struct intel_engine_execlists execlists;
 
diff --git a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
index 4294f146f13c..51b8718513bc 100644
--- a/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
+++ b/drivers/gpu/drm/i915/gt/intel_gpu_commands.h
@@ -7,6 +7,8 @@
 #ifndef _INTEL_GPU_COMMANDS_H_
 #define _INTEL_GPU_COMMANDS_H_
 
+#include <linux/bitops.h>
+
 /*
  * Target address alignments required for GPU access e.g.
  * MI_STORE_DWORD_IMM.
@@ -319,4 +321,31 @@
 #define COLOR_BLT     ((0x2<<29)|(0x40<<22))
 #define SRC_COPY_BLT  ((0x2<<29)|(0x43<<22))
 
+/*
+ * Used to convert any address to canonical form.
+ * Starting from gen8, some commands (e.g. STATE_BASE_ADDRESS,
+ * MI_LOAD_REGISTER_MEM and others, see Broadwell PRM Vol2a) require the
+ * addresses to be in a canonical form:
+ * "GraphicsAddress[63:48] are ignored by the HW and assumed to be in correct
+ * canonical form [63:48] == [47]."
+ */
+#define GEN8_HIGH_ADDRESS_BIT 47
+static inline u64 gen8_canonical_addr(u64 address)
+{
+	return sign_extend64(address, GEN8_HIGH_ADDRESS_BIT);
+}
+
+static inline u64 gen8_noncanonical_addr(u64 address)
+{
+	return address & GENMASK_ULL(GEN8_HIGH_ADDRESS_BIT, 0);
+}
+
+static inline u32 *__gen6_emit_bb_start(u32 *cs, u32 addr, unsigned int flags)
+{
+	*cs++ = MI_BATCH_BUFFER_START | flags;
+	*cs++ = addr;
+
+	return cs;
+}
+
 #endif /* _INTEL_GPU_COMMANDS_H_ */
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.c b/drivers/gpu/drm/i915/gt/intel_gt.c
index 4c26daf7ee46..ec84b5e62fef 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt.c
@@ -3,12 +3,15 @@
  * Copyright © 2019 Intel Corporation
  */
 
+#include "debugfs_gt.h"
 #include "i915_drv.h"
+#include "intel_context.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
 #include "intel_gt_requests.h"
 #include "intel_mocs.h"
 #include "intel_rc6.h"
+#include "intel_renderstate.h"
 #include "intel_rps.h"
 #include "intel_uncore.h"
 #include "intel_pm.h"
@@ -25,6 +28,7 @@ void intel_gt_init_early(struct intel_gt *gt, struct drm_i915_private *i915)
 
 	intel_gt_init_reset(gt);
 	intel_gt_init_requests(gt);
+	intel_gt_init_timelines(gt);
 	intel_gt_pm_init_early(gt);
 
 	intel_rps_init_early(&gt->rps);
@@ -73,7 +77,6 @@ int intel_gt_init_hw(struct intel_gt *gt)
 	struct intel_uncore *uncore = gt->uncore;
 	int ret;
 
-	BUG_ON(!i915->kernel_context);
 	ret = intel_gt_terminally_wedged(gt);
 	if (ret)
 		return ret;
@@ -303,7 +306,7 @@ void intel_gt_flush_ggtt_writes(struct intel_gt *gt)
 
 	intel_gt_chipset_flush(gt);
 
-	with_intel_runtime_pm(uncore->rpm, wakeref) {
+	with_intel_runtime_pm_if_in_use(uncore->rpm, wakeref) {
 		unsigned long flags;
 
 		spin_lock_irqsave(&uncore->lock, flags);
@@ -323,6 +326,8 @@ void intel_gt_chipset_flush(struct intel_gt *gt)
 void intel_gt_driver_register(struct intel_gt *gt)
 {
 	intel_rps_driver_register(&gt->rps);
+
+	debugfs_gt_register(gt);
 }
 
 static int intel_gt_init_scratch(struct intel_gt *gt, unsigned int size)
@@ -364,22 +369,273 @@ static void intel_gt_fini_scratch(struct intel_gt *gt)
 	i915_vma_unpin_and_release(&gt->scratch, 0);
 }
 
+static struct i915_address_space *kernel_vm(struct intel_gt *gt)
+{
+	if (INTEL_PPGTT(gt->i915) > INTEL_PPGTT_ALIASING)
+		return &i915_ppgtt_create(gt->i915)->vm;
+	else
+		return i915_vm_get(&gt->ggtt->vm);
+}
+
+static int __intel_context_flush_retire(struct intel_context *ce)
+{
+	struct intel_timeline *tl;
+
+	tl = intel_context_timeline_lock(ce);
+	if (IS_ERR(tl))
+		return PTR_ERR(tl);
+
+	intel_context_timeline_unlock(tl);
+	return 0;
+}
+
+static int __engines_record_defaults(struct intel_gt *gt)
+{
+	struct i915_request *requests[I915_NUM_ENGINES] = {};
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	/*
+	 * As we reset the gpu during very early sanitisation, the current
+	 * register state on the GPU should reflect its defaults values.
+	 * We load a context onto the hw (with restore-inhibit), then switch
+	 * over to a second context to save that default register state. We
+	 * can then prime every new context with that state so they all start
+	 * from the same default HW values.
+	 */
+
+	for_each_engine(engine, gt, id) {
+		struct intel_renderstate so;
+		struct intel_context *ce;
+		struct i915_request *rq;
+
+		err = intel_renderstate_init(&so, engine);
+		if (err)
+			goto out;
+
+		/* We must be able to switch to something! */
+		GEM_BUG_ON(!engine->kernel_context);
+		engine->serial++; /* force the kernel context switch */
+
+		ce = intel_context_create(engine);
+		if (IS_ERR(ce)) {
+			err = PTR_ERR(ce);
+			goto out;
+		}
+
+		rq = intel_context_create_request(ce);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			intel_context_put(ce);
+			goto out;
+		}
+
+		err = intel_engine_emit_ctx_wa(rq);
+		if (err)
+			goto err_rq;
+
+		err = intel_renderstate_emit(&so, rq);
+		if (err)
+			goto err_rq;
+
+err_rq:
+		requests[id] = i915_request_get(rq);
+		i915_request_add(rq);
+		intel_renderstate_fini(&so);
+		if (err)
+			goto out;
+	}
+
+	/* Flush the default context image to memory, and enable powersaving. */
+	if (intel_gt_wait_for_idle(gt, I915_GEM_IDLE_TIMEOUT) == -ETIME) {
+		err = -EIO;
+		goto out;
+	}
+
+	for (id = 0; id < ARRAY_SIZE(requests); id++) {
+		struct i915_request *rq;
+		struct i915_vma *state;
+		void *vaddr;
+
+		rq = requests[id];
+		if (!rq)
+			continue;
+
+		GEM_BUG_ON(!test_bit(CONTEXT_ALLOC_BIT, &rq->context->flags));
+		state = rq->context->state;
+		if (!state)
+			continue;
+
+		/* Serialise with retirement on another CPU */
+		GEM_BUG_ON(!i915_request_completed(rq));
+		err = __intel_context_flush_retire(rq->context);
+		if (err)
+			goto out;
+
+		/* We want to be able to unbind the state from the GGTT */
+		GEM_BUG_ON(intel_context_is_pinned(rq->context));
+
+		/*
+		 * As we will hold a reference to the logical state, it will
+		 * not be torn down with the context, and importantly the
+		 * object will hold onto its vma (making it possible for a
+		 * stray GTT write to corrupt our defaults). Unmap the vma
+		 * from the GTT to prevent such accidents and reclaim the
+		 * space.
+		 */
+		err = i915_vma_unbind(state);
+		if (err)
+			goto out;
+
+		i915_gem_object_lock(state->obj);
+		err = i915_gem_object_set_to_cpu_domain(state->obj, false);
+		i915_gem_object_unlock(state->obj);
+		if (err)
+			goto out;
+
+		i915_gem_object_set_cache_coherency(state->obj, I915_CACHE_LLC);
+
+		/* Check we can acquire the image of the context state */
+		vaddr = i915_gem_object_pin_map(state->obj, I915_MAP_FORCE_WB);
+		if (IS_ERR(vaddr)) {
+			err = PTR_ERR(vaddr);
+			goto out;
+		}
+
+		rq->engine->default_state = i915_gem_object_get(state->obj);
+		i915_gem_object_unpin_map(state->obj);
+	}
+
+out:
+	/*
+	 * If we have to abandon now, we expect the engines to be idle
+	 * and ready to be torn-down. The quickest way we can accomplish
+	 * this is by declaring ourselves wedged.
+	 */
+	if (err)
+		intel_gt_set_wedged(gt);
+
+	for (id = 0; id < ARRAY_SIZE(requests); id++) {
+		struct intel_context *ce;
+		struct i915_request *rq;
+
+		rq = requests[id];
+		if (!rq)
+			continue;
+
+		ce = rq->context;
+		i915_request_put(rq);
+		intel_context_put(ce);
+	}
+	return err;
+}
+
+static int __engines_verify_workarounds(struct intel_gt *gt)
+{
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	if (!IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
+		return 0;
+
+	for_each_engine(engine, gt, id) {
+		if (intel_engine_verify_workarounds(engine, "load"))
+			err = -EIO;
+	}
+
+	return err;
+}
+
+static void __intel_gt_disable(struct intel_gt *gt)
+{
+	intel_gt_set_wedged_on_init(gt);
+
+	intel_gt_suspend_prepare(gt);
+	intel_gt_suspend_late(gt);
+
+	GEM_BUG_ON(intel_gt_pm_is_awake(gt));
+}
+
 int intel_gt_init(struct intel_gt *gt)
 {
 	int err;
 
-	err = intel_gt_init_scratch(gt, IS_GEN(gt->i915, 2) ? SZ_256K : SZ_4K);
+	err = i915_inject_probe_error(gt->i915, -ENODEV);
 	if (err)
 		return err;
 
+	/*
+	 * This is just a security blanket to placate dragons.
+	 * On some systems, we very sporadically observe that the first TLBs
+	 * used by the CS may be stale, despite us poking the TLB reset. If
+	 * we hold the forcewake during initialisation these problems
+	 * just magically go away.
+	 */
+	intel_uncore_forcewake_get(gt->uncore, FORCEWAKE_ALL);
+
+	err = intel_gt_init_scratch(gt, IS_GEN(gt->i915, 2) ? SZ_256K : SZ_4K);
+	if (err)
+		goto out_fw;
+
 	intel_gt_pm_init(gt);
 
-	return 0;
+	gt->vm = kernel_vm(gt);
+	if (!gt->vm) {
+		err = -ENOMEM;
+		goto err_pm;
+	}
+
+	err = intel_engines_init(gt);
+	if (err)
+		goto err_engines;
+
+	intel_uc_init(&gt->uc);
+
+	err = intel_gt_resume(gt);
+	if (err)
+		goto err_uc_init;
+
+	err = __engines_record_defaults(gt);
+	if (err)
+		goto err_gt;
+
+	err = __engines_verify_workarounds(gt);
+	if (err)
+		goto err_gt;
+
+	err = i915_inject_probe_error(gt->i915, -EIO);
+	if (err)
+		goto err_gt;
+
+	goto out_fw;
+err_gt:
+	__intel_gt_disable(gt);
+	intel_uc_fini_hw(&gt->uc);
+err_uc_init:
+	intel_uc_fini(&gt->uc);
+err_engines:
+	intel_engines_release(gt);
+	i915_vm_put(fetch_and_zero(&gt->vm));
+err_pm:
+	intel_gt_pm_fini(gt);
+	intel_gt_fini_scratch(gt);
+out_fw:
+	if (err)
+		intel_gt_set_wedged_on_init(gt);
+	intel_uncore_forcewake_put(gt->uncore, FORCEWAKE_ALL);
+	return err;
 }
 
 void intel_gt_driver_remove(struct intel_gt *gt)
 {
-	GEM_BUG_ON(gt->awake);
+	__intel_gt_disable(gt);
+
+	intel_uc_fini_hw(&gt->uc);
+	intel_uc_fini(&gt->uc);
+
+	intel_engines_release(gt);
 }
 
 void intel_gt_driver_unregister(struct intel_gt *gt)
@@ -389,6 +645,12 @@ void intel_gt_driver_unregister(struct intel_gt *gt)
 
 void intel_gt_driver_release(struct intel_gt *gt)
 {
+	struct i915_address_space *vm;
+
+	vm = fetch_and_zero(&gt->vm);
+	if (vm) /* FIXME being called twice on error paths :( */
+		i915_vm_put(vm);
+
 	intel_gt_pm_fini(gt);
 	intel_gt_fini_scratch(gt);
 }
@@ -396,5 +658,8 @@ void intel_gt_driver_release(struct intel_gt *gt)
 void intel_gt_driver_late_release(struct intel_gt *gt)
 {
 	intel_uc_driver_late_release(&gt->uc);
+	intel_gt_fini_requests(gt);
 	intel_gt_fini_reset(gt);
+	intel_gt_fini_timelines(gt);
+	intel_engines_free(gt);
 }
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.h b/drivers/gpu/drm/i915/gt/intel_gt.h
index 5436f8c30708..2355cf129e9c 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt.h
@@ -12,6 +12,12 @@
 
 struct drm_i915_private;
 
+#define GT_TRACE(gt, fmt, ...) do {					\
+	const struct intel_gt *gt__ __maybe_unused = (gt);		\
+	GEM_TRACE("%s  " fmt, dev_name(gt__->i915->drm.dev),		\
+		  ##__VA_ARGS__);					\
+} while (0)
+
 static inline struct intel_gt *uc_to_gt(struct intel_uc *uc)
 {
 	return container_of(uc, struct intel_gt, uc);
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_irq.c b/drivers/gpu/drm/i915/gt/intel_gt_irq.c
index 973ee7eded64..f796bdf1ed30 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_irq.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt_irq.c
@@ -28,7 +28,7 @@ cs_irq_handler(struct intel_engine_cs *engine, u32 iir)
 		tasklet = true;
 
 	if (iir & GT_RENDER_USER_INTERRUPT) {
-		intel_engine_breadcrumbs_irq(engine);
+		intel_engine_signal_breadcrumbs(engine);
 		tasklet |= intel_engine_needs_breadcrumb_tasklet(engine);
 	}
 
@@ -245,9 +245,9 @@ void gen11_gt_irq_postinstall(struct intel_gt *gt)
 void gen5_gt_irq_handler(struct intel_gt *gt, u32 gt_iir)
 {
 	if (gt_iir & GT_RENDER_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(gt->engine_class[RENDER_CLASS][0]);
+		intel_engine_signal_breadcrumbs(gt->engine_class[RENDER_CLASS][0]);
 	if (gt_iir & ILK_BSD_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(gt->engine_class[VIDEO_DECODE_CLASS][0]);
+		intel_engine_signal_breadcrumbs(gt->engine_class[VIDEO_DECODE_CLASS][0]);
 }
 
 static void gen7_parity_error_irq_handler(struct intel_gt *gt, u32 iir)
@@ -271,11 +271,11 @@ static void gen7_parity_error_irq_handler(struct intel_gt *gt, u32 iir)
 void gen6_gt_irq_handler(struct intel_gt *gt, u32 gt_iir)
 {
 	if (gt_iir & GT_RENDER_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(gt->engine_class[RENDER_CLASS][0]);
+		intel_engine_signal_breadcrumbs(gt->engine_class[RENDER_CLASS][0]);
 	if (gt_iir & GT_BSD_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(gt->engine_class[VIDEO_DECODE_CLASS][0]);
+		intel_engine_signal_breadcrumbs(gt->engine_class[VIDEO_DECODE_CLASS][0]);
 	if (gt_iir & GT_BLT_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(gt->engine_class[COPY_ENGINE_CLASS][0]);
+		intel_engine_signal_breadcrumbs(gt->engine_class[COPY_ENGINE_CLASS][0]);
 
 	if (gt_iir & (GT_BLT_CS_ERROR_INTERRUPT |
 		      GT_BSD_CS_ERROR_INTERRUPT |
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_pm.c b/drivers/gpu/drm/i915/gt/intel_gt_pm.c
index a459a42ad5c2..45b68a17da4d 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_pm.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt_pm.c
@@ -43,7 +43,7 @@ static int __gt_unpark(struct intel_wakeref *wf)
 	struct intel_gt *gt = container_of(wf, typeof(*gt), wakeref);
 	struct drm_i915_private *i915 = gt->i915;
 
-	GEM_TRACE("\n");
+	GT_TRACE(gt, "\n");
 
 	i915_globals_unpark();
 
@@ -61,9 +61,7 @@ static int __gt_unpark(struct intel_wakeref *wf)
 	gt->awake = intel_display_power_get(i915, POWER_DOMAIN_GT_IRQ);
 	GEM_BUG_ON(!gt->awake);
 
-	if (NEEDS_RC6_CTX_CORRUPTION_WA(i915))
-		intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
-
+	intel_rc6_unpark(&gt->rc6);
 	intel_rps_unpark(&gt->rps);
 	i915_pmu_gt_unparked(i915);
 
@@ -78,24 +76,21 @@ static int __gt_park(struct intel_wakeref *wf)
 	intel_wakeref_t wakeref = fetch_and_zero(&gt->awake);
 	struct drm_i915_private *i915 = gt->i915;
 
-	GEM_TRACE("\n");
+	GT_TRACE(gt, "\n");
 
 	intel_gt_park_requests(gt);
 
 	i915_vma_parked(gt);
 	i915_pmu_gt_parked(i915);
 	intel_rps_park(&gt->rps);
+	intel_rc6_park(&gt->rc6);
 
 	/* Everything switched off, flush any residual interrupt just in case */
 	intel_synchronize_irq(i915);
 
-	if (NEEDS_RC6_CTX_CORRUPTION_WA(i915)) {
-		intel_rc6_ctx_wa_check(&i915->gt.rc6);
-		intel_uncore_forcewake_put(&i915->uncore, FORCEWAKE_ALL);
-	}
-
+	/* Defer dropping the display power well for 100ms, it's slow! */
 	GEM_BUG_ON(!wakeref);
-	intel_display_power_put(i915, POWER_DOMAIN_GT_IRQ, wakeref);
+	intel_display_power_put_async(i915, POWER_DOMAIN_GT_IRQ, wakeref);
 
 	i915_globals_park();
 
@@ -147,7 +142,7 @@ void intel_gt_sanitize(struct intel_gt *gt, bool force)
 	enum intel_engine_id id;
 	intel_wakeref_t wakeref;
 
-	GEM_TRACE("force:%s\n", yesno(force));
+	GT_TRACE(gt, "force:%s", yesno(force));
 
 	/* Use a raw wakeref to avoid calling intel_display_power_get early */
 	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
@@ -192,9 +187,9 @@ int intel_gt_resume(struct intel_gt *gt)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	int err = 0;
+	int err;
 
-	GEM_TRACE("\n");
+	GT_TRACE(gt, "\n");
 
 	/*
 	 * After resume, we may need to poke into the pinned kernel
@@ -207,6 +202,15 @@ int intel_gt_resume(struct intel_gt *gt)
 	intel_uncore_forcewake_get(gt->uncore, FORCEWAKE_ALL);
 	intel_rc6_sanitize(&gt->rc6);
 
+	/* Only when the HW is re-initialised, can we replay the requests */
+	err = intel_gt_init_hw(gt);
+	if (err) {
+		dev_err(gt->i915->drm.dev,
+			"Failed to initialize GPU, declaring it wedged!\n");
+		intel_gt_set_wedged(gt);
+		goto err_fw;
+	}
+
 	intel_rps_enable(&gt->rps);
 	intel_llc_enable(&gt->llc);
 
@@ -239,6 +243,7 @@ int intel_gt_resume(struct intel_gt *gt)
 
 	user_forcewake(gt, false);
 
+err_fw:
 	intel_uncore_forcewake_put(gt->uncore, FORCEWAKE_ALL);
 	intel_gt_pm_put(gt);
 
@@ -256,6 +261,7 @@ static void wait_for_suspend(struct intel_gt *gt)
 		 * the gpu quiet.
 		 */
 		intel_gt_set_wedged(gt);
+		intel_gt_retire_requests(gt);
 	}
 
 	intel_gt_pm_wait_for_idle(gt);
@@ -285,6 +291,11 @@ void intel_gt_suspend_late(struct intel_gt *gt)
 	/* We expect to be idle already; but also want to be independent */
 	wait_for_suspend(gt);
 
+	if (is_mock_gt(gt))
+		return;
+
+	GEM_BUG_ON(gt->awake);
+
 	/*
 	 * On disabling the device, we want to turn off HW access to memory
 	 * that we no longer own.
@@ -306,20 +317,19 @@ void intel_gt_suspend_late(struct intel_gt *gt)
 
 	intel_gt_sanitize(gt, false);
 
-	GEM_TRACE("\n");
+	GT_TRACE(gt, "\n");
 }
 
 void intel_gt_runtime_suspend(struct intel_gt *gt)
 {
 	intel_uc_runtime_suspend(&gt->uc);
 
-	GEM_TRACE("\n");
+	GT_TRACE(gt, "\n");
 }
 
 int intel_gt_runtime_resume(struct intel_gt *gt)
 {
-	GEM_TRACE("\n");
-
+	GT_TRACE(gt, "\n");
 	intel_gt_init_swizzling(gt);
 
 	return intel_uc_runtime_resume(&gt->uc);
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_pm.h b/drivers/gpu/drm/i915/gt/intel_gt_pm.h
index 990efc27a4e4..4a9e48c12bd4 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_pm.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_pm.h
@@ -22,6 +22,11 @@ static inline void intel_gt_pm_get(struct intel_gt *gt)
 	intel_wakeref_get(&gt->wakeref);
 }
 
+static inline void __intel_gt_pm_get(struct intel_gt *gt)
+{
+	__intel_wakeref_get(&gt->wakeref);
+}
+
 static inline bool intel_gt_pm_get_if_awake(struct intel_gt *gt)
 {
 	return intel_wakeref_get_if_active(&gt->wakeref);
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_requests.c b/drivers/gpu/drm/i915/gt/intel_gt_requests.c
index 3dc13ecf41bf..b4f04614230e 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_requests.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt_requests.c
@@ -8,6 +8,7 @@
 
 #include "i915_drv.h" /* for_each_engine() */
 #include "i915_request.h"
+#include "intel_engine_heartbeat.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
 #include "intel_gt_requests.h"
@@ -22,13 +23,18 @@ static void retire_requests(struct intel_timeline *tl)
 			break;
 }
 
-static void flush_submission(struct intel_gt *gt)
+static bool flush_submission(struct intel_gt *gt)
 {
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
+	bool active = false;
 
-	for_each_engine(engine, gt, id)
-		intel_engine_flush_submission(engine);
+	for_each_engine(engine, gt, id) {
+		active |= intel_engine_flush_submission(engine);
+		active |= flush_work(&engine->retire_work);
+	}
+
+	return active;
 }
 
 static void engine_retire(struct work_struct *work)
@@ -62,19 +68,16 @@ static void engine_retire(struct work_struct *work)
 static bool add_retire(struct intel_engine_cs *engine,
 		       struct intel_timeline *tl)
 {
+#define STUB ((struct intel_timeline *)1)
 	struct intel_timeline *first;
 
 	/*
 	 * We open-code a llist here to include the additional tag [BIT(0)]
 	 * so that we know when the timeline is already on a
 	 * retirement queue: either this engine or another.
-	 *
-	 * However, we rely on that a timeline can only be active on a single
-	 * engine at any one time and that add_retire() is called before the
-	 * engine releases the timeline and transferred to another to retire.
 	 */
 
-	if (READ_ONCE(tl->retire)) /* already queued */
+	if (cmpxchg(&tl->retire, NULL, STUB)) /* already queued */
 		return false;
 
 	intel_timeline_get(tl);
@@ -109,7 +112,6 @@ long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout)
 	struct intel_gt_timelines *timelines = &gt->timelines;
 	struct intel_timeline *tl, *tn;
 	unsigned long active_count = 0;
-	unsigned long flags;
 	bool interruptible;
 	LIST_HEAD(free);
 
@@ -119,7 +121,7 @@ long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout)
 
 	flush_submission(gt); /* kick the ksoftirqd tasklets */
 
-	spin_lock_irqsave(&timelines->lock, flags);
+	spin_lock(&timelines->lock);
 	list_for_each_entry_safe(tl, tn, &timelines->active_list, link) {
 		if (!mutex_trylock(&tl->mutex)) {
 			active_count++; /* report busy to caller, try again? */
@@ -129,7 +131,7 @@ long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout)
 		intel_timeline_get(tl);
 		GEM_BUG_ON(!atomic_read(&tl->active_count));
 		atomic_inc(&tl->active_count); /* pin the list element */
-		spin_unlock_irqrestore(&timelines->lock, flags);
+		spin_unlock(&timelines->lock);
 
 		if (timeout > 0) {
 			struct dma_fence *fence;
@@ -145,14 +147,14 @@ long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout)
 
 		retire_requests(tl);
 
-		spin_lock_irqsave(&timelines->lock, flags);
+		spin_lock(&timelines->lock);
 
 		/* Resume iteration after dropping lock */
 		list_safe_reset_next(tl, tn, link);
 		if (atomic_dec_and_test(&tl->active_count))
 			list_del(&tl->link);
 		else
-			active_count += !!rcu_access_pointer(tl->last_request.fence);
+			active_count += i915_active_fence_isset(&tl->last_request);
 
 		mutex_unlock(&tl->mutex);
 
@@ -162,11 +164,14 @@ long intel_gt_retire_requests_timeout(struct intel_gt *gt, long timeout)
 			list_add(&tl->link, &free);
 		}
 	}
-	spin_unlock_irqrestore(&timelines->lock, flags);
+	spin_unlock(&timelines->lock);
 
 	list_for_each_entry_safe(tl, tn, &free, link)
 		__intel_timeline_free(&tl->kref);
 
+	if (flush_submission(gt))
+		active_count++;
+
 	return active_count ? timeout : 0;
 }
 
@@ -190,9 +195,9 @@ static void retire_work_handler(struct work_struct *work)
 	struct intel_gt *gt =
 		container_of(work, typeof(*gt), requests.retire_work.work);
 
-	intel_gt_retire_requests(gt);
 	schedule_delayed_work(&gt->requests.retire_work,
 			      round_jiffies_up_relative(HZ));
+	intel_gt_retire_requests(gt);
 }
 
 void intel_gt_init_requests(struct intel_gt *gt)
@@ -210,3 +215,9 @@ void intel_gt_unpark_requests(struct intel_gt *gt)
 	schedule_delayed_work(&gt->requests.retire_work,
 			      round_jiffies_up_relative(HZ));
 }
+
+void intel_gt_fini_requests(struct intel_gt *gt)
+{
+	/* Wait until the work is marked as finished before unloading! */
+	cancel_delayed_work_sync(&gt->requests.retire_work);
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_requests.h b/drivers/gpu/drm/i915/gt/intel_gt_requests.h
index d626fb115386..dbac53baf1cb 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_requests.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_requests.h
@@ -27,5 +27,6 @@ int intel_gt_wait_for_idle(struct intel_gt *gt, long timeout);
 void intel_gt_init_requests(struct intel_gt *gt);
 void intel_gt_park_requests(struct intel_gt *gt);
 void intel_gt_unpark_requests(struct intel_gt *gt);
+void intel_gt_fini_requests(struct intel_gt *gt);
 
 #endif /* INTEL_GT_REQUESTS_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_types.h b/drivers/gpu/drm/i915/gt/intel_gt_types.h
index d4e14dbd172e..96890dd12b5f 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_types.h
@@ -90,6 +90,13 @@ struct intel_gt {
 	struct intel_engine_cs *engine[I915_NUM_ENGINES];
 	struct intel_engine_cs *engine_class[MAX_ENGINE_CLASS + 1]
 					    [MAX_ENGINE_INSTANCE + 1];
+
+	/*
+	 * Default address space (either GGTT or ppGTT depending on arch).
+	 *
+	 * Reserved for exclusive use by the kernel.
+	 */
+	struct i915_address_space *vm;
 };
 
 enum intel_gt_scratch_field {
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index 75dd0e0367b7..4fb70a7716e3 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -133,12 +133,11 @@
  */
 #include <linux/interrupt.h>
 
-#include "gem/i915_gem_context.h"
-
 #include "i915_drv.h"
 #include "i915_perf.h"
 #include "i915_trace.h"
 #include "i915_vgpu.h"
+#include "intel_context.h"
 #include "intel_engine_pm.h"
 #include "intel_gt.h"
 #include "intel_gt_pm.h"
@@ -880,7 +879,7 @@ __unwind_incomplete_requests(struct intel_engine_cs *engine)
 			list_move(&rq->sched.link, pl);
 			active = rq;
 		} else {
-			struct intel_engine_cs *owner = rq->hw_context->engine;
+			struct intel_engine_cs *owner = rq->context->engine;
 
 			/*
 			 * Decouple the virtual breadcrumb before moving it
@@ -983,6 +982,58 @@ static void intel_engine_context_out(struct intel_engine_cs *engine)
 	write_sequnlock_irqrestore(&engine->stats.lock, flags);
 }
 
+static int lrc_ring_mi_mode(const struct intel_engine_cs *engine)
+{
+	if (INTEL_GEN(engine->i915) >= 12)
+		return 0x60;
+	else if (INTEL_GEN(engine->i915) >= 9)
+		return 0x54;
+	else if (engine->class == RENDER_CLASS)
+		return 0x58;
+	else
+		return -1;
+}
+
+static void
+execlists_check_context(const struct intel_context *ce,
+			const struct intel_engine_cs *engine)
+{
+	const struct intel_ring *ring = ce->ring;
+	u32 *regs = ce->lrc_reg_state;
+	bool valid = true;
+	int x;
+
+	if (regs[CTX_RING_START] != i915_ggtt_offset(ring->vma)) {
+		pr_err("%s: context submitted with incorrect RING_START [%08x], expected %08x\n",
+		       engine->name,
+		       regs[CTX_RING_START],
+		       i915_ggtt_offset(ring->vma));
+		regs[CTX_RING_START] = i915_ggtt_offset(ring->vma);
+		valid = false;
+	}
+
+	if ((regs[CTX_RING_CTL] & ~(RING_WAIT | RING_WAIT_SEMAPHORE)) !=
+	    (RING_CTL_SIZE(ring->size) | RING_VALID)) {
+		pr_err("%s: context submitted with incorrect RING_CTL [%08x], expected %08x\n",
+		       engine->name,
+		       regs[CTX_RING_CTL],
+		       (u32)(RING_CTL_SIZE(ring->size) | RING_VALID));
+		regs[CTX_RING_CTL] = RING_CTL_SIZE(ring->size) | RING_VALID;
+		valid = false;
+	}
+
+	x = lrc_ring_mi_mode(engine);
+	if (x != -1 && regs[x + 1] & (regs[x + 1] >> 16) & STOP_RING) {
+		pr_err("%s: context submitted with STOP_RING [%08x] in RING_MI_MODE\n",
+		       engine->name, regs[x + 1]);
+		regs[x + 1] &= ~STOP_RING;
+		regs[x + 1] |= STOP_RING << 16;
+		valid = false;
+	}
+
+	WARN_ONCE(!valid, "Invalid lrc state found before submission\n");
+}
+
 static void restore_default_state(struct intel_context *ce,
 				  struct intel_engine_cs *engine)
 {
@@ -999,7 +1050,7 @@ static void restore_default_state(struct intel_context *ce,
 static void reset_active(struct i915_request *rq,
 			 struct intel_engine_cs *engine)
 {
-	struct intel_context * const ce = rq->hw_context;
+	struct intel_context * const ce = rq->context;
 	u32 head;
 
 	/*
@@ -1017,8 +1068,8 @@ static void reset_active(struct i915_request *rq,
 	 * remain correctly ordered. And we defer to __i915_request_submit()
 	 * so that all asynchronous waits are correctly handled.
 	 */
-	GEM_TRACE("%s(%s): { rq=%llx:%lld }\n",
-		  __func__, engine->name, rq->fence.context, rq->fence.seqno);
+	ENGINE_TRACE(engine, "{ rq=%llx:%lld }\n",
+		     rq->fence.context, rq->fence.seqno);
 
 	/* On resubmission of the active request, payload will be scrubbed */
 	if (i915_request_completed(rq))
@@ -1040,13 +1091,16 @@ static inline struct intel_engine_cs *
 __execlists_schedule_in(struct i915_request *rq)
 {
 	struct intel_engine_cs * const engine = rq->engine;
-	struct intel_context * const ce = rq->hw_context;
+	struct intel_context * const ce = rq->context;
 
 	intel_context_get(ce);
 
-	if (unlikely(i915_gem_context_is_banned(ce->gem_context)))
+	if (unlikely(intel_context_is_banned(ce)))
 		reset_active(rq, engine);
 
+	if (IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
+		execlists_check_context(ce, engine);
+
 	if (ce->tag) {
 		/* Use a fixed tag for OA and friends */
 		ce->lrc_desc |= (u64)ce->tag << 32;
@@ -1059,7 +1113,7 @@ __execlists_schedule_in(struct i915_request *rq)
 		BUILD_BUG_ON(NUM_CONTEXT_TAG > GEN12_MAX_CONTEXT_HW_ID);
 	}
 
-	intel_gt_pm_get(engine->gt);
+	__intel_gt_pm_get(engine->gt);
 	execlists_context_status_change(rq, INTEL_CONTEXT_SCHEDULE_IN);
 	intel_engine_context_in(engine);
 
@@ -1069,7 +1123,7 @@ __execlists_schedule_in(struct i915_request *rq)
 static inline struct i915_request *
 execlists_schedule_in(struct i915_request *rq, int idx)
 {
-	struct intel_context * const ce = rq->hw_context;
+	struct intel_context * const ce = rq->context;
 	struct intel_engine_cs *old;
 
 	GEM_BUG_ON(!intel_engine_pm_is_awake(rq->engine));
@@ -1100,7 +1154,7 @@ static inline void
 __execlists_schedule_out(struct i915_request *rq,
 			 struct intel_engine_cs * const engine)
 {
-	struct intel_context * const ce = rq->hw_context;
+	struct intel_context * const ce = rq->context;
 
 	/*
 	 * NB process_csb() is not under the engine->active.lock and hence
@@ -1138,7 +1192,7 @@ __execlists_schedule_out(struct i915_request *rq,
 static inline void
 execlists_schedule_out(struct i915_request *rq)
 {
-	struct intel_context * const ce = rq->hw_context;
+	struct intel_context * const ce = rq->context;
 	struct intel_engine_cs *cur, *old;
 
 	trace_i915_request_out(rq);
@@ -1155,7 +1209,7 @@ execlists_schedule_out(struct i915_request *rq)
 
 static u64 execlists_update_context(struct i915_request *rq)
 {
-	struct intel_context *ce = rq->hw_context;
+	struct intel_context *ce = rq->context;
 	u64 desc = ce->lrc_desc;
 	u32 tail;
 
@@ -1186,13 +1240,8 @@ static u64 execlists_update_context(struct i915_request *rq)
 	 * may not be visible to the HW prior to the completion of the UC
 	 * register write and that we may begin execution from the context
 	 * before its image is complete leading to invalid PD chasing.
-	 *
-	 * Furthermore, Braswell, at least, wants a full mb to be sure that
-	 * the writes are coherent in memory (visible to the GPU) prior to
-	 * execution, and not just visible to other CPUs (as is the result of
-	 * wmb).
 	 */
-	mb();
+	wmb();
 
 	/* Wa_1607138340:tgl */
 	if (IS_TGL_REVID(rq->i915, TGL_REVID_A0, TGL_REVID_A0))
@@ -1224,15 +1273,14 @@ trace_ports(const struct intel_engine_execlists *execlists,
 	if (!ports[0])
 		return;
 
-	GEM_TRACE("%s: %s { %llx:%lld%s, %llx:%lld }\n",
-		  engine->name, msg,
-		  ports[0]->fence.context,
-		  ports[0]->fence.seqno,
-		  i915_request_completed(ports[0]) ? "!" :
-		  i915_request_started(ports[0]) ? "*" :
-		  "",
-		  ports[1] ? ports[1]->fence.context : 0,
-		  ports[1] ? ports[1]->fence.seqno : 0);
+	ENGINE_TRACE(engine, "%s { %llx:%lld%s, %llx:%lld }\n", msg,
+		     ports[0]->fence.context,
+		     ports[0]->fence.seqno,
+		     i915_request_completed(ports[0]) ? "!" :
+		     i915_request_started(ports[0]) ? "*" :
+		     "",
+		     ports[1] ? ports[1]->fence.context : 0,
+		     ports[1] ? ports[1]->fence.seqno : 0);
 }
 
 static __maybe_unused bool
@@ -1256,33 +1304,56 @@ assert_pending_valid(const struct intel_engine_execlists *execlists,
 	}
 
 	for (port = execlists->pending; (rq = *port); port++) {
-		if (ce == rq->hw_context) {
-			GEM_TRACE_ERR("Duplicate context in pending[%zd]\n",
+		unsigned long flags;
+		bool ok = true;
+
+		GEM_BUG_ON(!kref_read(&rq->fence.refcount));
+		GEM_BUG_ON(!i915_request_is_active(rq));
+
+		if (ce == rq->context) {
+			GEM_TRACE_ERR("Dup context:%llx in pending[%zd]\n",
+				      ce->timeline->fence_context,
 				      port - execlists->pending);
 			return false;
 		}
+		ce = rq->context;
 
-		ce = rq->hw_context;
-		if (i915_request_completed(rq))
+		/* Hold tightly onto the lock to prevent concurrent retires! */
+		if (!spin_trylock_irqsave(&rq->lock, flags))
 			continue;
 
-		if (i915_active_is_idle(&ce->active)) {
-			GEM_TRACE_ERR("Inactive context in pending[%zd]\n",
+		if (i915_request_completed(rq))
+			goto unlock;
+
+		if (i915_active_is_idle(&ce->active) &&
+		    !intel_context_is_barrier(ce)) {
+			GEM_TRACE_ERR("Inactive context:%llx in pending[%zd]\n",
+				      ce->timeline->fence_context,
 				      port - execlists->pending);
-			return false;
+			ok = false;
+			goto unlock;
 		}
 
 		if (!i915_vma_is_pinned(ce->state)) {
-			GEM_TRACE_ERR("Unpinned context in pending[%zd]\n",
+			GEM_TRACE_ERR("Unpinned context:%llx in pending[%zd]\n",
+				      ce->timeline->fence_context,
 				      port - execlists->pending);
-			return false;
+			ok = false;
+			goto unlock;
 		}
 
 		if (!i915_vma_is_pinned(ce->ring->vma)) {
-			GEM_TRACE_ERR("Unpinned ringbuffer in pending[%zd]\n",
+			GEM_TRACE_ERR("Unpinned ring:%llx in pending[%zd]\n",
+				      ce->timeline->fence_context,
 				      port - execlists->pending);
-			return false;
+			ok = false;
+			goto unlock;
 		}
+
+unlock:
+		spin_unlock_irqrestore(&rq->lock, flags);
+		if (!ok)
+			return false;
 	}
 
 	return ce;
@@ -1327,7 +1398,7 @@ static void execlists_submit_ports(struct intel_engine_cs *engine)
 static bool ctx_single_port_submission(const struct intel_context *ce)
 {
 	return (IS_ENABLED(CONFIG_DRM_I915_GVT) &&
-		i915_gem_context_force_single_submission(ce->gem_context));
+		intel_context_force_single_submission(ce));
 }
 
 static bool can_merge_ctx(const struct intel_context *prev,
@@ -1363,7 +1434,7 @@ static bool can_merge_rq(const struct i915_request *prev,
 		     (I915_REQUEST_NOPREEMPT | I915_REQUEST_SENTINEL)))
 		return false;
 
-	if (!can_merge_ctx(prev->hw_context, next->hw_context))
+	if (!can_merge_ctx(prev->context, next->context))
 		return false;
 
 	return true;
@@ -1411,7 +1482,7 @@ static void virtual_xfer_breadcrumbs(struct virtual_engine *ve,
 	if (!list_empty(&ve->context.signal_link)) {
 		list_move_tail(&ve->context.signal_link,
 			       &engine->breadcrumbs.signalers);
-		intel_engine_queue_breadcrumbs(engine);
+		intel_engine_signal_breadcrumbs(engine);
 	}
 	spin_unlock(&old->breadcrumbs.irq_lock);
 }
@@ -1550,7 +1621,7 @@ static unsigned long active_preempt_timeout(struct intel_engine_cs *engine)
 		return 0;
 
 	/* Force a fast reset for terminated contexts (ignoring sysfs!) */
-	if (unlikely(i915_gem_context_is_banned(rq->gem_context)))
+	if (unlikely(intel_context_is_banned(rq->context)))
 		return 1;
 
 	return READ_ONCE(engine->props.preempt_timeout_ms);
@@ -1627,12 +1698,12 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 	last = last_active(execlists);
 	if (last) {
 		if (need_preempt(engine, last, rb)) {
-			GEM_TRACE("%s: preempting last=%llx:%lld, prio=%d, hint=%d\n",
-				  engine->name,
-				  last->fence.context,
-				  last->fence.seqno,
-				  last->sched.attr.priority,
-				  execlists->queue_priority_hint);
+			ENGINE_TRACE(engine,
+				     "preempting last=%llx:%lld, prio=%d, hint=%d\n",
+				     last->fence.context,
+				     last->fence.seqno,
+				     last->sched.attr.priority,
+				     execlists->queue_priority_hint);
 			record_preemption(execlists);
 
 			/*
@@ -1658,16 +1729,16 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 			 * tendency to ignore us rewinding the TAIL to the
 			 * end of an earlier request.
 			 */
-			last->hw_context->lrc_desc |= CTX_DESC_FORCE_RESTORE;
+			last->context->lrc_desc |= CTX_DESC_FORCE_RESTORE;
 			last = NULL;
 		} else if (need_timeslice(engine, last) &&
 			   timer_expired(&engine->execlists.timer)) {
-			GEM_TRACE("%s: expired last=%llx:%lld, prio=%d, hint=%d\n",
-				  engine->name,
-				  last->fence.context,
-				  last->fence.seqno,
-				  last->sched.attr.priority,
-				  execlists->queue_priority_hint);
+			ENGINE_TRACE(engine,
+				     "expired last=%llx:%lld, prio=%d, hint=%d\n",
+				     last->fence.context,
+				     last->fence.seqno,
+				     last->sched.attr.priority,
+				     execlists->queue_priority_hint);
 
 			ring_set_paused(engine, 1);
 			defer_active(engine);
@@ -1730,7 +1801,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 
 		GEM_BUG_ON(rq != ve->request);
 		GEM_BUG_ON(rq->engine != &ve->base);
-		GEM_BUG_ON(rq->hw_context != &ve->context);
+		GEM_BUG_ON(rq->context != &ve->context);
 
 		if (rq_prio(rq) >= queue_prio(execlists)) {
 			if (!virtual_matches(ve, rq, engine)) {
@@ -1744,14 +1815,14 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				return; /* leave this for another */
 			}
 
-			GEM_TRACE("%s: virtual rq=%llx:%lld%s, new engine? %s\n",
-				  engine->name,
-				  rq->fence.context,
-				  rq->fence.seqno,
-				  i915_request_completed(rq) ? "!" :
-				  i915_request_started(rq) ? "*" :
-				  "",
-				  yesno(engine != ve->siblings[0]));
+			ENGINE_TRACE(engine,
+				     "virtual rq=%llx:%lld%s, new engine? %s\n",
+				     rq->fence.context,
+				     rq->fence.seqno,
+				     i915_request_completed(rq) ? "!" :
+				     i915_request_started(rq) ? "*" :
+				     "",
+				     yesno(engine != ve->siblings[0]));
 
 			ve->request = NULL;
 			ve->base.execlists.queue_priority_hint = INT_MIN;
@@ -1849,7 +1920,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				 * same LRCA, i.e. we must submit 2 different
 				 * contexts if we submit 2 ELSP.
 				 */
-				if (last->hw_context == rq->hw_context)
+				if (last->context == rq->context)
 					goto done;
 
 				if (i915_request_has_sentinel(last))
@@ -1862,8 +1933,8 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				 * the same context (even though a different
 				 * request) to the second port.
 				 */
-				if (ctx_single_port_submission(last->hw_context) ||
-				    ctx_single_port_submission(rq->hw_context))
+				if (ctx_single_port_submission(last->context) ||
+				    ctx_single_port_submission(rq->context))
 					goto done;
 
 				merge = false;
@@ -1877,8 +1948,8 @@ static void execlists_dequeue(struct intel_engine_cs *engine)
 				}
 
 				GEM_BUG_ON(last &&
-					   !can_merge_ctx(last->hw_context,
-							  rq->hw_context));
+					   !can_merge_ctx(last->context,
+							  rq->context));
 
 				submit = true;
 				last = rq;
@@ -1907,9 +1978,6 @@ done:
 	 * interrupt for secondary ports).
 	 */
 	execlists->queue_priority_hint = queue_prio(execlists);
-	GEM_TRACE("%s: queue_priority_hint:%d, submit:%s\n",
-		  engine->name, execlists->queue_priority_hint,
-		  yesno(submit));
 
 	if (submit) {
 		*port = execlists_schedule_in(last, port - execlists->pending);
@@ -2058,7 +2126,7 @@ static void process_csb(struct intel_engine_cs *engine)
 	 */
 	head = execlists->csb_head;
 	tail = READ_ONCE(*execlists->csb_write);
-	GEM_TRACE("%s cs-irq head=%d, tail=%d\n", engine->name, head, tail);
+	ENGINE_TRACE(engine, "cs-irq head=%d, tail=%d\n", head, tail);
 	if (unlikely(head == tail))
 		return;
 
@@ -2096,9 +2164,8 @@ static void process_csb(struct intel_engine_cs *engine)
 		 * status notifier.
 		 */
 
-		GEM_TRACE("%s csb[%d]: status=0x%08x:0x%08x\n",
-			  engine->name, head,
-			  buf[2 * head + 0], buf[2 * head + 1]);
+		ENGINE_TRACE(engine, "csb[%d]: status=0x%08x:0x%08x\n",
+			     head, buf[2 * head + 0], buf[2 * head + 1]);
 
 		if (INTEL_GEN(engine->i915) >= 12)
 			promote = gen12_csb_parse(execlists, buf + 2 * head);
@@ -2189,10 +2256,9 @@ static noinline void preempt_reset(struct intel_engine_cs *engine)
 	/* Mark this tasklet as disabled to avoid waiting for it to complete */
 	tasklet_disable_nosync(&engine->execlists.tasklet);
 
-	GEM_TRACE("%s: preempt timeout %lu+%ums\n",
-		  engine->name,
-		  READ_ONCE(engine->props.preempt_timeout_ms),
-		  jiffies_to_msecs(jiffies - engine->execlists.preempt.expires));
+	ENGINE_TRACE(engine, "preempt timeout %lu+%ums\n",
+		     READ_ONCE(engine->props.preempt_timeout_ms),
+		     jiffies_to_msecs(jiffies - engine->execlists.preempt.expires));
 	intel_engine_reset(engine, "preemption time out");
 
 	tasklet_enable(&engine->execlists.tasklet);
@@ -2369,7 +2435,7 @@ __execlists_update_reg_state(const struct intel_context *ce,
 	GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->head));
 	GEM_BUG_ON(!intel_ring_offset_valid(ring, ring->tail));
 
-	regs[CTX_RING_BUFFER_START] = i915_ggtt_offset(ring->vma);
+	regs[CTX_RING_START] = i915_ggtt_offset(ring->vma);
 	regs[CTX_RING_HEAD] = ring->head;
 	regs[CTX_RING_TAIL] = ring->tail;
 
@@ -2497,7 +2563,7 @@ static int execlists_request_alloc(struct i915_request *request)
 {
 	int ret;
 
-	GEM_BUG_ON(!intel_context_is_pinned(request->hw_context));
+	GEM_BUG_ON(!intel_context_is_pinned(request->context));
 
 	/*
 	 * Flush enough space to reduce the likelihood of waiting after
@@ -2898,8 +2964,8 @@ static void execlists_reset_prepare(struct intel_engine_cs *engine)
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	unsigned long flags;
 
-	GEM_TRACE("%s: depth<-%d\n", engine->name,
-		  atomic_read(&execlists->tasklet.count));
+	ENGINE_TRACE(engine, "depth<-%d\n",
+		     atomic_read(&execlists->tasklet.count));
 
 	/*
 	 * Prevent request submission to the hardware until we have
@@ -2952,22 +3018,18 @@ static void reset_csb_pointers(struct intel_engine_cs *engine)
 	WRITE_ONCE(*execlists->csb_write, reset_value);
 	wmb(); /* Make sure this is visible to HW (paranoia?) */
 
+	/*
+	 * Sometimes Icelake forgets to reset its pointers on a GPU reset.
+	 * Bludgeon them with a mmio update to be sure.
+	 */
+	ENGINE_WRITE(engine, RING_CONTEXT_STATUS_PTR,
+		     reset_value << 8 | reset_value);
+	ENGINE_POSTING_READ(engine, RING_CONTEXT_STATUS_PTR);
+
 	invalidate_csb_entries(&execlists->csb_status[0],
 			       &execlists->csb_status[reset_value]);
 }
 
-static int lrc_ring_mi_mode(const struct intel_engine_cs *engine)
-{
-	if (INTEL_GEN(engine->i915) >= 12)
-		return 0x60;
-	else if (INTEL_GEN(engine->i915) >= 9)
-		return 0x54;
-	else if (engine->class == RENDER_CLASS)
-		return 0x58;
-	else
-		return -1;
-}
-
 static void __execlists_reset_reg_state(const struct intel_context *ce,
 					const struct intel_engine_cs *engine)
 {
@@ -3008,7 +3070,7 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 	/* We still have requests in-flight; the engine should be active */
 	GEM_BUG_ON(!intel_engine_pm_is_awake(engine));
 
-	ce = rq->hw_context;
+	ce = rq->context;
 	GEM_BUG_ON(!i915_vma_is_pinned(ce->state));
 
 	if (i915_request_completed(rq)) {
@@ -3065,8 +3127,8 @@ static void __execlists_reset(struct intel_engine_cs *engine, bool stalled)
 	restore_default_state(ce, engine);
 
 out_replay:
-	GEM_TRACE("%s replay {head:%04x, tail:%04x}\n",
-		  engine->name, ce->ring->head, ce->ring->tail);
+	ENGINE_TRACE(engine, "replay {head:%04x, tail:%04x}\n",
+		     ce->ring->head, ce->ring->tail);
 	intel_ring_update_space(ce->ring);
 	__execlists_reset_reg_state(ce, engine);
 	__execlists_update_reg_state(ce, engine);
@@ -3078,11 +3140,11 @@ unwind:
 	__unwind_incomplete_requests(engine);
 }
 
-static void execlists_reset(struct intel_engine_cs *engine, bool stalled)
+static void execlists_reset_rewind(struct intel_engine_cs *engine, bool stalled)
 {
 	unsigned long flags;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	spin_lock_irqsave(&engine->active.lock, flags);
 
@@ -3096,14 +3158,14 @@ static void nop_submission_tasklet(unsigned long data)
 	/* The driver is wedged; don't process any more events. */
 }
 
-static void execlists_cancel_requests(struct intel_engine_cs *engine)
+static void execlists_reset_cancel(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_request *rq, *rn;
 	struct rb_node *rb;
 	unsigned long flags;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	/*
 	 * Before we call engine->cancel_requests(), we should have exclusive
@@ -3190,13 +3252,13 @@ static void execlists_reset_finish(struct intel_engine_cs *engine)
 	if (__tasklet_enable(&execlists->tasklet))
 		/* And kick in case we missed a new request submission. */
 		tasklet_hi_schedule(&execlists->tasklet);
-	GEM_TRACE("%s: depth->%d\n", engine->name,
-		  atomic_read(&execlists->tasklet.count));
+	ENGINE_TRACE(engine, "depth->%d\n",
+		     atomic_read(&execlists->tasklet.count));
 }
 
-static int gen8_emit_bb_start(struct i915_request *rq,
-			      u64 offset, u32 len,
-			      const unsigned int flags)
+static int gen8_emit_bb_start_noarb(struct i915_request *rq,
+				    u64 offset, u32 len,
+				    const unsigned int flags)
 {
 	u32 *cs;
 
@@ -3230,7 +3292,7 @@ static int gen8_emit_bb_start(struct i915_request *rq,
 	return 0;
 }
 
-static int gen9_emit_bb_start(struct i915_request *rq,
+static int gen8_emit_bb_start(struct i915_request *rq,
 			      u64 offset, u32 len,
 			      const unsigned int flags)
 {
@@ -3685,12 +3747,12 @@ static void execlists_park(struct intel_engine_cs *engine)
 void intel_execlists_set_default_submission(struct intel_engine_cs *engine)
 {
 	engine->submit_request = execlists_submit_request;
-	engine->cancel_requests = execlists_cancel_requests;
 	engine->schedule = i915_schedule;
 	engine->execlists.tasklet.func = execlists_submission_tasklet;
 
 	engine->reset.prepare = execlists_reset_prepare;
-	engine->reset.reset = execlists_reset;
+	engine->reset.rewind = execlists_reset_rewind;
+	engine->reset.cancel = execlists_reset_cancel;
 	engine->reset.finish = execlists_reset_finish;
 
 	engine->park = execlists_park;
@@ -3705,13 +3767,27 @@ void intel_execlists_set_default_submission(struct intel_engine_cs *engine)
 
 	if (INTEL_GEN(engine->i915) >= 12)
 		engine->flags |= I915_ENGINE_HAS_RELATIVE_MMIO;
+
+	if (intel_engine_has_preemption(engine))
+		engine->emit_bb_start = gen8_emit_bb_start;
+	else
+		engine->emit_bb_start = gen8_emit_bb_start_noarb;
+}
+
+static void execlists_shutdown(struct intel_engine_cs *engine)
+{
+	/* Synchronise with residual timers and any softirq they raise */
+	del_timer_sync(&engine->execlists.timer);
+	del_timer_sync(&engine->execlists.preempt);
+	tasklet_kill(&engine->execlists.tasklet);
 }
 
-static void execlists_destroy(struct intel_engine_cs *engine)
+static void execlists_release(struct intel_engine_cs *engine)
 {
+	execlists_shutdown(engine);
+
 	intel_engine_cleanup_common(engine);
 	lrc_destroy_wa_ctx(engine);
-	kfree(engine);
 }
 
 static void
@@ -3719,13 +3795,9 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
 {
 	/* Default vfuncs which can be overriden by each engine. */
 
-	engine->destroy = execlists_destroy;
+	engine->release = execlists_release;
 	engine->resume = execlists_resume;
 
-	engine->reset.prepare = execlists_reset_prepare;
-	engine->reset.reset = execlists_reset;
-	engine->reset.finish = execlists_reset_finish;
-
 	engine->cops = &execlists_context_ops;
 	engine->request_alloc = execlists_request_alloc;
 
@@ -3748,10 +3820,6 @@ logical_ring_default_vfuncs(struct intel_engine_cs *engine)
 		 * until a more refined solution exists.
 		 */
 	}
-	if (IS_GEN(engine->i915, 8))
-		engine->emit_bb_start = gen8_emit_bb_start;
-	else
-		engine->emit_bb_start = gen9_emit_bb_start;
 }
 
 static inline void
@@ -3795,6 +3863,11 @@ static void rcs_submission_override(struct intel_engine_cs *engine)
 
 int intel_execlists_submission_setup(struct intel_engine_cs *engine)
 {
+	struct intel_engine_execlists * const execlists = &engine->execlists;
+	struct drm_i915_private *i915 = engine->i915;
+	struct intel_uncore *uncore = engine->uncore;
+	u32 base = engine->mmio_base;
+
 	tasklet_init(&engine->execlists.tasklet,
 		     execlists_submission_tasklet, (unsigned long)engine);
 	timer_setup(&engine->execlists.timer, execlists_timeslice, 0);
@@ -3806,21 +3879,6 @@ int intel_execlists_submission_setup(struct intel_engine_cs *engine)
 	if (engine->class == RENDER_CLASS)
 		rcs_submission_override(engine);
 
-	return 0;
-}
-
-int intel_execlists_submission_init(struct intel_engine_cs *engine)
-{
-	struct intel_engine_execlists * const execlists = &engine->execlists;
-	struct drm_i915_private *i915 = engine->i915;
-	struct intel_uncore *uncore = engine->uncore;
-	u32 base = engine->mmio_base;
-	int ret;
-
-	ret = intel_engine_init_common(engine);
-	if (ret)
-		return ret;
-
 	if (intel_init_workaround_bb(engine))
 		/*
 		 * We continue even if we fail to initialize WA batch
@@ -3901,7 +3959,7 @@ static void init_common_reg_state(u32 * const regs,
 			_MASKED_BIT_DISABLE(CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT |
 					    CTX_CTRL_RS_CTX_ENABLE);
 
-	regs[CTX_RING_BUFFER_CONTROL] = RING_CTL_SIZE(ring->size) | RING_VALID;
+	regs[CTX_RING_CTL] = RING_CTL_SIZE(ring->size) | RING_VALID;
 	regs[CTX_BB_STATE] = RING_BB_PPGTT;
 }
 
@@ -4019,6 +4077,7 @@ populate_lr_context(struct intel_context *ce,
 
 		memcpy(vaddr, defaults, engine->context_size);
 		i915_gem_object_unpin_map(engine->default_state);
+		__set_bit(CONTEXT_VALID_BIT, &ce->flags);
 		inhibit = false;
 	}
 
@@ -4166,6 +4225,13 @@ static void virtual_engine_initial_hint(struct virtual_engine *ve)
 						ve->siblings[0]);
 }
 
+static int virtual_context_alloc(struct intel_context *ce)
+{
+	struct virtual_engine *ve = container_of(ce, typeof(*ve), context);
+
+	return __execlists_context_alloc(ce, ve->siblings[0]);
+}
+
 static int virtual_context_pin(struct intel_context *ce)
 {
 	struct virtual_engine *ve = container_of(ce, typeof(*ve), context);
@@ -4203,6 +4269,8 @@ static void virtual_context_exit(struct intel_context *ce)
 }
 
 static const struct intel_context_ops virtual_context_ops = {
+	.alloc = virtual_context_alloc,
+
 	.pin = virtual_context_pin,
 	.unpin = execlists_context_unpin,
 
@@ -4229,10 +4297,9 @@ static intel_engine_mask_t virtual_submission_mask(struct virtual_engine *ve)
 		mask = ve->siblings[0]->mask;
 	}
 
-	GEM_TRACE("%s: rq=%llx:%lld, mask=%x, prio=%d\n",
-		  ve->base.name,
-		  rq->fence.context, rq->fence.seqno,
-		  mask, ve->base.execlists.queue_priority_hint);
+	ENGINE_TRACE(&ve->base, "rq=%llx:%lld, mask=%x, prio=%d\n",
+		     rq->fence.context, rq->fence.seqno,
+		     mask, ve->base.execlists.queue_priority_hint);
 
 	return mask;
 }
@@ -4323,10 +4390,9 @@ static void virtual_submit_request(struct i915_request *rq)
 	struct i915_request *old;
 	unsigned long flags;
 
-	GEM_TRACE("%s: rq=%llx:%lld\n",
-		  ve->base.name,
-		  rq->fence.context,
-		  rq->fence.seqno);
+	ENGINE_TRACE(&ve->base, "rq=%llx:%lld\n",
+		     rq->fence.context,
+		     rq->fence.seqno);
 
 	GEM_BUG_ON(ve->base.submit_request != virtual_submit_request);
 
@@ -4394,8 +4460,7 @@ virtual_bond_execute(struct i915_request *rq, struct dma_fence *signal)
 }
 
 struct intel_context *
-intel_execlists_create_virtual(struct i915_gem_context *ctx,
-			       struct intel_engine_cs **siblings,
+intel_execlists_create_virtual(struct intel_engine_cs **siblings,
 			       unsigned int count)
 {
 	struct virtual_engine *ve;
@@ -4406,13 +4471,13 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 		return ERR_PTR(-EINVAL);
 
 	if (count == 1)
-		return intel_context_create(ctx, siblings[0]);
+		return intel_context_create(siblings[0]);
 
 	ve = kzalloc(struct_size(ve, siblings, count), GFP_KERNEL);
 	if (!ve)
 		return ERR_PTR(-ENOMEM);
 
-	ve->base.i915 = ctx->i915;
+	ve->base.i915 = siblings[0]->i915;
 	ve->base.gt = siblings[0]->gt;
 	ve->base.uncore = siblings[0]->uncore;
 	ve->base.id = -1;
@@ -4439,7 +4504,6 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 
 	intel_engine_init_active(&ve->base, ENGINE_VIRTUAL);
 	intel_engine_init_breadcrumbs(&ve->base);
-
 	intel_engine_init_execlists(&ve->base);
 
 	ve->base.cops = &virtual_context_ops;
@@ -4455,7 +4519,7 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 		     virtual_submission_tasklet,
 		     (unsigned long)ve);
 
-	intel_context_init(&ve->context, ctx, &ve->base);
+	intel_context_init(&ve->context, &ve->base);
 
 	for (n = 0; n < count; n++) {
 		struct intel_engine_cs *sibling = siblings[n];
@@ -4522,12 +4586,6 @@ intel_execlists_create_virtual(struct i915_gem_context *ctx,
 
 	ve->base.flags |= I915_ENGINE_IS_VIRTUAL;
 
-	err = __execlists_context_alloc(&ve->context, siblings[0]);
-	if (err)
-		goto err_put;
-
-	__set_bit(CONTEXT_ALLOC_BIT, &ve->context.flags);
-
 	return &ve->context;
 
 err_put:
@@ -4536,14 +4594,12 @@ err_put:
 }
 
 struct intel_context *
-intel_execlists_clone_virtual(struct i915_gem_context *ctx,
-			      struct intel_engine_cs *src)
+intel_execlists_clone_virtual(struct intel_engine_cs *src)
 {
 	struct virtual_engine *se = to_virtual_engine(src);
 	struct intel_context *dst;
 
-	dst = intel_execlists_create_virtual(ctx,
-					     se->siblings,
+	dst = intel_execlists_create_virtual(se->siblings,
 					     se->num_siblings);
 	if (IS_ERR(dst))
 		return dst;
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.h b/drivers/gpu/drm/i915/gt/intel_lrc.h
index 04511d8ebdc1..dfbc214e14f5 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.h
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.h
@@ -83,7 +83,6 @@ enum {
 void intel_logical_ring_cleanup(struct intel_engine_cs *engine);
 
 int intel_execlists_submission_setup(struct intel_engine_cs *engine);
-int intel_execlists_submission_init(struct intel_engine_cs *engine);
 
 /* Logical Ring Contexts */
 /* At the start of the context image is its per-process HWS page */
@@ -111,13 +110,11 @@ void intel_execlists_show_requests(struct intel_engine_cs *engine,
 				   unsigned int max);
 
 struct intel_context *
-intel_execlists_create_virtual(struct i915_gem_context *ctx,
-			       struct intel_engine_cs **siblings,
+intel_execlists_create_virtual(struct intel_engine_cs **siblings,
 			       unsigned int count);
 
 struct intel_context *
-intel_execlists_clone_virtual(struct i915_gem_context *ctx,
-			      struct intel_engine_cs *src);
+intel_execlists_clone_virtual(struct intel_engine_cs *src);
 
 int intel_virtual_engine_attach_bond(struct intel_engine_cs *engine,
 				     const struct intel_engine_cs *master,
diff --git a/drivers/gpu/drm/i915/gt/intel_lrc_reg.h b/drivers/gpu/drm/i915/gt/intel_lrc_reg.h
index 06ab0276e10e..08a3be65f700 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc_reg.h
+++ b/drivers/gpu/drm/i915/gt/intel_lrc_reg.h
@@ -13,8 +13,8 @@
 #define CTX_CONTEXT_CONTROL		(0x02 + 1)
 #define CTX_RING_HEAD			(0x04 + 1)
 #define CTX_RING_TAIL			(0x06 + 1)
-#define CTX_RING_BUFFER_START		(0x08 + 1)
-#define CTX_RING_BUFFER_CONTROL		(0x0a + 1)
+#define CTX_RING_START			(0x08 + 1)
+#define CTX_RING_CTL			(0x0a + 1)
 #define CTX_BB_STATE			(0x10 + 1)
 #define CTX_BB_PER_CTX_PTR		(0x18 + 1)
 #define CTX_PDP3_UDW			(0x24 + 1)
diff --git a/drivers/gpu/drm/i915/gt/intel_mocs.c b/drivers/gpu/drm/i915/gt/intel_mocs.c
index 2b977991b785..893249ea48d4 100644
--- a/drivers/gpu/drm/i915/gt/intel_mocs.c
+++ b/drivers/gpu/drm/i915/gt/intel_mocs.c
@@ -283,65 +283,42 @@ static const struct drm_i915_mocs_entry icelake_mocs_table[] = {
 static bool get_mocs_settings(const struct drm_i915_private *i915,
 			      struct drm_i915_mocs_table *table)
 {
-	bool result = false;
-
 	if (INTEL_GEN(i915) >= 12) {
 		table->size  = ARRAY_SIZE(tigerlake_mocs_table);
 		table->table = tigerlake_mocs_table;
 		table->n_entries = GEN11_NUM_MOCS_ENTRIES;
-		result = true;
 	} else if (IS_GEN(i915, 11)) {
 		table->size  = ARRAY_SIZE(icelake_mocs_table);
 		table->table = icelake_mocs_table;
 		table->n_entries = GEN11_NUM_MOCS_ENTRIES;
-		result = true;
 	} else if (IS_GEN9_BC(i915) || IS_CANNONLAKE(i915)) {
 		table->size  = ARRAY_SIZE(skylake_mocs_table);
 		table->n_entries = GEN9_NUM_MOCS_ENTRIES;
 		table->table = skylake_mocs_table;
-		result = true;
 	} else if (IS_GEN9_LP(i915)) {
 		table->size  = ARRAY_SIZE(broxton_mocs_table);
 		table->n_entries = GEN9_NUM_MOCS_ENTRIES;
 		table->table = broxton_mocs_table;
-		result = true;
 	} else {
 		WARN_ONCE(INTEL_GEN(i915) >= 9,
 			  "Platform that should have a MOCS table does not.\n");
+		return false;
 	}
 
+	if (GEM_DEBUG_WARN_ON(table->size > table->n_entries))
+		return false;
+
 	/* WaDisableSkipCaching:skl,bxt,kbl,glk */
 	if (IS_GEN(i915, 9)) {
 		int i;
 
 		for (i = 0; i < table->size; i++)
-			if (WARN_ON(table->table[i].l3cc_value &
-				    (L3_ESC(1) | L3_SCC(0x7))))
+			if (GEM_DEBUG_WARN_ON(table->table[i].l3cc_value &
+					      (L3_ESC(1) | L3_SCC(0x7))))
 				return false;
 	}
 
-	return result;
-}
-
-static i915_reg_t mocs_register(const struct intel_engine_cs *engine, int index)
-{
-	switch (engine->id) {
-	case RCS0:
-		return GEN9_GFX_MOCS(index);
-	case VCS0:
-		return GEN9_MFX0_MOCS(index);
-	case BCS0:
-		return GEN9_BLT_MOCS(index);
-	case VECS0:
-		return GEN9_VEBOX_MOCS(index);
-	case VCS1:
-		return GEN9_MFX1_MOCS(index);
-	case VCS2:
-		return GEN11_MFX2_MOCS(index);
-	default:
-		MISSING_CASE(engine->id);
-		return INVALID_MMIO_REG;
-	}
+	return true;
 }
 
 /*
@@ -351,29 +328,47 @@ static i915_reg_t mocs_register(const struct intel_engine_cs *engine, int index)
 static u32 get_entry_control(const struct drm_i915_mocs_table *table,
 			     unsigned int index)
 {
-	if (table->table[index].used)
+	if (index < table->size && table->table[index].used)
 		return table->table[index].control_value;
 
 	return table->table[I915_MOCS_PTE].control_value;
 }
 
-static void init_mocs_table(struct intel_engine_cs *engine,
-			    const struct drm_i915_mocs_table *table)
+#define for_each_mocs(mocs, t, i) \
+	for (i = 0; \
+	     i < (t)->n_entries ? (mocs = get_entry_control((t), i)), 1 : 0;\
+	     i++)
+
+static void __init_mocs_table(struct intel_uncore *uncore,
+			      const struct drm_i915_mocs_table *table,
+			      u32 addr)
 {
-	struct intel_uncore *uncore = engine->uncore;
-	u32 unused_value = table->table[I915_MOCS_PTE].control_value;
 	unsigned int i;
+	u32 mocs;
+
+	for_each_mocs(mocs, table, i)
+		intel_uncore_write_fw(uncore, _MMIO(addr + i * 4), mocs);
+}
 
-	for (i = 0; i < table->size; i++)
-		intel_uncore_write_fw(uncore,
-				      mocs_register(engine, i),
-				      get_entry_control(table, i));
+static u32 mocs_offset(const struct intel_engine_cs *engine)
+{
+	static const u32 offset[] = {
+		[RCS0]  =  __GEN9_RCS0_MOCS0,
+		[VCS0]  =  __GEN9_VCS0_MOCS0,
+		[VCS1]  =  __GEN9_VCS1_MOCS0,
+		[VECS0] =  __GEN9_VECS0_MOCS0,
+		[BCS0]  =  __GEN9_BCS0_MOCS0,
+		[VCS2]  = __GEN11_VCS2_MOCS0,
+	};
+
+	GEM_BUG_ON(engine->id >= ARRAY_SIZE(offset));
+	return offset[engine->id];
+}
 
-	/* All remaining entries are unused */
-	for (; i < table->n_entries; i++)
-		intel_uncore_write_fw(uncore,
-				      mocs_register(engine, i),
-				      unused_value);
+static void init_mocs_table(struct intel_engine_cs *engine,
+			    const struct drm_i915_mocs_table *table)
+{
+	__init_mocs_table(engine->uncore, table, mocs_offset(engine));
 }
 
 /*
@@ -383,51 +378,34 @@ static void init_mocs_table(struct intel_engine_cs *engine,
 static u16 get_entry_l3cc(const struct drm_i915_mocs_table *table,
 			  unsigned int index)
 {
-	if (table->table[index].used)
+	if (index < table->size && table->table[index].used)
 		return table->table[index].l3cc_value;
 
 	return table->table[I915_MOCS_PTE].l3cc_value;
 }
 
-static inline u32 l3cc_combine(const struct drm_i915_mocs_table *table,
-			       u16 low,
-			       u16 high)
+static inline u32 l3cc_combine(u16 low, u16 high)
 {
 	return low | (u32)high << 16;
 }
 
+#define for_each_l3cc(l3cc, t, i) \
+	for (i = 0; \
+	     i < ((t)->n_entries + 1) / 2 ? \
+	     (l3cc = l3cc_combine(get_entry_l3cc((t), 2 * i), \
+				  get_entry_l3cc((t), 2 * i + 1))), 1 : \
+	     0; \
+	     i++)
+
 static void init_l3cc_table(struct intel_engine_cs *engine,
 			    const struct drm_i915_mocs_table *table)
 {
 	struct intel_uncore *uncore = engine->uncore;
-	u16 unused_value = table->table[I915_MOCS_PTE].l3cc_value;
 	unsigned int i;
+	u32 l3cc;
 
-	for (i = 0; i < table->size / 2; i++) {
-		u16 low = get_entry_l3cc(table, 2 * i);
-		u16 high = get_entry_l3cc(table, 2 * i + 1);
-
-		intel_uncore_write(uncore,
-				   GEN9_LNCFCMOCS(i),
-				   l3cc_combine(table, low, high));
-	}
-
-	/* Odd table size - 1 left over */
-	if (table->size & 1) {
-		u16 low = get_entry_l3cc(table, 2 * i);
-
-		intel_uncore_write(uncore,
-				   GEN9_LNCFCMOCS(i),
-				   l3cc_combine(table, low, unused_value));
-		i++;
-	}
-
-	/* All remaining entries are also unused */
-	for (; i < table->n_entries / 2; i++)
-		intel_uncore_write(uncore,
-				   GEN9_LNCFCMOCS(i),
-				   l3cc_combine(table, unused_value,
-						unused_value));
+	for_each_l3cc(l3cc, table, i)
+		intel_uncore_write_fw(uncore, GEN9_LNCFCMOCS(i), l3cc);
 }
 
 void intel_mocs_init_engine(struct intel_engine_cs *engine)
@@ -448,11 +426,14 @@ void intel_mocs_init_engine(struct intel_engine_cs *engine)
 		init_l3cc_table(engine, &table);
 }
 
-static void intel_mocs_init_global(struct intel_gt *gt)
+static u32 global_mocs_offset(void)
+{
+	return i915_mmio_reg_offset(GEN12_GLOBAL_MOCS(0));
+}
+
+static void init_global_mocs(struct intel_gt *gt)
 {
-	struct intel_uncore *uncore = gt->uncore;
 	struct drm_i915_mocs_table table;
-	unsigned int index;
 
 	/*
 	 * LLC and eDRAM control values are not applicable to dgfx
@@ -460,32 +441,18 @@ static void intel_mocs_init_global(struct intel_gt *gt)
 	if (IS_DGFX(gt->i915))
 		return;
 
-	GEM_BUG_ON(!HAS_GLOBAL_MOCS_REGISTERS(gt->i915));
-
 	if (!get_mocs_settings(gt->i915, &table))
 		return;
 
-	if (GEM_DEBUG_WARN_ON(table.size > table.n_entries))
-		return;
-
-	for (index = 0; index < table.size; index++)
-		intel_uncore_write(uncore,
-				   GEN12_GLOBAL_MOCS(index),
-				   table.table[index].control_value);
-
-	/*
-	 * Ok, now set the unused entries to the invalid entry (index 0). These
-	 * entries are officially undefined and no contract for the contents and
-	 * settings is given for these entries.
-	 */
-	for (; index < table.n_entries; index++)
-		intel_uncore_write(uncore,
-				   GEN12_GLOBAL_MOCS(index),
-				   table.table[0].control_value);
+	__init_mocs_table(gt->uncore, &table, global_mocs_offset());
 }
 
 void intel_mocs_init(struct intel_gt *gt)
 {
 	if (HAS_GLOBAL_MOCS_REGISTERS(gt->i915))
-		intel_mocs_init_global(gt);
+		init_global_mocs(gt);
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftest_mocs.c"
+#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6.c b/drivers/gpu/drm/i915/gt/intel_rc6.c
index 700104b90163..9e303c29d6e3 100644
--- a/drivers/gpu/drm/i915/gt/intel_rc6.c
+++ b/drivers/gpu/drm/i915/gt/intel_rc6.c
@@ -88,21 +88,18 @@ static void gen11_rc6_enable(struct intel_rc6 *rc6)
 	 * do not want the enable hysteresis to less than the wakeup latency.
 	 *
 	 * igt/gem_exec_nop/sequential provides a rough estimate for the
-	 * service latency, and puts it around 10us for Broadwell (and other
-	 * big core) and around 40us for Broxton (and other low power cores).
-	 * [Note that for legacy ringbuffer submission, this is less than 1us!]
-	 * However, the wakeup latency on Broxton is closer to 100us. To be
-	 * conservative, we have to factor in a context switch on top (due
-	 * to ksoftirqd).
+	 * service latency, and puts it under 10us for Icelake, similar to
+	 * Broadwell+, To be conservative, we want to factor in a context
+	 * switch on top (due to ksoftirqd).
 	 */
-	set(uncore, GEN9_MEDIA_PG_IDLE_HYSTERESIS, 250);
-	set(uncore, GEN9_RENDER_PG_IDLE_HYSTERESIS, 250);
+	set(uncore, GEN9_MEDIA_PG_IDLE_HYSTERESIS, 60);
+	set(uncore, GEN9_RENDER_PG_IDLE_HYSTERESIS, 60);
 
 	/* 3a: Enable RC6 */
-	set(uncore, GEN6_RC_CONTROL,
-	    GEN6_RC_CTL_HW_ENABLE |
-	    GEN6_RC_CTL_RC6_ENABLE |
-	    GEN6_RC_CTL_EI_MODE(1));
+	rc6->ctl_enable =
+		GEN6_RC_CTL_HW_ENABLE |
+		GEN6_RC_CTL_RC6_ENABLE |
+		GEN6_RC_CTL_EI_MODE(1);
 
 	set(uncore, GEN9_PG_ENABLE,
 	    GEN9_RENDER_PG_ENABLE |
@@ -173,10 +170,10 @@ static void gen9_rc6_enable(struct intel_rc6 *rc6)
 	else
 		rc6_mode = GEN6_RC_CTL_EI_MODE(1);
 
-	set(uncore, GEN6_RC_CONTROL,
-	    GEN6_RC_CTL_HW_ENABLE |
-	    GEN6_RC_CTL_RC6_ENABLE |
-	    rc6_mode);
+	rc6->ctl_enable =
+		GEN6_RC_CTL_HW_ENABLE |
+		GEN6_RC_CTL_RC6_ENABLE |
+		rc6_mode;
 
 	/*
 	 * WaRsDisableCoarsePowerGating:skl,cnl
@@ -203,10 +200,10 @@ static void gen8_rc6_enable(struct intel_rc6 *rc6)
 	set(uncore, GEN6_RC6_THRESHOLD, 625); /* 800us/1.28 for TO */
 
 	/* 3: Enable RC6 */
-	set(uncore, GEN6_RC_CONTROL,
+	rc6->ctl_enable =
 	    GEN6_RC_CTL_HW_ENABLE |
 	    GEN7_RC_CTL_TO_MODE |
-	    GEN6_RC_CTL_RC6_ENABLE);
+	    GEN6_RC_CTL_RC6_ENABLE;
 }
 
 static void gen6_rc6_enable(struct intel_rc6 *rc6)
@@ -242,10 +239,10 @@ static void gen6_rc6_enable(struct intel_rc6 *rc6)
 		rc6_mask |= GEN6_RC_CTL_RC6p_ENABLE;
 	if (HAS_RC6pp(i915))
 		rc6_mask |= GEN6_RC_CTL_RC6pp_ENABLE;
-	set(uncore, GEN6_RC_CONTROL,
+	rc6->ctl_enable =
 	    rc6_mask |
 	    GEN6_RC_CTL_EI_MODE(1) |
-	    GEN6_RC_CTL_HW_ENABLE);
+	    GEN6_RC_CTL_HW_ENABLE;
 
 	rc6vids = 0;
 	ret = sandybridge_pcode_read(i915, GEN6_PCODE_READ_RC6VIDS,
@@ -363,7 +360,7 @@ static void chv_rc6_enable(struct intel_rc6 *rc6)
 			       VLV_RENDER_RC6_COUNT_EN));
 
 	/* 3: Enable RC6 */
-	set(uncore, GEN6_RC_CONTROL, GEN7_RC_CTL_TO_MODE);
+	rc6->ctl_enable = GEN7_RC_CTL_TO_MODE;
 }
 
 static void vlv_rc6_enable(struct intel_rc6 *rc6)
@@ -389,8 +386,8 @@ static void vlv_rc6_enable(struct intel_rc6 *rc6)
 			       VLV_MEDIA_RC6_COUNT_EN |
 			       VLV_RENDER_RC6_COUNT_EN));
 
-	set(uncore, GEN6_RC_CONTROL,
-	    GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL);
+	rc6->ctl_enable =
+	    GEN7_RC_CTL_TO_MODE | VLV_RC_CTL_CTX_RST_PARALLEL;
 }
 
 static bool bxt_check_bios_rc6_setup(struct intel_rc6 *rc6)
@@ -491,64 +488,19 @@ static void rpm_put(struct intel_rc6 *rc6)
 	rc6->wakeref = false;
 }
 
-static bool intel_rc6_ctx_corrupted(struct intel_rc6 *rc6)
-{
-	return !intel_uncore_read(rc6_to_uncore(rc6), GEN8_RC6_CTX_INFO);
-}
-
-static void intel_rc6_ctx_wa_init(struct intel_rc6 *rc6)
+static bool pctx_corrupted(struct intel_rc6 *rc6)
 {
 	struct drm_i915_private *i915 = rc6_to_i915(rc6);
 
 	if (!NEEDS_RC6_CTX_CORRUPTION_WA(i915))
-		return;
-
-	if (intel_rc6_ctx_corrupted(rc6)) {
-		DRM_INFO("RC6 context corrupted, disabling runtime power management\n");
-		rc6->ctx_corrupted = true;
-	}
-}
-
-/**
- * intel_rc6_ctx_wa_resume - system resume sequence for the RC6 CTX WA
- * @rc6: rc6 state
- *
- * Perform any steps needed to re-init the RC6 CTX WA after system resume.
- */
-void intel_rc6_ctx_wa_resume(struct intel_rc6 *rc6)
-{
-	if (rc6->ctx_corrupted && !intel_rc6_ctx_corrupted(rc6)) {
-		DRM_INFO("RC6 context restored, re-enabling runtime power management\n");
-		rc6->ctx_corrupted = false;
-	}
-}
-
-/**
- * intel_rc6_ctx_wa_check - check for a new RC6 CTX corruption
- * @rc6: rc6 state
- *
- * Check if an RC6 CTX corruption has happened since the last check and if so
- * disable RC6 and runtime power management.
-*/
-void intel_rc6_ctx_wa_check(struct intel_rc6 *rc6)
-{
-	struct drm_i915_private *i915 = rc6_to_i915(rc6);
-
-	if (!NEEDS_RC6_CTX_CORRUPTION_WA(i915))
-		return;
-
-	if (rc6->ctx_corrupted)
-		return;
-
-	if (!intel_rc6_ctx_corrupted(rc6))
-		return;
-
-	DRM_NOTE("RC6 context corruption, disabling runtime power management\n");
+		return false;
 
-	intel_rc6_disable(rc6);
-	rc6->ctx_corrupted = true;
+	if (intel_uncore_read(rc6_to_uncore(rc6), GEN8_RC6_CTX_INFO))
+		return false;
 
-	return;
+	dev_notice(i915->drm.dev,
+		   "RC6 context corruption, disabling runtime power management\n");
+	return true;
 }
 
 static void __intel_rc6_disable(struct intel_rc6 *rc6)
@@ -575,8 +527,6 @@ void intel_rc6_init(struct intel_rc6 *rc6)
 	if (!rc6_supported(rc6))
 		return;
 
-	intel_rc6_ctx_wa_init(rc6);
-
 	if (IS_CHERRYVIEW(i915))
 		err = chv_rc6_init(rc6);
 	else if (IS_VALLEYVIEW(i915))
@@ -611,9 +561,6 @@ void intel_rc6_enable(struct intel_rc6 *rc6)
 
 	GEM_BUG_ON(rc6->enabled);
 
-	if (rc6->ctx_corrupted)
-		return;
-
 	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
 
 	if (IS_CHERRYVIEW(i915))
@@ -629,13 +576,51 @@ void intel_rc6_enable(struct intel_rc6 *rc6)
 	else if (INTEL_GEN(i915) >= 6)
 		gen6_rc6_enable(rc6);
 
+	rc6->manual = rc6->ctl_enable & GEN6_RC_CTL_RC6_ENABLE;
+	if (NEEDS_RC6_CTX_CORRUPTION_WA(i915))
+		rc6->ctl_enable = 0;
+
 	intel_uncore_forcewake_put(uncore, FORCEWAKE_ALL);
 
+	if (unlikely(pctx_corrupted(rc6)))
+		return;
+
 	/* rc6 is ready, runtime-pm is go! */
 	rpm_put(rc6);
 	rc6->enabled = true;
 }
 
+void intel_rc6_unpark(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+
+	if (!rc6->enabled)
+		return;
+
+	/* Restore HW timers for automatic RC6 entry while busy */
+	set(uncore, GEN6_RC_CONTROL, rc6->ctl_enable);
+}
+
+void intel_rc6_park(struct intel_rc6 *rc6)
+{
+	struct intel_uncore *uncore = rc6_to_uncore(rc6);
+
+	if (!rc6->enabled)
+		return;
+
+	if (unlikely(pctx_corrupted(rc6))) {
+		intel_rc6_disable(rc6);
+		return;
+	}
+
+	if (!rc6->manual)
+		return;
+
+	/* Turn off the HW timers and go directly to rc6 */
+	set(uncore, GEN6_RC_CONTROL, GEN6_RC_CTL_RC6_ENABLE);
+	set(uncore, GEN6_RC_STATE, 0x4 << RC_SW_TARGET_STATE_SHIFT);
+}
+
 void intel_rc6_disable(struct intel_rc6 *rc6)
 {
 	if (!rc6->enabled)
@@ -785,3 +770,7 @@ u64 intel_rc6_residency_us(struct intel_rc6 *rc6, i915_reg_t reg)
 {
 	return DIV_ROUND_UP_ULL(intel_rc6_residency_ns(rc6, reg), 1000);
 }
+
+#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
+#include "selftest_rc6.c"
+#endif
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6.h b/drivers/gpu/drm/i915/gt/intel_rc6.h
index 1370f6834a4c..9f0f23fca8af 100644
--- a/drivers/gpu/drm/i915/gt/intel_rc6.h
+++ b/drivers/gpu/drm/i915/gt/intel_rc6.h
@@ -15,6 +15,9 @@ struct intel_rc6;
 void intel_rc6_init(struct intel_rc6 *rc6);
 void intel_rc6_fini(struct intel_rc6 *rc6);
 
+void intel_rc6_unpark(struct intel_rc6 *rc6);
+void intel_rc6_park(struct intel_rc6 *rc6);
+
 void intel_rc6_sanitize(struct intel_rc6 *rc6);
 void intel_rc6_enable(struct intel_rc6 *rc6);
 void intel_rc6_disable(struct intel_rc6 *rc6);
@@ -22,7 +25,4 @@ void intel_rc6_disable(struct intel_rc6 *rc6);
 u64 intel_rc6_residency_ns(struct intel_rc6 *rc6, i915_reg_t reg);
 u64 intel_rc6_residency_us(struct intel_rc6 *rc6, i915_reg_t reg);
 
-void intel_rc6_ctx_wa_check(struct intel_rc6 *rc6);
-void intel_rc6_ctx_wa_resume(struct intel_rc6 *rc6);
-
 #endif /* INTEL_RC6_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_rc6_types.h b/drivers/gpu/drm/i915/gt/intel_rc6_types.h
index 89ad5697a8d4..bfbb623f7a4f 100644
--- a/drivers/gpu/drm/i915/gt/intel_rc6_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_rc6_types.h
@@ -18,12 +18,14 @@ struct intel_rc6 {
 	u64 prev_hw_residency[4];
 	u64 cur_residency[4];
 
+	u32 ctl_enable;
+
 	struct drm_i915_gem_object *pctx;
 
 	bool supported : 1;
 	bool enabled : 1;
+	bool manual : 1;
 	bool wakeref : 1;
-	bool ctx_corrupted : 1;
 };
 
 #endif /* INTEL_RC6_TYPES_H */
diff --git a/drivers/gpu/drm/i915/gt/intel_renderstate.c b/drivers/gpu/drm/i915/gt/intel_renderstate.c
index c4edc35e7d89..5954ecc3207f 100644
--- a/drivers/gpu/drm/i915/gt/intel_renderstate.c
+++ b/drivers/gpu/drm/i915/gt/intel_renderstate.c
@@ -29,16 +29,6 @@
 #include "intel_renderstate.h"
 #include "intel_ring.h"
 
-struct intel_renderstate {
-	const struct intel_renderstate_rodata *rodata;
-	struct drm_i915_gem_object *obj;
-	struct i915_vma *vma;
-	u32 batch_offset;
-	u32 batch_size;
-	u32 aux_offset;
-	u32 aux_size;
-};
-
 static const struct intel_renderstate_rodata *
 render_state_get_rodata(const struct intel_engine_cs *engine)
 {
@@ -84,11 +74,11 @@ static int render_state_setup(struct intel_renderstate *so,
 	u32 *d;
 	int ret;
 
-	ret = i915_gem_object_prepare_write(so->obj, &needs_clflush);
+	ret = i915_gem_object_prepare_write(so->vma->obj, &needs_clflush);
 	if (ret)
 		return ret;
 
-	d = kmap_atomic(i915_gem_object_get_dirty_page(so->obj, 0));
+	d = kmap_atomic(i915_gem_object_get_dirty_page(so->vma->obj, 0));
 
 	while (i < rodata->batch_items) {
 		u32 s = rodata->batch[i];
@@ -166,7 +156,7 @@ static int render_state_setup(struct intel_renderstate *so,
 
 	ret = 0;
 out:
-	i915_gem_object_finish_access(so->obj);
+	i915_gem_object_finish_access(so->vma->obj);
 	return ret;
 
 err:
@@ -177,61 +167,84 @@ err:
 
 #undef OUT_BATCH
 
-int intel_renderstate_emit(struct i915_request *rq)
+int intel_renderstate_init(struct intel_renderstate *so,
+			   struct intel_engine_cs *engine)
 {
-	struct intel_engine_cs *engine = rq->engine;
-	struct intel_renderstate so = {}; /* keep the compiler happy */
+	struct drm_i915_gem_object *obj;
 	int err;
 
-	so.rodata = render_state_get_rodata(engine);
-	if (!so.rodata)
+	memset(so, 0, sizeof(*so));
+
+	so->rodata = render_state_get_rodata(engine);
+	if (!so->rodata)
 		return 0;
 
-	if (so.rodata->batch_items * 4 > PAGE_SIZE)
+	if (so->rodata->batch_items * 4 > PAGE_SIZE)
 		return -EINVAL;
 
-	so.obj = i915_gem_object_create_internal(engine->i915, PAGE_SIZE);
-	if (IS_ERR(so.obj))
-		return PTR_ERR(so.obj);
+	obj = i915_gem_object_create_internal(engine->i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return PTR_ERR(obj);
 
-	so.vma = i915_vma_instance(so.obj, &engine->gt->ggtt->vm, NULL);
-	if (IS_ERR(so.vma)) {
-		err = PTR_ERR(so.vma);
+	so->vma = i915_vma_instance(obj, &engine->gt->ggtt->vm, NULL);
+	if (IS_ERR(so->vma)) {
+		err = PTR_ERR(so->vma);
 		goto err_obj;
 	}
 
-	err = i915_vma_pin(so.vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
+	err = i915_vma_pin(so->vma, 0, 0, PIN_GLOBAL | PIN_HIGH);
 	if (err)
 		goto err_vma;
 
-	err = render_state_setup(&so, rq->i915);
+	err = render_state_setup(so, engine->i915);
 	if (err)
 		goto err_unpin;
 
+	return 0;
+
+err_unpin:
+	i915_vma_unpin(so->vma);
+err_vma:
+	i915_vma_close(so->vma);
+err_obj:
+	i915_gem_object_put(obj);
+	so->vma = NULL;
+	return err;
+}
+
+int intel_renderstate_emit(struct intel_renderstate *so,
+			   struct i915_request *rq)
+{
+	struct intel_engine_cs *engine = rq->engine;
+	int err;
+
+	if (!so->vma)
+		return 0;
+
 	err = engine->emit_bb_start(rq,
-				    so.batch_offset, so.batch_size,
+				    so->batch_offset, so->batch_size,
 				    I915_DISPATCH_SECURE);
 	if (err)
-		goto err_unpin;
+		return err;
 
-	if (so.aux_size > 8) {
+	if (so->aux_size > 8) {
 		err = engine->emit_bb_start(rq,
-					    so.aux_offset, so.aux_size,
+					    so->aux_offset, so->aux_size,
 					    I915_DISPATCH_SECURE);
 		if (err)
-			goto err_unpin;
+			return err;
 	}
 
-	i915_vma_lock(so.vma);
-	err = i915_request_await_object(rq, so.vma->obj, false);
+	i915_vma_lock(so->vma);
+	err = i915_request_await_object(rq, so->vma->obj, false);
 	if (err == 0)
-		err = i915_vma_move_to_active(so.vma, rq, 0);
-	i915_vma_unlock(so.vma);
-err_unpin:
-	i915_vma_unpin(so.vma);
-err_vma:
-	i915_vma_close(so.vma);
-err_obj:
-	i915_gem_object_put(so.obj);
+		err = i915_vma_move_to_active(so->vma, rq, 0);
+	i915_vma_unlock(so->vma);
+
 	return err;
 }
+
+void intel_renderstate_fini(struct intel_renderstate *so)
+{
+	i915_vma_unpin_and_release(&so->vma, 0);
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_renderstate.h b/drivers/gpu/drm/i915/gt/intel_renderstate.h
index 8d5079145054..5700be69a05a 100644
--- a/drivers/gpu/drm/i915/gt/intel_renderstate.h
+++ b/drivers/gpu/drm/i915/gt/intel_renderstate.h
@@ -27,6 +27,8 @@
 #include <linux/types.h>
 
 struct i915_request;
+struct intel_engine_cs;
+struct i915_vma;
 
 struct intel_renderstate_rodata {
 	const u32 *reloc;
@@ -46,6 +48,19 @@ extern const struct intel_renderstate_rodata gen7_null_state;
 extern const struct intel_renderstate_rodata gen8_null_state;
 extern const struct intel_renderstate_rodata gen9_null_state;
 
-int intel_renderstate_emit(struct i915_request *rq);
+struct intel_renderstate {
+	const struct intel_renderstate_rodata *rodata;
+	struct i915_vma *vma;
+	u32 batch_offset;
+	u32 batch_size;
+	u32 aux_offset;
+	u32 aux_size;
+};
+
+int intel_renderstate_init(struct intel_renderstate *so,
+			   struct intel_engine_cs *engine);
+int intel_renderstate_emit(struct intel_renderstate *so,
+			   struct i915_request *rq);
+void intel_renderstate_fini(struct intel_renderstate *so);
 
 #endif /* _INTEL_RENDERSTATE_H_ */
diff --git a/drivers/gpu/drm/i915/gt/intel_reset.c b/drivers/gpu/drm/i915/gt/intel_reset.c
index c97423a76642..1c51296646e0 100644
--- a/drivers/gpu/drm/i915/gt/intel_reset.c
+++ b/drivers/gpu/drm/i915/gt/intel_reset.c
@@ -21,6 +21,7 @@
 #include "intel_reset.h"
 
 #include "uc/intel_guc.h"
+#include "uc/intel_guc_submission.h"
 
 #define RESET_MAX_RETRIES 3
 
@@ -40,27 +41,29 @@ static void rmw_clear_fw(struct intel_uncore *uncore, i915_reg_t reg, u32 clr)
 static void engine_skip_context(struct i915_request *rq)
 {
 	struct intel_engine_cs *engine = rq->engine;
-	struct i915_gem_context *hung_ctx = rq->gem_context;
+	struct intel_context *hung_ctx = rq->context;
 
 	if (!i915_request_is_active(rq))
 		return;
 
 	lockdep_assert_held(&engine->active.lock);
 	list_for_each_entry_continue(rq, &engine->active.requests, sched.link)
-		if (rq->gem_context == hung_ctx)
+		if (rq->context == hung_ctx)
 			i915_request_skip(rq, -EIO);
 }
 
-static void client_mark_guilty(struct drm_i915_file_private *file_priv,
-			       const struct i915_gem_context *ctx)
+static void client_mark_guilty(struct i915_gem_context *ctx, bool banned)
 {
-	unsigned int score;
+	struct drm_i915_file_private *file_priv = ctx->file_priv;
 	unsigned long prev_hang;
+	unsigned int score;
+
+	if (IS_ERR_OR_NULL(file_priv))
+		return;
 
-	if (i915_gem_context_is_banned(ctx))
+	score = 0;
+	if (banned)
 		score = I915_CLIENT_SCORE_CONTEXT_BAN;
-	else
-		score = 0;
 
 	prev_hang = xchg(&file_priv->hang_timestamp, jiffies);
 	if (time_before(jiffies, prev_hang + I915_CLIENT_FAST_HANG_JIFFIES))
@@ -75,17 +78,38 @@ static void client_mark_guilty(struct drm_i915_file_private *file_priv,
 	}
 }
 
-static bool context_mark_guilty(struct i915_gem_context *ctx)
+static bool mark_guilty(struct i915_request *rq)
 {
+	struct i915_gem_context *ctx;
 	unsigned long prev_hang;
 	bool banned;
 	int i;
 
+	rcu_read_lock();
+	ctx = rcu_dereference(rq->context->gem_context);
+	if (ctx && !kref_get_unless_zero(&ctx->ref))
+		ctx = NULL;
+	rcu_read_unlock();
+	if (!ctx)
+		return false;
+
+	if (i915_gem_context_is_closed(ctx)) {
+		intel_context_set_banned(rq->context);
+		banned = true;
+		goto out;
+	}
+
 	atomic_inc(&ctx->guilty_count);
 
 	/* Cool contexts are too cool to be banned! (Used for reset testing.) */
-	if (!i915_gem_context_is_bannable(ctx))
-		return false;
+	if (!i915_gem_context_is_bannable(ctx)) {
+		banned = false;
+		goto out;
+	}
+
+	dev_notice(ctx->i915->drm.dev,
+		   "%s context reset due to GPU hang\n",
+		   ctx->name);
 
 	/* Record the timestamp for the last N hangs */
 	prev_hang = ctx->hang_timestamp[0];
@@ -100,18 +124,25 @@ static bool context_mark_guilty(struct i915_gem_context *ctx)
 	if (banned) {
 		DRM_DEBUG_DRIVER("context %s: guilty %d, banned\n",
 				 ctx->name, atomic_read(&ctx->guilty_count));
-		i915_gem_context_set_banned(ctx);
+		intel_context_set_banned(rq->context);
 	}
 
-	if (!IS_ERR_OR_NULL(ctx->file_priv))
-		client_mark_guilty(ctx->file_priv, ctx);
+	client_mark_guilty(ctx, banned);
 
+out:
+	i915_gem_context_put(ctx);
 	return banned;
 }
 
-static void context_mark_innocent(struct i915_gem_context *ctx)
+static void mark_innocent(struct i915_request *rq)
 {
-	atomic_inc(&ctx->active_count);
+	struct i915_gem_context *ctx;
+
+	rcu_read_lock();
+	ctx = rcu_dereference(rq->context->gem_context);
+	if (ctx)
+		atomic_inc(&ctx->active_count);
+	rcu_read_unlock();
 }
 
 void __i915_request_reset(struct i915_request *rq, bool guilty)
@@ -124,14 +155,16 @@ void __i915_request_reset(struct i915_request *rq, bool guilty)
 
 	GEM_BUG_ON(i915_request_completed(rq));
 
+	rcu_read_lock(); /* protect the GEM context */
 	if (guilty) {
 		i915_request_skip(rq, -EIO);
-		if (context_mark_guilty(rq->gem_context))
+		if (mark_guilty(rq))
 			engine_skip_context(rq);
 	} else {
 		dma_fence_set_error(&rq->fence, -EAGAIN);
-		context_mark_innocent(rq->gem_context);
+		mark_innocent(rq);
 	}
+	rcu_read_unlock();
 }
 
 static bool i915_in_reset(struct pci_dev *pdev)
@@ -647,7 +680,8 @@ static void reset_prepare_engine(struct intel_engine_cs *engine)
 	 * GPU state upon resume, i.e. fail to restart after a reset.
 	 */
 	intel_uncore_forcewake_get(engine->uncore, FORCEWAKE_ALL);
-	engine->reset.prepare(engine);
+	if (engine->reset.prepare)
+		engine->reset.prepare(engine);
 }
 
 static void revoke_mmaps(struct intel_gt *gt)
@@ -667,8 +701,13 @@ static void revoke_mmaps(struct intel_gt *gt)
 			continue;
 
 		GEM_BUG_ON(vma->fence != &gt->ggtt->fence_regs[i]);
-		node = &vma->obj->base.vma_node;
+
+		if (!vma->mmo)
+			continue;
+
+		node = &vma->mmo->vma_node;
 		vma_offset = vma->ggtt_view.partial.offset << PAGE_SHIFT;
+
 		unmap_mapping_range(gt->i915->drm.anon_inode->i_mapping,
 				    drm_vma_node_offset_addr(node) + vma_offset,
 				    vma->size,
@@ -722,10 +761,11 @@ static int gt_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask)
 
 static void reset_finish_engine(struct intel_engine_cs *engine)
 {
-	engine->reset.finish(engine);
+	if (engine->reset.finish)
+		engine->reset.finish(engine);
 	intel_uncore_forcewake_put(engine->uncore, FORCEWAKE_ALL);
 
-	intel_engine_breadcrumbs_irq(engine);
+	intel_engine_signal_breadcrumbs(engine);
 }
 
 static void reset_finish(struct intel_gt *gt, intel_engine_mask_t awake)
@@ -754,7 +794,7 @@ static void nop_submit_request(struct i915_request *request)
 	i915_request_mark_complete(request);
 	spin_unlock_irqrestore(&engine->active.lock, flags);
 
-	intel_engine_queue_breadcrumbs(engine);
+	intel_engine_signal_breadcrumbs(engine);
 }
 
 static void __intel_gt_set_wedged(struct intel_gt *gt)
@@ -799,7 +839,8 @@ static void __intel_gt_set_wedged(struct intel_gt *gt)
 
 	/* Mark all executing requests as skipped */
 	for_each_engine(engine, gt, id)
-		engine->cancel_requests(engine);
+		if (engine->reset.cancel)
+			engine->reset.cancel(engine);
 
 	reset_finish(gt, awake);
 
@@ -820,7 +861,6 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 {
 	struct intel_gt_timelines *timelines = &gt->timelines;
 	struct intel_timeline *tl;
-	unsigned long flags;
 	bool ok;
 
 	if (!test_bit(I915_WEDGED, &gt->reset.flags))
@@ -842,7 +882,7 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 	 *
 	 * No more can be submitted until we reset the wedged bit.
 	 */
-	spin_lock_irqsave(&timelines->lock, flags);
+	spin_lock(&timelines->lock);
 	list_for_each_entry(tl, &timelines->active_list, link) {
 		struct dma_fence *fence;
 
@@ -850,7 +890,7 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 		if (!fence)
 			continue;
 
-		spin_unlock_irqrestore(&timelines->lock, flags);
+		spin_unlock(&timelines->lock);
 
 		/*
 		 * All internal dependencies (i915_requests) will have
@@ -863,10 +903,10 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)
 		dma_fence_put(fence);
 
 		/* Restart iteration after droping lock */
-		spin_lock_irqsave(&timelines->lock, flags);
+		spin_lock(&timelines->lock);
 		tl = list_entry(&timelines->active_list, typeof(*tl), link);
 	}
-	spin_unlock_irqrestore(&timelines->lock, flags);
+	spin_unlock(&timelines->lock);
 
 	/* We must reset pending GPU events before restoring our submission */
 	ok = !HAS_EXECLISTS(gt->i915); /* XXX better agnosticism desired */
@@ -1070,9 +1110,10 @@ static inline int intel_gt_reset_engine(struct intel_engine_cs *engine)
 int intel_engine_reset(struct intel_engine_cs *engine, const char *msg)
 {
 	struct intel_gt *gt = engine->gt;
+	bool uses_guc = intel_engine_in_guc_submission_mode(engine);
 	int ret;
 
-	GEM_TRACE("%s flags=%lx\n", engine->name, gt->reset.flags);
+	ENGINE_TRACE(engine, "flags=%lx\n", gt->reset.flags);
 	GEM_BUG_ON(!test_bit(I915_RESET_ENGINE + engine->id, &gt->reset.flags));
 
 	if (!intel_engine_pm_get_if_awake(engine))
@@ -1085,14 +1126,14 @@ int intel_engine_reset(struct intel_engine_cs *engine, const char *msg)
 			   "Resetting %s for %s\n", engine->name, msg);
 	atomic_inc(&engine->i915->gpu_error.reset_engine_count[engine->uabi_class]);
 
-	if (!engine->gt->uc.guc.execbuf_client)
+	if (!uses_guc)
 		ret = intel_gt_reset_engine(engine);
 	else
 		ret = intel_guc_reset_engine(&engine->gt->uc.guc, engine);
 	if (ret) {
 		/* If we fail here, we expect to fallback to a global reset */
 		DRM_DEBUG_DRIVER("%sFailed to reset %s, ret=%d\n",
-				 engine->gt->uc.guc.execbuf_client ? "GuC " : "",
+				 uses_guc ? "GuC " : "",
 				 engine->name, ret);
 		goto out;
 	}
diff --git a/drivers/gpu/drm/i915/gt/intel_ring_submission.c b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
index a47d5a7c32c9..81f872f9ef03 100644
--- a/drivers/gpu/drm/i915/gt/intel_ring_submission.c
+++ b/drivers/gpu/drm/i915/gt/intel_ring_submission.c
@@ -362,6 +362,12 @@ gen7_render_ring_flush(struct i915_request *rq, u32 mode)
 	 */
 	flags |= PIPE_CONTROL_CS_STALL;
 
+	/*
+	 * CS_STALL suggests at least a post-sync write.
+	 */
+	flags |= PIPE_CONTROL_QW_WRITE;
+	flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+
 	/* Just flush everything.  Experiments have shown that reducing the
 	 * number of bits based on the write domains has little performance
 	 * impact.
@@ -380,13 +386,6 @@ gen7_render_ring_flush(struct i915_request *rq, u32 mode)
 		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
 		flags |= PIPE_CONTROL_MEDIA_STATE_CLEAR;
-		/*
-		 * TLB invalidate requires a post-sync write.
-		 */
-		flags |= PIPE_CONTROL_QW_WRITE;
-		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
-
-		flags |= PIPE_CONTROL_STALL_AT_SCOREBOARD;
 
 		/* Workaround: we must issue a pipe_control with CS-stall bit
 		 * set before a pipe_control command that has the state cache
@@ -454,7 +453,8 @@ static u32 *gen7_xcs_emit_breadcrumb(struct i915_request *rq, u32 *cs)
 	GEM_BUG_ON(i915_request_active_timeline(rq)->hwsp_ggtt != rq->engine->status_page.vma);
 	GEM_BUG_ON(offset_in_page(i915_request_active_timeline(rq)->hwsp_offset) != I915_GEM_HWS_SEQNO_ADDR);
 
-	*cs++ = MI_FLUSH_DW | MI_FLUSH_DW_OP_STOREDW | MI_FLUSH_DW_STORE_INDEX;
+	*cs++ = MI_FLUSH_DW | MI_INVALIDATE_TLB |
+		MI_FLUSH_DW_OP_STOREDW | MI_FLUSH_DW_STORE_INDEX;
 	*cs++ = I915_GEM_HWS_SEQNO_ADDR | MI_FLUSH_DW_USE_GTT;
 	*cs++ = rq->fence.seqno;
 
@@ -496,14 +496,13 @@ static void set_hwstam(struct intel_engine_cs *engine, u32 mask)
 
 static void set_hws_pga(struct intel_engine_cs *engine, phys_addr_t phys)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
 	u32 addr;
 
 	addr = lower_32_bits(phys);
-	if (INTEL_GEN(dev_priv) >= 4)
+	if (INTEL_GEN(engine->i915) >= 4)
 		addr |= (phys >> 28) & 0xf0;
 
-	I915_WRITE(HWS_PGA, addr);
+	intel_uncore_write(engine->uncore, HWS_PGA, addr);
 }
 
 static struct page *status_page(struct intel_engine_cs *engine)
@@ -522,14 +521,13 @@ static void ring_setup_phys_status_page(struct intel_engine_cs *engine)
 
 static void set_hwsp(struct intel_engine_cs *engine, u32 offset)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
 	i915_reg_t hwsp;
 
 	/*
 	 * The ring status page addresses are no longer next to the rest of
 	 * the ring registers as of gen7.
 	 */
-	if (IS_GEN(dev_priv, 7)) {
+	if (IS_GEN(engine->i915, 7)) {
 		switch (engine->id) {
 		/*
 		 * No more rings exist on Gen7. Default case is only to shut up
@@ -551,14 +549,14 @@ static void set_hwsp(struct intel_engine_cs *engine, u32 offset)
 			hwsp = VEBOX_HWS_PGA_GEN7;
 			break;
 		}
-	} else if (IS_GEN(dev_priv, 6)) {
+	} else if (IS_GEN(engine->i915, 6)) {
 		hwsp = RING_HWS_PGA_GEN6(engine->mmio_base);
 	} else {
 		hwsp = RING_HWS_PGA(engine->mmio_base);
 	}
 
-	I915_WRITE(hwsp, offset);
-	POSTING_READ(hwsp);
+	intel_uncore_write(engine->uncore, hwsp, offset);
+	intel_uncore_posting_read(engine->uncore, hwsp);
 }
 
 static void flush_cs_tlb(struct intel_engine_cs *engine)
@@ -633,8 +631,8 @@ static int xcs_resume(struct intel_engine_cs *engine)
 	struct intel_ring *ring = engine->legacy.ring;
 	int ret = 0;
 
-	GEM_TRACE("%s: ring:{HEAD:%04x, TAIL:%04x}\n",
-		  engine->name, ring->head, ring->tail);
+	ENGINE_TRACE(engine, "ring:{HEAD:%04x, TAIL:%04x}\n",
+		     ring->head, ring->tail);
 
 	intel_uncore_forcewake_get(engine->uncore, FORCEWAKE_ALL);
 
@@ -721,7 +719,7 @@ static int xcs_resume(struct intel_engine_cs *engine)
 	}
 
 	/* Papering over lost _interrupts_ immediately following the restart */
-	intel_engine_queue_breadcrumbs(engine);
+	intel_engine_signal_breadcrumbs(engine);
 out:
 	intel_uncore_forcewake_put(engine->uncore, FORCEWAKE_ALL);
 
@@ -747,10 +745,10 @@ static void reset_prepare(struct intel_engine_cs *engine)
 	 *
 	 * FIXME: Wa for more modern gens needs to be validated
 	 */
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	if (intel_engine_stop_cs(engine))
-		GEM_TRACE("%s: timed out on STOP_RING\n", engine->name);
+		ENGINE_TRACE(engine, "timed out on STOP_RING\n");
 
 	intel_uncore_write_fw(uncore,
 			      RING_HEAD(base),
@@ -766,12 +764,11 @@ static void reset_prepare(struct intel_engine_cs *engine)
 
 	/* Check acts as a post */
 	if (intel_uncore_read_fw(uncore, RING_HEAD(base)))
-		GEM_TRACE("%s: ring head [%x] not parked\n",
-			  engine->name,
-			  intel_uncore_read_fw(uncore, RING_HEAD(base)));
+		ENGINE_TRACE(engine, "ring head [%x] not parked\n",
+			     intel_uncore_read_fw(uncore, RING_HEAD(base)));
 }
 
-static void reset_ring(struct intel_engine_cs *engine, bool stalled)
+static void reset_rewind(struct intel_engine_cs *engine, bool stalled)
 {
 	struct i915_request *pos, *rq;
 	unsigned long flags;
@@ -842,7 +839,8 @@ static void reset_finish(struct intel_engine_cs *engine)
 
 static int rcs_resume(struct intel_engine_cs *engine)
 {
-	struct drm_i915_private *dev_priv = engine->i915;
+	struct drm_i915_private *i915 = engine->i915;
+	struct intel_uncore *uncore = engine->uncore;
 
 	/*
 	 * Disable CONSTANT_BUFFER before it is loaded from the context
@@ -854,13 +852,14 @@ static int rcs_resume(struct intel_engine_cs *engine)
 	 * they are already accustomed to from before contexts were
 	 * enabled.
 	 */
-	if (IS_GEN(dev_priv, 4))
-		I915_WRITE(ECOSKPD,
+	if (IS_GEN(i915, 4))
+		intel_uncore_write(uncore, ECOSKPD,
 			   _MASKED_BIT_ENABLE(ECO_CONSTANT_BUFFER_SR_DISABLE));
 
 	/* WaTimedSingleVertexDispatch:cl,bw,ctg,elk,ilk,snb */
-	if (IS_GEN_RANGE(dev_priv, 4, 6))
-		I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(VS_TIMER_DISPATCH));
+	if (IS_GEN_RANGE(i915, 4, 6))
+		intel_uncore_write(uncore, MI_MODE,
+				   _MASKED_BIT_ENABLE(VS_TIMER_DISPATCH));
 
 	/* We need to disable the AsyncFlip performance optimisations in order
 	 * to use MI_WAIT_FOR_EVENT within the CS. It should already be
@@ -868,38 +867,40 @@ static int rcs_resume(struct intel_engine_cs *engine)
 	 *
 	 * WaDisableAsyncFlipPerfMode:snb,ivb,hsw,vlv
 	 */
-	if (IS_GEN_RANGE(dev_priv, 6, 7))
-		I915_WRITE(MI_MODE, _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
+	if (IS_GEN_RANGE(i915, 6, 7))
+		intel_uncore_write(uncore, MI_MODE,
+				   _MASKED_BIT_ENABLE(ASYNC_FLIP_PERF_DISABLE));
 
 	/* Required for the hardware to program scanline values for waiting */
 	/* WaEnableFlushTlbInvalidationMode:snb */
-	if (IS_GEN(dev_priv, 6))
-		I915_WRITE(GFX_MODE,
+	if (IS_GEN(i915, 6))
+		intel_uncore_write(uncore, GFX_MODE,
 			   _MASKED_BIT_ENABLE(GFX_TLB_INVALIDATE_EXPLICIT));
 
 	/* WaBCSVCSTlbInvalidationMode:ivb,vlv,hsw */
-	if (IS_GEN(dev_priv, 7))
-		I915_WRITE(GFX_MODE_GEN7,
+	if (IS_GEN(i915, 7))
+		intel_uncore_write(uncore, GFX_MODE_GEN7,
 			   _MASKED_BIT_ENABLE(GFX_TLB_INVALIDATE_EXPLICIT) |
 			   _MASKED_BIT_ENABLE(GFX_REPLAY_MODE));
 
-	if (IS_GEN(dev_priv, 6)) {
+	if (IS_GEN(i915, 6)) {
 		/* From the Sandybridge PRM, volume 1 part 3, page 24:
 		 * "If this bit is set, STCunit will have LRA as replacement
 		 *  policy. [...] This bit must be reset.  LRA replacement
 		 *  policy is not supported."
 		 */
-		I915_WRITE(CACHE_MODE_0,
+		intel_uncore_write(uncore, CACHE_MODE_0,
 			   _MASKED_BIT_DISABLE(CM0_STC_EVICT_DISABLE_LRA_SNB));
 	}
 
-	if (IS_GEN_RANGE(dev_priv, 6, 7))
-		I915_WRITE(INSTPM, _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));
+	if (IS_GEN_RANGE(i915, 6, 7))
+		intel_uncore_write(uncore, INSTPM,
+				   _MASKED_BIT_ENABLE(INSTPM_FORCE_ORDERING));
 
 	return xcs_resume(engine);
 }
 
-static void cancel_requests(struct intel_engine_cs *engine)
+static void reset_cancel(struct intel_engine_cs *engine)
 {
 	struct i915_request *request;
 	unsigned long flags;
@@ -1318,6 +1319,8 @@ static int ring_context_alloc(struct intel_context *ce)
 			return PTR_ERR(vma);
 
 		ce->state = vma;
+		if (engine->default_state)
+			__set_bit(CONTEXT_VALID_BIT, &ce->flags);
 	}
 
 	return 0;
@@ -1360,45 +1363,37 @@ static const struct intel_context_ops ring_context_ops = {
 	.destroy = ring_context_destroy,
 };
 
-static int load_pd_dir(struct i915_request *rq, const struct i915_ppgtt *ppgtt)
+static int load_pd_dir(struct i915_request *rq,
+		       const struct i915_ppgtt *ppgtt,
+		       u32 valid)
 {
 	const struct intel_engine_cs * const engine = rq->engine;
 	u32 *cs;
 
-	cs = intel_ring_begin(rq, 6);
+	cs = intel_ring_begin(rq, 12);
 	if (IS_ERR(cs))
 		return PTR_ERR(cs);
 
 	*cs++ = MI_LOAD_REGISTER_IMM(1);
 	*cs++ = i915_mmio_reg_offset(RING_PP_DIR_DCLV(engine->mmio_base));
-	*cs++ = PP_DIR_DCLV_2G;
+	*cs++ = valid;
 
 	*cs++ = MI_LOAD_REGISTER_IMM(1);
 	*cs++ = i915_mmio_reg_offset(RING_PP_DIR_BASE(engine->mmio_base));
 	*cs++ = px_base(ppgtt->pd)->ggtt_offset << 10;
 
-	intel_ring_advance(rq, cs);
-
-	return 0;
-}
-
-static int flush_pd_dir(struct i915_request *rq)
-{
-	const struct intel_engine_cs * const engine = rq->engine;
-	u32 *cs;
-
-	cs = intel_ring_begin(rq, 4);
-	if (IS_ERR(cs))
-		return PTR_ERR(cs);
-
-	/* Stall until the page table load is complete */
+	/* Stall until the page table load is complete? */
 	*cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT;
 	*cs++ = i915_mmio_reg_offset(RING_PP_DIR_BASE(engine->mmio_base));
-	*cs++ = intel_gt_scratch_offset(rq->engine->gt,
+	*cs++ = intel_gt_scratch_offset(engine->gt,
 					INTEL_GT_SCRATCH_FIELD_DEFAULT);
-	*cs++ = MI_NOOP;
+
+	*cs++ = MI_LOAD_REGISTER_IMM(1);
+	*cs++ = i915_mmio_reg_offset(RING_INSTPM(engine->mmio_base));
+	*cs++ = _MASKED_BIT_ENABLE(INSTPM_TLB_INVALIDATE);
 
 	intel_ring_advance(rq, cs);
+
 	return 0;
 }
 
@@ -1485,7 +1480,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags)
 
 	*cs++ = MI_NOOP;
 	*cs++ = MI_SET_CONTEXT;
-	*cs++ = i915_ggtt_offset(rq->hw_context->state) | flags;
+	*cs++ = i915_ggtt_offset(rq->context->state) | flags;
 	/*
 	 * w/a: MI_SET_CONTEXT must always be followed by MI_NOOP
 	 * WaMiSetContext_Hang:snb,ivb,vlv
@@ -1555,10 +1550,10 @@ static int remap_l3_slice(struct i915_request *rq, int slice)
 
 static int remap_l3(struct i915_request *rq)
 {
-	struct i915_gem_context *ctx = rq->gem_context;
+	struct i915_gem_context *ctx = i915_request_gem_context(rq);
 	int i, err;
 
-	if (!ctx->remap_slice)
+	if (!ctx || !ctx->remap_slice)
 		return 0;
 
 	for (i = 0; i < MAX_L3_SLICES; i++) {
@@ -1574,34 +1569,50 @@ static int remap_l3(struct i915_request *rq)
 	return 0;
 }
 
+static int switch_mm(struct i915_request *rq, struct i915_address_space *vm)
+{
+	int ret;
+
+	if (!vm)
+		return 0;
+
+	ret = rq->engine->emit_flush(rq, EMIT_FLUSH);
+	if (ret)
+		return ret;
+
+	/*
+	 * Not only do we need a full barrier (post-sync write) after
+	 * invalidating the TLBs, but we need to wait a little bit
+	 * longer. Whether this is merely delaying us, or the
+	 * subsequent flush is a key part of serialising with the
+	 * post-sync op, this extra pass appears vital before a
+	 * mm switch!
+	 */
+	ret = load_pd_dir(rq, i915_vm_to_ppgtt(vm), PP_DIR_DCLV_2G);
+	if (ret)
+		return ret;
+
+	return rq->engine->emit_flush(rq, EMIT_FLUSH);
+}
+
 static int switch_context(struct i915_request *rq)
 {
-	struct intel_context *ce = rq->hw_context;
-	struct i915_address_space *vm = vm_alias(ce);
+	struct intel_context *ce = rq->context;
 	int ret;
 
 	GEM_BUG_ON(HAS_EXECLISTS(rq->i915));
 
-	if (vm) {
-		ret = load_pd_dir(rq, i915_vm_to_ppgtt(vm));
-		if (ret)
-			return ret;
-	}
+	ret = switch_mm(rq, vm_alias(ce));
+	if (ret)
+		return ret;
 
 	if (ce->state) {
 		u32 hw_flags;
 
 		GEM_BUG_ON(rq->engine->id != RCS0);
 
-		/*
-		 * The kernel context(s) is treated as pure scratch and is not
-		 * expected to retain any state (as we sacrifice it during
-		 * suspend and on resume it may be corrupted). This is ok,
-		 * as nothing actually executes using the kernel context; it
-		 * is purely used for flushing user contexts.
-		 */
 		hw_flags = 0;
-		if (i915_gem_context_is_kernel(rq->gem_context))
+		if (!test_bit(CONTEXT_VALID_BIT, &ce->flags))
 			hw_flags = MI_RESTORE_INHIBIT;
 
 		ret = mi_set_context(rq, hw_flags);
@@ -1609,34 +1620,6 @@ static int switch_context(struct i915_request *rq)
 			return ret;
 	}
 
-	if (vm) {
-		struct intel_engine_cs *engine = rq->engine;
-
-		ret = engine->emit_flush(rq, EMIT_INVALIDATE);
-		if (ret)
-			return ret;
-
-		ret = flush_pd_dir(rq);
-		if (ret)
-			return ret;
-
-		/*
-		 * Not only do we need a full barrier (post-sync write) after
-		 * invalidating the TLBs, but we need to wait a little bit
-		 * longer. Whether this is merely delaying us, or the
-		 * subsequent flush is a key part of serialising with the
-		 * post-sync op, this extra pass appears vital before a
-		 * mm switch!
-		 */
-		ret = engine->emit_flush(rq, EMIT_INVALIDATE);
-		if (ret)
-			return ret;
-
-		ret = engine->emit_flush(rq, EMIT_FLUSH);
-		if (ret)
-			return ret;
-	}
-
 	ret = remap_l3(rq);
 	if (ret)
 		return ret;
@@ -1648,7 +1631,7 @@ static int ring_request_alloc(struct i915_request *request)
 {
 	int ret;
 
-	GEM_BUG_ON(!intel_context_is_pinned(request->hw_context));
+	GEM_BUG_ON(!intel_context_is_pinned(request->context));
 	GEM_BUG_ON(i915_request_timeline(request)->has_initial_breadcrumb);
 
 	/*
@@ -1804,7 +1787,6 @@ static int gen6_ring_flush(struct i915_request *rq, u32 mode)
 static void i9xx_set_default_submission(struct intel_engine_cs *engine)
 {
 	engine->submit_request = i9xx_submit_request;
-	engine->cancel_requests = cancel_requests;
 
 	engine->park = NULL;
 	engine->unpark = NULL;
@@ -1816,7 +1798,7 @@ static void gen6_bsd_set_default_submission(struct intel_engine_cs *engine)
 	engine->submit_request = gen6_bsd_submit_request;
 }
 
-static void ring_destroy(struct intel_engine_cs *engine)
+static void ring_release(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
 
@@ -1830,8 +1812,6 @@ static void ring_destroy(struct intel_engine_cs *engine)
 
 	intel_timeline_unpin(engine->legacy.timeline);
 	intel_timeline_put(engine->legacy.timeline);
-
-	kfree(engine);
 }
 
 static void setup_irq(struct intel_engine_cs *engine)
@@ -1862,11 +1842,12 @@ static void setup_common(struct intel_engine_cs *engine)
 
 	setup_irq(engine);
 
-	engine->destroy = ring_destroy;
+	engine->release = ring_release;
 
 	engine->resume = xcs_resume;
 	engine->reset.prepare = reset_prepare;
-	engine->reset.reset = reset_ring;
+	engine->reset.rewind = reset_rewind;
+	engine->reset.cancel = reset_cancel;
 	engine->reset.finish = reset_finish;
 
 	engine->cops = &ring_context_ops;
@@ -1977,6 +1958,10 @@ static void setup_vecs(struct intel_engine_cs *engine)
 
 int intel_ring_submission_setup(struct intel_engine_cs *engine)
 {
+	struct intel_timeline *timeline;
+	struct intel_ring *ring;
+	int err;
+
 	setup_common(engine);
 
 	switch (engine->class) {
@@ -1997,15 +1982,6 @@ int intel_ring_submission_setup(struct intel_engine_cs *engine)
 		return -ENODEV;
 	}
 
-	return 0;
-}
-
-int intel_ring_submission_init(struct intel_engine_cs *engine)
-{
-	struct intel_timeline *timeline;
-	struct intel_ring *ring;
-	int err;
-
 	timeline = intel_timeline_create(engine->gt, engine->status_page.vma);
 	if (IS_ERR(timeline)) {
 		err = PTR_ERR(timeline);
@@ -2031,16 +2007,10 @@ int intel_ring_submission_init(struct intel_engine_cs *engine)
 	engine->legacy.ring = ring;
 	engine->legacy.timeline = timeline;
 
-	err = intel_engine_init_common(engine);
-	if (err)
-		goto err_ring_unpin;
-
 	GEM_BUG_ON(timeline->hwsp_ggtt != engine->status_page.vma);
 
 	return 0;
 
-err_ring_unpin:
-	intel_ring_unpin(ring);
 err_ring:
 	intel_ring_put(ring);
 err_timeline_unpin:
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.c b/drivers/gpu/drm/i915/gt/intel_rps.c
index 20d6ee148afc..f232036c3c7a 100644
--- a/drivers/gpu/drm/i915/gt/intel_rps.c
+++ b/drivers/gpu/drm/i915/gt/intel_rps.c
@@ -37,6 +37,11 @@ static u32 rps_pm_sanitize_mask(struct intel_rps *rps, u32 mask)
 	return mask & ~rps->pm_intrmsk_mbz;
 }
 
+static inline void set(struct intel_uncore *uncore, i915_reg_t reg, u32 val)
+{
+	intel_uncore_write_fw(uncore, reg, val);
+}
+
 static u32 rps_pm_mask(struct intel_rps *rps, u8 val)
 {
 	u32 mask = 0;
@@ -78,8 +83,7 @@ static void rps_enable_interrupts(struct intel_rps *rps)
 	gen6_gt_pm_enable_irq(gt, rps->pm_events);
 	spin_unlock_irq(&gt->irq_lock);
 
-	intel_uncore_write(gt->uncore, GEN6_PMINTRMSK,
-			   rps_pm_mask(rps, rps->cur_freq));
+	set(gt->uncore, GEN6_PMINTRMSK, rps_pm_mask(rps, rps->cur_freq));
 }
 
 static void gen6_rps_reset_interrupts(struct intel_rps *rps)
@@ -113,8 +117,7 @@ static void rps_disable_interrupts(struct intel_rps *rps)
 
 	rps->pm_events = 0;
 
-	intel_uncore_write(gt->uncore, GEN6_PMINTRMSK,
-			   rps_pm_sanitize_mask(rps, ~0u));
+	set(gt->uncore, GEN6_PMINTRMSK, rps_pm_sanitize_mask(rps, ~0u));
 
 	spin_lock_irq(&gt->irq_lock);
 	gen6_gt_pm_disable_irq(gt, GEN6_PM_RPS_EVENTS);
@@ -573,25 +576,21 @@ static void rps_set_power(struct intel_rps *rps, int new_power)
 	if (IS_VALLEYVIEW(i915))
 		goto skip_hw_write;
 
-	intel_uncore_write(uncore, GEN6_RP_UP_EI,
-			   GT_INTERVAL_FROM_US(i915, ei_up));
-	intel_uncore_write(uncore, GEN6_RP_UP_THRESHOLD,
-			   GT_INTERVAL_FROM_US(i915,
-					       ei_up * threshold_up / 100));
-
-	intel_uncore_write(uncore, GEN6_RP_DOWN_EI,
-			   GT_INTERVAL_FROM_US(i915, ei_down));
-	intel_uncore_write(uncore, GEN6_RP_DOWN_THRESHOLD,
-			   GT_INTERVAL_FROM_US(i915,
-					       ei_down * threshold_down / 100));
-
-	intel_uncore_write(uncore, GEN6_RP_CONTROL,
-			   (INTEL_GEN(i915) > 9 ? 0 : GEN6_RP_MEDIA_TURBO) |
-			   GEN6_RP_MEDIA_HW_NORMAL_MODE |
-			   GEN6_RP_MEDIA_IS_GFX |
-			   GEN6_RP_ENABLE |
-			   GEN6_RP_UP_BUSY_AVG |
-			   GEN6_RP_DOWN_IDLE_AVG);
+	set(uncore, GEN6_RP_UP_EI, GT_INTERVAL_FROM_US(i915, ei_up));
+	set(uncore, GEN6_RP_UP_THRESHOLD,
+	    GT_INTERVAL_FROM_US(i915, ei_up * threshold_up / 100));
+
+	set(uncore, GEN6_RP_DOWN_EI, GT_INTERVAL_FROM_US(i915, ei_down));
+	set(uncore, GEN6_RP_DOWN_THRESHOLD,
+	    GT_INTERVAL_FROM_US(i915, ei_down * threshold_down / 100));
+
+	set(uncore, GEN6_RP_CONTROL,
+	    (INTEL_GEN(i915) > 9 ? 0 : GEN6_RP_MEDIA_TURBO) |
+	    GEN6_RP_MEDIA_HW_NORMAL_MODE |
+	    GEN6_RP_MEDIA_IS_GFX |
+	    GEN6_RP_ENABLE |
+	    GEN6_RP_UP_BUSY_AVG |
+	    GEN6_RP_DOWN_IDLE_AVG);
 
 skip_hw_write:
 	rps->power.mode = new_power;
@@ -666,7 +665,7 @@ static int gen6_rps_set(struct intel_rps *rps, u8 val)
 		swreq = (GEN6_FREQUENCY(val) |
 			 GEN6_OFFSET(0) |
 			 GEN6_AGGRESSIVE_TURBO);
-	intel_uncore_write(uncore, GEN6_RPNSWREQ, swreq);
+	set(uncore, GEN6_RPNSWREQ, swreq);
 
 	return 0;
 }
@@ -683,7 +682,7 @@ static int vlv_rps_set(struct intel_rps *rps, u8 val)
 	return err;
 }
 
-static int rps_set(struct intel_rps *rps, u8 val)
+static int rps_set(struct intel_rps *rps, u8 val, bool update)
 {
 	struct drm_i915_private *i915 = rps_to_i915(rps);
 	int err;
@@ -701,7 +700,8 @@ static int rps_set(struct intel_rps *rps, u8 val)
 	if (err)
 		return err;
 
-	gen6_rps_set_thresholds(rps, val);
+	if (update)
+		gen6_rps_set_thresholds(rps, val);
 	rps->last_freq = val;
 
 	return 0;
@@ -761,7 +761,7 @@ void intel_rps_park(struct intel_rps *rps)
 	 * power than the render powerwell.
 	 */
 	intel_uncore_forcewake_get(rps_to_uncore(rps), FORCEWAKE_MEDIA);
-	rps_set(rps, rps->idle_freq);
+	rps_set(rps, rps->idle_freq, false);
 	intel_uncore_forcewake_put(rps_to_uncore(rps), FORCEWAKE_MEDIA);
 }
 
@@ -790,14 +790,16 @@ void intel_rps_boost(struct i915_request *rq)
 
 int intel_rps_set(struct intel_rps *rps, u8 val)
 {
-	int err = 0;
+	int err;
 
 	lockdep_assert_held(&rps->lock);
 	GEM_BUG_ON(val > rps->max_freq);
 	GEM_BUG_ON(val < rps->min_freq);
 
 	if (rps->active) {
-		err = rps_set(rps, val);
+		err = rps_set(rps, val, true);
+		if (err)
+			return err;
 
 		/*
 		 * Make sure we continue to get interrupts
@@ -806,18 +808,15 @@ int intel_rps_set(struct intel_rps *rps, u8 val)
 		if (INTEL_GEN(rps_to_i915(rps)) >= 6) {
 			struct intel_uncore *uncore = rps_to_uncore(rps);
 
-			intel_uncore_write(uncore, GEN6_RP_INTERRUPT_LIMITS,
-					   rps_limits(rps, val));
+			set(uncore,
+			    GEN6_RP_INTERRUPT_LIMITS, rps_limits(rps, val));
 
-			intel_uncore_write(uncore, GEN6_PMINTRMSK,
-					   rps_pm_mask(rps, val));
+			set(uncore, GEN6_PMINTRMSK, rps_pm_mask(rps, val));
 		}
 	}
 
-	if (err == 0)
-		rps->cur_freq = val;
-
-	return err;
+	rps->cur_freq = val;
+	return 0;
 }
 
 static void gen6_rps_init(struct intel_rps *rps)
@@ -878,7 +877,7 @@ static bool rps_reset(struct intel_rps *rps)
 	rps->power.mode = -1;
 	rps->last_freq = -1;
 
-	if (rps_set(rps, rps->min_freq)) {
+	if (rps_set(rps, rps->min_freq, true)) {
 		DRM_ERROR("Failed to reset RPS to initial values\n");
 		return false;
 	}
@@ -1201,7 +1200,7 @@ void intel_rps_enable(struct intel_rps *rps)
 
 static void gen6_rps_disable(struct intel_rps *rps)
 {
-	intel_uncore_write(rps_to_uncore(rps), GEN6_RP_CONTROL, 0);
+	set(rps_to_uncore(rps), GEN6_RP_CONTROL, 0);
 }
 
 void intel_rps_disable(struct intel_rps *rps)
@@ -1566,7 +1565,7 @@ void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir)
 		return;
 
 	if (pm_iir & PM_VEBOX_USER_INTERRUPT)
-		intel_engine_breadcrumbs_irq(gt->engine[VECS0]);
+		intel_engine_signal_breadcrumbs(gt->engine[VECS0]);
 
 	if (pm_iir & PM_VEBOX_CS_ERROR_INTERRUPT)
 		DRM_DEBUG("Command parser error, pm_iir 0x%08x\n", pm_iir);
@@ -1663,23 +1662,53 @@ void intel_rps_init(struct intel_rps *rps)
 	if (INTEL_GEN(i915) <= 7)
 		rps->pm_intrmsk_mbz |= GEN6_PM_RP_UP_EI_EXPIRED;
 
-	if (INTEL_GEN(i915) >= 8)
+	if (INTEL_GEN(i915) >= 8 && INTEL_GEN(i915) < 11)
 		rps->pm_intrmsk_mbz |= GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC;
 }
 
-u32 intel_get_cagf(struct intel_rps *rps, u32 rpstat)
+u32 intel_rps_get_cagf(struct intel_rps *rps, u32 rpstat)
 {
 	struct drm_i915_private *i915 = rps_to_i915(rps);
 	u32 cagf;
 
-	if (INTEL_GEN(i915) >= 9)
+	if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915))
+		cagf = (rpstat >> 8) & 0xff;
+	else if (INTEL_GEN(i915) >= 9)
 		cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT;
 	else if (IS_HASWELL(i915) || IS_BROADWELL(i915))
 		cagf = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT;
 	else
 		cagf = (rpstat & GEN6_CAGF_MASK) >> GEN6_CAGF_SHIFT;
 
-	return  cagf;
+	return cagf;
+}
+
+static u32 read_cagf(struct intel_rps *rps)
+{
+	struct drm_i915_private *i915 = rps_to_i915(rps);
+	u32 freq;
+
+	if (IS_VALLEYVIEW(i915) || IS_CHERRYVIEW(i915)) {
+		vlv_punit_get(i915);
+		freq = vlv_punit_read(i915, PUNIT_REG_GPU_FREQ_STS);
+		vlv_punit_put(i915);
+	} else {
+		freq = intel_uncore_read(rps_to_gt(rps)->uncore, GEN6_RPSTAT1);
+	}
+
+	return intel_rps_get_cagf(rps, freq);
+}
+
+u32 intel_rps_read_actual_frequency(struct intel_rps *rps)
+{
+	struct intel_runtime_pm *rpm = rps_to_gt(rps)->uncore->rpm;
+	intel_wakeref_t wakeref;
+	u32 freq = 0;
+
+	with_intel_runtime_pm_if_in_use(rpm, wakeref)
+		freq = intel_gpu_freq(rps, read_cagf(rps));
+
+	return freq;
 }
 
 /* External interface for intel_ips.ko */
@@ -1715,6 +1744,7 @@ void intel_rps_driver_register(struct intel_rps *rps)
 	 * set up, to avoid intel-ips sneaking in and reading bogus values.
 	 */
 	if (IS_GEN(gt->i915, 5)) {
+		GEM_BUG_ON(ips_mchdev);
 		rcu_assign_pointer(ips_mchdev, gt->i915);
 		ips_ping_for_i915_load();
 	}
@@ -1722,7 +1752,8 @@ void intel_rps_driver_register(struct intel_rps *rps)
 
 void intel_rps_driver_unregister(struct intel_rps *rps)
 {
-	rcu_assign_pointer(ips_mchdev, NULL);
+	if (rcu_access_pointer(ips_mchdev) == rps_to_i915(rps))
+		rcu_assign_pointer(ips_mchdev, NULL);
 }
 
 static struct drm_i915_private *mchdev_get(void)
diff --git a/drivers/gpu/drm/i915/gt/intel_rps.h b/drivers/gpu/drm/i915/gt/intel_rps.h
index 9518c66c9792..dfa98194f3b2 100644
--- a/drivers/gpu/drm/i915/gt/intel_rps.h
+++ b/drivers/gpu/drm/i915/gt/intel_rps.h
@@ -29,7 +29,8 @@ void intel_rps_mark_interactive(struct intel_rps *rps, bool interactive);
 
 int intel_gpu_freq(struct intel_rps *rps, int val);
 int intel_freq_opcode(struct intel_rps *rps, int val);
-u32 intel_get_cagf(struct intel_rps *rps, u32 rpstat1);
+u32 intel_rps_get_cagf(struct intel_rps *rps, u32 rpstat1);
+u32 intel_rps_read_actual_frequency(struct intel_rps *rps);
 
 void gen5_rps_irq_handler(struct intel_rps *rps);
 void gen6_rps_irq_handler(struct intel_rps *rps, u32 pm_iir);
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline.c b/drivers/gpu/drm/i915/gt/intel_timeline.c
index 649798c184fb..ee5dc4fbdeb9 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline.c
+++ b/drivers/gpu/drm/i915/gt/intel_timeline.c
@@ -15,6 +15,9 @@
 #define ptr_set_bit(ptr, bit) ((typeof(ptr))((unsigned long)(ptr) | BIT(bit)))
 #define ptr_test_bit(ptr, bit) ((unsigned long)(ptr) & BIT(bit))
 
+#define CACHELINE_BITS 6
+#define CACHELINE_FREE CACHELINE_BITS
+
 struct intel_timeline_hwsp {
 	struct intel_gt *gt;
 	struct intel_gt_timelines *gt_timelines;
@@ -23,14 +26,6 @@ struct intel_timeline_hwsp {
 	u64 free_bitmap;
 };
 
-struct intel_timeline_cacheline {
-	struct i915_active active;
-	struct intel_timeline_hwsp *hwsp;
-	void *vaddr;
-#define CACHELINE_BITS 6
-#define CACHELINE_FREE CACHELINE_BITS
-};
-
 static struct i915_vma *__hwsp_alloc(struct intel_gt *gt)
 {
 	struct drm_i915_private *i915 = gt->i915;
@@ -133,7 +128,7 @@ static void __idle_cacheline_free(struct intel_timeline_cacheline *cl)
 	__idle_hwsp_free(cl->hwsp, ptr_unmask_bits(cl->vaddr, CACHELINE_BITS));
 
 	i915_active_fini(&cl->active);
-	kfree(cl);
+	kfree_rcu(cl, rcu);
 }
 
 __i915_active_call
@@ -254,7 +249,7 @@ int intel_timeline_init(struct intel_timeline *timeline,
 
 	mutex_init(&timeline->mutex);
 
-	INIT_ACTIVE_FENCE(&timeline->last_request, &timeline->mutex);
+	INIT_ACTIVE_FENCE(&timeline->last_request);
 	INIT_LIST_HEAD(&timeline->requests);
 
 	i915_syncmap_init(&timeline->sync);
@@ -262,7 +257,7 @@ int intel_timeline_init(struct intel_timeline *timeline,
 	return 0;
 }
 
-static void timelines_init(struct intel_gt *gt)
+void intel_gt_init_timelines(struct intel_gt *gt)
 {
 	struct intel_gt_timelines *timelines = &gt->timelines;
 
@@ -273,11 +268,6 @@ static void timelines_init(struct intel_gt *gt)
 	INIT_LIST_HEAD(&timelines->hwsp_free_list);
 }
 
-void intel_timelines_init(struct drm_i915_private *i915)
-{
-	timelines_init(&i915->gt);
-}
-
 void intel_timeline_fini(struct intel_timeline *timeline)
 {
 	GEM_BUG_ON(atomic_read(&timeline->pin_count));
@@ -338,7 +328,6 @@ int intel_timeline_pin(struct intel_timeline *tl)
 void intel_timeline_enter(struct intel_timeline *tl)
 {
 	struct intel_gt_timelines *timelines = &tl->gt->timelines;
-	unsigned long flags;
 
 	/*
 	 * Pretend we are serialised by the timeline->mutex.
@@ -364,16 +353,15 @@ void intel_timeline_enter(struct intel_timeline *tl)
 	if (atomic_add_unless(&tl->active_count, 1, 0))
 		return;
 
-	spin_lock_irqsave(&timelines->lock, flags);
+	spin_lock(&timelines->lock);
 	if (!atomic_fetch_inc(&tl->active_count))
 		list_add_tail(&tl->link, &timelines->active_list);
-	spin_unlock_irqrestore(&timelines->lock, flags);
+	spin_unlock(&timelines->lock);
 }
 
 void intel_timeline_exit(struct intel_timeline *tl)
 {
 	struct intel_gt_timelines *timelines = &tl->gt->timelines;
-	unsigned long flags;
 
 	/* See intel_timeline_enter() */
 	lockdep_assert_held(&tl->mutex);
@@ -382,10 +370,10 @@ void intel_timeline_exit(struct intel_timeline *tl)
 	if (atomic_add_unless(&tl->active_count, -1, 1))
 		return;
 
-	spin_lock_irqsave(&timelines->lock, flags);
+	spin_lock(&timelines->lock);
 	if (atomic_dec_and_test(&tl->active_count))
 		list_del(&tl->link);
-	spin_unlock_irqrestore(&timelines->lock, flags);
+	spin_unlock(&timelines->lock);
 
 	/*
 	 * Since this timeline is idle, all bariers upon which we were waiting
@@ -521,46 +509,35 @@ int intel_timeline_read_hwsp(struct i915_request *from,
 			     struct i915_request *to,
 			     u32 *hwsp)
 {
-	struct intel_timeline *tl;
+	struct intel_timeline_cacheline *cl;
 	int err;
 
+	GEM_BUG_ON(!rcu_access_pointer(from->hwsp_cacheline));
+
 	rcu_read_lock();
-	tl = rcu_dereference(from->timeline);
-	if (i915_request_completed(from) || !kref_get_unless_zero(&tl->kref))
-		tl = NULL;
+	cl = rcu_dereference(from->hwsp_cacheline);
+	if (unlikely(!i915_active_acquire_if_busy(&cl->active)))
+		goto unlock; /* seqno wrapped and completed! */
+	if (unlikely(i915_request_completed(from)))
+		goto release;
 	rcu_read_unlock();
-	if (!tl) /* already completed */
-		return 1;
-
-	GEM_BUG_ON(rcu_access_pointer(to->timeline) == tl);
 
-	err = -EBUSY;
-	if (mutex_trylock(&tl->mutex)) {
-		struct intel_timeline_cacheline *cl = from->hwsp_cacheline;
-
-		if (i915_request_completed(from)) {
-			err = 1;
-			goto unlock;
-		}
+	err = cacheline_ref(cl, to);
+	if (err)
+		goto out;
 
-		err = cacheline_ref(cl, to);
-		if (err)
-			goto unlock;
+	*hwsp = i915_ggtt_offset(cl->hwsp->vma) +
+		ptr_unmask_bits(cl->vaddr, CACHELINE_BITS) * CACHELINE_BYTES;
 
-		if (likely(cl == tl->hwsp_cacheline)) {
-			*hwsp = tl->hwsp_offset;
-		} else { /* across a seqno wrap, recover the original offset */
-			*hwsp = i915_ggtt_offset(cl->hwsp->vma) +
-				ptr_unmask_bits(cl->vaddr, CACHELINE_BITS) *
-				CACHELINE_BYTES;
-		}
+out:
+	i915_active_release(&cl->active);
+	return err;
 
+release:
+	i915_active_release(&cl->active);
 unlock:
-		mutex_unlock(&tl->mutex);
-	}
-	intel_timeline_put(tl);
-
-	return err;
+	rcu_read_unlock();
+	return 1;
 }
 
 void intel_timeline_unpin(struct intel_timeline *tl)
@@ -583,7 +560,7 @@ void __intel_timeline_free(struct kref *kref)
 	kfree_rcu(timeline, rcu);
 }
 
-static void timelines_fini(struct intel_gt *gt)
+void intel_gt_fini_timelines(struct intel_gt *gt)
 {
 	struct intel_gt_timelines *timelines = &gt->timelines;
 
@@ -591,11 +568,6 @@ static void timelines_fini(struct intel_gt *gt)
 	GEM_BUG_ON(!list_empty(&timelines->hwsp_free_list));
 }
 
-void intel_timelines_fini(struct drm_i915_private *i915)
-{
-	timelines_fini(&i915->gt);
-}
-
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
 #include "gt/selftests/mock_timeline.c"
 #include "gt/selftest_timeline.c"
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline.h b/drivers/gpu/drm/i915/gt/intel_timeline.h
index f583af1ba18d..f5b7eade3809 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline.h
+++ b/drivers/gpu/drm/i915/gt/intel_timeline.h
@@ -88,7 +88,7 @@ int intel_timeline_read_hwsp(struct i915_request *from,
 			     struct i915_request *until,
 			     u32 *hwsp_offset);
 
-void intel_timelines_init(struct drm_i915_private *i915);
-void intel_timelines_fini(struct drm_i915_private *i915);
+void intel_gt_init_timelines(struct intel_gt *gt);
+void intel_gt_fini_timelines(struct intel_gt *gt);
 
 #endif
diff --git a/drivers/gpu/drm/i915/gt/intel_timeline_types.h b/drivers/gpu/drm/i915/gt/intel_timeline_types.h
index aaf15cbe1ce1..02181c5020db 100644
--- a/drivers/gpu/drm/i915/gt/intel_timeline_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_timeline_types.h
@@ -10,14 +10,15 @@
 #include <linux/list.h>
 #include <linux/kref.h>
 #include <linux/mutex.h>
+#include <linux/rcupdate.h>
 #include <linux/types.h>
 
 #include "i915_active_types.h"
 
-struct drm_i915_private;
 struct i915_vma;
-struct intel_timeline_cacheline;
 struct i915_syncmap;
+struct intel_gt;
+struct intel_timeline_hwsp;
 
 struct intel_timeline {
 	u64 fence_context;
@@ -87,4 +88,13 @@ struct intel_timeline {
 	struct rcu_head rcu;
 };
 
+struct intel_timeline_cacheline {
+	struct i915_active active;
+
+	struct intel_timeline_hwsp *hwsp;
+	void *vaddr;
+
+	struct rcu_head rcu;
+};
+
 #endif /* __I915_TIMELINE_TYPES_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_workarounds.c b/drivers/gpu/drm/i915/gt/intel_workarounds.c
index e4bccc14602f..195ccf7db272 100644
--- a/drivers/gpu/drm/i915/gt/intel_workarounds.c
+++ b/drivers/gpu/drm/i915/gt/intel_workarounds.c
@@ -6,6 +6,7 @@
 
 #include "i915_drv.h"
 #include "intel_context.h"
+#include "intel_engine_pm.h"
 #include "intel_gt.h"
 #include "intel_ring.h"
 #include "intel_workarounds.h"
@@ -146,21 +147,27 @@ static void _wa_add(struct i915_wa_list *wal, const struct i915_wa *wa)
 	}
 }
 
-static void
-wa_write_masked_or(struct i915_wa_list *wal, i915_reg_t reg, u32 mask,
-		   u32 val)
+static void wa_add(struct i915_wa_list *wal, i915_reg_t reg, u32 mask,
+		   u32 val, u32 read_mask)
 {
 	struct i915_wa wa = {
 		.reg  = reg,
 		.mask = mask,
 		.val  = val,
-		.read = mask,
+		.read = read_mask,
 	};
 
 	_wa_add(wal, &wa);
 }
 
 static void
+wa_write_masked_or(struct i915_wa_list *wal, i915_reg_t reg, u32 mask,
+		   u32 val)
+{
+	wa_add(wal, reg, mask, val, mask);
+}
+
+static void
 wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val)
 {
 	wa_write_masked_or(wal, reg, val, _MASKED_BIT_ENABLE(val));
@@ -568,9 +575,24 @@ static void icl_ctx_workarounds_init(struct intel_engine_cs *engine,
 static void tgl_ctx_workarounds_init(struct intel_engine_cs *engine,
 				     struct i915_wa_list *wal)
 {
+	u32 val;
+
 	/* Wa_1409142259:tgl */
 	WA_SET_BIT_MASKED(GEN11_COMMON_SLICE_CHICKEN3,
 			  GEN12_DISABLE_CPS_AWARE_COLOR_PIPE);
+
+	/* Wa_1604555607:tgl */
+	val = intel_uncore_read(engine->uncore, FF_MODE2);
+	val &= ~FF_MODE2_TDS_TIMER_MASK;
+	val |= FF_MODE2_TDS_TIMER_128;
+	/*
+	 * FIXME: FF_MODE2 register is not readable till TGL B0. We can
+	 * enable verification of WA from the later steppings, which enables
+	 * the read of FF_MODE2.
+	 */
+	wa_add(wal, FF_MODE2, FF_MODE2_TDS_TIMER_MASK, val,
+	       IS_TGL_REVID(engine->i915, TGL_REVID_A0, TGL_REVID_A0) ? 0 :
+			    FF_MODE2_TDS_TIMER_MASK);
 }
 
 static void
@@ -1315,6 +1337,14 @@ rcs_engine_wa_init(struct intel_engine_cs *engine, struct i915_wa_list *wal)
 			     GEN6_RC_SLEEP_PSMI_CONTROL,
 			     GEN12_WAIT_FOR_EVENT_POWER_DOWN_DISABLE |
 			     GEN8_RC_SEMA_IDLE_MSG_DISABLE);
+
+		/*
+		 * Wa_1606679103:tgl
+		 * (see also Wa_1606682166:icl)
+		 */
+		wa_write_or(wal,
+			    GEN7_SARCHKMD,
+			    GEN7_DISABLE_SAMPLER_PREFETCH);
 	}
 
 	if (IS_GEN(i915, 11)) {
@@ -1574,7 +1604,9 @@ static int engine_wa_list_verify(struct intel_context *ce,
 	if (IS_ERR(vma))
 		return PTR_ERR(vma);
 
+	intel_engine_pm_get(ce->engine);
 	rq = intel_context_create_request(ce);
+	intel_engine_pm_put(ce->engine);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto err_vma;
@@ -1584,16 +1616,17 @@ static int engine_wa_list_verify(struct intel_context *ce,
 	if (err)
 		goto err_vma;
 
+	i915_request_get(rq);
 	i915_request_add(rq);
 	if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 		err = -ETIME;
-		goto err_vma;
+		goto err_rq;
 	}
 
 	results = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
 	if (IS_ERR(results)) {
 		err = PTR_ERR(results);
-		goto err_vma;
+		goto err_rq;
 	}
 
 	err = 0;
@@ -1607,6 +1640,8 @@ static int engine_wa_list_verify(struct intel_context *ce,
 
 	i915_gem_object_unpin_map(vma->obj);
 
+err_rq:
+	i915_request_put(rq);
 err_vma:
 	i915_vma_unpin(vma);
 	i915_vma_put(vma);
diff --git a/drivers/gpu/drm/i915/gt/mock_engine.c b/drivers/gpu/drm/i915/gt/mock_engine.c
index 83f549d203a0..4e1eafa94be9 100644
--- a/drivers/gpu/drm/i915/gt/mock_engine.c
+++ b/drivers/gpu/drm/i915/gt/mock_engine.c
@@ -77,7 +77,7 @@ static void advance(struct i915_request *request)
 	i915_request_mark_complete(request);
 	GEM_BUG_ON(!i915_request_completed(request));
 
-	intel_engine_queue_breadcrumbs(request->engine);
+	intel_engine_signal_breadcrumbs(request->engine);
 }
 
 static void hw_delay_complete(struct timer_list *t)
@@ -207,16 +207,12 @@ static void mock_reset_prepare(struct intel_engine_cs *engine)
 {
 }
 
-static void mock_reset(struct intel_engine_cs *engine, bool stalled)
+static void mock_reset_rewind(struct intel_engine_cs *engine, bool stalled)
 {
 	GEM_BUG_ON(stalled);
 }
 
-static void mock_reset_finish(struct intel_engine_cs *engine)
-{
-}
-
-static void mock_cancel_requests(struct intel_engine_cs *engine)
+static void mock_reset_cancel(struct intel_engine_cs *engine)
 {
 	struct i915_request *request;
 	unsigned long flags;
@@ -234,6 +230,24 @@ static void mock_cancel_requests(struct intel_engine_cs *engine)
 	spin_unlock_irqrestore(&engine->active.lock, flags);
 }
 
+static void mock_reset_finish(struct intel_engine_cs *engine)
+{
+}
+
+static void mock_engine_release(struct intel_engine_cs *engine)
+{
+	struct mock_engine *mock =
+		container_of(engine, typeof(*mock), base);
+
+	GEM_BUG_ON(timer_pending(&mock->hw_delay));
+
+	intel_context_unpin(engine->kernel_context);
+	intel_context_put(engine->kernel_context);
+
+	intel_engine_fini_retire(engine);
+	intel_engine_fini_breadcrumbs(engine);
+}
+
 struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 				    const char *name,
 				    int id)
@@ -265,9 +279,11 @@ struct intel_engine_cs *mock_engine(struct drm_i915_private *i915,
 	engine->base.submit_request = mock_submit_request;
 
 	engine->base.reset.prepare = mock_reset_prepare;
-	engine->base.reset.reset = mock_reset;
+	engine->base.reset.rewind = mock_reset_rewind;
+	engine->base.reset.cancel = mock_reset_cancel;
 	engine->base.reset.finish = mock_reset_finish;
-	engine->base.cancel_requests = mock_cancel_requests;
+
+	engine->base.release = mock_engine_release;
 
 	i915->gt.engine[id] = &engine->base;
 	i915->gt.engine_class[0][id] = &engine->base;
@@ -290,6 +306,7 @@ int mock_engine_init(struct intel_engine_cs *engine)
 	intel_engine_init_breadcrumbs(engine);
 	intel_engine_init_execlists(engine);
 	intel_engine_init__pm(engine);
+	intel_engine_init_retire(engine);
 	intel_engine_pool_init(&engine->pool);
 
 	ce = create_kernel_context(engine);
@@ -321,18 +338,3 @@ void mock_engine_flush(struct intel_engine_cs *engine)
 void mock_engine_reset(struct intel_engine_cs *engine)
 {
 }
-
-void mock_engine_free(struct intel_engine_cs *engine)
-{
-	struct mock_engine *mock =
-		container_of(engine, typeof(*mock), base);
-
-	GEM_BUG_ON(timer_pending(&mock->hw_delay));
-
-	intel_context_unpin(engine->kernel_context);
-	intel_context_put(engine->kernel_context);
-
-	intel_engine_fini_breadcrumbs(engine);
-
-	kfree(engine);
-}
diff --git a/drivers/gpu/drm/i915/gt/selftest_context.c b/drivers/gpu/drm/i915/gt/selftest_context.c
index bc720defc6b8..e874dfaa5316 100644
--- a/drivers/gpu/drm/i915/gt/selftest_context.c
+++ b/drivers/gpu/drm/i915/gt/selftest_context.c
@@ -5,6 +5,7 @@
  */
 
 #include "i915_selftest.h"
+#include "intel_engine_heartbeat.h"
 #include "intel_engine_pm.h"
 #include "intel_gt.h"
 
@@ -47,35 +48,36 @@ static int context_sync(struct intel_context *ce)
 
 	mutex_lock(&tl->mutex);
 	do {
-		struct dma_fence *fence;
+		struct i915_request *rq;
 		long timeout;
 
-		fence = i915_active_fence_get(&tl->last_request);
-		if (!fence)
+		if (list_empty(&tl->requests))
 			break;
 
-		timeout = dma_fence_wait_timeout(fence, false, HZ / 10);
+		rq = list_last_entry(&tl->requests, typeof(*rq), link);
+		i915_request_get(rq);
+
+		timeout = i915_request_wait(rq, 0, HZ / 10);
 		if (timeout < 0)
 			err = timeout;
 		else
-			i915_request_retire_upto(to_request(fence));
+			i915_request_retire_upto(rq);
 
-		dma_fence_put(fence);
+		i915_request_put(rq);
 	} while (!err);
 	mutex_unlock(&tl->mutex);
 
 	return err;
 }
 
-static int __live_context_size(struct intel_engine_cs *engine,
-			       struct i915_gem_context *fixme)
+static int __live_context_size(struct intel_engine_cs *engine)
 {
 	struct intel_context *ce;
 	struct i915_request *rq;
 	void *vaddr;
 	int err;
 
-	ce = intel_context_create(fixme, engine);
+	ce = intel_context_create(engine);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
 
@@ -118,7 +120,7 @@ static int __live_context_size(struct intel_engine_cs *engine,
 		goto err_unpin;
 
 	/* Force the context switch */
-	rq = i915_request_create(engine->kernel_context);
+	rq = intel_engine_create_kernel_request(engine);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto err_unpin;
@@ -143,7 +145,6 @@ static int live_context_size(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *fixme;
 	enum intel_engine_id id;
 	int err = 0;
 
@@ -152,10 +153,6 @@ static int live_context_size(void *arg)
 	 * HW tries to write past the end of one.
 	 */
 
-	fixme = kernel_context(gt->i915);
-	if (IS_ERR(fixme))
-		return PTR_ERR(fixme);
-
 	for_each_engine(engine, gt, id) {
 		struct {
 			struct drm_i915_gem_object *state;
@@ -180,7 +177,7 @@ static int live_context_size(void *arg)
 		/* Overlaps with the execlists redzone */
 		engine->context_size += I915_GTT_PAGE_SIZE;
 
-		err = __live_context_size(engine, fixme);
+		err = __live_context_size(engine);
 
 		engine->context_size -= I915_GTT_PAGE_SIZE;
 
@@ -193,13 +190,12 @@ static int live_context_size(void *arg)
 			break;
 	}
 
-	kernel_context_close(fixme);
 	return err;
 }
 
-static int __live_active_context(struct intel_engine_cs *engine,
-				 struct i915_gem_context *fixme)
+static int __live_active_context(struct intel_engine_cs *engine)
 {
+	unsigned long saved_heartbeat;
 	struct intel_context *ce;
 	int pass;
 	int err;
@@ -223,40 +219,55 @@ static int __live_active_context(struct intel_engine_cs *engine,
 		return -EINVAL;
 	}
 
-	ce = intel_context_create(fixme, engine);
+	ce = intel_context_create(engine);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
 
+	saved_heartbeat = engine->props.heartbeat_interval_ms;
+	engine->props.heartbeat_interval_ms = 0;
+
 	for (pass = 0; pass <= 2; pass++) {
 		struct i915_request *rq;
 
+		intel_engine_pm_get(engine);
+
 		rq = intel_context_create_request(ce);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
-			goto err;
+			goto out_engine;
 		}
 
 		err = request_sync(rq);
 		if (err)
-			goto err;
+			goto out_engine;
 
 		/* Context will be kept active until after an idle-barrier. */
 		if (i915_active_is_idle(&ce->active)) {
 			pr_err("context is not active; expected idle-barrier (%s pass %d)\n",
 			       engine->name, pass);
 			err = -EINVAL;
-			goto err;
+			goto out_engine;
 		}
 
 		if (!intel_engine_pm_is_awake(engine)) {
 			pr_err("%s is asleep before idle-barrier\n",
 			       engine->name);
 			err = -EINVAL;
-			goto err;
+			goto out_engine;
 		}
+
+out_engine:
+		intel_engine_pm_put(engine);
+		if (err)
+			goto err;
 	}
 
 	/* Now make sure our idle-barriers are flushed */
+	err = intel_engine_flush_barriers(engine);
+	if (err)
+		goto err;
+
+	/* Wait for the barrier and in the process wait for engine to park */
 	err = context_sync(engine->kernel_context);
 	if (err)
 		goto err;
@@ -266,12 +277,15 @@ static int __live_active_context(struct intel_engine_cs *engine,
 		err = -EINVAL;
 	}
 
+	intel_engine_pm_flush(engine);
+
 	if (intel_engine_pm_is_awake(engine)) {
 		struct drm_printer p = drm_debug_printer(__func__);
 
 		intel_engine_dump(engine, &p,
-				  "%s is still awake after idle-barriers\n",
-				  engine->name);
+				  "%s is still awake:%d after idle-barriers\n",
+				  engine->name,
+				  atomic_read(&engine->wakeref.count));
 		GEM_TRACE_DUMP();
 
 		err = -EINVAL;
@@ -279,6 +293,7 @@ static int __live_active_context(struct intel_engine_cs *engine,
 	}
 
 err:
+	engine->props.heartbeat_interval_ms = saved_heartbeat;
 	intel_context_put(ce);
 	return err;
 }
@@ -287,23 +302,11 @@ static int live_active_context(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *fixme;
 	enum intel_engine_id id;
-	struct drm_file *file;
 	int err = 0;
 
-	file = mock_file(gt->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	fixme = live_context(gt->i915, file);
-	if (IS_ERR(fixme)) {
-		err = PTR_ERR(fixme);
-		goto out_file;
-	}
-
 	for_each_engine(engine, gt, id) {
-		err = __live_active_context(engine, fixme);
+		err = __live_active_context(engine);
 		if (err)
 			break;
 
@@ -312,8 +315,6 @@ static int live_active_context(void *arg)
 			break;
 	}
 
-out_file:
-	mock_file_free(gt->i915, file);
 	return err;
 }
 
@@ -345,10 +346,10 @@ unpin:
 	return err;
 }
 
-static int __live_remote_context(struct intel_engine_cs *engine,
-				 struct i915_gem_context *fixme)
+static int __live_remote_context(struct intel_engine_cs *engine)
 {
 	struct intel_context *local, *remote;
+	unsigned long saved_heartbeat;
 	int pass;
 	int err;
 
@@ -360,16 +361,26 @@ static int __live_remote_context(struct intel_engine_cs *engine,
 	 * clobber the idle-barrier.
 	 */
 
-	remote = intel_context_create(fixme, engine);
+	if (intel_engine_pm_is_awake(engine)) {
+		pr_err("%s is awake before starting %s!\n",
+		       engine->name, __func__);
+		return -EINVAL;
+	}
+
+	remote = intel_context_create(engine);
 	if (IS_ERR(remote))
 		return PTR_ERR(remote);
 
-	local = intel_context_create(fixme, engine);
+	local = intel_context_create(engine);
 	if (IS_ERR(local)) {
 		err = PTR_ERR(local);
 		goto err_remote;
 	}
 
+	saved_heartbeat = engine->props.heartbeat_interval_ms;
+	engine->props.heartbeat_interval_ms = 0;
+	intel_engine_pm_get(engine);
+
 	for (pass = 0; pass <= 2; pass++) {
 		err = __remote_sync(local, remote);
 		if (err)
@@ -387,6 +398,9 @@ static int __live_remote_context(struct intel_engine_cs *engine,
 		}
 	}
 
+	intel_engine_pm_put(engine);
+	engine->props.heartbeat_interval_ms = saved_heartbeat;
+
 	intel_context_put(local);
 err_remote:
 	intel_context_put(remote);
@@ -397,23 +411,11 @@ static int live_remote_context(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *fixme;
 	enum intel_engine_id id;
-	struct drm_file *file;
 	int err = 0;
 
-	file = mock_file(gt->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	fixme = live_context(gt->i915, file);
-	if (IS_ERR(fixme)) {
-		err = PTR_ERR(fixme);
-		goto out_file;
-	}
-
 	for_each_engine(engine, gt, id) {
-		err = __live_remote_context(engine, fixme);
+		err = __live_remote_context(engine);
 		if (err)
 			break;
 
@@ -422,8 +424,6 @@ static int live_remote_context(void *arg)
 			break;
 	}
 
-out_file:
-	mock_file_free(gt->i915, file);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_engine_cs.c b/drivers/gpu/drm/i915/gt/selftest_engine_cs.c
index 3880f07c29b8..f88e445a1cae 100644
--- a/drivers/gpu/drm/i915/gt/selftest_engine_cs.c
+++ b/drivers/gpu/drm/i915/gt/selftest_engine_cs.c
@@ -4,7 +4,365 @@
  * Copyright © 2018 Intel Corporation
  */
 
-#include "../i915_selftest.h"
+#include <linux/sort.h>
+
+#include "intel_gt_pm.h"
+#include "intel_rps.h"
+
+#include "i915_selftest.h"
+#include "selftests/igt_flush_test.h"
+
+#define COUNT 5
+
+static int cmp_u32(const void *A, const void *B)
+{
+	const u32 *a = A, *b = B;
+
+	return *a - *b;
+}
+
+static void perf_begin(struct intel_gt *gt)
+{
+	intel_gt_pm_get(gt);
+
+	/* Boost gpufreq to max [waitboost] and keep it fixed */
+	atomic_inc(&gt->rps.num_waiters);
+	schedule_work(&gt->rps.work);
+	flush_work(&gt->rps.work);
+}
+
+static int perf_end(struct intel_gt *gt)
+{
+	atomic_dec(&gt->rps.num_waiters);
+	intel_gt_pm_put(gt);
+
+	return igt_flush_test(gt->i915);
+}
+
+static int write_timestamp(struct i915_request *rq, int slot)
+{
+	u32 cmd;
+	u32 *cs;
+
+	cs = intel_ring_begin(rq, 4);
+	if (IS_ERR(cs))
+		return PTR_ERR(cs);
+
+	cmd = MI_STORE_REGISTER_MEM | MI_USE_GGTT;
+	if (INTEL_GEN(rq->i915) >= 8)
+		cmd++;
+	*cs++ = cmd;
+	*cs++ = i915_mmio_reg_offset(RING_TIMESTAMP(rq->engine->mmio_base));
+	*cs++ = i915_request_timeline(rq)->hwsp_offset + slot * sizeof(u32);
+	*cs++ = 0;
+
+	intel_ring_advance(rq, cs);
+
+	return 0;
+}
+
+static struct i915_vma *create_empty_batch(struct intel_context *ce)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	u32 *cs;
+	int err;
+
+	obj = i915_gem_object_create_internal(ce->engine->i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	cs = i915_gem_object_pin_map(obj, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_put;
+	}
+
+	cs[0] = MI_BATCH_BUFFER_END;
+
+	i915_gem_object_flush_map(obj);
+
+	vma = i915_vma_instance(obj, ce->vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_unpin;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER);
+	if (err)
+		goto err_unpin;
+
+	i915_gem_object_unpin_map(obj);
+	return vma;
+
+err_unpin:
+	i915_gem_object_unpin_map(obj);
+err_put:
+	i915_gem_object_put(obj);
+	return ERR_PTR(err);
+}
+
+static u32 trifilter(u32 *a)
+{
+	u64 sum;
+
+	sort(a, COUNT, sizeof(*a), cmp_u32, NULL);
+
+	sum = mul_u32_u32(a[2], 2);
+	sum += a[1];
+	sum += a[3];
+
+	return sum >> 2;
+}
+
+static int perf_mi_bb_start(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	if (INTEL_GEN(gt->i915) < 7) /* for per-engine CS_TIMESTAMP */
+		return 0;
+
+	perf_begin(gt);
+	for_each_engine(engine, gt, id) {
+		struct intel_context *ce = engine->kernel_context;
+		struct i915_vma *batch;
+		u32 cycles[COUNT];
+		int i;
+
+		intel_engine_pm_get(engine);
+
+		batch = create_empty_batch(ce);
+		if (IS_ERR(batch)) {
+			err = PTR_ERR(batch);
+			intel_engine_pm_put(engine);
+			break;
+		}
+
+		err = i915_vma_sync(batch);
+		if (err) {
+			intel_engine_pm_put(engine);
+			i915_vma_put(batch);
+			break;
+		}
+
+		for (i = 0; i < ARRAY_SIZE(cycles); i++) {
+			struct i915_request *rq;
+
+			rq = i915_request_create(ce);
+			if (IS_ERR(rq)) {
+				err = PTR_ERR(rq);
+				break;
+			}
+
+			err = write_timestamp(rq, 2);
+			if (err)
+				goto out;
+
+			err = rq->engine->emit_bb_start(rq,
+							batch->node.start, 8,
+							0);
+			if (err)
+				goto out;
+
+			err = write_timestamp(rq, 3);
+			if (err)
+				goto out;
+
+out:
+			i915_request_get(rq);
+			i915_request_add(rq);
+
+			if (i915_request_wait(rq, 0, HZ / 5) < 0)
+				err = -EIO;
+			i915_request_put(rq);
+			if (err)
+				break;
+
+			cycles[i] = rq->hwsp_seqno[3] - rq->hwsp_seqno[2];
+		}
+		i915_vma_put(batch);
+		intel_engine_pm_put(engine);
+		if (err)
+			break;
+
+		pr_info("%s: MI_BB_START cycles: %u\n",
+			engine->name, trifilter(cycles));
+	}
+	if (perf_end(gt))
+		err = -EIO;
+
+	return err;
+}
+
+static struct i915_vma *create_nop_batch(struct intel_context *ce)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	u32 *cs;
+	int err;
+
+	obj = i915_gem_object_create_internal(ce->engine->i915, SZ_64K);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	cs = i915_gem_object_pin_map(obj, I915_MAP_WB);
+	if (IS_ERR(cs)) {
+		err = PTR_ERR(cs);
+		goto err_put;
+	}
+
+	memset(cs, 0, SZ_64K);
+	cs[SZ_64K / sizeof(*cs) - 1] = MI_BATCH_BUFFER_END;
+
+	i915_gem_object_flush_map(obj);
+
+	vma = i915_vma_instance(obj, ce->vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_unpin;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER);
+	if (err)
+		goto err_unpin;
+
+	i915_gem_object_unpin_map(obj);
+	return vma;
+
+err_unpin:
+	i915_gem_object_unpin_map(obj);
+err_put:
+	i915_gem_object_put(obj);
+	return ERR_PTR(err);
+}
+
+static int perf_mi_noop(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	int err = 0;
+
+	if (INTEL_GEN(gt->i915) < 7) /* for per-engine CS_TIMESTAMP */
+		return 0;
+
+	perf_begin(gt);
+	for_each_engine(engine, gt, id) {
+		struct intel_context *ce = engine->kernel_context;
+		struct i915_vma *base, *nop;
+		u32 cycles[COUNT];
+		int i;
+
+		intel_engine_pm_get(engine);
+
+		base = create_empty_batch(ce);
+		if (IS_ERR(base)) {
+			err = PTR_ERR(base);
+			intel_engine_pm_put(engine);
+			break;
+		}
+
+		err = i915_vma_sync(base);
+		if (err) {
+			i915_vma_put(base);
+			intel_engine_pm_put(engine);
+			break;
+		}
+
+		nop = create_nop_batch(ce);
+		if (IS_ERR(nop)) {
+			err = PTR_ERR(nop);
+			i915_vma_put(base);
+			intel_engine_pm_put(engine);
+			break;
+		}
+
+		err = i915_vma_sync(nop);
+		if (err) {
+			i915_vma_put(nop);
+			i915_vma_put(base);
+			intel_engine_pm_put(engine);
+			break;
+		}
+
+		for (i = 0; i < ARRAY_SIZE(cycles); i++) {
+			struct i915_request *rq;
+
+			rq = i915_request_create(ce);
+			if (IS_ERR(rq)) {
+				err = PTR_ERR(rq);
+				break;
+			}
+
+			err = write_timestamp(rq, 2);
+			if (err)
+				goto out;
+
+			err = rq->engine->emit_bb_start(rq,
+							base->node.start, 8,
+							0);
+			if (err)
+				goto out;
+
+			err = write_timestamp(rq, 3);
+			if (err)
+				goto out;
+
+			err = rq->engine->emit_bb_start(rq,
+							nop->node.start,
+							nop->node.size,
+							0);
+			if (err)
+				goto out;
+
+			err = write_timestamp(rq, 4);
+			if (err)
+				goto out;
+
+out:
+			i915_request_get(rq);
+			i915_request_add(rq);
+
+			if (i915_request_wait(rq, 0, HZ / 5) < 0)
+				err = -EIO;
+			i915_request_put(rq);
+			if (err)
+				break;
+
+			cycles[i] =
+				(rq->hwsp_seqno[4] - rq->hwsp_seqno[3]) -
+				(rq->hwsp_seqno[3] - rq->hwsp_seqno[2]);
+		}
+		i915_vma_put(nop);
+		i915_vma_put(base);
+		intel_engine_pm_put(engine);
+		if (err)
+			break;
+
+		pr_info("%s: 16K MI_NOOP cycles: %u\n",
+			engine->name, trifilter(cycles));
+	}
+	if (perf_end(gt))
+		err = -EIO;
+
+	return err;
+}
+
+int intel_engine_cs_perf_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(perf_mi_bb_start),
+		SUBTEST(perf_mi_noop),
+	};
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	return intel_gt_live_subtests(tests, &i915->gt);
+}
 
 static int intel_mmio_bases_check(void *arg)
 {
diff --git a/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c b/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
index e864406bd2d9..43d4d589749f 100644
--- a/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
+++ b/drivers/gpu/drm/i915/gt/selftest_engine_heartbeat.c
@@ -11,6 +11,28 @@
 #include "intel_gt_requests.h"
 #include "i915_selftest.h"
 
+static int timeline_sync(struct intel_timeline *tl)
+{
+	struct dma_fence *fence;
+	long timeout;
+
+	fence = i915_active_fence_get(&tl->last_request);
+	if (!fence)
+		return 0;
+
+	timeout = dma_fence_wait_timeout(fence, true, HZ / 2);
+	dma_fence_put(fence);
+	if (timeout < 0)
+		return timeout;
+
+	return 0;
+}
+
+static int engine_sync_barrier(struct intel_engine_cs *engine)
+{
+	return timeline_sync(engine->kernel_context->timeline);
+}
+
 struct pulse {
 	struct i915_active active;
 	struct kref kref;
@@ -53,9 +75,7 @@ static struct pulse *pulse_create(void)
 
 static void pulse_unlock_wait(struct pulse *p)
 {
-	mutex_lock(&p->active.mutex);
-	mutex_unlock(&p->active.mutex);
-	flush_work(&p->active.work);
+	i915_active_unlock_wait(&p->active);
 }
 
 static int __live_idle_pulse(struct intel_engine_cs *engine,
@@ -92,7 +112,12 @@ static int __live_idle_pulse(struct intel_engine_cs *engine,
 
 	GEM_BUG_ON(!llist_empty(&engine->barrier_tasks));
 
-	if (intel_gt_retire_requests_timeout(engine->gt, HZ / 5)) {
+	if (engine_sync_barrier(engine)) {
+		struct drm_printer m = drm_err_printer("pulse");
+
+		pr_err("%s: no heartbeat pulse?\n", engine->name);
+		intel_engine_dump(engine, &m, "%s", engine->name);
+
 		err = -ETIME;
 		goto out;
 	}
@@ -175,8 +200,7 @@ static int __live_heartbeat_fast(struct intel_engine_cs *engine)
 	int err;
 	int i;
 
-	ce = intel_context_create(engine->kernel_context->gem_context,
-				  engine);
+	ce = intel_context_create(engine);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_gt_pm.c b/drivers/gpu/drm/i915/gt/selftest_gt_pm.c
index d1752f15702a..09ff8e4f88af 100644
--- a/drivers/gpu/drm/i915/gt/selftest_gt_pm.c
+++ b/drivers/gpu/drm/i915/gt/selftest_gt_pm.c
@@ -6,6 +6,7 @@
  */
 
 #include "selftest_llc.h"
+#include "selftest_rc6.h"
 
 static int live_gt_resume(void *arg)
 {
@@ -50,6 +51,7 @@ static int live_gt_resume(void *arg)
 int intel_gt_pm_live_selftests(struct drm_i915_private *i915)
 {
 	static const struct i915_subtest tests[] = {
+		SUBTEST(live_rc6_manual),
 		SUBTEST(live_gt_resume),
 	};
 
@@ -58,3 +60,20 @@ int intel_gt_pm_live_selftests(struct drm_i915_private *i915)
 
 	return intel_gt_live_subtests(tests, &i915->gt);
 }
+
+int intel_gt_pm_late_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		/*
+		 * These tests may leave the system in an undesirable state.
+		 * They are intended to be run last in CI and the system
+		 * rebooted afterwards.
+		 */
+		SUBTEST(live_rc6_ctx_wa),
+	};
+
+	if (intel_gt_is_wedged(&i915->gt))
+		return 0;
+
+	return intel_gt_live_subtests(tests, &i915->gt);
+}
diff --git a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
index 85e9ccf5c304..5dbda2a74272 100644
--- a/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
+++ b/drivers/gpu/drm/i915/gt/selftest_hangcheck.c
@@ -25,7 +25,9 @@
 #include <linux/kthread.h>
 
 #include "gem/i915_gem_context.h"
-#include "gt/intel_gt.h"
+
+#include "intel_gt.h"
+#include "intel_engine_heartbeat.h"
 #include "intel_engine_pm.h"
 
 #include "i915_selftest.h"
@@ -308,6 +310,24 @@ static bool wait_until_running(struct hang *h, struct i915_request *rq)
 			  1000));
 }
 
+static void engine_heartbeat_disable(struct intel_engine_cs *engine,
+				     unsigned long *saved)
+{
+	*saved = engine->props.heartbeat_interval_ms;
+	engine->props.heartbeat_interval_ms = 0;
+
+	intel_engine_pm_get(engine);
+	intel_engine_park_heartbeat(engine);
+}
+
+static void engine_heartbeat_enable(struct intel_engine_cs *engine,
+				    unsigned long saved)
+{
+	intel_engine_pm_put(engine);
+
+	engine->props.heartbeat_interval_ms = saved;
+}
+
 static int igt_hang_sanitycheck(void *arg)
 {
 	struct intel_gt *gt = arg;
@@ -377,36 +397,30 @@ static int igt_reset_nop(void *arg)
 	struct intel_gt *gt = arg;
 	struct i915_gpu_error *global = &gt->i915->gpu_error;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *ctx;
 	unsigned int reset_count, count;
 	enum intel_engine_id id;
-	struct drm_file *file;
 	IGT_TIMEOUT(end_time);
 	int err = 0;
 
 	/* Check that we can reset during non-user portions of requests */
 
-	file = mock_file(gt->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	ctx = live_context(gt->i915, file);
-	if (IS_ERR(ctx)) {
-		err = PTR_ERR(ctx);
-		goto out;
-	}
-
-	i915_gem_context_clear_bannable(ctx);
 	reset_count = i915_reset_count(global);
 	count = 0;
 	do {
 		for_each_engine(engine, gt, id) {
+			struct intel_context *ce;
 			int i;
 
+			ce = intel_context_create(engine);
+			if (IS_ERR(ce)) {
+				err = PTR_ERR(ce);
+				break;
+			}
+
 			for (i = 0; i < 16; i++) {
 				struct i915_request *rq;
 
-				rq = igt_request_alloc(ctx, engine);
+				rq = intel_context_create_request(ce);
 				if (IS_ERR(rq)) {
 					err = PTR_ERR(rq);
 					break;
@@ -414,6 +428,8 @@ static int igt_reset_nop(void *arg)
 
 				i915_request_add(rq);
 			}
+
+			intel_context_put(ce);
 		}
 
 		igt_global_reset_lock(gt);
@@ -437,10 +453,7 @@ static int igt_reset_nop(void *arg)
 	} while (time_before(jiffies, end_time));
 	pr_info("%s: %d resets\n", __func__, count);
 
-	err = igt_flush_test(gt->i915);
-out:
-	mock_file_free(gt->i915, file);
-	if (intel_gt_is_wedged(gt))
+	if (igt_flush_test(gt->i915))
 		err = -EIO;
 	return err;
 }
@@ -450,36 +463,29 @@ static int igt_reset_nop_engine(void *arg)
 	struct intel_gt *gt = arg;
 	struct i915_gpu_error *global = &gt->i915->gpu_error;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *ctx;
 	enum intel_engine_id id;
-	struct drm_file *file;
-	int err = 0;
 
 	/* Check that we can engine-reset during non-user portions */
 
 	if (!intel_has_reset_engine(gt))
 		return 0;
 
-	file = mock_file(gt->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	ctx = live_context(gt->i915, file);
-	if (IS_ERR(ctx)) {
-		err = PTR_ERR(ctx);
-		goto out;
-	}
-
-	i915_gem_context_clear_bannable(ctx);
 	for_each_engine(engine, gt, id) {
-		unsigned int reset_count, reset_engine_count;
-		unsigned int count;
+		unsigned int reset_count, reset_engine_count, count;
+		struct intel_context *ce;
+		unsigned long heartbeat;
 		IGT_TIMEOUT(end_time);
+		int err;
+
+		ce = intel_context_create(engine);
+		if (IS_ERR(ce))
+			return PTR_ERR(ce);
 
 		reset_count = i915_reset_count(global);
 		reset_engine_count = i915_reset_engine_count(global, engine);
 		count = 0;
 
+		engine_heartbeat_disable(engine, &heartbeat);
 		set_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 		do {
 			int i;
@@ -494,7 +500,7 @@ static int igt_reset_nop_engine(void *arg)
 			for (i = 0; i < 16; i++) {
 				struct i915_request *rq;
 
-				rq = igt_request_alloc(ctx, engine);
+				rq = intel_context_create_request(ce);
 				if (IS_ERR(rq)) {
 					err = PTR_ERR(rq);
 					break;
@@ -523,22 +529,18 @@ static int igt_reset_nop_engine(void *arg)
 			}
 		} while (time_before(jiffies, end_time));
 		clear_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
-		pr_info("%s(%s): %d resets\n", __func__, engine->name, count);
+		engine_heartbeat_enable(engine, heartbeat);
 
-		if (err)
-			break;
+		pr_info("%s(%s): %d resets\n", __func__, engine->name, count);
 
-		err = igt_flush_test(gt->i915);
+		intel_context_put(ce);
+		if (igt_flush_test(gt->i915))
+			err = -EIO;
 		if (err)
-			break;
+			return err;
 	}
 
-	err = igt_flush_test(gt->i915);
-out:
-	mock_file_free(gt->i915, file);
-	if (intel_gt_is_wedged(gt))
-		err = -EIO;
-	return err;
+	return 0;
 }
 
 static int __igt_reset_engine(struct intel_gt *gt, bool active)
@@ -562,6 +564,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 
 	for_each_engine(engine, gt, id) {
 		unsigned int reset_count, reset_engine_count;
+		unsigned long heartbeat;
 		IGT_TIMEOUT(end_time);
 
 		if (active && !intel_engine_can_store_dword(engine))
@@ -577,7 +580,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 		reset_count = i915_reset_count(global);
 		reset_engine_count = i915_reset_engine_count(global, engine);
 
-		intel_engine_pm_get(engine);
+		engine_heartbeat_disable(engine, &heartbeat);
 		set_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 		do {
 			if (active) {
@@ -629,7 +632,7 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active)
 			}
 		} while (time_before(jiffies, end_time));
 		clear_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
-		intel_engine_pm_put(engine);
+		engine_heartbeat_enable(engine, heartbeat);
 
 		if (err)
 			break;
@@ -699,43 +702,43 @@ static int active_engine(void *data)
 	struct active_engine *arg = data;
 	struct intel_engine_cs *engine = arg->engine;
 	struct i915_request *rq[8] = {};
-	struct i915_gem_context *ctx[ARRAY_SIZE(rq)];
-	struct drm_file *file;
-	unsigned long count = 0;
+	struct intel_context *ce[ARRAY_SIZE(rq)];
+	unsigned long count;
 	int err = 0;
 
-	file = mock_file(engine->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	for (count = 0; count < ARRAY_SIZE(ctx); count++) {
-		ctx[count] = live_context(engine->i915, file);
-		if (IS_ERR(ctx[count])) {
-			err = PTR_ERR(ctx[count]);
+	for (count = 0; count < ARRAY_SIZE(ce); count++) {
+		ce[count] = intel_context_create(engine);
+		if (IS_ERR(ce[count])) {
+			err = PTR_ERR(ce[count]);
 			while (--count)
-				i915_gem_context_put(ctx[count]);
-			goto err_file;
+				intel_context_put(ce[count]);
+			return err;
 		}
 	}
 
+	count = 0;
 	while (!kthread_should_stop()) {
 		unsigned int idx = count++ & (ARRAY_SIZE(rq) - 1);
 		struct i915_request *old = rq[idx];
 		struct i915_request *new;
 
-		new = igt_request_alloc(ctx[idx], engine);
+		new = intel_context_create_request(ce[idx]);
 		if (IS_ERR(new)) {
 			err = PTR_ERR(new);
 			break;
 		}
 
-		if (arg->flags & TEST_PRIORITY)
-			ctx[idx]->sched.priority =
-				i915_prandom_u32_max_state(512, &prng);
-
 		rq[idx] = i915_request_get(new);
 		i915_request_add(new);
 
+		if (engine->schedule && arg->flags & TEST_PRIORITY) {
+			struct i915_sched_attr attr = {
+				.priority =
+					i915_prandom_u32_max_state(512, &prng),
+			};
+			engine->schedule(rq[idx], &attr);
+		}
+
 		err = active_request_put(old);
 		if (err)
 			break;
@@ -749,10 +752,10 @@ static int active_engine(void *data)
 		/* Keep the first error */
 		if (!err)
 			err = err__;
+
+		intel_context_put(ce[count]);
 	}
 
-err_file:
-	mock_file_free(engine->i915, file);
 	return err;
 }
 
@@ -786,6 +789,7 @@ static int __igt_reset_engines(struct intel_gt *gt,
 		struct active_engine threads[I915_NUM_ENGINES] = {};
 		unsigned long device = i915_reset_count(global);
 		unsigned long count = 0, reported;
+		unsigned long heartbeat;
 		IGT_TIMEOUT(end_time);
 
 		if (flags & TEST_ACTIVE &&
@@ -828,7 +832,7 @@ static int __igt_reset_engines(struct intel_gt *gt,
 
 		yield(); /* start all threads before we begin */
 
-		intel_engine_pm_get(engine);
+		engine_heartbeat_disable(engine, &heartbeat);
 		set_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
 		do {
 			struct i915_request *rq = NULL;
@@ -902,7 +906,8 @@ static int __igt_reset_engines(struct intel_gt *gt,
 			}
 		} while (time_before(jiffies, end_time));
 		clear_bit(I915_RESET_ENGINE + id, &gt->reset.flags);
-		intel_engine_pm_put(engine);
+		engine_heartbeat_enable(engine, heartbeat);
+
 		pr_info("i915_reset_engine(%s:%s): %lu resets\n",
 			engine->name, test_name, count);
 
@@ -1300,32 +1305,21 @@ static int igt_reset_evict_ggtt(void *arg)
 static int igt_reset_evict_ppgtt(void *arg)
 {
 	struct intel_gt *gt = arg;
-	struct i915_gem_context *ctx;
-	struct i915_address_space *vm;
-	struct drm_file *file;
+	struct i915_ppgtt *ppgtt;
 	int err;
 
-	file = mock_file(gt->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
+	/* aliasing == global gtt locking, covered above */
+	if (INTEL_PPGTT(gt->i915) < INTEL_PPGTT_FULL)
+		return 0;
 
-	ctx = live_context(gt->i915, file);
-	if (IS_ERR(ctx)) {
-		err = PTR_ERR(ctx);
-		goto out;
-	}
+	ppgtt = i915_ppgtt_create(gt->i915);
+	if (IS_ERR(ppgtt))
+		return PTR_ERR(ppgtt);
 
-	err = 0;
-	vm = i915_gem_context_get_vm_rcu(ctx);
-	if (!i915_is_ggtt(vm)) {
-		/* aliasing == global gtt locking, covered above */
-		err = __igt_reset_evict_vma(gt, vm,
-					    evict_vma, EXEC_OBJECT_WRITE);
-	}
-	i915_vm_put(vm);
+	err = __igt_reset_evict_vma(gt, &ppgtt->vm,
+				    evict_vma, EXEC_OBJECT_WRITE);
+	i915_vm_put(&ppgtt->vm);
 
-out:
-	mock_file_free(gt->i915, file);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c
index eb71ac2f992c..9ec9833c9c7b 100644
--- a/drivers/gpu/drm/i915/gt/selftest_lrc.c
+++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c
@@ -50,14 +50,31 @@ static struct i915_vma *create_scratch(struct intel_gt *gt)
 	return vma;
 }
 
+static void engine_heartbeat_disable(struct intel_engine_cs *engine,
+				     unsigned long *saved)
+{
+	*saved = engine->props.heartbeat_interval_ms;
+	engine->props.heartbeat_interval_ms = 0;
+
+	intel_engine_pm_get(engine);
+	intel_engine_park_heartbeat(engine);
+}
+
+static void engine_heartbeat_enable(struct intel_engine_cs *engine,
+				    unsigned long saved)
+{
+	intel_engine_pm_put(engine);
+
+	engine->props.heartbeat_interval_ms = saved;
+}
+
 static int live_sanitycheck(void *arg)
 {
 	struct intel_gt *gt = arg;
-	struct i915_gem_engines_iter it;
-	struct i915_gem_context *ctx;
-	struct intel_context *ce;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
 	struct igt_spinner spin;
-	int err = -ENOMEM;
+	int err = 0;
 
 	if (!HAS_LOGICAL_RING_CONTEXTS(gt->i915))
 		return 0;
@@ -65,17 +82,20 @@ static int live_sanitycheck(void *arg)
 	if (igt_spinner_init(&spin, gt))
 		return -ENOMEM;
 
-	ctx = kernel_context(gt->i915);
-	if (!ctx)
-		goto err_spin;
-
-	for_each_gem_engine(ce, i915_gem_context_lock_engines(ctx), it) {
+	for_each_engine(engine, gt, id) {
+		struct intel_context *ce;
 		struct i915_request *rq;
 
+		ce = intel_context_create(engine);
+		if (IS_ERR(ce)) {
+			err = PTR_ERR(ce);
+			break;
+		}
+
 		rq = igt_spinner_create_request(&spin, ce, MI_NOOP);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
-			goto err_ctx;
+			goto out_ctx;
 		}
 
 		i915_request_add(rq);
@@ -84,21 +104,21 @@ static int live_sanitycheck(void *arg)
 			GEM_TRACE_DUMP();
 			intel_gt_set_wedged(gt);
 			err = -EIO;
-			goto err_ctx;
+			goto out_ctx;
 		}
 
 		igt_spinner_end(&spin);
 		if (igt_flush_test(gt->i915)) {
 			err = -EIO;
-			goto err_ctx;
+			goto out_ctx;
 		}
+
+out_ctx:
+		intel_context_put(ce);
+		if (err)
+			break;
 	}
 
-	err = 0;
-err_ctx:
-	i915_gem_context_unlock_engines(ctx);
-	kernel_context_close(ctx);
-err_spin:
 	igt_spinner_fini(&spin);
 	return err;
 }
@@ -106,7 +126,6 @@ err_spin:
 static int live_unlite_restore(struct intel_gt *gt, int prio)
 {
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *ctx;
 	enum intel_engine_id id;
 	struct igt_spinner spin;
 	int err = -ENOMEM;
@@ -119,15 +138,12 @@ static int live_unlite_restore(struct intel_gt *gt, int prio)
 	if (igt_spinner_init(&spin, gt))
 		return err;
 
-	ctx = kernel_context(gt->i915);
-	if (!ctx)
-		goto err_spin;
-
 	err = 0;
 	for_each_engine(engine, gt, id) {
 		struct intel_context *ce[2] = {};
 		struct i915_request *rq[2];
 		struct igt_live_test t;
+		unsigned long saved;
 		int n;
 
 		if (prio && !intel_engine_has_preemption(engine))
@@ -140,11 +156,12 @@ static int live_unlite_restore(struct intel_gt *gt, int prio)
 			err = -EIO;
 			break;
 		}
+		engine_heartbeat_disable(engine, &saved);
 
 		for (n = 0; n < ARRAY_SIZE(ce); n++) {
 			struct intel_context *tmp;
 
-			tmp = intel_context_create(ctx, engine);
+			tmp = intel_context_create(engine);
 			if (IS_ERR(tmp)) {
 				err = PTR_ERR(tmp);
 				goto err_ce;
@@ -247,14 +264,13 @@ err_ce:
 			intel_context_put(ce[n]);
 		}
 
+		engine_heartbeat_enable(engine, saved);
 		if (igt_live_test_end(&t))
 			err = -EIO;
 		if (err)
 			break;
 	}
 
-	kernel_context_close(ctx);
-err_spin:
 	igt_spinner_fini(&spin);
 	return err;
 }
@@ -309,17 +325,17 @@ emit_semaphore_chain(struct i915_request *rq, struct i915_vma *vma, int idx)
 static struct i915_request *
 semaphore_queue(struct intel_engine_cs *engine, struct i915_vma *vma, int idx)
 {
-	struct i915_gem_context *ctx;
+	struct intel_context *ce;
 	struct i915_request *rq;
 	int err;
 
-	ctx = kernel_context(engine->i915);
-	if (!ctx)
-		return ERR_PTR(-ENOMEM);
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return ERR_CAST(ce);
 
-	rq = igt_request_alloc(ctx, engine);
+	rq = intel_context_create_request(ce);
 	if (IS_ERR(rq))
-		goto out_ctx;
+		goto out_ce;
 
 	err = 0;
 	if (rq->engine->emit_init_breadcrumb)
@@ -332,8 +348,8 @@ semaphore_queue(struct intel_engine_cs *engine, struct i915_vma *vma, int idx)
 	if (err)
 		rq = ERR_PTR(err);
 
-out_ctx:
-	kernel_context_close(ctx);
+out_ce:
+	intel_context_put(ce);
 	return rq;
 }
 
@@ -348,7 +364,7 @@ release_queue(struct intel_engine_cs *engine,
 	struct i915_request *rq;
 	u32 *cs;
 
-	rq = i915_request_create(engine->kernel_context);
+	rq = intel_engine_create_kernel_request(engine);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
@@ -468,12 +484,16 @@ static int live_timeslice_preempt(void *arg)
 		enum intel_engine_id id;
 
 		for_each_engine(engine, gt, id) {
+			unsigned long saved;
+
 			if (!intel_engine_has_preemption(engine))
 				continue;
 
 			memset(vaddr, 0, PAGE_SIZE);
 
+			engine_heartbeat_disable(engine, &saved);
 			err = slice_semaphore_queue(engine, vma, count);
+			engine_heartbeat_enable(engine, saved);
 			if (err)
 				goto err_pin;
 
@@ -497,7 +517,7 @@ static struct i915_request *nop_request(struct intel_engine_cs *engine)
 {
 	struct i915_request *rq;
 
-	rq = i915_request_create(engine->kernel_context);
+	rq = intel_engine_create_kernel_request(engine);
 	if (IS_ERR(rq))
 		return rq;
 
@@ -566,17 +586,19 @@ static int live_timeslice_queue(void *arg)
 			.priority = I915_USER_PRIORITY(I915_PRIORITY_MAX),
 		};
 		struct i915_request *rq, *nop;
+		unsigned long saved;
 
 		if (!intel_engine_has_preemption(engine))
 			continue;
 
+		engine_heartbeat_disable(engine, &saved);
 		memset(vaddr, 0, PAGE_SIZE);
 
 		/* ELSP[0]: semaphore wait */
 		rq = semaphore_queue(engine, vma, 0);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
-			goto err_pin;
+			goto err_heartbeat;
 		}
 		engine->schedule(rq, &attr);
 		wait_for_submit(engine, rq);
@@ -585,8 +607,7 @@ static int live_timeslice_queue(void *arg)
 		nop = nop_request(engine);
 		if (IS_ERR(nop)) {
 			err = PTR_ERR(nop);
-			i915_request_put(rq);
-			goto err_pin;
+			goto err_rq;
 		}
 		wait_for_submit(engine, nop);
 		i915_request_put(nop);
@@ -596,10 +617,8 @@ static int live_timeslice_queue(void *arg)
 
 		/* Queue: semaphore signal, matching priority as semaphore */
 		err = release_queue(engine, vma, 1, effective_prio(rq));
-		if (err) {
-			i915_request_put(rq);
-			goto err_pin;
-		}
+		if (err)
+			goto err_rq;
 
 		intel_engine_flush_submission(engine);
 		if (!READ_ONCE(engine->execlists.timer.expires) &&
@@ -630,12 +649,14 @@ static int live_timeslice_queue(void *arg)
 			memset(vaddr, 0xff, PAGE_SIZE);
 			err = -EIO;
 		}
+err_rq:
 		i915_request_put(rq);
+err_heartbeat:
+		engine_heartbeat_enable(engine, saved);
 		if (err)
 			break;
 	}
 
-err_pin:
 	i915_vma_unpin(vma);
 err_map:
 	i915_gem_object_unpin_map(obj);
@@ -748,15 +769,19 @@ static int live_busywait_preempt(void *arg)
 		*cs++ = 0;
 
 		intel_ring_advance(lo, cs);
+
+		i915_request_get(lo);
 		i915_request_add(lo);
 
 		if (wait_for(READ_ONCE(*map), 10)) {
+			i915_request_put(lo);
 			err = -ETIMEDOUT;
 			goto err_vma;
 		}
 
 		/* Low priority request should be busywaiting now */
 		if (i915_request_wait(lo, 0, 1) != -ETIME) {
+			i915_request_put(lo);
 			pr_err("%s: Busywaiting request did not!\n",
 			       engine->name);
 			err = -EIO;
@@ -766,6 +791,7 @@ static int live_busywait_preempt(void *arg)
 		hi = igt_request_alloc(ctx_hi, engine);
 		if (IS_ERR(hi)) {
 			err = PTR_ERR(hi);
+			i915_request_put(lo);
 			goto err_vma;
 		}
 
@@ -773,6 +799,7 @@ static int live_busywait_preempt(void *arg)
 		if (IS_ERR(cs)) {
 			err = PTR_ERR(cs);
 			i915_request_add(hi);
+			i915_request_put(lo);
 			goto err_vma;
 		}
 
@@ -793,11 +820,13 @@ static int live_busywait_preempt(void *arg)
 			intel_engine_dump(engine, &p, "%s\n", engine->name);
 			GEM_TRACE_DUMP();
 
+			i915_request_put(lo);
 			intel_gt_set_wedged(gt);
 			err = -EIO;
 			goto err_vma;
 		}
 		GEM_BUG_ON(READ_ONCE(*map));
+		i915_request_put(lo);
 
 		if (igt_live_test_end(&t)) {
 			err = -EIO;
@@ -1187,13 +1216,13 @@ static int __cancel_active0(struct live_preempt_cancel *arg)
 				__func__, arg->engine->name))
 		return -EIO;
 
-	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
 	rq = spinner_create_request(&arg->a.spin,
 				    arg->a.ctx, arg->engine,
 				    MI_ARB_CHECK);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
+	clear_bit(CONTEXT_BANNED, &rq->context->flags);
 	i915_request_get(rq);
 	i915_request_add(rq);
 	if (!igt_wait_for_spinner(&arg->a.spin, rq)) {
@@ -1201,7 +1230,7 @@ static int __cancel_active0(struct live_preempt_cancel *arg)
 		goto out;
 	}
 
-	i915_gem_context_set_banned(arg->a.ctx);
+	intel_context_set_banned(rq->context);
 	err = intel_engine_pulse(arg->engine);
 	if (err)
 		goto out;
@@ -1236,13 +1265,13 @@ static int __cancel_active1(struct live_preempt_cancel *arg)
 				__func__, arg->engine->name))
 		return -EIO;
 
-	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
 	rq[0] = spinner_create_request(&arg->a.spin,
 				       arg->a.ctx, arg->engine,
 				       MI_NOOP); /* no preemption */
 	if (IS_ERR(rq[0]))
 		return PTR_ERR(rq[0]);
 
+	clear_bit(CONTEXT_BANNED, &rq[0]->context->flags);
 	i915_request_get(rq[0]);
 	i915_request_add(rq[0]);
 	if (!igt_wait_for_spinner(&arg->a.spin, rq[0])) {
@@ -1250,7 +1279,6 @@ static int __cancel_active1(struct live_preempt_cancel *arg)
 		goto out;
 	}
 
-	clear_bit(CONTEXT_BANNED, &arg->b.ctx->flags);
 	rq[1] = spinner_create_request(&arg->b.spin,
 				       arg->b.ctx, arg->engine,
 				       MI_ARB_CHECK);
@@ -1259,13 +1287,14 @@ static int __cancel_active1(struct live_preempt_cancel *arg)
 		goto out;
 	}
 
+	clear_bit(CONTEXT_BANNED, &rq[1]->context->flags);
 	i915_request_get(rq[1]);
 	err = i915_request_await_dma_fence(rq[1], &rq[0]->fence);
 	i915_request_add(rq[1]);
 	if (err)
 		goto out;
 
-	i915_gem_context_set_banned(arg->b.ctx);
+	intel_context_set_banned(rq[1]->context);
 	err = intel_engine_pulse(arg->engine);
 	if (err)
 		goto out;
@@ -1308,13 +1337,13 @@ static int __cancel_queued(struct live_preempt_cancel *arg)
 				__func__, arg->engine->name))
 		return -EIO;
 
-	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
 	rq[0] = spinner_create_request(&arg->a.spin,
 				       arg->a.ctx, arg->engine,
 				       MI_ARB_CHECK);
 	if (IS_ERR(rq[0]))
 		return PTR_ERR(rq[0]);
 
+	clear_bit(CONTEXT_BANNED, &rq[0]->context->flags);
 	i915_request_get(rq[0]);
 	i915_request_add(rq[0]);
 	if (!igt_wait_for_spinner(&arg->a.spin, rq[0])) {
@@ -1322,13 +1351,13 @@ static int __cancel_queued(struct live_preempt_cancel *arg)
 		goto out;
 	}
 
-	clear_bit(CONTEXT_BANNED, &arg->b.ctx->flags);
 	rq[1] = igt_request_alloc(arg->b.ctx, arg->engine);
 	if (IS_ERR(rq[1])) {
 		err = PTR_ERR(rq[1]);
 		goto out;
 	}
 
+	clear_bit(CONTEXT_BANNED, &rq[1]->context->flags);
 	i915_request_get(rq[1]);
 	err = i915_request_await_dma_fence(rq[1], &rq[0]->fence);
 	i915_request_add(rq[1]);
@@ -1349,7 +1378,7 @@ static int __cancel_queued(struct live_preempt_cancel *arg)
 	if (err)
 		goto out;
 
-	i915_gem_context_set_banned(arg->a.ctx);
+	intel_context_set_banned(rq[2]->context);
 	err = intel_engine_pulse(arg->engine);
 	if (err)
 		goto out;
@@ -1396,13 +1425,13 @@ static int __cancel_hostile(struct live_preempt_cancel *arg)
 		return 0;
 
 	GEM_TRACE("%s(%s)\n", __func__, arg->engine->name);
-	clear_bit(CONTEXT_BANNED, &arg->a.ctx->flags);
 	rq = spinner_create_request(&arg->a.spin,
 				    arg->a.ctx, arg->engine,
 				    MI_NOOP); /* preemption disabled */
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
+	clear_bit(CONTEXT_BANNED, &rq->context->flags);
 	i915_request_get(rq);
 	i915_request_add(rq);
 	if (!igt_wait_for_spinner(&arg->a.spin, rq)) {
@@ -1410,7 +1439,7 @@ static int __cancel_hostile(struct live_preempt_cancel *arg)
 		goto out;
 	}
 
-	i915_gem_context_set_banned(arg->a.ctx);
+	intel_context_set_banned(rq->context);
 	err = intel_engine_pulse(arg->engine); /* force reset */
 	if (err)
 		goto out;
@@ -1665,6 +1694,7 @@ static int live_suppress_wait_preempt(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct preempt_client client[4];
+	struct i915_request *rq[ARRAY_SIZE(client)] = {};
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
 	int err = -ENOMEM;
@@ -1698,7 +1728,6 @@ static int live_suppress_wait_preempt(void *arg)
 			continue;
 
 		for (depth = 0; depth < ARRAY_SIZE(client); depth++) {
-			struct i915_request *rq[ARRAY_SIZE(client)];
 			struct i915_request *dummy;
 
 			engine->execlists.preempt_hang.count = 0;
@@ -1708,18 +1737,22 @@ static int live_suppress_wait_preempt(void *arg)
 				goto err_client_3;
 
 			for (i = 0; i < ARRAY_SIZE(client); i++) {
-				rq[i] = spinner_create_request(&client[i].spin,
-							       client[i].ctx, engine,
-							       MI_NOOP);
-				if (IS_ERR(rq[i])) {
-					err = PTR_ERR(rq[i]);
+				struct i915_request *this;
+
+				this = spinner_create_request(&client[i].spin,
+							      client[i].ctx, engine,
+							      MI_NOOP);
+				if (IS_ERR(this)) {
+					err = PTR_ERR(this);
 					goto err_wedged;
 				}
 
 				/* Disable NEWCLIENT promotion */
-				__i915_active_fence_set(&i915_request_timeline(rq[i])->last_request,
+				__i915_active_fence_set(&i915_request_timeline(this)->last_request,
 							&dummy->fence);
-				i915_request_add(rq[i]);
+
+				rq[i] = i915_request_get(this);
+				i915_request_add(this);
 			}
 
 			dummy_request_free(dummy);
@@ -1740,8 +1773,11 @@ static int live_suppress_wait_preempt(void *arg)
 				goto err_wedged;
 			}
 
-			for (i = 0; i < ARRAY_SIZE(client); i++)
+			for (i = 0; i < ARRAY_SIZE(client); i++) {
 				igt_spinner_end(&client[i].spin);
+				i915_request_put(rq[i]);
+				rq[i] = NULL;
+			}
 
 			if (igt_flush_test(gt->i915))
 				goto err_wedged;
@@ -1769,8 +1805,10 @@ err_client_0:
 	return err;
 
 err_wedged:
-	for (i = 0; i < ARRAY_SIZE(client); i++)
+	for (i = 0; i < ARRAY_SIZE(client); i++) {
 		igt_spinner_end(&client[i].spin);
+		i915_request_put(rq[i]);
+	}
 	intel_gt_set_wedged(gt);
 	err = -EIO;
 	goto err_client_3;
@@ -1815,6 +1853,8 @@ static int live_chain_preempt(void *arg)
 					    MI_ARB_CHECK);
 		if (IS_ERR(rq))
 			goto err_wedged;
+
+		i915_request_get(rq);
 		i915_request_add(rq);
 
 		ring_size = rq->wa_tail - rq->head;
@@ -1827,8 +1867,10 @@ static int live_chain_preempt(void *arg)
 		igt_spinner_end(&lo.spin);
 		if (i915_request_wait(rq, 0, HZ / 2) < 0) {
 			pr_err("Timed out waiting to flush %s\n", engine->name);
+			i915_request_put(rq);
 			goto err_wedged;
 		}
+		i915_request_put(rq);
 
 		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name)) {
 			err = -EIO;
@@ -1862,6 +1904,8 @@ static int live_chain_preempt(void *arg)
 			rq = igt_request_alloc(hi.ctx, engine);
 			if (IS_ERR(rq))
 				goto err_wedged;
+
+			i915_request_get(rq);
 			i915_request_add(rq);
 			engine->schedule(rq, &attr);
 
@@ -1874,14 +1918,19 @@ static int live_chain_preempt(void *arg)
 				       count);
 				intel_engine_dump(engine, &p,
 						  "%s\n", engine->name);
+				i915_request_put(rq);
 				goto err_wedged;
 			}
 			igt_spinner_end(&lo.spin);
+			i915_request_put(rq);
 
 			rq = igt_request_alloc(lo.ctx, engine);
 			if (IS_ERR(rq))
 				goto err_wedged;
+
+			i915_request_get(rq);
 			i915_request_add(rq);
+
 			if (i915_request_wait(rq, 0, HZ / 5) < 0) {
 				struct drm_printer p =
 					drm_info_printer(gt->i915->drm.dev);
@@ -1890,8 +1939,11 @@ static int live_chain_preempt(void *arg)
 				       count);
 				intel_engine_dump(engine, &p,
 						  "%s\n", engine->name);
+
+				i915_request_put(rq);
 				goto err_wedged;
 			}
+			i915_request_put(rq);
 		}
 
 		if (igt_live_test_end(&t)) {
@@ -1915,6 +1967,201 @@ err_wedged:
 	goto err_client_lo;
 }
 
+static int create_gang(struct intel_engine_cs *engine,
+		       struct i915_request **prev)
+{
+	struct drm_i915_gem_object *obj;
+	struct intel_context *ce;
+	struct i915_request *rq;
+	struct i915_vma *vma;
+	u32 *cs;
+	int err;
+
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
+
+	obj = i915_gem_object_create_internal(engine->i915, 4096);
+	if (IS_ERR(obj)) {
+		err = PTR_ERR(obj);
+		goto err_ce;
+	}
+
+	vma = i915_vma_instance(obj, ce->vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_obj;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER);
+	if (err)
+		goto err_obj;
+
+	cs = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	if (IS_ERR(cs))
+		goto err_obj;
+
+	/* Semaphore target: spin until zero */
+	*cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE;
+
+	*cs++ = MI_SEMAPHORE_WAIT |
+		MI_SEMAPHORE_POLL |
+		MI_SEMAPHORE_SAD_EQ_SDD;
+	*cs++ = 0;
+	*cs++ = lower_32_bits(vma->node.start);
+	*cs++ = upper_32_bits(vma->node.start);
+
+	if (*prev) {
+		u64 offset = (*prev)->batch->node.start;
+
+		/* Terminate the spinner in the next lower priority batch. */
+		*cs++ = MI_STORE_DWORD_IMM_GEN4;
+		*cs++ = lower_32_bits(offset);
+		*cs++ = upper_32_bits(offset);
+		*cs++ = 0;
+	}
+
+	*cs++ = MI_BATCH_BUFFER_END;
+	i915_gem_object_flush_map(obj);
+	i915_gem_object_unpin_map(obj);
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		goto err_obj;
+
+	rq->batch = vma;
+	i915_request_get(rq);
+
+	i915_vma_lock(vma);
+	err = i915_request_await_object(rq, vma->obj, false);
+	if (!err)
+		err = i915_vma_move_to_active(vma, rq, 0);
+	if (!err)
+		err = rq->engine->emit_bb_start(rq,
+						vma->node.start,
+						PAGE_SIZE, 0);
+	i915_vma_unlock(vma);
+	i915_request_add(rq);
+	if (err)
+		goto err_rq;
+
+	i915_gem_object_put(obj);
+	intel_context_put(ce);
+
+	rq->client_link.next = &(*prev)->client_link;
+	*prev = rq;
+	return 0;
+
+err_rq:
+	i915_request_put(rq);
+err_obj:
+	i915_gem_object_put(obj);
+err_ce:
+	intel_context_put(ce);
+	return err;
+}
+
+static int live_preempt_gang(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	if (!HAS_LOGICAL_RING_PREEMPTION(gt->i915))
+		return 0;
+
+	/*
+	 * Build as long a chain of preempters as we can, with each
+	 * request higher priority than the last. Once we are ready, we release
+	 * the last batch which then precolates down the chain, each releasing
+	 * the next oldest in turn. The intent is to simply push as hard as we
+	 * can with the number of preemptions, trying to exceed narrow HW
+	 * limits. At a minimum, we insist that we can sort all the user
+	 * high priority levels into execution order.
+	 */
+
+	for_each_engine(engine, gt, id) {
+		struct i915_request *rq = NULL;
+		struct igt_live_test t;
+		IGT_TIMEOUT(end_time);
+		int prio = 0;
+		int err = 0;
+		u32 *cs;
+
+		if (!intel_engine_has_preemption(engine))
+			continue;
+
+		if (igt_live_test_begin(&t, gt->i915, __func__, engine->name))
+			return -EIO;
+
+		do {
+			struct i915_sched_attr attr = {
+				.priority = I915_USER_PRIORITY(prio++),
+			};
+
+			err = create_gang(engine, &rq);
+			if (err)
+				break;
+
+			/* Submit each spinner at increasing priority */
+			engine->schedule(rq, &attr);
+
+			if (prio <= I915_PRIORITY_MAX)
+				continue;
+
+			if (prio > (INT_MAX >> I915_USER_PRIORITY_SHIFT))
+				break;
+
+			if (__igt_timeout(end_time, NULL))
+				break;
+		} while (1);
+		pr_debug("%s: Preempt chain of %d requests\n",
+			 engine->name, prio);
+
+		/*
+		 * Such that the last spinner is the highest priority and
+		 * should execute first. When that spinner completes,
+		 * it will terminate the next lowest spinner until there
+		 * are no more spinners and the gang is complete.
+		 */
+		cs = i915_gem_object_pin_map(rq->batch->obj, I915_MAP_WC);
+		if (!IS_ERR(cs)) {
+			*cs = 0;
+			i915_gem_object_unpin_map(rq->batch->obj);
+		} else {
+			err = PTR_ERR(cs);
+			intel_gt_set_wedged(gt);
+		}
+
+		while (rq) { /* wait for each rq from highest to lowest prio */
+			struct i915_request *n =
+				list_next_entry(rq, client_link);
+
+			if (err == 0 && i915_request_wait(rq, 0, HZ / 5) < 0) {
+				struct drm_printer p =
+					drm_info_printer(engine->i915->drm.dev);
+
+				pr_err("Failed to flush chain of %d requests, at %d\n",
+				       prio, rq_prio(rq) >> I915_USER_PRIORITY_SHIFT);
+				intel_engine_dump(engine, &p,
+						  "%s\n", engine->name);
+
+				err = -ETIME;
+			}
+
+			i915_request_put(rq);
+			rq = n;
+		}
+
+		if (igt_live_test_end(&t))
+			err = -EIO;
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 static int live_preempt_hang(void *arg)
 {
 	struct intel_gt *gt = arg;
@@ -2391,28 +2638,18 @@ static int nop_virtual_engine(struct intel_gt *gt,
 #define CHAIN BIT(0)
 {
 	IGT_TIMEOUT(end_time);
-	struct i915_request *request[16];
-	struct i915_gem_context *ctx[16];
+	struct i915_request *request[16] = {};
 	struct intel_context *ve[16];
 	unsigned long n, prime, nc;
 	struct igt_live_test t;
 	ktime_t times[2] = {};
 	int err;
 
-	GEM_BUG_ON(!nctx || nctx > ARRAY_SIZE(ctx));
+	GEM_BUG_ON(!nctx || nctx > ARRAY_SIZE(ve));
 
 	for (n = 0; n < nctx; n++) {
-		ctx[n] = kernel_context(gt->i915);
-		if (!ctx[n]) {
-			err = -ENOMEM;
-			nctx = n;
-			goto out;
-		}
-
-		ve[n] = intel_execlists_create_virtual(ctx[n],
-						       siblings, nsibling);
+		ve[n] = intel_execlists_create_virtual(siblings, nsibling);
 		if (IS_ERR(ve[n])) {
-			kernel_context_close(ctx[n]);
 			err = PTR_ERR(ve[n]);
 			nctx = n;
 			goto out;
@@ -2421,7 +2658,6 @@ static int nop_virtual_engine(struct intel_gt *gt,
 		err = intel_context_pin(ve[n]);
 		if (err) {
 			intel_context_put(ve[n]);
-			kernel_context_close(ctx[n]);
 			nctx = n;
 			goto out;
 		}
@@ -2437,27 +2673,35 @@ static int nop_virtual_engine(struct intel_gt *gt,
 		if (flags & CHAIN) {
 			for (nc = 0; nc < nctx; nc++) {
 				for (n = 0; n < prime; n++) {
-					request[nc] =
-						i915_request_create(ve[nc]);
-					if (IS_ERR(request[nc])) {
-						err = PTR_ERR(request[nc]);
+					struct i915_request *rq;
+
+					rq = i915_request_create(ve[nc]);
+					if (IS_ERR(rq)) {
+						err = PTR_ERR(rq);
 						goto out;
 					}
 
-					i915_request_add(request[nc]);
+					if (request[nc])
+						i915_request_put(request[nc]);
+					request[nc] = i915_request_get(rq);
+					i915_request_add(rq);
 				}
 			}
 		} else {
 			for (n = 0; n < prime; n++) {
 				for (nc = 0; nc < nctx; nc++) {
-					request[nc] =
-						i915_request_create(ve[nc]);
-					if (IS_ERR(request[nc])) {
-						err = PTR_ERR(request[nc]);
+					struct i915_request *rq;
+
+					rq = i915_request_create(ve[nc]);
+					if (IS_ERR(rq)) {
+						err = PTR_ERR(rq);
 						goto out;
 					}
 
-					i915_request_add(request[nc]);
+					if (request[nc])
+						i915_request_put(request[nc]);
+					request[nc] = i915_request_get(rq);
+					i915_request_add(rq);
 				}
 			}
 		}
@@ -2483,6 +2727,11 @@ static int nop_virtual_engine(struct intel_gt *gt,
 		if (prime == 1)
 			times[0] = times[1];
 
+		for (nc = 0; nc < nctx; nc++) {
+			i915_request_put(request[nc]);
+			request[nc] = NULL;
+		}
+
 		if (__igt_timeout(end_time, NULL))
 			break;
 	}
@@ -2500,9 +2749,9 @@ out:
 		err = -EIO;
 
 	for (nc = 0; nc < nctx; nc++) {
+		i915_request_put(request[nc]);
 		intel_context_unpin(ve[nc]);
 		intel_context_put(ve[nc]);
-		kernel_context_close(ctx[nc]);
 	}
 	return err;
 }
@@ -2561,7 +2810,6 @@ static int mask_virtual_engine(struct intel_gt *gt,
 			       unsigned int nsibling)
 {
 	struct i915_request *request[MAX_ENGINE_INSTANCE + 1];
-	struct i915_gem_context *ctx;
 	struct intel_context *ve;
 	struct igt_live_test t;
 	unsigned int n;
@@ -2572,11 +2820,7 @@ static int mask_virtual_engine(struct intel_gt *gt,
 	 * restrict it to our desired engine within the virtual engine.
 	 */
 
-	ctx = kernel_context(gt->i915);
-	if (!ctx)
-		return -ENOMEM;
-
-	ve = intel_execlists_create_virtual(ctx, siblings, nsibling);
+	ve = intel_execlists_create_virtual(siblings, nsibling);
 	if (IS_ERR(ve)) {
 		err = PTR_ERR(ve);
 		goto out_close;
@@ -2644,7 +2888,6 @@ out_unpin:
 out_put:
 	intel_context_put(ve);
 out_close:
-	kernel_context_close(ctx);
 	return err;
 }
 
@@ -2684,7 +2927,6 @@ static int preserved_virtual_engine(struct intel_gt *gt,
 				    unsigned int nsibling)
 {
 	struct i915_request *last = NULL;
-	struct i915_gem_context *ctx;
 	struct intel_context *ve;
 	struct i915_vma *scratch;
 	struct igt_live_test t;
@@ -2692,17 +2934,11 @@ static int preserved_virtual_engine(struct intel_gt *gt,
 	int err = 0;
 	u32 *cs;
 
-	ctx = kernel_context(gt->i915);
-	if (!ctx)
-		return -ENOMEM;
-
 	scratch = create_scratch(siblings[0]->gt);
-	if (IS_ERR(scratch)) {
-		err = PTR_ERR(scratch);
-		goto out_close;
-	}
+	if (IS_ERR(scratch))
+		return PTR_ERR(scratch);
 
-	ve = intel_execlists_create_virtual(ctx, siblings, nsibling);
+	ve = intel_execlists_create_virtual(siblings, nsibling);
 	if (IS_ERR(ve)) {
 		err = PTR_ERR(ve);
 		goto out_scratch;
@@ -2785,8 +3021,6 @@ out_put:
 	intel_context_put(ve);
 out_scratch:
 	i915_vma_unpin_and_release(&scratch, 0);
-out_close:
-	kernel_context_close(ctx);
 	return err;
 }
 
@@ -2838,16 +3072,54 @@ static int bond_virtual_engine(struct intel_gt *gt,
 #define BOND_SCHEDULE BIT(0)
 {
 	struct intel_engine_cs *master;
-	struct i915_gem_context *ctx;
 	struct i915_request *rq[16];
 	enum intel_engine_id id;
+	struct igt_spinner spin;
 	unsigned long n;
 	int err;
 
+	/*
+	 * A set of bonded requests is intended to be run concurrently
+	 * across a number of engines. We use one request per-engine
+	 * and a magic fence to schedule each of the bonded requests
+	 * at the same time. A consequence of our current scheduler is that
+	 * we only move requests to the HW ready queue when the request
+	 * becomes ready, that is when all of its prerequisite fences have
+	 * been signaled. As one of those fences is the master submit fence,
+	 * there is a delay on all secondary fences as the HW may be
+	 * currently busy. Equally, as all the requests are independent,
+	 * they may have other fences that delay individual request
+	 * submission to HW. Ergo, we do not guarantee that all requests are
+	 * immediately submitted to HW at the same time, just that if the
+	 * rules are abided by, they are ready at the same time as the
+	 * first is submitted. Userspace can embed semaphores in its batch
+	 * to ensure parallel execution of its phases as it requires.
+	 * Though naturally it gets requested that perhaps the scheduler should
+	 * take care of parallel execution, even across preemption events on
+	 * different HW. (The proper answer is of course "lalalala".)
+	 *
+	 * With the submit-fence, we have identified three possible phases
+	 * of synchronisation depending on the master fence: queued (not
+	 * ready), executing, and signaled. The first two are quite simple
+	 * and checked below. However, the signaled master fence handling is
+	 * contentious. Currently we do not distinguish between a signaled
+	 * fence and an expired fence, as once signaled it does not convey
+	 * any information about the previous execution. It may even be freed
+	 * and hence checking later it may not exist at all. Ergo we currently
+	 * do not apply the bonding constraint for an already signaled fence,
+	 * as our expectation is that it should not constrain the secondaries
+	 * and is outside of the scope of the bonded request API (i.e. all
+	 * userspace requests are meant to be running in parallel). As
+	 * it imposes no constraint, and is effectively a no-op, we do not
+	 * check below as normal execution flows are checked extensively above.
+	 *
+	 * XXX Is the degenerate handling of signaled submit fences the
+	 * expected behaviour for userpace?
+	 */
+
 	GEM_BUG_ON(nsibling >= ARRAY_SIZE(rq) - 1);
 
-	ctx = kernel_context(gt->i915);
-	if (!ctx)
+	if (igt_spinner_init(&spin, gt))
 		return -ENOMEM;
 
 	err = 0;
@@ -2860,7 +3132,9 @@ static int bond_virtual_engine(struct intel_gt *gt,
 
 		memset_p((void *)rq, ERR_PTR(-EINVAL), ARRAY_SIZE(rq));
 
-		rq[0] = igt_request_alloc(ctx, master);
+		rq[0] = igt_spinner_create_request(&spin,
+						   master->kernel_context,
+						   MI_NOOP);
 		if (IS_ERR(rq[0])) {
 			err = PTR_ERR(rq[0]);
 			goto out;
@@ -2873,16 +3147,21 @@ static int bond_virtual_engine(struct intel_gt *gt,
 							       &fence,
 							       GFP_KERNEL);
 		}
+
 		i915_request_add(rq[0]);
 		if (err < 0)
 			goto out;
 
+		if (!(flags & BOND_SCHEDULE) &&
+		    !igt_wait_for_spinner(&spin, rq[0])) {
+			err = -EIO;
+			goto out;
+		}
+
 		for (n = 0; n < nsibling; n++) {
 			struct intel_context *ve;
 
-			ve = intel_execlists_create_virtual(ctx,
-							    siblings,
-							    nsibling);
+			ve = intel_execlists_create_virtual(siblings, nsibling);
 			if (IS_ERR(ve)) {
 				err = PTR_ERR(ve);
 				onstack_fence_fini(&fence);
@@ -2924,6 +3203,8 @@ static int bond_virtual_engine(struct intel_gt *gt,
 			}
 		}
 		onstack_fence_fini(&fence);
+		intel_engine_flush_submission(master);
+		igt_spinner_end(&spin);
 
 		if (i915_request_wait(rq[0], 0, HZ / 10) < 0) {
 			pr_err("Master request did not execute (on %s)!\n",
@@ -2960,7 +3241,7 @@ out:
 	if (igt_flush_test(gt->i915))
 		err = -EIO;
 
-	kernel_context_close(ctx);
+	igt_spinner_fini(&spin);
 	return err;
 }
 
@@ -3028,6 +3309,7 @@ int intel_execlists_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(live_suppress_self_preempt),
 		SUBTEST(live_suppress_wait_preempt),
 		SUBTEST(live_chain_preempt),
+		SUBTEST(live_preempt_gang),
 		SUBTEST(live_preempt_hang),
 		SUBTEST(live_preempt_timeout),
 		SUBTEST(live_preempt_smoke),
@@ -3207,12 +3489,12 @@ static int live_lrc_fixed(void *arg)
 		} tbl[] = {
 			{
 				i915_mmio_reg_offset(RING_START(engine->mmio_base)),
-				CTX_RING_BUFFER_START - 1,
+				CTX_RING_START - 1,
 				"RING_START"
 			},
 			{
 				i915_mmio_reg_offset(RING_CTL(engine->mmio_base)),
-				CTX_RING_BUFFER_CONTROL - 1,
+				CTX_RING_CTL - 1,
 				"RING_CTL"
 			},
 			{
@@ -3231,7 +3513,7 @@ static int live_lrc_fixed(void *arg)
 				"RING_MI_MODE"
 			},
 			{
-				engine->mmio_base + 0x110,
+				i915_mmio_reg_offset(RING_BBSTATE(engine->mmio_base)),
 				CTX_BB_STATE - 1,
 				"BB_STATE"
 			},
@@ -3270,8 +3552,7 @@ static int live_lrc_fixed(void *arg)
 	return err;
 }
 
-static int __live_lrc_state(struct i915_gem_context *fixme,
-			    struct intel_engine_cs *engine,
+static int __live_lrc_state(struct intel_engine_cs *engine,
 			    struct i915_vma *scratch)
 {
 	struct intel_context *ce;
@@ -3286,7 +3567,7 @@ static int __live_lrc_state(struct i915_gem_context *fixme,
 	int err;
 	int n;
 
-	ce = intel_context_create(fixme, engine);
+	ce = intel_context_create(engine);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
 
@@ -3360,7 +3641,6 @@ static int live_lrc_state(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *fixme;
 	struct i915_vma *scratch;
 	enum intel_engine_id id;
 	int err = 0;
@@ -3370,18 +3650,12 @@ static int live_lrc_state(void *arg)
 	 * intel_context.
 	 */
 
-	fixme = kernel_context(gt->i915);
-	if (!fixme)
-		return -ENOMEM;
-
 	scratch = create_scratch(gt);
-	if (IS_ERR(scratch)) {
-		err = PTR_ERR(scratch);
-		goto out_close;
-	}
+	if (IS_ERR(scratch))
+		return PTR_ERR(scratch);
 
 	for_each_engine(engine, gt, id) {
-		err = __live_lrc_state(fixme, engine, scratch);
+		err = __live_lrc_state(engine, scratch);
 		if (err)
 			break;
 	}
@@ -3390,8 +3664,6 @@ static int live_lrc_state(void *arg)
 		err = -EIO;
 
 	i915_vma_unpin_and_release(&scratch, 0);
-out_close:
-	kernel_context_close(fixme);
 	return err;
 }
 
@@ -3401,7 +3673,7 @@ static int gpr_make_dirty(struct intel_engine_cs *engine)
 	u32 *cs;
 	int n;
 
-	rq = i915_request_create(engine->kernel_context);
+	rq = intel_engine_create_kernel_request(engine);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
@@ -3424,8 +3696,7 @@ static int gpr_make_dirty(struct intel_engine_cs *engine)
 	return 0;
 }
 
-static int __live_gpr_clear(struct i915_gem_context *fixme,
-			    struct intel_engine_cs *engine,
+static int __live_gpr_clear(struct intel_engine_cs *engine,
 			    struct i915_vma *scratch)
 {
 	struct intel_context *ce;
@@ -3441,7 +3712,7 @@ static int __live_gpr_clear(struct i915_gem_context *fixme,
 	if (err)
 		return err;
 
-	ce = intel_context_create(fixme, engine);
+	ce = intel_context_create(engine);
 	if (IS_ERR(ce))
 		return PTR_ERR(ce);
 
@@ -3503,7 +3774,6 @@ static int live_gpr_clear(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *fixme;
 	struct i915_vma *scratch;
 	enum intel_engine_id id;
 	int err = 0;
@@ -3513,18 +3783,12 @@ static int live_gpr_clear(void *arg)
 	 * to avoid leaking any information from previous contexts.
 	 */
 
-	fixme = kernel_context(gt->i915);
-	if (!fixme)
-		return -ENOMEM;
-
 	scratch = create_scratch(gt);
-	if (IS_ERR(scratch)) {
-		err = PTR_ERR(scratch);
-		goto out_close;
-	}
+	if (IS_ERR(scratch))
+		return PTR_ERR(scratch);
 
 	for_each_engine(engine, gt, id) {
-		err = __live_gpr_clear(fixme, engine, scratch);
+		err = __live_gpr_clear(engine, scratch);
 		if (err)
 			break;
 	}
@@ -3533,8 +3797,6 @@ static int live_gpr_clear(void *arg)
 		err = -EIO;
 
 	i915_vma_unpin_and_release(&scratch, 0);
-out_close:
-	kernel_context_close(fixme);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/gt/selftest_mocs.c b/drivers/gpu/drm/i915/gt/selftest_mocs.c
new file mode 100644
index 000000000000..de1f83100fb6
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_mocs.c
@@ -0,0 +1,419 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "gt/intel_engine_pm.h"
+#include "i915_selftest.h"
+
+#include "gem/selftests/mock_context.h"
+#include "selftests/igt_reset.h"
+#include "selftests/igt_spinner.h"
+
+struct live_mocs {
+	struct drm_i915_mocs_table table;
+	struct i915_vma *scratch;
+	void *vaddr;
+};
+
+static int request_add_sync(struct i915_request *rq, int err)
+{
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (i915_request_wait(rq, 0, HZ / 5) < 0)
+		err = -ETIME;
+	i915_request_put(rq);
+
+	return err;
+}
+
+static int request_add_spin(struct i915_request *rq, struct igt_spinner *spin)
+{
+	int err = 0;
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (spin && !igt_wait_for_spinner(spin, rq))
+		err = -ETIME;
+	i915_request_put(rq);
+
+	return err;
+}
+
+static struct i915_vma *create_scratch(struct intel_gt *gt)
+{
+	struct drm_i915_gem_object *obj;
+	struct i915_vma *vma;
+	int err;
+
+	obj = i915_gem_object_create_internal(gt->i915, PAGE_SIZE);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	i915_gem_object_set_cache_coherency(obj, I915_CACHING_CACHED);
+
+	vma = i915_vma_instance(obj, &gt->ggtt->vm, NULL);
+	if (IS_ERR(vma)) {
+		i915_gem_object_put(obj);
+		return vma;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL);
+	if (err) {
+		i915_gem_object_put(obj);
+		return ERR_PTR(err);
+	}
+
+	return vma;
+}
+
+static int live_mocs_init(struct live_mocs *arg, struct intel_gt *gt)
+{
+	int err;
+
+	if (!get_mocs_settings(gt->i915, &arg->table))
+		return -EINVAL;
+
+	arg->scratch = create_scratch(gt);
+	if (IS_ERR(arg->scratch))
+		return PTR_ERR(arg->scratch);
+
+	arg->vaddr = i915_gem_object_pin_map(arg->scratch->obj, I915_MAP_WB);
+	if (IS_ERR(arg->vaddr)) {
+		err = PTR_ERR(arg->vaddr);
+		goto err_scratch;
+	}
+
+	return 0;
+
+err_scratch:
+	i915_vma_unpin_and_release(&arg->scratch, 0);
+	return err;
+}
+
+static void live_mocs_fini(struct live_mocs *arg)
+{
+	i915_vma_unpin_and_release(&arg->scratch, I915_VMA_RELEASE_MAP);
+}
+
+static int read_regs(struct i915_request *rq,
+		     u32 addr, unsigned int count,
+		     uint32_t *offset)
+{
+	unsigned int i;
+	u32 *cs;
+
+	GEM_BUG_ON(!IS_ALIGNED(*offset, sizeof(u32)));
+
+	cs = intel_ring_begin(rq, 4 * count);
+	if (IS_ERR(cs))
+		return PTR_ERR(cs);
+
+	for (i = 0; i < count; i++) {
+		*cs++ = MI_STORE_REGISTER_MEM_GEN8 | MI_USE_GGTT;
+		*cs++ = addr;
+		*cs++ = *offset;
+		*cs++ = 0;
+
+		addr += sizeof(u32);
+		*offset += sizeof(u32);
+	}
+
+	intel_ring_advance(rq, cs);
+
+	return 0;
+}
+
+static int read_mocs_table(struct i915_request *rq,
+			   const struct drm_i915_mocs_table *table,
+			   uint32_t *offset)
+{
+	u32 addr;
+
+	if (HAS_GLOBAL_MOCS_REGISTERS(rq->i915))
+		addr = global_mocs_offset();
+	else
+		addr = mocs_offset(rq->engine);
+
+	return read_regs(rq, addr, table->n_entries, offset);
+}
+
+static int read_l3cc_table(struct i915_request *rq,
+			   const struct drm_i915_mocs_table *table,
+			   uint32_t *offset)
+{
+	u32 addr = i915_mmio_reg_offset(GEN9_LNCFCMOCS(0));
+
+	return read_regs(rq, addr, (table->n_entries + 1) / 2, offset);
+}
+
+static int check_mocs_table(struct intel_engine_cs *engine,
+			    const struct drm_i915_mocs_table *table,
+			    uint32_t **vaddr)
+{
+	unsigned int i;
+	u32 expect;
+
+	for_each_mocs(expect, table, i) {
+		if (**vaddr != expect) {
+			pr_err("%s: Invalid MOCS[%d] entry, found %08x, expected %08x\n",
+			       engine->name, i, **vaddr, expect);
+			return -EINVAL;
+		}
+		++*vaddr;
+	}
+
+	return 0;
+}
+
+static bool mcr_range(struct drm_i915_private *i915, u32 offset)
+{
+	/*
+	 * Registers in this range are affected by the MCR selector
+	 * which only controls CPU initiated MMIO. Routing does not
+	 * work for CS access so we cannot verify them on this path.
+	 */
+	return INTEL_GEN(i915) >= 8 && offset >= 0xb000 && offset <= 0xb4ff;
+}
+
+static int check_l3cc_table(struct intel_engine_cs *engine,
+			    const struct drm_i915_mocs_table *table,
+			    uint32_t **vaddr)
+{
+	/* Can we read the MCR range 0xb00 directly? See intel_workarounds! */
+	u32 reg = i915_mmio_reg_offset(GEN9_LNCFCMOCS(0));
+	unsigned int i;
+	u32 expect;
+
+	for_each_l3cc(expect, table, i) {
+		if (!mcr_range(engine->i915, reg) && **vaddr != expect) {
+			pr_err("%s: Invalid L3CC[%d] entry, found %08x, expected %08x\n",
+			       engine->name, i, **vaddr, expect);
+			return -EINVAL;
+		}
+		++*vaddr;
+		reg += 4;
+	}
+
+	return 0;
+}
+
+static int check_mocs_engine(struct live_mocs *arg,
+			     struct intel_context *ce)
+{
+	struct i915_vma *vma = arg->scratch;
+	struct i915_request *rq;
+	u32 offset;
+	u32 *vaddr;
+	int err;
+
+	memset32(arg->vaddr, STACK_MAGIC, PAGE_SIZE / sizeof(u32));
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		return PTR_ERR(rq);
+
+	i915_vma_lock(vma);
+	err = i915_request_await_object(rq, vma->obj, true);
+	if (!err)
+		err = i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
+	i915_vma_unlock(vma);
+
+	/* Read the mocs tables back using SRM */
+	offset = i915_ggtt_offset(vma);
+	if (!err)
+		err = read_mocs_table(rq, &arg->table, &offset);
+	if (!err && ce->engine->class == RENDER_CLASS)
+		err = read_l3cc_table(rq, &arg->table, &offset);
+	offset -= i915_ggtt_offset(vma);
+	GEM_BUG_ON(offset > PAGE_SIZE);
+
+	err = request_add_sync(rq, err);
+	if (err)
+		return err;
+
+	/* Compare the results against the expected tables */
+	vaddr = arg->vaddr;
+	if (!err)
+		err = check_mocs_table(ce->engine, &arg->table, &vaddr);
+	if (!err && ce->engine->class == RENDER_CLASS)
+		err = check_l3cc_table(ce->engine, &arg->table, &vaddr);
+	if (err)
+		return err;
+
+	GEM_BUG_ON(arg->vaddr + offset != vaddr);
+	return 0;
+}
+
+static int live_mocs_kernel(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	struct live_mocs mocs;
+	int err;
+
+	/* Basic check the system is configured with the expected mocs table */
+
+	err = live_mocs_init(&mocs, gt);
+	if (err)
+		return err;
+
+	for_each_engine(engine, gt, id) {
+		intel_engine_pm_get(engine);
+		err = check_mocs_engine(&mocs, engine->kernel_context);
+		intel_engine_pm_put(engine);
+		if (err)
+			break;
+	}
+
+	live_mocs_fini(&mocs);
+	return err;
+}
+
+static int live_mocs_clean(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	struct live_mocs mocs;
+	int err;
+
+	/* Every new context should see the same mocs table */
+
+	err = live_mocs_init(&mocs, gt);
+	if (err)
+		return err;
+
+	for_each_engine(engine, gt, id) {
+		struct intel_context *ce;
+
+		ce = intel_context_create(engine);
+		if (IS_ERR(ce)) {
+			err = PTR_ERR(ce);
+			break;
+		}
+
+		err = check_mocs_engine(&mocs, ce);
+		intel_context_put(ce);
+		if (err)
+			break;
+	}
+
+	live_mocs_fini(&mocs);
+	return err;
+}
+
+static int active_engine_reset(struct intel_context *ce,
+			       const char *reason)
+{
+	struct igt_spinner spin;
+	struct i915_request *rq;
+	int err;
+
+	err = igt_spinner_init(&spin, ce->engine->gt);
+	if (err)
+		return err;
+
+	rq = igt_spinner_create_request(&spin, ce, MI_NOOP);
+	if (IS_ERR(rq)) {
+		igt_spinner_fini(&spin);
+		return PTR_ERR(rq);
+	}
+
+	err = request_add_spin(rq, &spin);
+	if (err == 0)
+		err = intel_engine_reset(ce->engine, reason);
+
+	igt_spinner_end(&spin);
+	igt_spinner_fini(&spin);
+
+	return err;
+}
+
+static int __live_mocs_reset(struct live_mocs *mocs,
+			     struct intel_context *ce)
+{
+	int err;
+
+	err = intel_engine_reset(ce->engine, "mocs");
+	if (err)
+		return err;
+
+	err = check_mocs_engine(mocs, ce);
+	if (err)
+		return err;
+
+	err = active_engine_reset(ce, "mocs");
+	if (err)
+		return err;
+
+	err = check_mocs_engine(mocs, ce);
+	if (err)
+		return err;
+
+	intel_gt_reset(ce->engine->gt, ce->engine->mask, "mocs");
+
+	err = check_mocs_engine(mocs, ce);
+	if (err)
+		return err;
+
+	return 0;
+}
+
+static int live_mocs_reset(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	struct live_mocs mocs;
+	int err = 0;
+
+	/* Check the mocs setup is retained over per-engine and global resets */
+
+	if (!intel_has_reset_engine(gt))
+		return 0;
+
+	err = live_mocs_init(&mocs, gt);
+	if (err)
+		return err;
+
+	igt_global_reset_lock(gt);
+	for_each_engine(engine, gt, id) {
+		struct intel_context *ce;
+
+		ce = intel_context_create(engine);
+		if (IS_ERR(ce)) {
+			err = PTR_ERR(ce);
+			break;
+		}
+
+		intel_engine_pm_get(engine);
+		err = __live_mocs_reset(&mocs, ce);
+		intel_engine_pm_put(engine);
+
+		intel_context_put(ce);
+		if (err)
+			break;
+	}
+	igt_global_reset_unlock(gt);
+
+	live_mocs_fini(&mocs);
+	return err;
+}
+
+int intel_mocs_live_selftests(struct drm_i915_private *i915)
+{
+	static const struct i915_subtest tests[] = {
+		SUBTEST(live_mocs_kernel),
+		SUBTEST(live_mocs_clean),
+		SUBTEST(live_mocs_reset),
+	};
+	struct drm_i915_mocs_table table;
+
+	if (!get_mocs_settings(i915, &table))
+		return 0;
+
+	return intel_gt_live_subtests(tests, &i915->gt);
+}
diff --git a/drivers/gpu/drm/i915/gt/selftest_rc6.c b/drivers/gpu/drm/i915/gt/selftest_rc6.c
new file mode 100644
index 000000000000..8cc55a0e9e06
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_rc6.c
@@ -0,0 +1,203 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include "intel_context.h"
+#include "intel_engine_pm.h"
+#include "intel_gt_requests.h"
+#include "intel_ring.h"
+#include "selftest_rc6.h"
+
+#include "selftests/i915_random.h"
+
+int live_rc6_manual(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_rc6 *rc6 = &gt->rc6;
+	intel_wakeref_t wakeref;
+	u64 res[2];
+	int err = 0;
+
+	/*
+	 * Our claim is that we can "encourage" the GPU to enter rc6 at will.
+	 * Let's try it!
+	 */
+
+	if (!rc6->enabled)
+		return 0;
+
+	/* bsw/byt use a PCU and decouple RC6 from our manual control */
+	if (IS_VALLEYVIEW(gt->i915) || IS_CHERRYVIEW(gt->i915))
+		return 0;
+
+	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
+
+	/* Force RC6 off for starters */
+	__intel_rc6_disable(rc6);
+	msleep(1); /* wakeup is not immediate, takes about 100us on icl */
+
+	res[0] = intel_rc6_residency_ns(rc6, GEN6_GT_GFX_RC6);
+	msleep(250);
+	res[1] = intel_rc6_residency_ns(rc6, GEN6_GT_GFX_RC6);
+	if ((res[1] - res[0]) >> 10) {
+		pr_err("RC6 residency increased by %lldus while disabled for 250ms!\n",
+		       (res[1] - res[0]) >> 10);
+		err = -EINVAL;
+		goto out_unlock;
+	}
+
+	/* Manually enter RC6 */
+	intel_rc6_park(rc6);
+
+	res[0] = intel_rc6_residency_ns(rc6, GEN6_GT_GFX_RC6);
+	msleep(100);
+	res[1] = intel_rc6_residency_ns(rc6, GEN6_GT_GFX_RC6);
+
+	if (res[1] == res[0]) {
+		pr_err("Did not enter RC6! RC6_STATE=%08x, RC6_CONTROL=%08x\n",
+		       intel_uncore_read_fw(gt->uncore, GEN6_RC_STATE),
+		       intel_uncore_read_fw(gt->uncore, GEN6_RC_CONTROL));
+		err = -EINVAL;
+	}
+
+	/* Restore what should have been the original state! */
+	intel_rc6_unpark(rc6);
+
+out_unlock:
+	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
+	return err;
+}
+
+static const u32 *__live_rc6_ctx(struct intel_context *ce)
+{
+	struct i915_request *rq;
+	const u32 *result;
+	u32 cmd;
+	u32 *cs;
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		return ERR_CAST(rq);
+
+	cs = intel_ring_begin(rq, 4);
+	if (IS_ERR(cs)) {
+		i915_request_add(rq);
+		return cs;
+	}
+
+	cmd = MI_STORE_REGISTER_MEM | MI_USE_GGTT;
+	if (INTEL_GEN(rq->i915) >= 8)
+		cmd++;
+
+	*cs++ = cmd;
+	*cs++ = i915_mmio_reg_offset(GEN8_RC6_CTX_INFO);
+	*cs++ = ce->timeline->hwsp_offset + 8;
+	*cs++ = 0;
+	intel_ring_advance(rq, cs);
+
+	result = rq->hwsp_seqno + 2;
+	i915_request_add(rq);
+
+	return result;
+}
+
+static struct intel_engine_cs **
+randomised_engines(struct intel_gt *gt,
+		   struct rnd_state *prng,
+		   unsigned int *count)
+{
+	struct intel_engine_cs *engine, **engines;
+	enum intel_engine_id id;
+	int n;
+
+	n = 0;
+	for_each_engine(engine, gt, id)
+		n++;
+	if (!n)
+		return NULL;
+
+	engines = kmalloc_array(n, sizeof(*engines), GFP_KERNEL);
+	if (!engines)
+		return NULL;
+
+	n = 0;
+	for_each_engine(engine, gt, id)
+		engines[n++] = engine;
+
+	i915_prandom_shuffle(engines, sizeof(*engines), n, prng);
+
+	*count = n;
+	return engines;
+}
+
+int live_rc6_ctx_wa(void *arg)
+{
+	struct intel_gt *gt = arg;
+	struct intel_engine_cs **engines;
+	unsigned int n, count;
+	I915_RND_STATE(prng);
+	int err = 0;
+
+	/* A read of CTX_INFO upsets rc6. Poke the bear! */
+	if (INTEL_GEN(gt->i915) < 8)
+		return 0;
+
+	engines = randomised_engines(gt, &prng, &count);
+	if (!engines)
+		return 0;
+
+	for (n = 0; n < count; n++) {
+		struct intel_engine_cs *engine = engines[n];
+		int pass;
+
+		for (pass = 0; pass < 2; pass++) {
+			struct intel_context *ce;
+			unsigned int resets =
+				i915_reset_engine_count(&gt->i915->gpu_error,
+							engine);
+			const u32 *res;
+
+			/* Use a sacrifical context */
+			ce = intel_context_create(engine);
+			if (IS_ERR(ce)) {
+				err = PTR_ERR(ce);
+				goto out;
+			}
+
+			intel_engine_pm_get(engine);
+			res = __live_rc6_ctx(ce);
+			intel_engine_pm_put(engine);
+			intel_context_put(ce);
+			if (IS_ERR(res)) {
+				err = PTR_ERR(res);
+				goto out;
+			}
+
+			if (intel_gt_wait_for_idle(gt, HZ / 5) == -ETIME) {
+				intel_gt_set_wedged(gt);
+				err = -ETIME;
+				goto out;
+			}
+
+			intel_gt_pm_wait_for_idle(gt);
+			pr_debug("%s: CTX_INFO=%0x\n",
+				 engine->name, READ_ONCE(*res));
+
+			if (resets !=
+			    i915_reset_engine_count(&gt->i915->gpu_error,
+						    engine)) {
+				pr_err("%s: GPU reset required\n",
+				       engine->name);
+				add_taint_for_CI(TAINT_WARN);
+				err = -EIO;
+				goto out;
+			}
+		}
+	}
+
+out:
+	kfree(engines);
+	return err;
+}
diff --git a/drivers/gpu/drm/i915/gt/selftest_rc6.h b/drivers/gpu/drm/i915/gt/selftest_rc6.h
new file mode 100644
index 000000000000..762fd442d7b2
--- /dev/null
+++ b/drivers/gpu/drm/i915/gt/selftest_rc6.h
@@ -0,0 +1,13 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef SELFTEST_RC6_H
+#define SELFTEST_RC6_H
+
+int live_rc6_ctx_wa(void *arg);
+int live_rc6_manual(void *arg);
+
+#endif /* SELFTEST_RC6_H */
diff --git a/drivers/gpu/drm/i915/gt/selftest_timeline.c b/drivers/gpu/drm/i915/gt/selftest_timeline.c
index f04a59fe5d2c..e2d78cc22fb4 100644
--- a/drivers/gpu/drm/i915/gt/selftest_timeline.c
+++ b/drivers/gpu/drm/i915/gt/selftest_timeline.c
@@ -458,7 +458,7 @@ tl_write(struct intel_timeline *tl, struct intel_engine_cs *engine, u32 value)
 		goto out;
 	}
 
-	rq = i915_request_create(engine->kernel_context);
+	rq = intel_engine_create_kernel_request(engine);
 	if (IS_ERR(rq))
 		goto out_unpin;
 
@@ -675,9 +675,7 @@ static int live_hwsp_wrap(void *arg)
 		if (!intel_engine_can_store_dword(engine))
 			continue;
 
-		intel_engine_pm_get(engine);
-		rq = i915_request_create(engine->kernel_context);
-		intel_engine_pm_put(engine);
+		rq = intel_engine_create_kernel_request(engine);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
 			goto out;
diff --git a/drivers/gpu/drm/i915/gt/selftest_workarounds.c b/drivers/gpu/drm/i915/gt/selftest_workarounds.c
index abce6e4ec9c0..ac1921854cbf 100644
--- a/drivers/gpu/drm/i915/gt/selftest_workarounds.c
+++ b/drivers/gpu/drm/i915/gt/selftest_workarounds.c
@@ -264,22 +264,15 @@ static int
 switch_to_scratch_context(struct intel_engine_cs *engine,
 			  struct igt_spinner *spin)
 {
-	struct i915_gem_context *ctx;
 	struct intel_context *ce;
 	struct i915_request *rq;
 	int err = 0;
 
-	ctx = kernel_context(engine->i915);
-	if (IS_ERR(ctx))
-		return PTR_ERR(ctx);
-
-	GEM_BUG_ON(i915_gem_context_is_bannable(ctx));
-
-	ce = i915_gem_context_get_engine(ctx, engine->legacy_idx);
-	GEM_BUG_ON(IS_ERR(ce));
+	ce = intel_context_create(engine);
+	if (IS_ERR(ce))
+		return PTR_ERR(ce);
 
 	rq = igt_spinner_create_request(spin, ce, MI_NOOP);
-
 	intel_context_put(ce);
 
 	if (IS_ERR(rq)) {
@@ -293,7 +286,6 @@ err:
 	if (err && spin)
 		igt_spinner_end(spin);
 
-	kernel_context_close(ctx);
 	return err;
 }
 
@@ -367,20 +359,17 @@ out_ctx:
 	return err;
 }
 
-static struct i915_vma *create_batch(struct i915_gem_context *ctx)
+static struct i915_vma *create_batch(struct i915_address_space *vm)
 {
 	struct drm_i915_gem_object *obj;
-	struct i915_address_space *vm;
 	struct i915_vma *vma;
 	int err;
 
-	obj = i915_gem_object_create_internal(ctx->i915, 16 * PAGE_SIZE);
+	obj = i915_gem_object_create_internal(vm->i915, 16 * PAGE_SIZE);
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
-	vm = i915_gem_context_get_vm_rcu(ctx);
 	vma = i915_vma_instance(obj, vm, NULL);
-	i915_vm_put(vm);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		goto err_obj;
@@ -452,8 +441,7 @@ static int whitelist_writable_count(struct intel_engine_cs *engine)
 	return count;
 }
 
-static int check_dirty_whitelist(struct i915_gem_context *ctx,
-				 struct intel_engine_cs *engine)
+static int check_dirty_whitelist(struct intel_context *ce)
 {
 	const u32 values[] = {
 		0x00000000,
@@ -481,19 +469,17 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 		0xffff00ff,
 		0xffffffff,
 	};
-	struct i915_address_space *vm;
+	struct intel_engine_cs *engine = ce->engine;
 	struct i915_vma *scratch;
 	struct i915_vma *batch;
 	int err = 0, i, v;
 	u32 *cs, *results;
 
-	vm = i915_gem_context_get_vm_rcu(ctx);
-	scratch = create_scratch(vm, 2 * ARRAY_SIZE(values) + 1);
-	i915_vm_put(vm);
+	scratch = create_scratch(ce->vm, 2 * ARRAY_SIZE(values) + 1);
 	if (IS_ERR(scratch))
 		return PTR_ERR(scratch);
 
-	batch = create_batch(ctx);
+	batch = create_batch(ce->vm);
 	if (IS_ERR(batch)) {
 		err = PTR_ERR(batch);
 		goto out_scratch;
@@ -518,7 +504,7 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 
 		srm = MI_STORE_REGISTER_MEM;
 		lrm = MI_LOAD_REGISTER_MEM;
-		if (INTEL_GEN(ctx->i915) >= 8)
+		if (INTEL_GEN(engine->i915) >= 8)
 			lrm++, srm++;
 
 		pr_debug("%s: Writing garbage to %x\n",
@@ -577,7 +563,7 @@ static int check_dirty_whitelist(struct i915_gem_context *ctx,
 		i915_gem_object_unpin_map(batch->obj);
 		intel_gt_chipset_flush(engine->gt);
 
-		rq = igt_request_alloc(ctx, engine);
+		rq = intel_context_create_request(ce);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
 			goto out_batch;
@@ -696,7 +682,7 @@ out_unpin:
 			break;
 	}
 
-	if (igt_flush_test(ctx->i915))
+	if (igt_flush_test(engine->i915))
 		err = -EIO;
 out_batch:
 	i915_vma_unpin_and_release(&batch, 0);
@@ -709,38 +695,31 @@ static int live_dirty_whitelist(void *arg)
 {
 	struct intel_gt *gt = arg;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *ctx;
 	enum intel_engine_id id;
-	struct drm_file *file;
-	int err = 0;
 
 	/* Can the user write to the whitelisted registers? */
 
 	if (INTEL_GEN(gt->i915) < 7) /* minimum requirement for LRI, SRM, LRM */
 		return 0;
 
-	file = mock_file(gt->i915);
-	if (IS_ERR(file))
-		return PTR_ERR(file);
-
-	ctx = live_context(gt->i915, file);
-	if (IS_ERR(ctx)) {
-		err = PTR_ERR(ctx);
-		goto out_file;
-	}
-
 	for_each_engine(engine, gt, id) {
+		struct intel_context *ce;
+		int err;
+
 		if (engine->whitelist.count == 0)
 			continue;
 
-		err = check_dirty_whitelist(ctx, engine);
+		ce = intel_context_create(engine);
+		if (IS_ERR(ce))
+			return PTR_ERR(ce);
+
+		err = check_dirty_whitelist(ce);
+		intel_context_put(ce);
 		if (err)
-			goto out_file;
+			return err;
 	}
 
-out_file:
-	mock_file_free(gt->i915, file);
-	return err;
+	return 0;
 }
 
 static int live_reset_whitelist(void *arg)
@@ -830,12 +809,15 @@ err_req:
 static int scrub_whitelisted_registers(struct i915_gem_context *ctx,
 				       struct intel_engine_cs *engine)
 {
+	struct i915_address_space *vm;
 	struct i915_request *rq;
 	struct i915_vma *batch;
 	int i, err = 0;
 	u32 *cs;
 
-	batch = create_batch(ctx);
+	vm = i915_gem_context_get_vm_rcu(ctx);
+	batch = create_batch(vm);
+	i915_vm_put(vm);
 	if (IS_ERR(batch))
 		return PTR_ERR(batch);
 
diff --git a/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c b/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c
index 2a77c051f36a..aeb1d1f616e8 100644
--- a/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c
+++ b/drivers/gpu/drm/i915/gt/selftests/mock_timeline.c
@@ -15,7 +15,7 @@ void mock_timeline_init(struct intel_timeline *timeline, u64 context)
 
 	mutex_init(&timeline->mutex);
 
-	INIT_ACTIVE_FENCE(&timeline->last_request, &timeline->mutex);
+	INIT_ACTIVE_FENCE(&timeline->last_request);
 	INIT_LIST_HEAD(&timeline->requests);
 
 	i915_syncmap_init(&timeline->sync);
diff --git a/drivers/gpu/drm/i915/gt/selftests/mock_timeline.h b/drivers/gpu/drm/i915/gt/selftests/mock_timeline.h
index 689efc66c908..d2bcc3df6183 100644
--- a/drivers/gpu/drm/i915/gt/selftests/mock_timeline.h
+++ b/drivers/gpu/drm/i915/gt/selftests/mock_timeline.h
@@ -7,6 +7,8 @@
 #ifndef __MOCK_TIMELINE__
 #define __MOCK_TIMELINE__
 
+#include <linux/types.h>
+
 struct intel_timeline;
 
 void mock_timeline_init(struct intel_timeline *timeline, u64 context);
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.c b/drivers/gpu/drm/i915/gt/uc/intel_guc.c
index 3ee4a4e7689d..5d00a3b2d914 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.c
@@ -32,18 +32,17 @@
  * just the HuC, but more are expected to land in the future).
  */
 
-static void gen8_guc_raise_irq(struct intel_guc *guc)
+void intel_guc_notify(struct intel_guc *guc)
 {
 	struct intel_gt *gt = guc_to_gt(guc);
 
-	intel_uncore_write(gt->uncore, GUC_SEND_INTERRUPT, GUC_SEND_TRIGGER);
-}
-
-static void gen11_guc_raise_irq(struct intel_guc *guc)
-{
-	struct intel_gt *gt = guc_to_gt(guc);
-
-	intel_uncore_write(gt->uncore, GEN11_GUC_HOST_INTERRUPT, 0);
+	/*
+	 * On Gen11+, the value written to the register is passes as a payload
+	 * to the FW. However, the FW currently treats all values the same way
+	 * (H2G interrupt), so we can just write the value that the HW expects
+	 * on older gens.
+	 */
+	intel_uncore_write(gt->uncore, guc->notify_reg, GUC_SEND_TRIGGER);
 }
 
 static inline i915_reg_t guc_send_reg(struct intel_guc *guc, u32 i)
@@ -177,15 +176,13 @@ void intel_guc_init_early(struct intel_guc *guc)
 
 	mutex_init(&guc->send_mutex);
 	spin_lock_init(&guc->irq_lock);
-	guc->send = intel_guc_send_nop;
-	guc->handler = intel_guc_to_host_event_handler_nop;
 	if (INTEL_GEN(i915) >= 11) {
-		guc->notify = gen11_guc_raise_irq;
+		guc->notify_reg = GEN11_GUC_HOST_INTERRUPT;
 		guc->interrupts.reset = gen11_reset_guc_interrupts;
 		guc->interrupts.enable = gen11_enable_guc_interrupts;
 		guc->interrupts.disable = gen11_disable_guc_interrupts;
 	} else {
-		guc->notify = gen8_guc_raise_irq;
+		guc->notify_reg = GUC_SEND_INTERRUPT;
 		guc->interrupts.reset = gen9_reset_guc_interrupts;
 		guc->interrupts.enable = gen9_enable_guc_interrupts;
 		guc->interrupts.disable = gen9_disable_guc_interrupts;
@@ -401,18 +398,8 @@ void intel_guc_fini(struct intel_guc *guc)
 	intel_guc_log_destroy(&guc->log);
 	intel_uc_fw_fini(&guc->fw);
 	intel_uc_fw_cleanup_fetch(&guc->fw);
-}
 
-int intel_guc_send_nop(struct intel_guc *guc, const u32 *action, u32 len,
-		       u32 *response_buf, u32 response_buf_size)
-{
-	WARN(1, "Unexpected send: action=%#x\n", *action);
-	return -ENODEV;
-}
-
-void intel_guc_to_host_event_handler_nop(struct intel_guc *guc)
-{
-	WARN(1, "Unexpected event: no suitable handler\n");
+	intel_uc_fw_change_status(&guc->fw, INTEL_UC_FIRMWARE_DISABLED);
 }
 
 /*
@@ -704,3 +691,37 @@ err:
 	i915_gem_object_put(obj);
 	return vma;
 }
+
+/**
+ * intel_guc_allocate_and_map_vma() - Allocate and map VMA for GuC usage
+ * @guc:	the guc
+ * @size:	size of area to allocate (both virtual space and memory)
+ * @out_vma:	return variable for the allocated vma pointer
+ * @out_vaddr:	return variable for the obj mapping
+ *
+ * This wrapper calls intel_guc_allocate_vma() and then maps the allocated
+ * object with I915_MAP_WB.
+ *
+ * Return:	0 if successful, a negative errno code otherwise.
+ */
+int intel_guc_allocate_and_map_vma(struct intel_guc *guc, u32 size,
+				   struct i915_vma **out_vma, void **out_vaddr)
+{
+	struct i915_vma *vma;
+	void *vaddr;
+
+	vma = intel_guc_allocate_vma(guc, size);
+	if (IS_ERR(vma))
+		return PTR_ERR(vma);
+
+	vaddr = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
+	if (IS_ERR(vaddr)) {
+		i915_vma_unpin_and_release(&vma, 0);
+		return PTR_ERR(vaddr);
+	}
+
+	*out_vma = vma;
+	*out_vaddr = vaddr;
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc.h b/drivers/gpu/drm/i915/gt/uc/intel_guc.h
index e6400204a2bd..910d49590068 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc.h
@@ -20,8 +20,8 @@ struct __guc_ads_blob;
 
 /*
  * Top level structure of GuC. It handles firmware loading and manages client
- * pool and doorbells. intel_guc owns a intel_guc_client to replace the legacy
- * ExecList submission.
+ * pool. intel_guc owns a intel_guc_client to replace the legacy ExecList
+ * submission.
  */
 struct intel_guc {
 	struct intel_uc_fw fw;
@@ -46,13 +46,13 @@ struct intel_guc {
 
 	struct i915_vma *stage_desc_pool;
 	void *stage_desc_pool_vaddr;
-	struct ida stage_ids;
 
-	struct intel_guc_client *execbuf_client;
+	struct i915_vma *workqueue;
+	void *workqueue_vaddr;
+	spinlock_t wq_lock;
 
-	DECLARE_BITMAP(doorbell_bitmap, GUC_NUM_DOORBELLS);
-	/* Cyclic counter mod pagesize	*/
-	u32 db_cacheline;
+	struct i915_vma *proc_desc;
+	void *proc_desc_vaddr;
 
 	/* Control params for fw initialization */
 	u32 params[GUC_CTL_MAX_DWORDS];
@@ -64,44 +64,33 @@ struct intel_guc {
 		enum forcewake_domains fw_domains;
 	} send_regs;
 
+	/* register used to send interrupts to the GuC FW */
+	i915_reg_t notify_reg;
+
 	/* Store msg (e.g. log flush) that we see while CTBs are disabled */
 	u32 mmio_msg;
 
 	/* To serialize the intel_guc_send actions */
 	struct mutex send_mutex;
-
-	/* GuC's FW specific send function */
-	int (*send)(struct intel_guc *guc, const u32 *data, u32 len,
-		    u32 *response_buf, u32 response_buf_size);
-
-	/* GuC's FW specific event handler function */
-	void (*handler)(struct intel_guc *guc);
-
-	/* GuC's FW specific notify function */
-	void (*notify)(struct intel_guc *guc);
 };
 
 static
 inline int intel_guc_send(struct intel_guc *guc, const u32 *action, u32 len)
 {
-	return guc->send(guc, action, len, NULL, 0);
+	return intel_guc_ct_send(&guc->ct, action, len, NULL, 0);
 }
 
 static inline int
 intel_guc_send_and_receive(struct intel_guc *guc, const u32 *action, u32 len,
 			   u32 *response_buf, u32 response_buf_size)
 {
-	return guc->send(guc, action, len, response_buf, response_buf_size);
-}
-
-static inline void intel_guc_notify(struct intel_guc *guc)
-{
-	guc->notify(guc);
+	return intel_guc_ct_send(&guc->ct, action, len,
+				 response_buf, response_buf_size);
 }
 
 static inline void intel_guc_to_host_event_handler(struct intel_guc *guc)
 {
-	guc->handler(guc);
+	intel_guc_ct_event_handler(&guc->ct);
 }
 
 /* GuC addresses above GUC_GGTT_TOP also don't map through the GTT */
@@ -136,12 +125,9 @@ void intel_guc_init_send_regs(struct intel_guc *guc);
 void intel_guc_write_params(struct intel_guc *guc);
 int intel_guc_init(struct intel_guc *guc);
 void intel_guc_fini(struct intel_guc *guc);
-int intel_guc_send_nop(struct intel_guc *guc, const u32 *action, u32 len,
-		       u32 *response_buf, u32 response_buf_size);
+void intel_guc_notify(struct intel_guc *guc);
 int intel_guc_send_mmio(struct intel_guc *guc, const u32 *action, u32 len,
 			u32 *response_buf, u32 response_buf_size);
-void intel_guc_to_host_event_handler(struct intel_guc *guc);
-void intel_guc_to_host_event_handler_nop(struct intel_guc *guc);
 int intel_guc_to_host_process_recv_msg(struct intel_guc *guc,
 				       const u32 *payload, u32 len);
 int intel_guc_sample_forcewake(struct intel_guc *guc);
@@ -149,6 +135,8 @@ int intel_guc_auth_huc(struct intel_guc *guc, u32 rsa_offset);
 int intel_guc_suspend(struct intel_guc *guc);
 int intel_guc_resume(struct intel_guc *guc);
 struct i915_vma *intel_guc_allocate_vma(struct intel_guc *guc, u32 size);
+int intel_guc_allocate_and_map_vma(struct intel_guc *guc, u32 size,
+				   struct i915_vma **out_vma, void **out_vaddr);
 
 static inline bool intel_guc_is_supported(struct intel_guc *guc)
 {
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c
index ca6674b8e00c..101728006ae9 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ads.c
@@ -93,7 +93,8 @@ static void __guc_ads_init(struct intel_guc *guc)
 		 */
 		blob->ads.golden_context_lrca[engine_class] = 0;
 		blob->ads.eng_state_size[engine_class] =
-			intel_engine_context_size(dev_priv, engine_class) -
+			intel_engine_context_size(guc_to_gt(guc),
+						  engine_class) -
 			skipped_size;
 	}
 
@@ -135,32 +136,19 @@ static void __guc_ads_init(struct intel_guc *guc)
 int intel_guc_ads_create(struct intel_guc *guc)
 {
 	const u32 size = PAGE_ALIGN(sizeof(struct __guc_ads_blob));
-	struct i915_vma *vma;
-	void *blob;
 	int ret;
 
 	GEM_BUG_ON(guc->ads_vma);
 
-	vma = intel_guc_allocate_vma(guc, size);
-	if (IS_ERR(vma))
-		return PTR_ERR(vma);
+	ret = intel_guc_allocate_and_map_vma(guc, size, &guc->ads_vma,
+					     (void **)&guc->ads_blob);
 
-	blob = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
-	if (IS_ERR(blob)) {
-		ret = PTR_ERR(blob);
-		goto err_vma;
-	}
-
-	guc->ads_vma = vma;
-	guc->ads_blob = blob;
+	if (ret)
+		return ret;
 
 	__guc_ads_init(guc);
 
 	return 0;
-
-err_vma:
-	i915_vma_unpin_and_release(&guc->ads_vma, 0);
-	return ret;
 }
 
 void intel_guc_ads_destroy(struct intel_guc *guc)
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
index b49115517510..c6f971a049f9 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
@@ -37,13 +37,10 @@ static void ct_incoming_request_worker_func(struct work_struct *w);
  */
 void intel_guc_ct_init_early(struct intel_guc_ct *ct)
 {
-	/* we're using static channel owners */
-	ct->host_channel.owner = CTB_OWNER_HOST;
-
-	spin_lock_init(&ct->lock);
-	INIT_LIST_HEAD(&ct->pending_requests);
-	INIT_LIST_HEAD(&ct->incoming_requests);
-	INIT_WORK(&ct->worker, ct_incoming_request_worker_func);
+	spin_lock_init(&ct->requests.lock);
+	INIT_LIST_HEAD(&ct->requests.pending);
+	INIT_LIST_HEAD(&ct->requests.incoming);
+	INIT_WORK(&ct->requests.worker, ct_incoming_request_worker_func);
 }
 
 static inline struct intel_guc *ct_to_guc(struct intel_guc_ct *ct)
@@ -64,14 +61,13 @@ static inline const char *guc_ct_buffer_type_to_str(u32 type)
 }
 
 static void guc_ct_buffer_desc_init(struct guc_ct_buffer_desc *desc,
-				    u32 cmds_addr, u32 size, u32 owner)
+				    u32 cmds_addr, u32 size)
 {
-	CT_DEBUG_DRIVER("CT: desc %p init addr=%#x size=%u owner=%u\n",
-			desc, cmds_addr, size, owner);
+	CT_DEBUG_DRIVER("CT: init addr=%#x size=%u\n", cmds_addr, size);
 	memset(desc, 0, sizeof(*desc));
 	desc->addr = cmds_addr;
 	desc->size = size;
-	desc->owner = owner;
+	desc->owner = CTB_OWNER_HOST;
 }
 
 static void guc_ct_buffer_desc_reset(struct guc_ct_buffer_desc *desc)
@@ -104,12 +100,11 @@ static int guc_action_register_ct_buffer(struct intel_guc *guc,
 }
 
 static int guc_action_deregister_ct_buffer(struct intel_guc *guc,
-					   u32 owner,
 					   u32 type)
 {
 	u32 action[] = {
 		INTEL_GUC_ACTION_DEREGISTER_COMMAND_TRANSPORT_BUFFER,
-		owner,
+		CTB_OWNER_HOST,
 		type
 	};
 	int err;
@@ -117,20 +112,27 @@ static int guc_action_deregister_ct_buffer(struct intel_guc *guc,
 	/* Can't use generic send(), CT deregistration must go over MMIO */
 	err = intel_guc_send_mmio(guc, action, ARRAY_SIZE(action), NULL, 0);
 	if (err)
-		DRM_ERROR("CT: deregister %s buffer failed; owner=%d err=%d\n",
-			  guc_ct_buffer_type_to_str(type), owner, err);
+		DRM_ERROR("CT: deregister %s buffer failed; err=%d\n",
+			  guc_ct_buffer_type_to_str(type), err);
 	return err;
 }
 
-static int ctch_init(struct intel_guc *guc,
-		     struct intel_guc_ct_channel *ctch)
+/**
+ * intel_guc_ct_init - Init buffer-based communication
+ * @ct: pointer to CT struct
+ *
+ * Allocate memory required for buffer-based communication.
+ *
+ * Return: 0 on success, a negative errno code on failure.
+ */
+int intel_guc_ct_init(struct intel_guc_ct *ct)
 {
-	struct i915_vma *vma;
+	struct intel_guc *guc = ct_to_guc(ct);
 	void *blob;
 	int err;
 	int i;
 
-	GEM_BUG_ON(ctch->vma);
+	GEM_BUG_ON(ct->vma);
 
 	/* We allocate 1 page to hold both descriptors and both buffers.
 	 *       ___________.....................
@@ -154,71 +156,65 @@ static int ctch_init(struct intel_guc *guc,
 	 * other code will need updating as well.
 	 */
 
-	/* allocate vma */
-	vma = intel_guc_allocate_vma(guc, PAGE_SIZE);
-	if (IS_ERR(vma)) {
-		err = PTR_ERR(vma);
-		goto err_out;
+	err = intel_guc_allocate_and_map_vma(guc, PAGE_SIZE, &ct->vma, &blob);
+	if (err) {
+		DRM_ERROR("CT: channel allocation failed; err=%d\n", err);
+		return err;
 	}
-	ctch->vma = vma;
 
-	/* map first page */
-	blob = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
-	if (IS_ERR(blob)) {
-		err = PTR_ERR(blob);
-		goto err_vma;
-	}
 	CT_DEBUG_DRIVER("CT: vma base=%#x\n",
-			intel_guc_ggtt_offset(guc, ctch->vma));
+			intel_guc_ggtt_offset(guc, ct->vma));
 
 	/* store pointers to desc and cmds */
-	for (i = 0; i < ARRAY_SIZE(ctch->ctbs); i++) {
-		GEM_BUG_ON((i != CTB_SEND) && (i != CTB_RECV));
-		ctch->ctbs[i].desc = blob + PAGE_SIZE/4 * i;
-		ctch->ctbs[i].cmds = blob + PAGE_SIZE/4 * i + PAGE_SIZE/2;
+	for (i = 0; i < ARRAY_SIZE(ct->ctbs); i++) {
+		GEM_BUG_ON((i !=  CTB_SEND) && (i != CTB_RECV));
+		ct->ctbs[i].desc = blob + PAGE_SIZE/4 * i;
+		ct->ctbs[i].cmds = blob + PAGE_SIZE/4 * i + PAGE_SIZE/2;
 	}
 
 	return 0;
-
-err_vma:
-	i915_vma_unpin_and_release(&ctch->vma, 0);
-err_out:
-	CT_DEBUG_DRIVER("CT: channel %d initialization failed; err=%d\n",
-			ctch->owner, err);
-	return err;
 }
 
-static void ctch_fini(struct intel_guc *guc,
-		      struct intel_guc_ct_channel *ctch)
+/**
+ * intel_guc_ct_fini - Fini buffer-based communication
+ * @ct: pointer to CT struct
+ *
+ * Deallocate memory required for buffer-based communication.
+ */
+void intel_guc_ct_fini(struct intel_guc_ct *ct)
 {
-	GEM_BUG_ON(ctch->enabled);
+	GEM_BUG_ON(ct->enabled);
 
-	i915_vma_unpin_and_release(&ctch->vma, I915_VMA_RELEASE_MAP);
+	i915_vma_unpin_and_release(&ct->vma, I915_VMA_RELEASE_MAP);
 }
 
-static int ctch_enable(struct intel_guc *guc,
-		       struct intel_guc_ct_channel *ctch)
+/**
+ * intel_guc_ct_enable - Enable buffer based command transport.
+ * @ct: pointer to CT struct
+ *
+ * Return: 0 on success, a negative errno code on failure.
+ */
+int intel_guc_ct_enable(struct intel_guc_ct *ct)
 {
+	struct intel_guc *guc = ct_to_guc(ct);
 	u32 base;
 	int err;
 	int i;
 
-	GEM_BUG_ON(!ctch->vma);
-
-	GEM_BUG_ON(ctch->enabled);
+	GEM_BUG_ON(ct->enabled);
 
 	/* vma should be already allocated and map'ed */
-	base = intel_guc_ggtt_offset(guc, ctch->vma);
+	GEM_BUG_ON(!ct->vma);
+	base = intel_guc_ggtt_offset(guc, ct->vma);
 
 	/* (re)initialize descriptors
 	 * cmds buffers are in the second half of the blob page
 	 */
-	for (i = 0; i < ARRAY_SIZE(ctch->ctbs); i++) {
+	for (i = 0; i < ARRAY_SIZE(ct->ctbs); i++) {
 		GEM_BUG_ON((i != CTB_SEND) && (i != CTB_RECV));
-		guc_ct_buffer_desc_init(ctch->ctbs[i].desc,
+		guc_ct_buffer_desc_init(ct->ctbs[i].desc,
 					base + PAGE_SIZE/4 * i + PAGE_SIZE/2,
-					PAGE_SIZE/4,
-					ctch->owner);
+					PAGE_SIZE/4);
 	}
 
 	/* register buffers, starting wirh RECV buffer
@@ -236,38 +232,42 @@ static int ctch_enable(struct intel_guc *guc,
 	if (unlikely(err))
 		goto err_deregister;
 
-	ctch->enabled = true;
+	ct->enabled = true;
 
 	return 0;
 
 err_deregister:
 	guc_action_deregister_ct_buffer(guc,
-					ctch->owner,
 					INTEL_GUC_CT_BUFFER_TYPE_RECV);
 err_out:
-	DRM_ERROR("CT: can't open channel %d; err=%d\n", ctch->owner, err);
+	DRM_ERROR("CT: can't open channel; err=%d\n", err);
 	return err;
 }
 
-static void ctch_disable(struct intel_guc *guc,
-			 struct intel_guc_ct_channel *ctch)
+/**
+ * intel_guc_ct_disable - Disable buffer based command transport.
+ * @ct: pointer to CT struct
+ */
+void intel_guc_ct_disable(struct intel_guc_ct *ct)
 {
-	GEM_BUG_ON(!ctch->enabled);
+	struct intel_guc *guc = ct_to_guc(ct);
 
-	ctch->enabled = false;
+	GEM_BUG_ON(!ct->enabled);
 
-	guc_action_deregister_ct_buffer(guc,
-					ctch->owner,
-					INTEL_GUC_CT_BUFFER_TYPE_SEND);
-	guc_action_deregister_ct_buffer(guc,
-					ctch->owner,
-					INTEL_GUC_CT_BUFFER_TYPE_RECV);
+	ct->enabled = false;
+
+	if (intel_guc_is_running(guc)) {
+		guc_action_deregister_ct_buffer(guc,
+						INTEL_GUC_CT_BUFFER_TYPE_SEND);
+		guc_action_deregister_ct_buffer(guc,
+						INTEL_GUC_CT_BUFFER_TYPE_RECV);
+	}
 }
 
-static u32 ctch_get_next_fence(struct intel_guc_ct_channel *ctch)
+static u32 ct_get_next_fence(struct intel_guc_ct *ct)
 {
 	/* For now it's trivial */
-	return ++ctch->next_fence;
+	return ++ct->requests.next_fence;
 }
 
 /**
@@ -440,35 +440,34 @@ static int wait_for_ct_request_update(struct ct_request *req, u32 *status)
 	return err;
 }
 
-static int ctch_send(struct intel_guc_ct *ct,
-		     struct intel_guc_ct_channel *ctch,
-		     const u32 *action,
-		     u32 len,
-		     u32 *response_buf,
-		     u32 response_buf_size,
-		     u32 *status)
+static int ct_send(struct intel_guc_ct *ct,
+		   const u32 *action,
+		   u32 len,
+		   u32 *response_buf,
+		   u32 response_buf_size,
+		   u32 *status)
 {
-	struct intel_guc_ct_buffer *ctb = &ctch->ctbs[CTB_SEND];
+	struct intel_guc_ct_buffer *ctb = &ct->ctbs[CTB_SEND];
 	struct guc_ct_buffer_desc *desc = ctb->desc;
 	struct ct_request request;
 	unsigned long flags;
 	u32 fence;
 	int err;
 
-	GEM_BUG_ON(!ctch->enabled);
+	GEM_BUG_ON(!ct->enabled);
 	GEM_BUG_ON(!len);
 	GEM_BUG_ON(len & ~GUC_CT_MSG_LEN_MASK);
 	GEM_BUG_ON(!response_buf && response_buf_size);
 
-	fence = ctch_get_next_fence(ctch);
+	fence = ct_get_next_fence(ct);
 	request.fence = fence;
 	request.status = 0;
 	request.response_len = response_buf_size;
 	request.response_buf = response_buf;
 
-	spin_lock_irqsave(&ct->lock, flags);
-	list_add_tail(&request.link, &ct->pending_requests);
-	spin_unlock_irqrestore(&ct->lock, flags);
+	spin_lock_irqsave(&ct->requests.lock, flags);
+	list_add_tail(&request.link, &ct->requests.pending);
+	spin_unlock_irqrestore(&ct->requests.lock, flags);
 
 	err = ctb_write(ctb, action, len, fence, !!response_buf);
 	if (unlikely(err))
@@ -501,9 +500,9 @@ static int ctch_send(struct intel_guc_ct *ct,
 	}
 
 unlink:
-	spin_lock_irqsave(&ct->lock, flags);
+	spin_lock_irqsave(&ct->requests.lock, flags);
 	list_del(&request.link);
-	spin_unlock_irqrestore(&ct->lock, flags);
+	spin_unlock_irqrestore(&ct->requests.lock, flags);
 
 	return err;
 }
@@ -511,18 +510,21 @@ unlink:
 /*
  * Command Transport (CT) buffer based GuC send function.
  */
-int intel_guc_send_ct(struct intel_guc *guc, const u32 *action, u32 len,
+int intel_guc_ct_send(struct intel_guc_ct *ct, const u32 *action, u32 len,
 		      u32 *response_buf, u32 response_buf_size)
 {
-	struct intel_guc_ct *ct = &guc->ct;
-	struct intel_guc_ct_channel *ctch = &ct->host_channel;
+	struct intel_guc *guc = ct_to_guc(ct);
 	u32 status = ~0; /* undefined */
 	int ret;
 
+	if (unlikely(!ct->enabled)) {
+		WARN(1, "Unexpected send: action=%#x\n", *action);
+		return -ENODEV;
+	}
+
 	mutex_lock(&guc->send_mutex);
 
-	ret = ctch_send(ct, ctch, action, len, response_buf, response_buf_size,
-			&status);
+	ret = ct_send(ct, action, len, response_buf, response_buf_size, &status);
 	if (unlikely(ret < 0)) {
 		DRM_ERROR("CT: send action %#X failed; err=%d status=%#X\n",
 			  action[0], ret, status);
@@ -653,8 +655,8 @@ static int ct_handle_response(struct intel_guc_ct *ct, const u32 *msg)
 
 	CT_DEBUG_DRIVER("CT: response fence %u status %#x\n", fence, status);
 
-	spin_lock(&ct->lock);
-	list_for_each_entry(req, &ct->pending_requests, link) {
+	spin_lock(&ct->requests.lock);
+	list_for_each_entry(req, &ct->requests.pending, link) {
 		if (unlikely(fence != req->fence)) {
 			CT_DEBUG_DRIVER("CT: request %u awaits response\n",
 					req->fence);
@@ -672,7 +674,7 @@ static int ct_handle_response(struct intel_guc_ct *ct, const u32 *msg)
 		found = true;
 		break;
 	}
-	spin_unlock(&ct->lock);
+	spin_unlock(&ct->requests.lock);
 
 	if (!found)
 		DRM_ERROR("CT: unsolicited response %*ph\n", 4 * msglen, msg);
@@ -710,13 +712,13 @@ static bool ct_process_incoming_requests(struct intel_guc_ct *ct)
 	u32 *payload;
 	bool done;
 
-	spin_lock_irqsave(&ct->lock, flags);
-	request = list_first_entry_or_null(&ct->incoming_requests,
+	spin_lock_irqsave(&ct->requests.lock, flags);
+	request = list_first_entry_or_null(&ct->requests.incoming,
 					   struct ct_incoming_request, link);
 	if (request)
 		list_del(&request->link);
-	done = !!list_empty(&ct->incoming_requests);
-	spin_unlock_irqrestore(&ct->lock, flags);
+	done = !!list_empty(&ct->requests.incoming);
+	spin_unlock_irqrestore(&ct->requests.lock, flags);
 
 	if (!request)
 		return true;
@@ -734,12 +736,13 @@ static bool ct_process_incoming_requests(struct intel_guc_ct *ct)
 
 static void ct_incoming_request_worker_func(struct work_struct *w)
 {
-	struct intel_guc_ct *ct = container_of(w, struct intel_guc_ct, worker);
+	struct intel_guc_ct *ct =
+		container_of(w, struct intel_guc_ct, requests.worker);
 	bool done;
 
 	done = ct_process_incoming_requests(ct);
 	if (!done)
-		queue_work(system_unbound_wq, &ct->worker);
+		queue_work(system_unbound_wq, &ct->requests.worker);
 }
 
 /**
@@ -777,23 +780,28 @@ static int ct_handle_request(struct intel_guc_ct *ct, const u32 *msg)
 	}
 	memcpy(request->msg, msg, 4 * msglen);
 
-	spin_lock_irqsave(&ct->lock, flags);
-	list_add_tail(&request->link, &ct->incoming_requests);
-	spin_unlock_irqrestore(&ct->lock, flags);
+	spin_lock_irqsave(&ct->requests.lock, flags);
+	list_add_tail(&request->link, &ct->requests.incoming);
+	spin_unlock_irqrestore(&ct->requests.lock, flags);
 
-	queue_work(system_unbound_wq, &ct->worker);
+	queue_work(system_unbound_wq, &ct->requests.worker);
 	return 0;
 }
 
-static void ct_process_host_channel(struct intel_guc_ct *ct)
+/*
+ * When we're communicating with the GuC over CT, GuC uses events
+ * to notify us about new messages being posted on the RECV buffer.
+ */
+void intel_guc_ct_event_handler(struct intel_guc_ct *ct)
 {
-	struct intel_guc_ct_channel *ctch = &ct->host_channel;
-	struct intel_guc_ct_buffer *ctb = &ctch->ctbs[CTB_RECV];
+	struct intel_guc_ct_buffer *ctb = &ct->ctbs[CTB_RECV];
 	u32 msg[GUC_CT_MSG_LEN_MASK + 1]; /* one extra dw for the header */
 	int err = 0;
 
-	if (!ctch->enabled)
+	if (unlikely(!ct->enabled)) {
+		WARN(1, "Unexpected GuC event received while CT disabled!\n");
 		return;
+	}
 
 	do {
 		err = ctb_read(ctb, msg);
@@ -812,86 +820,3 @@ static void ct_process_host_channel(struct intel_guc_ct *ct)
 	}
 }
 
-/*
- * When we're communicating with the GuC over CT, GuC uses events
- * to notify us about new messages being posted on the RECV buffer.
- */
-void intel_guc_to_host_event_handler_ct(struct intel_guc *guc)
-{
-	struct intel_guc_ct *ct = &guc->ct;
-
-	ct_process_host_channel(ct);
-}
-
-/**
- * intel_guc_ct_init - Init CT communication
- * @ct: pointer to CT struct
- *
- * Allocate memory required for communication via
- * the CT channel.
- *
- * Return: 0 on success, a negative errno code on failure.
- */
-int intel_guc_ct_init(struct intel_guc_ct *ct)
-{
-	struct intel_guc *guc = ct_to_guc(ct);
-	struct intel_guc_ct_channel *ctch = &ct->host_channel;
-	int err;
-
-	err = ctch_init(guc, ctch);
-	if (unlikely(err)) {
-		DRM_ERROR("CT: can't open channel %d; err=%d\n",
-			  ctch->owner, err);
-		return err;
-	}
-
-	GEM_BUG_ON(!ctch->vma);
-	return 0;
-}
-
-/**
- * intel_guc_ct_fini - Fini CT communication
- * @ct: pointer to CT struct
- *
- * Deallocate memory required for communication via
- * the CT channel.
- */
-void intel_guc_ct_fini(struct intel_guc_ct *ct)
-{
-	struct intel_guc *guc = ct_to_guc(ct);
-	struct intel_guc_ct_channel *ctch = &ct->host_channel;
-
-	ctch_fini(guc, ctch);
-}
-
-/**
- * intel_guc_ct_enable - Enable buffer based command transport.
- * @ct: pointer to CT struct
- *
- * Return: 0 on success, a negative errno code on failure.
- */
-int intel_guc_ct_enable(struct intel_guc_ct *ct)
-{
-	struct intel_guc *guc = ct_to_guc(ct);
-	struct intel_guc_ct_channel *ctch = &ct->host_channel;
-
-	if (ctch->enabled)
-		return 0;
-
-	return ctch_enable(guc, ctch);
-}
-
-/**
- * intel_guc_ct_disable - Disable buffer based command transport.
- * @ct: pointer to CT struct
- */
-void intel_guc_ct_disable(struct intel_guc_ct *ct)
-{
-	struct intel_guc *guc = ct_to_guc(ct);
-	struct intel_guc_ct_channel *ctch = &ct->host_channel;
-
-	if (!ctch->enabled)
-		return;
-
-	ctch_disable(guc, ctch);
-}
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h
index 7c24d83f5c24..3e7fe237cfa5 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h
@@ -35,44 +35,28 @@ struct intel_guc_ct_buffer {
 	u32 *cmds;
 };
 
-/** Represents pair of command transport buffers.
- *
- * Buffers go in pairs to allow bi-directional communication.
- * To simplify the code we place both of them in the same vma.
- * Buffers from the same pair must share unique owner id.
- *
- * @vma: pointer to the vma with pair of CT buffers
- * @ctbs: buffers for sending(0) and receiving(1) commands
- * @owner: unique identifier
- * @next_fence: fence to be used with next send command
- */
-struct intel_guc_ct_channel {
-	struct i915_vma *vma;
-	struct intel_guc_ct_buffer ctbs[2];
-	u32 owner;
-	u32 next_fence;
-	bool enabled;
-};
 
-/** Holds all command transport channels.
+/** Top-level structure for Command Transport related data
  *
- * @host_channel: main channel used by the host
+ * Includes a pair of CT buffers for bi-directional communication and tracking
+ * for the H2G and G2H requests sent and received through the buffers.
  */
 struct intel_guc_ct {
-	struct intel_guc_ct_channel host_channel;
-	/* other channels are tbd */
+	struct i915_vma *vma;
+	bool enabled;
 
-	/** @lock: protects pending requests list */
-	spinlock_t lock;
+	/* buffers for sending(0) and receiving(1) commands */
+	struct intel_guc_ct_buffer ctbs[2];
 
-	/** @pending_requests: list of requests waiting for response */
-	struct list_head pending_requests;
+	struct {
+		u32 next_fence; /* fence to be used with next request to send */
 
-	/** @incoming_requests: list of incoming requests */
-	struct list_head incoming_requests;
+		spinlock_t lock; /* protects pending requests list */
+		struct list_head pending; /* requests waiting for response */
 
-	/** @worker: worker for handling incoming requests */
-	struct work_struct worker;
+		struct list_head incoming; /* incoming requests */
+		struct work_struct worker; /* handler for incoming requests */
+	} requests;
 };
 
 void intel_guc_ct_init_early(struct intel_guc_ct *ct);
@@ -81,13 +65,13 @@ void intel_guc_ct_fini(struct intel_guc_ct *ct);
 int intel_guc_ct_enable(struct intel_guc_ct *ct);
 void intel_guc_ct_disable(struct intel_guc_ct *ct);
 
-static inline void intel_guc_ct_stop(struct intel_guc_ct *ct)
+static inline bool intel_guc_ct_enabled(struct intel_guc_ct *ct)
 {
-	ct->host_channel.enabled = false;
+	return ct->enabled;
 }
 
-int intel_guc_send_ct(struct intel_guc *guc, const u32 *action, u32 len,
+int intel_guc_ct_send(struct intel_guc_ct *ct, const u32 *action, u32 len,
 		      u32 *response_buf, u32 response_buf_size);
-void intel_guc_to_host_event_handler_ct(struct intel_guc *guc);
+void intel_guc_ct_event_handler(struct intel_guc_ct *ct);
 
 #endif /* _INTEL_GUC_CT_H_ */
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c
index 5528224448f6..3a1c47d600ea 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_fw.c
@@ -149,7 +149,7 @@ int intel_guc_fw_upload(struct intel_guc *guc)
 	 * Current uCode expects the code to be loaded at 8k; locations below
 	 * this are used for the stack.
 	 */
-	ret = intel_uc_fw_upload(&guc->fw, gt, 0x2000, UOS_MOVE);
+	ret = intel_uc_fw_upload(&guc->fw, 0x2000, UOS_MOVE);
 	if (ret)
 		goto out;
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h
index a26a85d50209..a6b733c146c9 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_fwif.h
@@ -31,7 +31,6 @@
 
 #define GUC_DOORBELL_INVALID		256
 
-#define GUC_DB_SIZE			(PAGE_SIZE)
 #define GUC_WQ_SIZE			(PAGE_SIZE * 2)
 
 /* Work queue item header definitions */
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
index 2498c55e0ea5..9e42324fdecd 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c
@@ -18,15 +18,6 @@
 #include "i915_drv.h"
 #include "i915_trace.h"
 
-enum {
-	GUC_PREEMPT_NONE = 0,
-	GUC_PREEMPT_INPROGRESS,
-	GUC_PREEMPT_FINISHED,
-};
-#define GUC_PREEMPT_BREADCRUMB_DWORDS	0x8
-#define GUC_PREEMPT_BREADCRUMB_BYTES	\
-	(sizeof(u32) * GUC_PREEMPT_BREADCRUMB_DWORDS)
-
 /**
  * DOC: GuC-based command submission
  *
@@ -36,25 +27,14 @@ enum {
  * code) matches the old submission model and will be updated as part of the
  * upgrade to the new flow.
  *
- * GuC client:
- * A intel_guc_client refers to a submission path through GuC. Currently, there
- * is only one client, which is charged with all submissions to the GuC. This
- * struct is the owner of a doorbell, a process descriptor and a workqueue (all
- * of them inside a single gem object that contains all required pages for these
- * elements).
- *
  * GuC stage descriptor:
  * During initialization, the driver allocates a static pool of 1024 such
- * descriptors, and shares them with the GuC.
- * Currently, there exists a 1:1 mapping between a intel_guc_client and a
- * guc_stage_desc (via the client's stage_id), so effectively only one
- * gets used. This stage descriptor lets the GuC know about the doorbell,
- * workqueue and process descriptor. Theoretically, it also lets the GuC
- * know about our HW contexts (context ID, etc...), but we actually
- * employ a kind of submission where the GuC uses the LRCA sent via the work
- * item instead (the single guc_stage_desc associated to execbuf client
- * contains information about the default kernel context only, but this is
- * essentially unused). This is called a "proxy" submission.
+ * descriptors, and shares them with the GuC. Currently, we only use one
+ * descriptor. This stage descriptor lets the GuC know about the workqueue and
+ * process descriptor. Theoretically, it also lets the GuC know about our HW
+ * contexts (context ID, etc...), but we actually employ a kind of submission
+ * where the GuC uses the LRCA sent via the work item instead. This is called
+ * a "proxy" submission.
  *
  * The Scratch registers:
  * There are 16 MMIO-based registers start from 0xC180. The kernel driver writes
@@ -63,11 +43,6 @@ enum {
  * Firmware writes a success/fail code back to the action register after
  * processes the request. The kernel driver polls waiting for this update and
  * then proceeds.
- * See intel_guc_send()
- *
- * Doorbells:
- * Doorbells are interrupts to uKernel. A doorbell is a single cache line (QW)
- * mapped into process space.
  *
  * Work Items:
  * There are several types of work items that the host may place into a
@@ -84,213 +59,45 @@ static inline struct i915_priolist *to_priolist(struct rb_node *rb)
 	return rb_entry(rb, struct i915_priolist, node);
 }
 
-static inline bool is_high_priority(struct intel_guc_client *client)
+static struct guc_stage_desc *__get_stage_desc(struct intel_guc *guc, u32 id)
 {
-	return (client->priority == GUC_CLIENT_PRIORITY_KMD_HIGH ||
-		client->priority == GUC_CLIENT_PRIORITY_HIGH);
-}
-
-static int reserve_doorbell(struct intel_guc_client *client)
-{
-	unsigned long offset;
-	unsigned long end;
-	u16 id;
-
-	GEM_BUG_ON(client->doorbell_id != GUC_DOORBELL_INVALID);
+	struct guc_stage_desc *base = guc->stage_desc_pool_vaddr;
 
-	/*
-	 * The bitmap tracks which doorbell registers are currently in use.
-	 * It is split into two halves; the first half is used for normal
-	 * priority contexts, the second half for high-priority ones.
-	 */
-	offset = 0;
-	end = GUC_NUM_DOORBELLS / 2;
-	if (is_high_priority(client)) {
-		offset = end;
-		end += offset;
-	}
-
-	id = find_next_zero_bit(client->guc->doorbell_bitmap, end, offset);
-	if (id == end)
-		return -ENOSPC;
-
-	__set_bit(id, client->guc->doorbell_bitmap);
-	client->doorbell_id = id;
-	DRM_DEBUG_DRIVER("client %u (high prio=%s) reserved doorbell: %d\n",
-			 client->stage_id, yesno(is_high_priority(client)),
-			 id);
-	return 0;
+	return &base[id];
 }
 
-static bool has_doorbell(struct intel_guc_client *client)
+static int guc_workqueue_create(struct intel_guc *guc)
 {
-	if (client->doorbell_id == GUC_DOORBELL_INVALID)
-		return false;
-
-	return test_bit(client->doorbell_id, client->guc->doorbell_bitmap);
+	return intel_guc_allocate_and_map_vma(guc, GUC_WQ_SIZE, &guc->workqueue,
+					      &guc->workqueue_vaddr);
 }
 
-static void unreserve_doorbell(struct intel_guc_client *client)
+static void guc_workqueue_destroy(struct intel_guc *guc)
 {
-	GEM_BUG_ON(!has_doorbell(client));
-
-	__clear_bit(client->doorbell_id, client->guc->doorbell_bitmap);
-	client->doorbell_id = GUC_DOORBELL_INVALID;
+	i915_vma_unpin_and_release(&guc->workqueue, I915_VMA_RELEASE_MAP);
 }
 
 /*
- * Tell the GuC to allocate or deallocate a specific doorbell
- */
-
-static int __guc_allocate_doorbell(struct intel_guc *guc, u32 stage_id)
-{
-	u32 action[] = {
-		INTEL_GUC_ACTION_ALLOCATE_DOORBELL,
-		stage_id
-	};
-
-	return intel_guc_send(guc, action, ARRAY_SIZE(action));
-}
-
-static int __guc_deallocate_doorbell(struct intel_guc *guc, u32 stage_id)
-{
-	u32 action[] = {
-		INTEL_GUC_ACTION_DEALLOCATE_DOORBELL,
-		stage_id
-	};
-
-	return intel_guc_send(guc, action, ARRAY_SIZE(action));
-}
-
-static struct guc_stage_desc *__get_stage_desc(struct intel_guc_client *client)
-{
-	struct guc_stage_desc *base = client->guc->stage_desc_pool_vaddr;
-
-	return &base[client->stage_id];
-}
-
-/*
- * Initialise, update, or clear doorbell data shared with the GuC
- *
- * These functions modify shared data and so need access to the mapped
- * client object which contains the page being used for the doorbell
+ * Initialise the process descriptor shared with the GuC firmware.
  */
-
-static void __update_doorbell_desc(struct intel_guc_client *client, u16 new_id)
-{
-	struct guc_stage_desc *desc;
-
-	/* Update the GuC's idea of the doorbell ID */
-	desc = __get_stage_desc(client);
-	desc->db_id = new_id;
-}
-
-static struct guc_doorbell_info *__get_doorbell(struct intel_guc_client *client)
-{
-	return client->vaddr + client->doorbell_offset;
-}
-
-static bool __doorbell_valid(struct intel_guc *guc, u16 db_id)
-{
-	struct intel_uncore *uncore = guc_to_gt(guc)->uncore;
-
-	GEM_BUG_ON(db_id >= GUC_NUM_DOORBELLS);
-	return intel_uncore_read(uncore, GEN8_DRBREGL(db_id)) & GEN8_DRB_VALID;
-}
-
-static void __init_doorbell(struct intel_guc_client *client)
-{
-	struct guc_doorbell_info *doorbell;
-
-	doorbell = __get_doorbell(client);
-	doorbell->db_status = GUC_DOORBELL_ENABLED;
-	doorbell->cookie = 0;
-}
-
-static void __fini_doorbell(struct intel_guc_client *client)
-{
-	struct guc_doorbell_info *doorbell;
-	u16 db_id = client->doorbell_id;
-
-	doorbell = __get_doorbell(client);
-	doorbell->db_status = GUC_DOORBELL_DISABLED;
-
-	/* Doorbell release flow requires that we wait for GEN8_DRB_VALID bit
-	 * to go to zero after updating db_status before we call the GuC to
-	 * release the doorbell
-	 */
-	if (wait_for_us(!__doorbell_valid(client->guc, db_id), 10))
-		WARN_ONCE(true, "Doorbell never became invalid after disable\n");
-}
-
-static int create_doorbell(struct intel_guc_client *client)
+static int guc_proc_desc_create(struct intel_guc *guc)
 {
-	int ret;
-
-	if (WARN_ON(!has_doorbell(client)))
-		return -ENODEV; /* internal setup error, should never happen */
-
-	__update_doorbell_desc(client, client->doorbell_id);
-	__init_doorbell(client);
-
-	ret = __guc_allocate_doorbell(client->guc, client->stage_id);
-	if (ret) {
-		__fini_doorbell(client);
-		__update_doorbell_desc(client, GUC_DOORBELL_INVALID);
-		DRM_DEBUG_DRIVER("Couldn't create client %u doorbell: %d\n",
-				 client->stage_id, ret);
-		return ret;
-	}
+	const u32 size = PAGE_ALIGN(sizeof(struct guc_process_desc));
 
-	return 0;
+	return intel_guc_allocate_and_map_vma(guc, size, &guc->proc_desc,
+					      &guc->proc_desc_vaddr);
 }
 
-static int destroy_doorbell(struct intel_guc_client *client)
+static void guc_proc_desc_destroy(struct intel_guc *guc)
 {
-	int ret;
-
-	GEM_BUG_ON(!has_doorbell(client));
-
-	__fini_doorbell(client);
-	ret = __guc_deallocate_doorbell(client->guc, client->stage_id);
-	if (ret)
-		DRM_ERROR("Couldn't destroy client %u doorbell: %d\n",
-			  client->stage_id, ret);
-
-	__update_doorbell_desc(client, GUC_DOORBELL_INVALID);
-
-	return ret;
+	i915_vma_unpin_and_release(&guc->proc_desc, I915_VMA_RELEASE_MAP);
 }
 
-static unsigned long __select_cacheline(struct intel_guc *guc)
-{
-	unsigned long offset;
-
-	/* Doorbell uses a single cache line within a page */
-	offset = offset_in_page(guc->db_cacheline);
-
-	/* Moving to next cache line to reduce contention */
-	guc->db_cacheline += cache_line_size();
-
-	DRM_DEBUG_DRIVER("reserved cacheline 0x%lx, next 0x%x, linesize %u\n",
-			 offset, guc->db_cacheline, cache_line_size());
-	return offset;
-}
-
-static inline struct guc_process_desc *
-__get_process_desc(struct intel_guc_client *client)
-{
-	return client->vaddr + client->proc_desc_offset;
-}
-
-/*
- * Initialise the process descriptor shared with the GuC firmware.
- */
-static void guc_proc_desc_init(struct intel_guc_client *client)
+static void guc_proc_desc_init(struct intel_guc *guc)
 {
 	struct guc_process_desc *desc;
 
-	desc = memset(__get_process_desc(client), 0, sizeof(*desc));
+	desc = memset(guc->proc_desc_vaddr, 0, sizeof(*desc));
 
 	/*
 	 * XXX: pDoorbell and WQVBaseAddress are pointers in process address
@@ -301,47 +108,27 @@ static void guc_proc_desc_init(struct intel_guc_client *client)
 	desc->wq_base_addr = 0;
 	desc->db_base_addr = 0;
 
-	desc->stage_id = client->stage_id;
 	desc->wq_size_bytes = GUC_WQ_SIZE;
 	desc->wq_status = WQ_STATUS_ACTIVE;
-	desc->priority = client->priority;
+	desc->priority = GUC_CLIENT_PRIORITY_KMD_NORMAL;
 }
 
-static void guc_proc_desc_fini(struct intel_guc_client *client)
+static void guc_proc_desc_fini(struct intel_guc *guc)
 {
-	struct guc_process_desc *desc;
-
-	desc = __get_process_desc(client);
-	memset(desc, 0, sizeof(*desc));
+	memset(guc->proc_desc_vaddr, 0, sizeof(struct guc_process_desc));
 }
 
 static int guc_stage_desc_pool_create(struct intel_guc *guc)
 {
-	struct i915_vma *vma;
-	void *vaddr;
-
-	vma = intel_guc_allocate_vma(guc,
-				     PAGE_ALIGN(sizeof(struct guc_stage_desc) *
-				     GUC_MAX_STAGE_DESCRIPTORS));
-	if (IS_ERR(vma))
-		return PTR_ERR(vma);
-
-	vaddr = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
-	if (IS_ERR(vaddr)) {
-		i915_vma_unpin_and_release(&vma, 0);
-		return PTR_ERR(vaddr);
-	}
-
-	guc->stage_desc_pool = vma;
-	guc->stage_desc_pool_vaddr = vaddr;
-	ida_init(&guc->stage_ids);
+	u32 size = PAGE_ALIGN(sizeof(struct guc_stage_desc) *
+			      GUC_MAX_STAGE_DESCRIPTORS);
 
-	return 0;
+	return intel_guc_allocate_and_map_vma(guc, size, &guc->stage_desc_pool,
+					      &guc->stage_desc_pool_vaddr);
 }
 
 static void guc_stage_desc_pool_destroy(struct intel_guc *guc)
 {
-	ida_destroy(&guc->stage_ids);
 	i915_vma_unpin_and_release(&guc->stage_desc_pool, I915_VMA_RELEASE_MAP);
 }
 
@@ -349,63 +136,49 @@ static void guc_stage_desc_pool_destroy(struct intel_guc *guc)
  * Initialise/clear the stage descriptor shared with the GuC firmware.
  *
  * This descriptor tells the GuC where (in GGTT space) to find the important
- * data structures relating to this client (doorbell, process descriptor,
- * write queue, etc).
+ * data structures related to work submission (process descriptor, write queue,
+ * etc).
  */
-static void guc_stage_desc_init(struct intel_guc_client *client)
+static void guc_stage_desc_init(struct intel_guc *guc)
 {
-	struct intel_guc *guc = client->guc;
 	struct guc_stage_desc *desc;
-	u32 gfx_addr;
 
-	desc = __get_stage_desc(client);
+	/* we only use 1 stage desc, so hardcode it to 0 */
+	desc = __get_stage_desc(guc, 0);
 	memset(desc, 0, sizeof(*desc));
 
 	desc->attribute = GUC_STAGE_DESC_ATTR_ACTIVE |
 			  GUC_STAGE_DESC_ATTR_KERNEL;
-	if (is_high_priority(client))
-		desc->attribute |= GUC_STAGE_DESC_ATTR_PREEMPT;
-	desc->stage_id = client->stage_id;
-	desc->priority = client->priority;
-	desc->db_id = client->doorbell_id;
 
-	/*
-	 * The doorbell, process descriptor, and workqueue are all parts
-	 * of the client object, which the GuC will reference via the GGTT
-	 */
-	gfx_addr = intel_guc_ggtt_offset(guc, client->vma);
-	desc->db_trigger_phy = sg_dma_address(client->vma->pages->sgl) +
-				client->doorbell_offset;
-	desc->db_trigger_cpu = ptr_to_u64(__get_doorbell(client));
-	desc->db_trigger_uk = gfx_addr + client->doorbell_offset;
-	desc->process_desc = gfx_addr + client->proc_desc_offset;
-	desc->wq_addr = gfx_addr + GUC_DB_SIZE;
-	desc->wq_size = GUC_WQ_SIZE;
+	desc->stage_id = 0;
+	desc->priority = GUC_CLIENT_PRIORITY_KMD_NORMAL;
 
-	desc->desc_private = ptr_to_u64(client);
+	desc->process_desc = intel_guc_ggtt_offset(guc, guc->proc_desc);
+	desc->wq_addr = intel_guc_ggtt_offset(guc, guc->workqueue);
+	desc->wq_size = GUC_WQ_SIZE;
 }
 
-static void guc_stage_desc_fini(struct intel_guc_client *client)
+static void guc_stage_desc_fini(struct intel_guc *guc)
 {
 	struct guc_stage_desc *desc;
 
-	desc = __get_stage_desc(client);
+	desc = __get_stage_desc(guc, 0);
 	memset(desc, 0, sizeof(*desc));
 }
 
 /* Construct a Work Item and append it to the GuC's Work Queue */
-static void guc_wq_item_append(struct intel_guc_client *client,
+static void guc_wq_item_append(struct intel_guc *guc,
 			       u32 target_engine, u32 context_desc,
 			       u32 ring_tail, u32 fence_id)
 {
 	/* wqi_len is in DWords, and does not include the one-word header */
 	const size_t wqi_size = sizeof(struct guc_wq_item);
 	const u32 wqi_len = wqi_size / sizeof(u32) - 1;
-	struct guc_process_desc *desc = __get_process_desc(client);
+	struct guc_process_desc *desc = guc->proc_desc_vaddr;
 	struct guc_wq_item *wqi;
 	u32 wq_off;
 
-	lockdep_assert_held(&client->wq_lock);
+	lockdep_assert_held(&guc->wq_lock);
 
 	/* For now workqueue item is 4 DWs; workqueue buffer is 2 pages. So we
 	 * should not have the case where structure wqi is across page, neither
@@ -425,58 +198,30 @@ static void guc_wq_item_append(struct intel_guc_client *client,
 			      GUC_WQ_SIZE) < wqi_size);
 	GEM_BUG_ON(wq_off & (wqi_size - 1));
 
-	/* WQ starts from the page after doorbell / process_desc */
-	wqi = client->vaddr + wq_off + GUC_DB_SIZE;
-
-	if (I915_SELFTEST_ONLY(client->use_nop_wqi)) {
-		wqi->header = WQ_TYPE_NOOP | (wqi_len << WQ_LEN_SHIFT);
-	} else {
-		/* Now fill in the 4-word work queue item */
-		wqi->header = WQ_TYPE_INORDER |
-			      (wqi_len << WQ_LEN_SHIFT) |
-			      (target_engine << WQ_TARGET_SHIFT) |
-			      WQ_NO_WCFLUSH_WAIT;
-		wqi->context_desc = context_desc;
-		wqi->submit_element_info = ring_tail << WQ_RING_TAIL_SHIFT;
-		GEM_BUG_ON(ring_tail > WQ_RING_TAIL_MAX);
-		wqi->fence_id = fence_id;
-	}
+	wqi = guc->workqueue_vaddr + wq_off;
+
+	/* Now fill in the 4-word work queue item */
+	wqi->header = WQ_TYPE_INORDER |
+		      (wqi_len << WQ_LEN_SHIFT) |
+		      (target_engine << WQ_TARGET_SHIFT) |
+		      WQ_NO_WCFLUSH_WAIT;
+	wqi->context_desc = context_desc;
+	wqi->submit_element_info = ring_tail << WQ_RING_TAIL_SHIFT;
+	GEM_BUG_ON(ring_tail > WQ_RING_TAIL_MAX);
+	wqi->fence_id = fence_id;
 
 	/* Make the update visible to GuC */
 	WRITE_ONCE(desc->tail, (wq_off + wqi_size) & (GUC_WQ_SIZE - 1));
 }
 
-static void guc_ring_doorbell(struct intel_guc_client *client)
-{
-	struct guc_doorbell_info *db;
-	u32 cookie;
-
-	lockdep_assert_held(&client->wq_lock);
-
-	/* pointer of current doorbell cacheline */
-	db = __get_doorbell(client);
-
-	/*
-	 * We're not expecting the doorbell cookie to change behind our back,
-	 * we also need to treat 0 as a reserved value.
-	 */
-	cookie = READ_ONCE(db->cookie);
-	WARN_ON_ONCE(xchg(&db->cookie, cookie + 1 ?: cookie + 2) != cookie);
-
-	/* XXX: doorbell was lost and need to acquire it again */
-	GEM_BUG_ON(db->db_status != GUC_DOORBELL_ENABLED);
-}
-
 static void guc_add_request(struct intel_guc *guc, struct i915_request *rq)
 {
-	struct intel_guc_client *client = guc->execbuf_client;
 	struct intel_engine_cs *engine = rq->engine;
-	u32 ctx_desc = lower_32_bits(rq->hw_context->lrc_desc);
+	u32 ctx_desc = lower_32_bits(rq->context->lrc_desc);
 	u32 ring_tail = intel_ring_set_tail(rq->ring, rq->tail) / sizeof(u64);
 
-	guc_wq_item_append(client, engine->guc_id, ctx_desc,
+	guc_wq_item_append(guc, engine->guc_id, ctx_desc,
 			   ring_tail, rq->fence.seqno);
-	guc_ring_doorbell(client);
 }
 
 /*
@@ -488,10 +233,9 @@ static void guc_add_request(struct intel_guc *guc, struct i915_request *rq)
  */
 static void flush_ggtt_writes(struct i915_vma *vma)
 {
-	struct drm_i915_private *i915 = vma->vm->i915;
-
 	if (i915_vma_is_map_and_fenceable(vma))
-		intel_uncore_posting_read_fw(&i915->uncore, GUC_STATUS);
+		intel_uncore_posting_read_fw(vma->vm->gt->uncore,
+					     GUC_STATUS);
 }
 
 static void guc_submit(struct intel_engine_cs *engine,
@@ -499,9 +243,8 @@ static void guc_submit(struct intel_engine_cs *engine,
 		       struct i915_request **end)
 {
 	struct intel_guc *guc = &engine->gt->uc.guc;
-	struct intel_guc_client *client = guc->execbuf_client;
 
-	spin_lock(&client->wq_lock);
+	spin_lock(&guc->wq_lock);
 
 	do {
 		struct i915_request *rq = *out++;
@@ -510,7 +253,7 @@ static void guc_submit(struct intel_engine_cs *engine,
 		guc_add_request(guc, rq);
 	} while (out != end);
 
-	spin_unlock(&client->wq_lock);
+	spin_unlock(&guc->wq_lock);
 }
 
 static inline int rq_prio(const struct i915_request *rq)
@@ -529,7 +272,7 @@ static struct i915_request *schedule_in(struct i915_request *rq, int idx)
 	 * required if we generalise the inflight tracking.
 	 */
 
-	intel_gt_pm_get(rq->engine->gt);
+	__intel_gt_pm_get(rq->engine->gt);
 	return i915_request_get(rq);
 }
 
@@ -537,7 +280,7 @@ static void schedule_out(struct i915_request *rq)
 {
 	trace_i915_request_out(rq);
 
-	intel_gt_pm_put(rq->engine->gt);
+	intel_gt_pm_put_async(rq->engine->gt);
 	i915_request_put(rq);
 }
 
@@ -572,7 +315,7 @@ static void __guc_dequeue(struct intel_engine_cs *engine)
 		int i;
 
 		priolist_for_each_request_consume(rq, rn, p, i) {
-			if (last && rq->hw_context != last->hw_context) {
+			if (last && rq->context != last->context) {
 				if (port == last_port)
 					goto done;
 
@@ -631,7 +374,7 @@ static void guc_reset_prepare(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	/*
 	 * Prevent request submission to the hardware until we have
@@ -658,7 +401,7 @@ cancel_port_requests(struct intel_engine_execlists * const execlists)
 		memset(execlists->inflight, 0, sizeof(execlists->inflight));
 }
 
-static void guc_reset(struct intel_engine_cs *engine, bool stalled)
+static void guc_reset_rewind(struct intel_engine_cs *engine, bool stalled)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_request *rq;
@@ -677,20 +420,20 @@ static void guc_reset(struct intel_engine_cs *engine, bool stalled)
 		stalled = false;
 
 	__i915_request_reset(rq, stalled);
-	intel_lr_context_reset(engine, rq->hw_context, rq->head, stalled);
+	intel_lr_context_reset(engine, rq->context, rq->head, stalled);
 
 out_unlock:
 	spin_unlock_irqrestore(&engine->active.lock, flags);
 }
 
-static void guc_cancel_requests(struct intel_engine_cs *engine)
+static void guc_reset_cancel(struct intel_engine_cs *engine)
 {
 	struct intel_engine_execlists * const execlists = &engine->execlists;
 	struct i915_request *rq, *rn;
 	struct rb_node *rb;
 	unsigned long flags;
 
-	GEM_TRACE("%s\n", engine->name);
+	ENGINE_TRACE(engine, "\n");
 
 	/*
 	 * Before we call engine->cancel_requests(), we should have exclusive
@@ -751,8 +494,8 @@ static void guc_reset_finish(struct intel_engine_cs *engine)
 		/* And kick in case we missed a new request submission. */
 		tasklet_hi_schedule(&execlists->tasklet);
 
-	GEM_TRACE("%s: depth->%d\n", engine->name,
-		  atomic_read(&execlists->tasklet.count));
+	ENGINE_TRACE(engine, "depth->%d\n",
+		     atomic_read(&execlists->tasklet.count));
 }
 
 /*
@@ -761,213 +504,6 @@ static void guc_reset_finish(struct intel_engine_cs *engine)
  * path of guc_submit() above.
  */
 
-/* Check that a doorbell register is in the expected state */
-static bool doorbell_ok(struct intel_guc *guc, u16 db_id)
-{
-	bool valid;
-
-	GEM_BUG_ON(db_id >= GUC_NUM_DOORBELLS);
-
-	valid = __doorbell_valid(guc, db_id);
-
-	if (test_bit(db_id, guc->doorbell_bitmap) == valid)
-		return true;
-
-	DRM_DEBUG_DRIVER("Doorbell %u has unexpected state: valid=%s\n",
-			 db_id, yesno(valid));
-
-	return false;
-}
-
-static bool guc_verify_doorbells(struct intel_guc *guc)
-{
-	bool doorbells_ok = true;
-	u16 db_id;
-
-	for (db_id = 0; db_id < GUC_NUM_DOORBELLS; ++db_id)
-		if (!doorbell_ok(guc, db_id))
-			doorbells_ok = false;
-
-	return doorbells_ok;
-}
-
-/**
- * guc_client_alloc() - Allocate an intel_guc_client
- * @guc:	the intel_guc structure
- * @priority:	four levels priority _CRITICAL, _HIGH, _NORMAL and _LOW
- *		The kernel client to replace ExecList submission is created with
- *		NORMAL priority. Priority of a client for scheduler can be HIGH,
- *		while a preemption context can use CRITICAL.
- *
- * Return:	An intel_guc_client object if success, else NULL.
- */
-static struct intel_guc_client *
-guc_client_alloc(struct intel_guc *guc, u32 priority)
-{
-	struct intel_guc_client *client;
-	struct i915_vma *vma;
-	void *vaddr;
-	int ret;
-
-	client = kzalloc(sizeof(*client), GFP_KERNEL);
-	if (!client)
-		return ERR_PTR(-ENOMEM);
-
-	client->guc = guc;
-	client->priority = priority;
-	client->doorbell_id = GUC_DOORBELL_INVALID;
-	spin_lock_init(&client->wq_lock);
-
-	ret = ida_simple_get(&guc->stage_ids, 0, GUC_MAX_STAGE_DESCRIPTORS,
-			     GFP_KERNEL);
-	if (ret < 0)
-		goto err_client;
-
-	client->stage_id = ret;
-
-	/* The first page is doorbell/proc_desc. Two followed pages are wq. */
-	vma = intel_guc_allocate_vma(guc, GUC_DB_SIZE + GUC_WQ_SIZE);
-	if (IS_ERR(vma)) {
-		ret = PTR_ERR(vma);
-		goto err_id;
-	}
-
-	/* We'll keep just the first (doorbell/proc) page permanently kmap'd. */
-	client->vma = vma;
-
-	vaddr = i915_gem_object_pin_map(vma->obj, I915_MAP_WB);
-	if (IS_ERR(vaddr)) {
-		ret = PTR_ERR(vaddr);
-		goto err_vma;
-	}
-	client->vaddr = vaddr;
-
-	ret = reserve_doorbell(client);
-	if (ret)
-		goto err_vaddr;
-
-	client->doorbell_offset = __select_cacheline(guc);
-
-	/*
-	 * Since the doorbell only requires a single cacheline, we can save
-	 * space by putting the application process descriptor in the same
-	 * page. Use the half of the page that doesn't include the doorbell.
-	 */
-	if (client->doorbell_offset >= (GUC_DB_SIZE / 2))
-		client->proc_desc_offset = 0;
-	else
-		client->proc_desc_offset = (GUC_DB_SIZE / 2);
-
-	DRM_DEBUG_DRIVER("new priority %u client %p: stage_id %u\n",
-			 priority, client, client->stage_id);
-	DRM_DEBUG_DRIVER("doorbell id %u, cacheline offset 0x%lx\n",
-			 client->doorbell_id, client->doorbell_offset);
-
-	return client;
-
-err_vaddr:
-	i915_gem_object_unpin_map(client->vma->obj);
-err_vma:
-	i915_vma_unpin_and_release(&client->vma, 0);
-err_id:
-	ida_simple_remove(&guc->stage_ids, client->stage_id);
-err_client:
-	kfree(client);
-	return ERR_PTR(ret);
-}
-
-static void guc_client_free(struct intel_guc_client *client)
-{
-	unreserve_doorbell(client);
-	i915_vma_unpin_and_release(&client->vma, I915_VMA_RELEASE_MAP);
-	ida_simple_remove(&client->guc->stage_ids, client->stage_id);
-	kfree(client);
-}
-
-static inline bool ctx_save_restore_disabled(struct intel_context *ce)
-{
-	u32 sr = ce->lrc_reg_state[CTX_CONTEXT_CONTROL + 1];
-
-#define SR_DISABLED \
-	_MASKED_BIT_ENABLE(CTX_CTRL_ENGINE_CTX_RESTORE_INHIBIT | \
-			   CTX_CTRL_ENGINE_CTX_SAVE_INHIBIT)
-
-	return (sr & SR_DISABLED) == SR_DISABLED;
-
-#undef SR_DISABLED
-}
-
-static int guc_clients_create(struct intel_guc *guc)
-{
-	struct intel_guc_client *client;
-
-	GEM_BUG_ON(guc->execbuf_client);
-
-	client = guc_client_alloc(guc, GUC_CLIENT_PRIORITY_KMD_NORMAL);
-	if (IS_ERR(client)) {
-		DRM_ERROR("Failed to create GuC client for submission!\n");
-		return PTR_ERR(client);
-	}
-	guc->execbuf_client = client;
-
-	return 0;
-}
-
-static void guc_clients_destroy(struct intel_guc *guc)
-{
-	struct intel_guc_client *client;
-
-	client = fetch_and_zero(&guc->execbuf_client);
-	if (client)
-		guc_client_free(client);
-}
-
-static int __guc_client_enable(struct intel_guc_client *client)
-{
-	int ret;
-
-	guc_proc_desc_init(client);
-	guc_stage_desc_init(client);
-
-	ret = create_doorbell(client);
-	if (ret)
-		goto fail;
-
-	return 0;
-
-fail:
-	guc_stage_desc_fini(client);
-	guc_proc_desc_fini(client);
-	return ret;
-}
-
-static void __guc_client_disable(struct intel_guc_client *client)
-{
-	/*
-	 * By the time we're here, GuC may have already been reset. if that is
-	 * the case, instead of trying (in vain) to communicate with it, let's
-	 * just cleanup the doorbell HW and our internal state.
-	 */
-	if (intel_guc_is_running(client->guc))
-		destroy_doorbell(client);
-	else
-		__fini_doorbell(client);
-
-	guc_stage_desc_fini(client);
-	guc_proc_desc_fini(client);
-}
-
-static int guc_clients_enable(struct intel_guc *guc)
-{
-	return __guc_client_enable(guc->execbuf_client);
-}
-
-static void guc_clients_disable(struct intel_guc *guc)
-{
-	if (guc->execbuf_client)
-		__guc_client_disable(guc->execbuf_client);
-}
-
 /*
  * Set up the memory resources to be shared with the GuC (via the GGTT)
  * at firmware loading time.
@@ -988,13 +524,20 @@ int intel_guc_submission_init(struct intel_guc *guc)
 	 */
 	GEM_BUG_ON(!guc->stage_desc_pool);
 
-	WARN_ON(!guc_verify_doorbells(guc));
-	ret = guc_clients_create(guc);
+	ret = guc_workqueue_create(guc);
 	if (ret)
 		goto err_pool;
 
+	ret = guc_proc_desc_create(guc);
+	if (ret)
+		goto err_workqueue;
+
+	spin_lock_init(&guc->wq_lock);
+
 	return 0;
 
+err_workqueue:
+	guc_workqueue_destroy(guc);
 err_pool:
 	guc_stage_desc_pool_destroy(guc);
 	return ret;
@@ -1002,83 +545,37 @@ err_pool:
 
 void intel_guc_submission_fini(struct intel_guc *guc)
 {
-	guc_clients_destroy(guc);
-	WARN_ON(!guc_verify_doorbells(guc));
-
-	if (guc->stage_desc_pool)
+	if (guc->stage_desc_pool) {
+		guc_proc_desc_destroy(guc);
+		guc_workqueue_destroy(guc);
 		guc_stage_desc_pool_destroy(guc);
+	}
 }
 
 static void guc_interrupts_capture(struct intel_gt *gt)
 {
-	struct intel_rps *rps = &gt->rps;
 	struct intel_uncore *uncore = gt->uncore;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	int irqs;
+	u32 irqs = GT_CONTEXT_SWITCH_INTERRUPT;
+	u32 dmask = irqs << 16 | irqs;
 
-	/* tell all command streamers to forward interrupts (but not vblank)
-	 * to GuC
-	 */
-	irqs = _MASKED_BIT_ENABLE(GFX_INTERRUPT_STEERING);
-	for_each_engine(engine, gt, id)
-		ENGINE_WRITE(engine, RING_MODE_GEN7, irqs);
-
-	/* route USER_INTERRUPT to Host, all others are sent to GuC. */
-	irqs = GT_RENDER_USER_INTERRUPT << GEN8_RCS_IRQ_SHIFT |
-	       GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT;
-	/* These three registers have the same bit definitions */
-	intel_uncore_write(uncore, GUC_BCS_RCS_IER, ~irqs);
-	intel_uncore_write(uncore, GUC_VCS2_VCS1_IER, ~irqs);
-	intel_uncore_write(uncore, GUC_WD_VECS_IER, ~irqs);
+	GEM_BUG_ON(INTEL_GEN(gt->i915) < 11);
 
-	/*
-	 * The REDIRECT_TO_GUC bit of the PMINTRMSK register directs all
-	 * (unmasked) PM interrupts to the GuC. All other bits of this
-	 * register *disable* generation of a specific interrupt.
-	 *
-	 * 'pm_intrmsk_mbz' indicates bits that are NOT to be set when
-	 * writing to the PM interrupt mask register, i.e. interrupts
-	 * that must not be disabled.
-	 *
-	 * If the GuC is handling these interrupts, then we must not let
-	 * the PM code disable ANY interrupt that the GuC is expecting.
-	 * So for each ENABLED (0) bit in this register, we must SET the
-	 * bit in pm_intrmsk_mbz so that it's left enabled for the GuC.
-	 * GuC needs ARAT expired interrupt unmasked hence it is set in
-	 * pm_intrmsk_mbz.
-	 *
-	 * Here we CLEAR REDIRECT_TO_GUC bit in pm_intrmsk_mbz, which will
-	 * result in the register bit being left SET!
-	 */
-	rps->pm_intrmsk_mbz |= ARAT_EXPIRED_INTRMSK;
-	rps->pm_intrmsk_mbz &= ~GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC;
+	/* Don't handle the ctx switch interrupt in GuC submission mode */
+	intel_uncore_rmw(uncore, GEN11_RENDER_COPY_INTR_ENABLE, dmask, 0);
+	intel_uncore_rmw(uncore, GEN11_VCS_VECS_INTR_ENABLE, dmask, 0);
 }
 
 static void guc_interrupts_release(struct intel_gt *gt)
 {
-	struct intel_rps *rps = &gt->rps;
 	struct intel_uncore *uncore = gt->uncore;
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	int irqs;
+	u32 irqs = GT_CONTEXT_SWITCH_INTERRUPT;
+	u32 dmask = irqs << 16 | irqs;
 
-	/*
-	 * tell all command streamers NOT to forward interrupts or vblank
-	 * to GuC.
-	 */
-	irqs = _MASKED_FIELD(GFX_FORWARD_VBLANK_MASK, GFX_FORWARD_VBLANK_NEVER);
-	irqs |= _MASKED_BIT_DISABLE(GFX_INTERRUPT_STEERING);
-	for_each_engine(engine, gt, id)
-		ENGINE_WRITE(engine, RING_MODE_GEN7, irqs);
-
-	/* route all GT interrupts to the host */
-	intel_uncore_write(uncore, GUC_BCS_RCS_IER, 0);
-	intel_uncore_write(uncore, GUC_VCS2_VCS1_IER, 0);
-	intel_uncore_write(uncore, GUC_WD_VECS_IER, 0);
-
-	rps->pm_intrmsk_mbz |= GEN8_PMINTR_DISABLE_REDIRECT_TO_GUC;
-	rps->pm_intrmsk_mbz &= ~ARAT_EXPIRED_INTRMSK;
+	GEM_BUG_ON(INTEL_GEN(gt->i915) < 11);
+
+	/* Handle ctx switch interrupts again */
+	intel_uncore_rmw(uncore, GEN11_RENDER_COPY_INTR_ENABLE, 0, dmask);
+	intel_uncore_rmw(uncore, GEN11_VCS_VECS_INTR_ENABLE, 0, dmask);
 }
 
 static void guc_set_default_submission(struct intel_engine_cs *engine)
@@ -1102,11 +599,10 @@ static void guc_set_default_submission(struct intel_engine_cs *engine)
 	engine->park = engine->unpark = NULL;
 
 	engine->reset.prepare = guc_reset_prepare;
-	engine->reset.reset = guc_reset;
+	engine->reset.rewind = guc_reset_rewind;
+	engine->reset.cancel = guc_reset_cancel;
 	engine->reset.finish = guc_reset_finish;
 
-	engine->cancel_requests = guc_cancel_requests;
-
 	engine->flags &= ~I915_ENGINE_SUPPORTS_STATS;
 	engine->flags |= I915_ENGINE_NEEDS_BREADCRUMB_TASKLET;
 
@@ -1119,16 +615,11 @@ static void guc_set_default_submission(struct intel_engine_cs *engine)
 	GEM_BUG_ON(engine->irq_enable || engine->irq_disable);
 }
 
-int intel_guc_submission_enable(struct intel_guc *guc)
+void intel_guc_submission_enable(struct intel_guc *guc)
 {
 	struct intel_gt *gt = guc_to_gt(guc);
 	struct intel_engine_cs *engine;
 	enum intel_engine_id id;
-	int err;
-
-	err = i915_inject_probe_error(gt->i915, -ENXIO);
-	if (err)
-		return err;
 
 	/*
 	 * We're using GuC work items for submitting work through GuC. Since
@@ -1143,11 +634,8 @@ int intel_guc_submission_enable(struct intel_guc *guc)
 		     sizeof(struct guc_wq_item) *
 		     I915_NUM_ENGINES > GUC_WQ_SIZE);
 
-	GEM_BUG_ON(!guc->execbuf_client);
-
-	err = guc_clients_enable(guc);
-	if (err)
-		return err;
+	guc_proc_desc_init(guc);
+	guc_stage_desc_init(guc);
 
 	/* Take over from manual control of ELSP (execlists) */
 	guc_interrupts_capture(gt);
@@ -1156,8 +644,6 @@ int intel_guc_submission_enable(struct intel_guc *guc)
 		engine->set_default_submission = guc_set_default_submission;
 		engine->set_default_submission(engine);
 	}
-
-	return 0;
 }
 
 void intel_guc_submission_disable(struct intel_guc *guc)
@@ -1166,8 +652,12 @@ void intel_guc_submission_disable(struct intel_guc *guc)
 
 	GEM_BUG_ON(gt->awake); /* GT should be parked first */
 
+	/* Note: By the time we're here, GuC may have already been reset */
+
 	guc_interrupts_release(gt);
-	guc_clients_disable(guc);
+
+	guc_stage_desc_fini(guc);
+	guc_proc_desc_fini(guc);
 }
 
 static bool __guc_submission_support(struct intel_guc *guc)
@@ -1186,6 +676,7 @@ void intel_guc_submission_init_early(struct intel_guc *guc)
 	guc->submission_supported = __guc_submission_support(guc);
 }
 
-#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
-#include "selftest_guc.c"
-#endif
+bool intel_engine_in_guc_submission_mode(const struct intel_engine_cs *engine)
+{
+	return engine->set_default_submission == guc_set_default_submission;
+}
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.h
index 54d716828352..e402a2932592 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_submission.h
@@ -6,62 +6,18 @@
 #ifndef _INTEL_GUC_SUBMISSION_H_
 #define _INTEL_GUC_SUBMISSION_H_
 
-#include <linux/spinlock.h>
+#include <linux/types.h>
 
-#include "gt/intel_engine_types.h"
-
-#include "i915_gem.h"
-#include "i915_selftest.h"
-
-struct drm_i915_private;
-
-/*
- * This structure primarily describes the GEM object shared with the GuC.
- * The specs sometimes refer to this object as a "GuC context", but we use
- * the term "client" to avoid confusion with hardware contexts. This
- * GEM object is held for the entire lifetime of our interaction with
- * the GuC, being allocated before the GuC is loaded with its firmware.
- * Because there's no way to update the address used by the GuC after
- * initialisation, the shared object must stay pinned into the GGTT as
- * long as the GuC is in use. We also keep the first page (only) mapped
- * into kernel address space, as it includes shared data that must be
- * updated on every request submission.
- *
- * The single GEM object described here is actually made up of several
- * separate areas, as far as the GuC is concerned. The first page (kept
- * kmap'd) includes the "process descriptor" which holds sequence data for
- * the doorbell, and one cacheline which actually *is* the doorbell; a
- * write to this will "ring the doorbell" (i.e. send an interrupt to the
- * GuC). The subsequent  pages of the client object constitute the work
- * queue (a circular array of work items), again described in the process
- * descriptor. Work queue pages are mapped momentarily as required.
- */
-struct intel_guc_client {
-	struct i915_vma *vma;
-	void *vaddr;
-	struct intel_guc *guc;
-
-	/* bitmap of (host) engine ids */
-	u32 priority;
-	u32 stage_id;
-	u32 proc_desc_offset;
-
-	u16 doorbell_id;
-	unsigned long doorbell_offset;
-
-	/* Protects GuC client's WQ access */
-	spinlock_t wq_lock;
-
-	/* For testing purposes, use nop WQ items instead of real ones */
-	I915_SELFTEST_DECLARE(bool use_nop_wqi);
-};
+struct intel_guc;
+struct intel_engine_cs;
 
 void intel_guc_submission_init_early(struct intel_guc *guc);
 int intel_guc_submission_init(struct intel_guc *guc);
-int intel_guc_submission_enable(struct intel_guc *guc);
+void intel_guc_submission_enable(struct intel_guc *guc);
 void intel_guc_submission_disable(struct intel_guc *guc);
 void intel_guc_submission_fini(struct intel_guc *guc);
 int intel_guc_preempt_work_create(struct intel_guc *guc);
 void intel_guc_preempt_work_destroy(struct intel_guc *guc);
+bool intel_engine_in_guc_submission_mode(const struct intel_engine_cs *engine);
 
 #endif
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c
index d654340d4d03..eee193bf2cc4 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_huc_fw.c
@@ -39,5 +39,5 @@ void intel_huc_fw_init_early(struct intel_huc *huc)
 int intel_huc_fw_upload(struct intel_huc *huc)
 {
 	/* HW doesn't look at destination address for HuC, so set it to 0 */
-	return intel_uc_fw_upload(&huc->fw, huc_to_gt(huc), 0, HUC_UKERNEL);
+	return intel_uc_fw_upload(&huc->fw, 0, HUC_UKERNEL);
 }
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc.c b/drivers/gpu/drm/i915/gt/uc/intel_uc.c
index 629b19377a29..3ffc6267f96e 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc.c
@@ -123,6 +123,11 @@ static void __uc_free_load_err_log(struct intel_uc *uc)
 		i915_gem_object_put(log);
 }
 
+static inline bool guc_communication_enabled(struct intel_guc *guc)
+{
+	return intel_guc_ct_enabled(&guc->ct);
+}
+
 /*
  * Events triggered while CT buffers are disabled are logged in the SCRATCH_15
  * register using the same bits used in the CT message payload. Since our
@@ -158,7 +163,7 @@ static void guc_handle_mmio_msg(struct intel_guc *guc)
 	struct drm_i915_private *i915 = guc_to_gt(guc)->i915;
 
 	/* we need communication to be enabled to reply to GuC */
-	GEM_BUG_ON(guc->handler == intel_guc_to_host_event_handler_nop);
+	GEM_BUG_ON(!guc_communication_enabled(guc));
 
 	if (!guc->mmio_msg)
 		return;
@@ -185,11 +190,6 @@ static void guc_disable_interrupts(struct intel_guc *guc)
 	guc->interrupts.disable(guc);
 }
 
-static inline bool guc_communication_enabled(struct intel_guc *guc)
-{
-	return guc->send != intel_guc_send_nop;
-}
-
 static int guc_enable_communication(struct intel_guc *guc)
 {
 	struct drm_i915_private *i915 = guc_to_gt(guc)->i915;
@@ -205,9 +205,6 @@ static int guc_enable_communication(struct intel_guc *guc)
 	if (ret)
 		return ret;
 
-	guc->send = intel_guc_send_ct;
-	guc->handler = intel_guc_to_host_event_handler_ct;
-
 	/* check for mmio messages received before/during the CT enable */
 	guc_get_mmio_msg(guc);
 	guc_handle_mmio_msg(guc);
@@ -216,7 +213,7 @@ static int guc_enable_communication(struct intel_guc *guc)
 
 	/* check for CT messages received before we enabled interrupts */
 	spin_lock_irq(&i915->irq_lock);
-	intel_guc_to_host_event_handler_ct(guc);
+	intel_guc_ct_event_handler(&guc->ct);
 	spin_unlock_irq(&i915->irq_lock);
 
 	DRM_INFO("GuC communication enabled\n");
@@ -224,7 +221,7 @@ static int guc_enable_communication(struct intel_guc *guc)
 	return 0;
 }
 
-static void __guc_stop_communication(struct intel_guc *guc)
+static void guc_disable_communication(struct intel_guc *guc)
 {
 	/*
 	 * Events generated during or after CT disable are logged by guc in
@@ -235,23 +232,6 @@ static void __guc_stop_communication(struct intel_guc *guc)
 
 	guc_disable_interrupts(guc);
 
-	guc->send = intel_guc_send_nop;
-	guc->handler = intel_guc_to_host_event_handler_nop;
-}
-
-static void guc_stop_communication(struct intel_guc *guc)
-{
-	intel_guc_ct_stop(&guc->ct);
-
-	__guc_stop_communication(guc);
-
-	DRM_INFO("GuC communication stopped\n");
-}
-
-static void guc_disable_communication(struct intel_guc *guc)
-{
-	__guc_stop_communication(guc);
-
 	intel_guc_ct_disable(&guc->ct);
 
 	/*
@@ -267,28 +247,22 @@ static void guc_disable_communication(struct intel_guc *guc)
 
 void intel_uc_fetch_firmwares(struct intel_uc *uc)
 {
-	struct drm_i915_private *i915 = uc_to_gt(uc)->i915;
 	int err;
 
 	if (!intel_uc_uses_guc(uc))
 		return;
 
-	err = intel_uc_fw_fetch(&uc->guc.fw, i915);
+	err = intel_uc_fw_fetch(&uc->guc.fw);
 	if (err)
 		return;
 
 	if (intel_uc_uses_huc(uc))
-		intel_uc_fw_fetch(&uc->huc.fw, i915);
+		intel_uc_fw_fetch(&uc->huc.fw);
 }
 
 void intel_uc_cleanup_firmwares(struct intel_uc *uc)
 {
-	if (!intel_uc_uses_guc(uc))
-		return;
-
-	if (intel_uc_uses_huc(uc))
-		intel_uc_fw_cleanup_fetch(&uc->huc.fw);
-
+	intel_uc_fw_cleanup_fetch(&uc->huc.fw);
 	intel_uc_fw_cleanup_fetch(&uc->guc.fw);
 }
 
@@ -316,15 +290,8 @@ void intel_uc_init(struct intel_uc *uc)
 
 void intel_uc_fini(struct intel_uc *uc)
 {
-	struct intel_guc *guc = &uc->guc;
-
-	if (!intel_uc_uses_guc(uc))
-		return;
-
-	if (intel_uc_uses_huc(uc))
-		intel_huc_fini(&uc->huc);
-
-	intel_guc_fini(guc);
+	intel_huc_fini(&uc->huc);
+	intel_guc_fini(&uc->guc);
 
 	__uc_free_load_err_log(uc);
 }
@@ -486,11 +453,8 @@ int intel_uc_init_hw(struct intel_uc *uc)
 	if (ret)
 		goto err_communication;
 
-	if (intel_uc_supports_guc_submission(uc)) {
-		ret = intel_guc_submission_enable(guc);
-		if (ret)
-			goto err_communication;
-	}
+	if (intel_uc_supports_guc_submission(uc))
+		intel_guc_submission_enable(guc);
 
 	dev_info(i915->drm.dev, "%s firmware %s version %u.%u %s:%s\n",
 		 intel_uc_fw_type_repr(INTEL_UC_FW_TYPE_GUC), guc->fw.path,
@@ -560,7 +524,7 @@ void intel_uc_reset_prepare(struct intel_uc *uc)
 	if (!intel_guc_is_running(guc))
 		return;
 
-	guc_stop_communication(guc);
+	guc_disable_communication(guc);
 	__uc_sanitize(uc);
 }
 
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
index 66a30ab7044a..8ee0a0c7f447 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.c
@@ -11,7 +11,6 @@
 #include "intel_uc_fw_abi.h"
 #include "i915_drv.h"
 
-#ifdef CONFIG_DRM_I915_DEBUG_GUC
 static inline struct intel_gt *__uc_fw_to_gt(struct intel_uc_fw *uc_fw)
 {
 	GEM_BUG_ON(uc_fw->status == INTEL_UC_FIRMWARE_UNINITIALIZED);
@@ -22,6 +21,7 @@ static inline struct intel_gt *__uc_fw_to_gt(struct intel_uc_fw *uc_fw)
 	return container_of(uc_fw, struct intel_gt, uc.huc.fw);
 }
 
+#ifdef CONFIG_DRM_I915_DEBUG_GUC
 void intel_uc_fw_change_status(struct intel_uc_fw *uc_fw,
 			       enum intel_uc_fw_status status)
 {
@@ -219,10 +219,9 @@ void intel_uc_fw_init_early(struct intel_uc_fw *uc_fw,
 				  INTEL_UC_FIRMWARE_NOT_SUPPORTED);
 }
 
-static void __force_fw_fetch_failures(struct intel_uc_fw *uc_fw,
-				      struct drm_i915_private *i915,
-				      int e)
+static void __force_fw_fetch_failures(struct intel_uc_fw *uc_fw, int e)
 {
+	struct drm_i915_private *i915 = __uc_fw_to_gt(uc_fw)->i915;
 	bool user = e == -EINVAL;
 
 	if (i915_inject_probe_error(i915, e)) {
@@ -260,14 +259,14 @@ static void __force_fw_fetch_failures(struct intel_uc_fw *uc_fw,
 /**
  * intel_uc_fw_fetch - fetch uC firmware
  * @uc_fw: uC firmware
- * @i915: device private
  *
  * Fetch uC firmware into GEM obj.
  *
  * Return: 0 on success, a negative errno code on failure.
  */
-int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw, struct drm_i915_private *i915)
+int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw)
 {
+	struct drm_i915_private *i915 = __uc_fw_to_gt(uc_fw)->i915;
 	struct device *dev = i915->drm.dev;
 	struct drm_i915_gem_object *obj;
 	const struct firmware *fw = NULL;
@@ -282,8 +281,8 @@ int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw, struct drm_i915_private *i915)
 	if (err)
 		return err;
 
-	__force_fw_fetch_failures(uc_fw, i915, -EINVAL);
-	__force_fw_fetch_failures(uc_fw, i915, -ESTALE);
+	__force_fw_fetch_failures(uc_fw, -EINVAL);
+	__force_fw_fetch_failures(uc_fw, -ESTALE);
 
 	err = request_firmware(&fw, uc_fw->path, dev);
 	if (err)
@@ -390,8 +389,9 @@ fail:
 	return err;
 }
 
-static u32 uc_fw_ggtt_offset(struct intel_uc_fw *uc_fw, struct i915_ggtt *ggtt)
+static u32 uc_fw_ggtt_offset(struct intel_uc_fw *uc_fw)
 {
+	struct i915_ggtt *ggtt = __uc_fw_to_gt(uc_fw)->ggtt;
 	struct drm_mm_node *node = &ggtt->uc_fw;
 
 	GEM_BUG_ON(!drm_mm_node_allocated(node));
@@ -401,13 +401,12 @@ static u32 uc_fw_ggtt_offset(struct intel_uc_fw *uc_fw, struct i915_ggtt *ggtt)
 	return lower_32_bits(node->start);
 }
 
-static void intel_uc_fw_ggtt_bind(struct intel_uc_fw *uc_fw,
-				  struct intel_gt *gt)
+static void uc_fw_bind_ggtt(struct intel_uc_fw *uc_fw)
 {
 	struct drm_i915_gem_object *obj = uc_fw->obj;
-	struct i915_ggtt *ggtt = gt->ggtt;
+	struct i915_ggtt *ggtt = __uc_fw_to_gt(uc_fw)->ggtt;
 	struct i915_vma dummy = {
-		.node.start = uc_fw_ggtt_offset(uc_fw, ggtt),
+		.node.start = uc_fw_ggtt_offset(uc_fw),
 		.node.size = obj->base.size,
 		.pages = obj->mm.pages,
 		.vm = &ggtt->vm,
@@ -422,19 +421,18 @@ static void intel_uc_fw_ggtt_bind(struct intel_uc_fw *uc_fw,
 	ggtt->vm.insert_entries(&ggtt->vm, &dummy, I915_CACHE_NONE, 0);
 }
 
-static void intel_uc_fw_ggtt_unbind(struct intel_uc_fw *uc_fw,
-				    struct intel_gt *gt)
+static void uc_fw_unbind_ggtt(struct intel_uc_fw *uc_fw)
 {
 	struct drm_i915_gem_object *obj = uc_fw->obj;
-	struct i915_ggtt *ggtt = gt->ggtt;
-	u64 start = uc_fw_ggtt_offset(uc_fw, ggtt);
+	struct i915_ggtt *ggtt = __uc_fw_to_gt(uc_fw)->ggtt;
+	u64 start = uc_fw_ggtt_offset(uc_fw);
 
 	ggtt->vm.clear_range(&ggtt->vm, start, obj->base.size);
 }
 
-static int uc_fw_xfer(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
-		      u32 wopcm_offset, u32 dma_flags)
+static int uc_fw_xfer(struct intel_uc_fw *uc_fw, u32 dst_offset, u32 dma_flags)
 {
+	struct intel_gt *gt = __uc_fw_to_gt(uc_fw);
 	struct intel_uncore *uncore = gt->uncore;
 	u64 offset;
 	int ret;
@@ -446,13 +444,13 @@ static int uc_fw_xfer(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
 	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
 
 	/* Set the source address for the uCode */
-	offset = uc_fw_ggtt_offset(uc_fw, gt->ggtt);
+	offset = uc_fw_ggtt_offset(uc_fw);
 	GEM_BUG_ON(upper_32_bits(offset) & 0xFFFF0000);
 	intel_uncore_write_fw(uncore, DMA_ADDR_0_LOW, lower_32_bits(offset));
 	intel_uncore_write_fw(uncore, DMA_ADDR_0_HIGH, upper_32_bits(offset));
 
 	/* Set the DMA destination */
-	intel_uncore_write_fw(uncore, DMA_ADDR_1_LOW, wopcm_offset);
+	intel_uncore_write_fw(uncore, DMA_ADDR_1_LOW, dst_offset);
 	intel_uncore_write_fw(uncore, DMA_ADDR_1_HIGH, DMA_ADDRESS_SPACE_WOPCM);
 
 	/*
@@ -484,17 +482,16 @@ static int uc_fw_xfer(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
 /**
  * intel_uc_fw_upload - load uC firmware using custom loader
  * @uc_fw: uC firmware
- * @gt: the intel_gt structure
- * @wopcm_offset: destination offset in wopcm
+ * @dst_offset: destination offset
  * @dma_flags: flags for flags for dma ctrl
  *
  * Loads uC firmware and updates internal flags.
  *
  * Return: 0 on success, non-zero on failure.
  */
-int intel_uc_fw_upload(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
-		       u32 wopcm_offset, u32 dma_flags)
+int intel_uc_fw_upload(struct intel_uc_fw *uc_fw, u32 dst_offset, u32 dma_flags)
 {
+	struct intel_gt *gt = __uc_fw_to_gt(uc_fw);
 	int err;
 
 	/* make sure the status was cleared the last time we reset the uc */
@@ -508,9 +505,9 @@ int intel_uc_fw_upload(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
 		return -ENOEXEC;
 
 	/* Call custom loader */
-	intel_uc_fw_ggtt_bind(uc_fw, gt);
-	err = uc_fw_xfer(uc_fw, gt, wopcm_offset, dma_flags);
-	intel_uc_fw_ggtt_unbind(uc_fw, gt);
+	uc_fw_bind_ggtt(uc_fw);
+	err = uc_fw_xfer(uc_fw, dst_offset, dma_flags);
+	uc_fw_unbind_ggtt(uc_fw);
 	if (err)
 		goto fail;
 
@@ -547,10 +544,7 @@ int intel_uc_fw_init(struct intel_uc_fw *uc_fw)
 
 void intel_uc_fw_fini(struct intel_uc_fw *uc_fw)
 {
-	if (!intel_uc_fw_is_available(uc_fw))
-		return;
-
-	i915_gem_object_unpin_pages(uc_fw->obj);
+	intel_uc_fw_cleanup_fetch(uc_fw);
 }
 
 /**
diff --git a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h
index 7a0a5989afc9..1f30543d0d2d 100644
--- a/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h
+++ b/drivers/gpu/drm/i915/gt/uc/intel_uc_fw.h
@@ -229,10 +229,9 @@ static inline u32 intel_uc_fw_get_upload_size(struct intel_uc_fw *uc_fw)
 void intel_uc_fw_init_early(struct intel_uc_fw *uc_fw,
 			    enum intel_uc_fw_type type, bool supported,
 			    enum intel_platform platform, u8 rev);
-int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw, struct drm_i915_private *i915);
+int intel_uc_fw_fetch(struct intel_uc_fw *uc_fw);
 void intel_uc_fw_cleanup_fetch(struct intel_uc_fw *uc_fw);
-int intel_uc_fw_upload(struct intel_uc_fw *uc_fw, struct intel_gt *gt,
-		       u32 wopcm_offset, u32 dma_flags);
+int intel_uc_fw_upload(struct intel_uc_fw *uc_fw, u32 offset, u32 dma_flags);
 int intel_uc_fw_init(struct intel_uc_fw *uc_fw);
 void intel_uc_fw_fini(struct intel_uc_fw *uc_fw);
 size_t intel_uc_fw_copy_rsa(struct intel_uc_fw *uc_fw, void *dst, u32 max_len);
diff --git a/drivers/gpu/drm/i915/gt/uc/selftest_guc.c b/drivers/gpu/drm/i915/gt/uc/selftest_guc.c
deleted file mode 100644
index d8a80388bd31..000000000000
--- a/drivers/gpu/drm/i915/gt/uc/selftest_guc.c
+++ /dev/null
@@ -1,299 +0,0 @@
-// SPDX-License-Identifier: MIT
-/*
- * Copyright © 2017 Intel Corporation
- */
-
-#include "i915_selftest.h"
-#include "gem/i915_gem_pm.h"
-
-/* max doorbell number + negative test for each client type */
-#define ATTEMPTS (GUC_NUM_DOORBELLS + GUC_CLIENT_PRIORITY_NUM)
-
-static struct intel_guc_client *clients[ATTEMPTS];
-
-static bool available_dbs(struct intel_guc *guc, u32 priority)
-{
-	unsigned long offset;
-	unsigned long end;
-	u16 id;
-
-	/* first half is used for normal priority, second half for high */
-	offset = 0;
-	end = GUC_NUM_DOORBELLS / 2;
-	if (priority <= GUC_CLIENT_PRIORITY_HIGH) {
-		offset = end;
-		end += offset;
-	}
-
-	id = find_next_zero_bit(guc->doorbell_bitmap, end, offset);
-	if (id < end)
-		return true;
-
-	return false;
-}
-
-static int check_all_doorbells(struct intel_guc *guc)
-{
-	u16 db_id;
-
-	pr_info_once("Max number of doorbells: %d", GUC_NUM_DOORBELLS);
-	for (db_id = 0; db_id < GUC_NUM_DOORBELLS; ++db_id) {
-		if (!doorbell_ok(guc, db_id)) {
-			pr_err("doorbell %d, not ok\n", db_id);
-			return -EIO;
-		}
-	}
-
-	return 0;
-}
-
-static int ring_doorbell_nop(struct intel_guc_client *client)
-{
-	struct guc_process_desc *desc = __get_process_desc(client);
-	int err;
-
-	client->use_nop_wqi = true;
-
-	spin_lock_irq(&client->wq_lock);
-
-	guc_wq_item_append(client, 0, 0, 0, 0);
-	guc_ring_doorbell(client);
-
-	spin_unlock_irq(&client->wq_lock);
-
-	client->use_nop_wqi = false;
-
-	/* if there are no issues GuC will update the WQ head and keep the
-	 * WQ in active status
-	 */
-	err = wait_for(READ_ONCE(desc->head) == READ_ONCE(desc->tail), 10);
-	if (err) {
-		pr_err("doorbell %u ring failed!\n", client->doorbell_id);
-		return -EIO;
-	}
-
-	if (desc->wq_status != WQ_STATUS_ACTIVE) {
-		pr_err("doorbell %u ring put WQ in bad state (%u)!\n",
-		       client->doorbell_id, desc->wq_status);
-		return -EIO;
-	}
-
-	return 0;
-}
-
-/*
- * Basic client sanity check, handy to validate create_clients.
- */
-static int validate_client(struct intel_guc_client *client, int client_priority)
-{
-	if (client->priority != client_priority ||
-	    client->doorbell_id == GUC_DOORBELL_INVALID)
-		return -EINVAL;
-	else
-		return 0;
-}
-
-static bool client_doorbell_in_sync(struct intel_guc_client *client)
-{
-	return !client || doorbell_ok(client->guc, client->doorbell_id);
-}
-
-/*
- * Check that we're able to synchronize guc_clients with their doorbells
- *
- * We're creating clients and reserving doorbells once, at module load. During
- * module lifetime, GuC, doorbell HW, and i915 state may go out of sync due to
- * GuC being reset. In other words - GuC clients are still around, but the
- * status of their doorbells may be incorrect. This is the reason behind
- * validating that the doorbells status expected by the driver matches what the
- * GuC/HW have.
- */
-static int igt_guc_clients(void *arg)
-{
-	struct intel_gt *gt = arg;
-	struct intel_guc *guc = &gt->uc.guc;
-	intel_wakeref_t wakeref;
-	int err = 0;
-
-	GEM_BUG_ON(!HAS_GT_UC(gt->i915));
-	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
-
-	err = check_all_doorbells(guc);
-	if (err)
-		goto unlock;
-
-	/*
-	 * Get rid of clients created during driver load because the test will
-	 * recreate them.
-	 */
-	guc_clients_disable(guc);
-	guc_clients_destroy(guc);
-	if (guc->execbuf_client) {
-		pr_err("guc_clients_destroy lied!\n");
-		err = -EINVAL;
-		goto unlock;
-	}
-
-	err = guc_clients_create(guc);
-	if (err) {
-		pr_err("Failed to create clients\n");
-		goto unlock;
-	}
-	GEM_BUG_ON(!guc->execbuf_client);
-
-	err = validate_client(guc->execbuf_client,
-			      GUC_CLIENT_PRIORITY_KMD_NORMAL);
-	if (err) {
-		pr_err("execbug client validation failed\n");
-		goto out;
-	}
-
-	/* the client should now have reserved a doorbell */
-	if (!has_doorbell(guc->execbuf_client)) {
-		pr_err("guc_clients_create didn't reserve doorbells\n");
-		err = -EINVAL;
-		goto out;
-	}
-
-	/* Now enable the clients */
-	guc_clients_enable(guc);
-
-	/* each client should now have received a doorbell */
-	if (!client_doorbell_in_sync(guc->execbuf_client)) {
-		pr_err("failed to initialize the doorbells\n");
-		err = -EINVAL;
-		goto out;
-	}
-
-	/*
-	 * Basic test - an attempt to reallocate a valid doorbell to the
-	 * client it is currently assigned should not cause a failure.
-	 */
-	err = create_doorbell(guc->execbuf_client);
-
-out:
-	/*
-	 * Leave clean state for other test, plus the driver always destroy the
-	 * clients during unload.
-	 */
-	guc_clients_disable(guc);
-	guc_clients_destroy(guc);
-	guc_clients_create(guc);
-	guc_clients_enable(guc);
-unlock:
-	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
-	return err;
-}
-
-/*
- * Create as many clients as number of doorbells. Note that there's already
- * client(s)/doorbell(s) created during driver load, but this test creates
- * its own and do not interact with the existing ones.
- */
-static int igt_guc_doorbells(void *arg)
-{
-	struct intel_gt *gt = arg;
-	struct intel_guc *guc = &gt->uc.guc;
-	intel_wakeref_t wakeref;
-	int i, err = 0;
-	u16 db_id;
-
-	GEM_BUG_ON(!HAS_GT_UC(gt->i915));
-	wakeref = intel_runtime_pm_get(gt->uncore->rpm);
-
-	err = check_all_doorbells(guc);
-	if (err)
-		goto unlock;
-
-	for (i = 0; i < ATTEMPTS; i++) {
-		clients[i] = guc_client_alloc(guc, i % GUC_CLIENT_PRIORITY_NUM);
-
-		if (!clients[i]) {
-			pr_err("[%d] No guc client\n", i);
-			err = -EINVAL;
-			goto out;
-		}
-
-		if (IS_ERR(clients[i])) {
-			if (PTR_ERR(clients[i]) != -ENOSPC) {
-				pr_err("[%d] unexpected error\n", i);
-				err = PTR_ERR(clients[i]);
-				goto out;
-			}
-
-			if (available_dbs(guc, i % GUC_CLIENT_PRIORITY_NUM)) {
-				pr_err("[%d] non-db related alloc fail\n", i);
-				err = -EINVAL;
-				goto out;
-			}
-
-			/* expected, ran out of dbs for this client type */
-			continue;
-		}
-
-		/*
-		 * The check below is only valid because we keep a doorbell
-		 * assigned during the whole life of the client.
-		 */
-		if (clients[i]->stage_id >= GUC_NUM_DOORBELLS) {
-			pr_err("[%d] more clients than doorbells (%d >= %d)\n",
-			       i, clients[i]->stage_id, GUC_NUM_DOORBELLS);
-			err = -EINVAL;
-			goto out;
-		}
-
-		err = validate_client(clients[i], i % GUC_CLIENT_PRIORITY_NUM);
-		if (err) {
-			pr_err("[%d] client_alloc sanity check failed!\n", i);
-			err = -EINVAL;
-			goto out;
-		}
-
-		db_id = clients[i]->doorbell_id;
-
-		err = __guc_client_enable(clients[i]);
-		if (err) {
-			pr_err("[%d] Failed to create a doorbell\n", i);
-			goto out;
-		}
-
-		/* doorbell id shouldn't change, we are holding the mutex */
-		if (db_id != clients[i]->doorbell_id) {
-			pr_err("[%d] doorbell id changed (%d != %d)\n",
-			       i, db_id, clients[i]->doorbell_id);
-			err = -EINVAL;
-			goto out;
-		}
-
-		err = check_all_doorbells(guc);
-		if (err)
-			goto out;
-
-		err = ring_doorbell_nop(clients[i]);
-		if (err)
-			goto out;
-	}
-
-out:
-	for (i = 0; i < ATTEMPTS; i++)
-		if (!IS_ERR_OR_NULL(clients[i])) {
-			__guc_client_disable(clients[i]);
-			guc_client_free(clients[i]);
-		}
-unlock:
-	intel_runtime_pm_put(gt->uncore->rpm, wakeref);
-	return err;
-}
-
-int intel_guc_live_selftest(struct drm_i915_private *i915)
-{
-	static const struct i915_subtest tests[] = {
-		SUBTEST(igt_guc_clients),
-		SUBTEST(igt_guc_doorbells),
-	};
-
-	if (!USES_GUC_SUBMISSION(i915))
-		return 0;
-
-	return intel_gt_live_subtests(tests, &i915->gt);
-}
diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.h b/drivers/gpu/drm/i915/gvt/cmd_parser.h
index 286703643002..ab25d151932a 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.h
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.h
@@ -38,6 +38,10 @@
 
 #define GVT_CMD_HASH_BITS 7
 
+struct intel_gvt;
+struct intel_shadow_wa_ctx;
+struct intel_vgpu_workload;
+
 void intel_gvt_clean_cmd_parser(struct intel_gvt *gvt);
 
 int intel_gvt_init_cmd_parser(struct intel_gvt *gvt);
diff --git a/drivers/gpu/drm/i915/gvt/display.h b/drivers/gpu/drm/i915/gvt/display.h
index a87f33e6a23c..b59b34046e1e 100644
--- a/drivers/gpu/drm/i915/gvt/display.h
+++ b/drivers/gpu/drm/i915/gvt/display.h
@@ -35,6 +35,11 @@
 #ifndef _GVT_DISPLAY_H_
 #define _GVT_DISPLAY_H_
 
+#include <linux/types.h>
+
+struct intel_gvt;
+struct intel_vgpu;
+
 #define SBI_REG_MAX	20
 #define DPCD_SIZE	0x700
 
diff --git a/drivers/gpu/drm/i915/gvt/edid.h b/drivers/gpu/drm/i915/gvt/edid.h
index f6dfc8b795ec..dfe0cbc6aad8 100644
--- a/drivers/gpu/drm/i915/gvt/edid.h
+++ b/drivers/gpu/drm/i915/gvt/edid.h
@@ -35,6 +35,10 @@
 #ifndef _GVT_EDID_H_
 #define _GVT_EDID_H_
 
+#include <linux/types.h>
+
+struct intel_vgpu;
+
 #define EDID_SIZE		128
 #define EDID_ADDR		0x50 /* Linux hvm EDID addr */
 
diff --git a/drivers/gpu/drm/i915/gvt/execlist.h b/drivers/gpu/drm/i915/gvt/execlist.h
index 5ccc2c695848..5c0c1fd30c83 100644
--- a/drivers/gpu/drm/i915/gvt/execlist.h
+++ b/drivers/gpu/drm/i915/gvt/execlist.h
@@ -35,6 +35,8 @@
 #ifndef _GVT_EXECLIST_H_
 #define _GVT_EXECLIST_H_
 
+#include <linux/types.h>
+
 struct execlist_ctx_descriptor_format {
 	union {
 		u32 ldw;
diff --git a/drivers/gpu/drm/i915/gvt/fb_decoder.h b/drivers/gpu/drm/i915/gvt/fb_decoder.h
index 60c155085029..67b6ede9e707 100644
--- a/drivers/gpu/drm/i915/gvt/fb_decoder.h
+++ b/drivers/gpu/drm/i915/gvt/fb_decoder.h
@@ -36,6 +36,8 @@
 #ifndef _GVT_FB_DECODER_H_
 #define _GVT_FB_DECODER_H_
 
+#include <linux/types.h>
+
 #define _PLANE_CTL_FORMAT_SHIFT		24
 #define _PLANE_CTL_TILED_SHIFT		10
 #define _PIPE_V_SRCSZ_SHIFT		0
diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
index 4b04af569c05..34cb404ba4b7 100644
--- a/drivers/gpu/drm/i915/gvt/gtt.c
+++ b/drivers/gpu/drm/i915/gvt/gtt.c
@@ -1282,7 +1282,7 @@ static int ppgtt_populate_shadow_entry(struct intel_vgpu *vgpu,
 		return -EINVAL;
 	default:
 		GEM_BUG_ON(1);
-	};
+	}
 
 	/* direct shadow */
 	ret = intel_gvt_hypervisor_dma_map_guest_page(vgpu, gfn, page_size,
diff --git a/drivers/gpu/drm/i915/gvt/hypercall.h b/drivers/gpu/drm/i915/gvt/hypercall.h
index 4862fb12778e..9599c0a762b2 100644
--- a/drivers/gpu/drm/i915/gvt/hypercall.h
+++ b/drivers/gpu/drm/i915/gvt/hypercall.h
@@ -33,6 +33,10 @@
 #ifndef _GVT_HYPERCALL_H_
 #define _GVT_HYPERCALL_H_
 
+#include <linux/types.h>
+
+struct device;
+
 enum hypervisor_type {
 	INTEL_GVT_HYPERVISOR_XEN = 0,
 	INTEL_GVT_HYPERVISOR_KVM,
diff --git a/drivers/gpu/drm/i915/gvt/interrupt.h b/drivers/gpu/drm/i915/gvt/interrupt.h
index 5313fb1b33e1..fcd663811d37 100644
--- a/drivers/gpu/drm/i915/gvt/interrupt.h
+++ b/drivers/gpu/drm/i915/gvt/interrupt.h
@@ -32,6 +32,8 @@
 #ifndef _GVT_INTERRUPT_H_
 #define _GVT_INTERRUPT_H_
 
+#include <linux/types.h>
+
 enum intel_gvt_event_type {
 	RCS_MI_USER_INTERRUPT = 0,
 	RCS_DEBUG,
@@ -135,6 +137,7 @@ enum intel_gvt_event_type {
 
 struct intel_gvt_irq;
 struct intel_gvt;
+struct intel_vgpu;
 
 typedef void (*gvt_event_virt_handler_t)(struct intel_gvt_irq *irq,
 	enum intel_gvt_event_type event, struct intel_vgpu *vgpu);
diff --git a/drivers/gpu/drm/i915/gvt/mmio.h b/drivers/gpu/drm/i915/gvt/mmio.h
index 5874f1cb4306..2e68f4b02c94 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.h
+++ b/drivers/gpu/drm/i915/gvt/mmio.h
@@ -36,6 +36,8 @@
 #ifndef _GVT_MMIO_H_
 #define _GVT_MMIO_H_
 
+#include <linux/types.h>
+
 struct intel_gvt;
 struct intel_vgpu;
 
diff --git a/drivers/gpu/drm/i915/gvt/page_track.h b/drivers/gpu/drm/i915/gvt/page_track.h
index fa607a71c3c0..f6eb7135583c 100644
--- a/drivers/gpu/drm/i915/gvt/page_track.h
+++ b/drivers/gpu/drm/i915/gvt/page_track.h
@@ -25,6 +25,9 @@
 #ifndef _GVT_PAGE_TRACK_H_
 #define _GVT_PAGE_TRACK_H_
 
+#include <linux/types.h>
+
+struct intel_vgpu;
 struct intel_vgpu_page_track;
 
 typedef int (*gvt_page_track_handler_t)(
diff --git a/drivers/gpu/drm/i915/gvt/sched_policy.h b/drivers/gpu/drm/i915/gvt/sched_policy.h
index 7b59e3e88b8b..3dacdad5f529 100644
--- a/drivers/gpu/drm/i915/gvt/sched_policy.h
+++ b/drivers/gpu/drm/i915/gvt/sched_policy.h
@@ -34,6 +34,9 @@
 #ifndef __GVT_SCHED_POLICY__
 #define __GVT_SCHED_POLICY__
 
+struct intel_gvt;
+struct intel_vgpu;
+
 struct intel_gvt_sched_policy_ops {
 	int (*init)(struct intel_gvt *gvt);
 	void (*clean)(struct intel_gvt *gvt);
diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c
index 5b2a7d072ec9..b3299f88e24e 100644
--- a/drivers/gpu/drm/i915/gvt/scheduler.c
+++ b/drivers/gpu/drm/i915/gvt/scheduler.c
@@ -35,12 +35,12 @@
 
 #include <linux/kthread.h>
 
-#include "gem/i915_gem_context.h"
 #include "gem/i915_gem_pm.h"
 #include "gt/intel_context.h"
 #include "gt/intel_ring.h"
 
 #include "i915_drv.h"
+#include "i915_gem_gtt.h"
 #include "gvt.h"
 
 #define RING_CTX_OFF(x) \
@@ -59,7 +59,7 @@ static void set_context_pdp_root_pointer(
 static void update_shadow_pdps(struct intel_vgpu_workload *workload)
 {
 	struct drm_i915_gem_object *ctx_obj =
-		workload->req->hw_context->state->obj;
+		workload->req->context->state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 
@@ -130,7 +130,7 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	struct intel_gvt *gvt = vgpu->gvt;
 	int ring_id = workload->ring_id;
 	struct drm_i915_gem_object *ctx_obj =
-		workload->req->hw_context->state->obj;
+		workload->req->context->state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *dst;
@@ -205,9 +205,9 @@ static int populate_shadow_context(struct intel_vgpu_workload *workload)
 	return 0;
 }
 
-static inline bool is_gvt_request(struct i915_request *req)
+static inline bool is_gvt_request(struct i915_request *rq)
 {
-	return i915_gem_context_force_single_submission(req->gem_context);
+	return intel_context_force_single_submission(rq->context);
 }
 
 static void save_ring_hw_state(struct intel_vgpu *vgpu, int ring_id)
@@ -307,7 +307,7 @@ static int copy_workload_to_ring_buffer(struct intel_vgpu_workload *workload)
 	u32 *cs;
 	int err;
 
-	if (IS_GEN(req->i915, 9) && is_inhibit_context(req->hw_context))
+	if (IS_GEN(req->i915, 9) && is_inhibit_context(req->context))
 		intel_vgpu_restore_inhibit_context(vgpu, req);
 
 	/*
@@ -363,11 +363,10 @@ static void release_shadow_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 }
 
 static void set_context_ppgtt_from_shadow(struct intel_vgpu_workload *workload,
-					  struct i915_gem_context *ctx)
+					  struct intel_context *ce)
 {
 	struct intel_vgpu_mm *mm = workload->shadow_mm;
-	struct i915_ppgtt *ppgtt =
-		i915_vm_to_ppgtt(i915_gem_context_get_vm_rcu(ctx));
+	struct i915_ppgtt *ppgtt = i915_vm_to_ppgtt(ce->vm);
 	int i = 0;
 
 	if (mm->ppgtt_mm.root_entry_type == GTT_TYPE_PPGTT_ROOT_L4_ENTRY) {
@@ -380,8 +379,6 @@ static void set_context_ppgtt_from_shadow(struct intel_vgpu_workload *workload,
 			px_dma(pd) = mm->ppgtt_mm.shadow_pdps[i];
 		}
 	}
-
-	i915_vm_put(&ppgtt->vm);
 }
 
 static int
@@ -529,7 +526,7 @@ static void update_wa_ctx_2_shadow_ctx(struct intel_shadow_wa_ctx *wa_ctx)
 		container_of(wa_ctx, struct intel_vgpu_workload, wa_ctx);
 	struct i915_request *rq = workload->req;
 	struct execlist_ring_context *shadow_ring_context =
-		(struct execlist_ring_context *)rq->hw_context->lrc_reg_state;
+		(struct execlist_ring_context *)rq->context->lrc_reg_state;
 
 	shadow_ring_context->bb_per_ctx_ptr.val =
 		(shadow_ring_context->bb_per_ctx_ptr.val &
@@ -628,7 +625,7 @@ static int prepare_workload(struct intel_vgpu_workload *workload)
 
 	update_shadow_pdps(workload);
 
-	set_context_ppgtt_from_shadow(workload, s->shadow[ring]->gem_context);
+	set_context_ppgtt_from_shadow(workload, s->shadow[ring]);
 
 	ret = intel_vgpu_sync_oos_pages(workload->vgpu);
 	if (ret) {
@@ -787,7 +784,7 @@ static void update_guest_context(struct intel_vgpu_workload *workload)
 	struct i915_request *rq = workload->req;
 	struct intel_vgpu *vgpu = workload->vgpu;
 	struct intel_gvt *gvt = vgpu->gvt;
-	struct drm_i915_gem_object *ctx_obj = rq->hw_context->state->obj;
+	struct drm_i915_gem_object *ctx_obj = rq->context->state->obj;
 	struct execlist_ring_context *shadow_ring_context;
 	struct page *page;
 	void *src;
@@ -1223,18 +1220,14 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 	struct drm_i915_private *i915 = vgpu->gvt->dev_priv;
 	struct intel_vgpu_submission *s = &vgpu->submission;
 	struct intel_engine_cs *engine;
-	struct i915_gem_context *ctx;
 	struct i915_ppgtt *ppgtt;
 	enum intel_engine_id i;
 	int ret;
 
-	ctx = i915_gem_context_create_kernel(i915, I915_PRIORITY_MAX);
-	if (IS_ERR(ctx))
-		return PTR_ERR(ctx);
+	ppgtt = i915_ppgtt_create(i915);
+	if (IS_ERR(ppgtt))
+		return PTR_ERR(ppgtt);
 
-	i915_gem_context_set_force_single_submission(ctx);
-
-	ppgtt = i915_vm_to_ppgtt(i915_gem_context_get_vm_rcu(ctx));
 	i915_context_ppgtt_root_save(s, ppgtt);
 
 	for_each_engine(engine, i915, i) {
@@ -1243,12 +1236,16 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 		INIT_LIST_HEAD(&s->workload_q_head[i]);
 		s->shadow[i] = ERR_PTR(-EINVAL);
 
-		ce = intel_context_create(ctx, engine);
+		ce = intel_context_create(engine);
 		if (IS_ERR(ce)) {
 			ret = PTR_ERR(ce);
 			goto out_shadow_ctx;
 		}
 
+		i915_vm_put(ce->vm);
+		ce->vm = i915_vm_get(&ppgtt->vm);
+		intel_context_set_single_submission(ce);
+
 		if (!USES_GUC_SUBMISSION(i915)) { /* Max ring buffer size */
 			const unsigned int ring_size = 512 * SZ_4K;
 
@@ -1281,7 +1278,6 @@ int intel_vgpu_setup_submission(struct intel_vgpu *vgpu)
 	bitmap_zero(s->tlb_handle_pending, I915_NUM_ENGINES);
 
 	i915_vm_put(&ppgtt->vm);
-	i915_gem_context_put(ctx);
 	return 0;
 
 out_shadow_ctx:
@@ -1294,7 +1290,6 @@ out_shadow_ctx:
 		intel_context_put(s->shadow[i]);
 	}
 	i915_vm_put(&ppgtt->vm);
-	i915_gem_context_put(ctx);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_active.c b/drivers/gpu/drm/i915/i915_active.c
index a19e7d89bc8a..cfe09964622b 100644
--- a/drivers/gpu/drm/i915/i915_active.c
+++ b/drivers/gpu/drm/i915/i915_active.c
@@ -6,6 +6,7 @@
 
 #include <linux/debugobjects.h>
 
+#include "gt/intel_context.h"
 #include "gt/intel_engine_pm.h"
 #include "gt/intel_ring.h"
 
@@ -91,10 +92,9 @@ static void debug_active_init(struct i915_active *ref)
 
 static void debug_active_activate(struct i915_active *ref)
 {
-	spin_lock_irq(&ref->tree_lock);
+	lockdep_assert_held(&ref->tree_lock);
 	if (!atomic_read(&ref->count)) /* before the first inc */
 		debug_object_activate(ref, &active_debug_desc);
-	spin_unlock_irq(&ref->tree_lock);
 }
 
 static void debug_active_deactivate(struct i915_active *ref)
@@ -186,18 +186,33 @@ active_retire(struct i915_active *ref)
 	__active_retire(ref);
 }
 
+static inline struct dma_fence **
+__active_fence_slot(struct i915_active_fence *active)
+{
+	return (struct dma_fence ** __force)&active->fence;
+}
+
+static inline bool
+active_fence_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
+{
+	struct i915_active_fence *active =
+		container_of(cb, typeof(*active), cb);
+
+	return cmpxchg(__active_fence_slot(active), fence, NULL) == fence;
+}
+
 static void
 node_retire(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
-	i915_active_fence_cb(fence, cb);
-	active_retire(container_of(cb, struct active_node, base.cb)->ref);
+	if (active_fence_cb(fence, cb))
+		active_retire(container_of(cb, struct active_node, base.cb)->ref);
 }
 
 static void
 excl_retire(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
-	i915_active_fence_cb(fence, cb);
-	active_retire(container_of(cb, struct i915_active, excl.cb));
+	if (active_fence_cb(fence, cb))
+		active_retire(container_of(cb, struct i915_active, excl.cb));
 }
 
 static struct i915_active_fence *
@@ -244,7 +259,7 @@ active_instance(struct i915_active *ref, struct intel_timeline *tl)
 	}
 
 	node = prealloc;
-	__i915_active_fence_init(&node->base, &tl->mutex, NULL, node_retire);
+	__i915_active_fence_init(&node->base, NULL, node_retire);
 	node->ref = ref;
 	node->timeline = idx;
 
@@ -262,7 +277,8 @@ out:
 void __i915_active_init(struct i915_active *ref,
 			int (*active)(struct i915_active *ref),
 			void (*retire)(struct i915_active *ref),
-			struct lock_class_key *key)
+			struct lock_class_key *mkey,
+			struct lock_class_key *wkey)
 {
 	unsigned long bits;
 
@@ -280,9 +296,12 @@ void __i915_active_init(struct i915_active *ref,
 
 	init_llist_head(&ref->preallocated_barriers);
 	atomic_set(&ref->count, 0);
-	__mutex_init(&ref->mutex, "i915_active", key);
-	__i915_active_fence_init(&ref->excl, &ref->mutex, NULL, excl_retire);
+	__mutex_init(&ref->mutex, "i915_active", mkey);
+	__i915_active_fence_init(&ref->excl, NULL, excl_retire);
 	INIT_WORK(&ref->work, active_work);
+#if IS_ENABLED(CONFIG_LOCKDEP)
+	lockdep_init_map(&ref->work.lockdep_map, "i915_active.work", wkey, 0);
+#endif
 }
 
 static bool ____active_del_barrier(struct i915_active *ref,
@@ -376,15 +395,8 @@ void i915_active_set_exclusive(struct i915_active *ref, struct dma_fence *f)
 	/* We expect the caller to manage the exclusive timeline ordering */
 	GEM_BUG_ON(i915_active_is_idle(ref));
 
-	/*
-	 * As we don't know which mutex the caller is using, we told a small
-	 * lie to the debug code that it is using the i915_active.mutex;
-	 * and now we must stick to that lie.
-	 */
-	mutex_acquire(&ref->mutex.dep_map, 0, 0, _THIS_IP_);
 	if (!__i915_active_fence_set(&ref->excl, f))
 		atomic_inc(&ref->count);
-	mutex_release(&ref->mutex.dep_map, _THIS_IP_);
 }
 
 bool i915_active_acquire_if_busy(struct i915_active *ref)
@@ -407,8 +419,10 @@ int i915_active_acquire(struct i915_active *ref)
 	if (!atomic_read(&ref->count) && ref->active)
 		err = ref->active(ref);
 	if (!err) {
+		spin_lock_irq(&ref->tree_lock); /* vs __active_retire() */
 		debug_active_activate(ref);
 		atomic_inc(&ref->count);
+		spin_unlock_irq(&ref->tree_lock);
 	}
 
 	mutex_unlock(&ref->mutex);
@@ -461,6 +475,7 @@ int i915_active_wait(struct i915_active *ref)
 	if (wait_var_event_interruptible(ref, i915_active_is_idle(ref)))
 		return -EINTR;
 
+	flush_work(&ref->work);
 	return 0;
 }
 
@@ -615,10 +630,6 @@ int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 				goto unwind;
 			}
 
-#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
-			node->base.lock =
-				&engine->kernel_context->timeline->mutex;
-#endif
 			RCU_INIT_POINTER(node->base.fence, NULL);
 			node->base.cb.func = node_retire;
 			node->timeline = idx;
@@ -639,6 +650,7 @@ int i915_active_acquire_preallocate_barrier(struct i915_active *ref,
 			node->base.cb.node.prev = (void *)engine;
 			atomic_inc(&ref->count);
 		}
+		GEM_BUG_ON(rcu_access_pointer(node->base.fence) != ERR_PTR(-EAGAIN));
 
 		GEM_BUG_ON(barrier_to_engine(node) != engine);
 		llist_add(barrier_to_ll(node), &ref->preallocated_barriers);
@@ -702,12 +714,18 @@ void i915_active_acquire_barrier(struct i915_active *ref)
 	}
 }
 
+static struct dma_fence **ll_to_fence_slot(struct llist_node *node)
+{
+	return __active_fence_slot(&barrier_from_ll(node)->base);
+}
+
 void i915_request_add_active_barriers(struct i915_request *rq)
 {
 	struct intel_engine_cs *engine = rq->engine;
 	struct llist_node *node, *next;
 	unsigned long flags;
 
+	GEM_BUG_ON(!intel_context_is_barrier(rq->context));
 	GEM_BUG_ON(intel_engine_is_virtual(engine));
 	GEM_BUG_ON(i915_request_timeline(rq) != engine->kernel_context->timeline);
 
@@ -721,19 +739,13 @@ void i915_request_add_active_barriers(struct i915_request *rq)
 	 */
 	spin_lock_irqsave(&rq->lock, flags);
 	llist_for_each_safe(node, next, node) {
-		RCU_INIT_POINTER(barrier_from_ll(node)->base.fence, &rq->fence);
-		smp_wmb(); /* serialise with reuse_idle_barrier */
+		/* serialise with reuse_idle_barrier */
+		smp_store_mb(*ll_to_fence_slot(node), &rq->fence);
 		list_add_tail((struct list_head *)node, &rq->fence.cb_list);
 	}
 	spin_unlock_irqrestore(&rq->lock, flags);
 }
 
-#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
-#define active_is_held(active) lockdep_is_held((active)->lock)
-#else
-#define active_is_held(active) true
-#endif
-
 /*
  * __i915_active_fence_set: Update the last active fence along its timeline
  * @active: the active tracker
@@ -744,7 +756,7 @@ void i915_request_add_active_barriers(struct i915_request *rq)
  * fence onto this one. Returns the previous fence (if not already completed),
  * which the caller must ensure is executed before the new fence. To ensure
  * that the order of fences within the timeline of the i915_active_fence is
- * maintained, it must be locked by the caller.
+ * understood, it should be locked by the caller.
  */
 struct dma_fence *
 __i915_active_fence_set(struct i915_active_fence *active,
@@ -753,34 +765,41 @@ __i915_active_fence_set(struct i915_active_fence *active,
 	struct dma_fence *prev;
 	unsigned long flags;
 
-	/* NB: must be serialised by an outer timeline mutex (active->lock) */
-	spin_lock_irqsave(fence->lock, flags);
+	if (fence == rcu_access_pointer(active->fence))
+		return fence;
+
 	GEM_BUG_ON(test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags));
 
-	prev = rcu_dereference_protected(active->fence, active_is_held(active));
+	/*
+	 * Consider that we have two threads arriving (A and B), with
+	 * C already resident as the active->fence.
+	 *
+	 * A does the xchg first, and so it sees C or NULL depending
+	 * on the timing of the interrupt handler. If it is NULL, the
+	 * previous fence must have been signaled and we know that
+	 * we are first on the timeline. If it is still present,
+	 * we acquire the lock on that fence and serialise with the interrupt
+	 * handler, in the process removing it from any future interrupt
+	 * callback. A will then wait on C before executing (if present).
+	 *
+	 * As B is second, it sees A as the previous fence and so waits for
+	 * it to complete its transition and takes over the occupancy for
+	 * itself -- remembering that it needs to wait on A before executing.
+	 *
+	 * Note the strong ordering of the timeline also provides consistent
+	 * nesting rules for the fence->lock; the inner lock is always the
+	 * older lock.
+	 */
+	spin_lock_irqsave(fence->lock, flags);
+	prev = xchg(__active_fence_slot(active), fence);
 	if (prev) {
 		GEM_BUG_ON(prev == fence);
 		spin_lock_nested(prev->lock, SINGLE_DEPTH_NESTING);
 		__list_del_entry(&active->cb.node);
 		spin_unlock(prev->lock); /* serialise with prev->cb_list */
-
-		/*
-		 * active->fence is reset by the callback from inside
-		 * interrupt context. We need to serialise our list
-		 * manipulation with the fence->lock to prevent the prev
-		 * being lost inside an interrupt (it can't be replaced as
-		 * no other caller is allowed to enter __i915_active_fence_set
-		 * as we hold the timeline lock). After serialising with
-		 * the callback, we need to double check which ran first,
-		 * our list_del() [decoupling prev from the callback] or
-		 * the callback...
-		 */
-		prev = rcu_access_pointer(active->fence);
 	}
-
-	rcu_assign_pointer(active->fence, fence);
+	GEM_BUG_ON(rcu_access_pointer(active->fence) != fence);
 	list_add_tail(&active->cb.node, &fence->cb_list);
-
 	spin_unlock_irqrestore(fence->lock, flags);
 
 	return prev;
@@ -792,10 +811,6 @@ int i915_active_fence_set(struct i915_active_fence *active,
 	struct dma_fence *fence;
 	int err = 0;
 
-#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
-	lockdep_assert_held(active->lock);
-#endif
-
 	/* Must maintain timeline ordering wrt previous active requests */
 	rcu_read_lock();
 	fence = __i915_active_fence_set(active, &rq->fence);
@@ -812,7 +827,7 @@ int i915_active_fence_set(struct i915_active_fence *active,
 
 void i915_active_noop(struct dma_fence *fence, struct dma_fence_cb *cb)
 {
-	i915_active_fence_cb(fence, cb);
+	active_fence_cb(fence, cb);
 }
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
diff --git a/drivers/gpu/drm/i915/i915_active.h b/drivers/gpu/drm/i915/i915_active.h
index 44859356ce97..b571f675c795 100644
--- a/drivers/gpu/drm/i915/i915_active.h
+++ b/drivers/gpu/drm/i915/i915_active.h
@@ -61,19 +61,15 @@ void i915_active_noop(struct dma_fence *fence, struct dma_fence_cb *cb);
  */
 static inline void
 __i915_active_fence_init(struct i915_active_fence *active,
-			 struct mutex *lock,
 			 void *fence,
 			 dma_fence_func_t fn)
 {
 	RCU_INIT_POINTER(active->fence, fence);
 	active->cb.func = fn ?: i915_active_noop;
-#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
-	active->lock = lock;
-#endif
 }
 
-#define INIT_ACTIVE_FENCE(A, LOCK) \
-	__i915_active_fence_init((A), (LOCK), NULL, NULL)
+#define INIT_ACTIVE_FENCE(A) \
+	__i915_active_fence_init((A), NULL, NULL)
 
 struct dma_fence *
 __i915_active_fence_set(struct i915_active_fence *active,
@@ -127,15 +123,6 @@ i915_active_fence_isset(const struct i915_active_fence *active)
 	return rcu_access_pointer(active->fence);
 }
 
-static inline void
-i915_active_fence_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
-{
-	struct i915_active_fence *active =
-		container_of(cb, typeof(*active), cb);
-
-	RCU_INIT_POINTER(active->fence, NULL);
-}
-
 /*
  * GPU activity tracking
  *
@@ -165,11 +152,15 @@ i915_active_fence_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
 void __i915_active_init(struct i915_active *ref,
 			int (*active)(struct i915_active *ref),
 			void (*retire)(struct i915_active *ref),
-			struct lock_class_key *key);
+			struct lock_class_key *mkey,
+			struct lock_class_key *wkey);
+
+/* Specialise each class of i915_active to avoid impossible lockdep cycles. */
 #define i915_active_init(ref, active, retire) do {		\
-	static struct lock_class_key __key;				\
+	static struct lock_class_key __mkey;				\
+	static struct lock_class_key __wkey;				\
 									\
-	__i915_active_init(ref, active, retire, &__key);		\
+	__i915_active_init(ref, active, retire, &__mkey, &__wkey);	\
 } while (0)
 
 int i915_active_ref(struct i915_active *ref,
@@ -215,5 +206,6 @@ void i915_active_acquire_barrier(struct i915_active *ref);
 void i915_request_add_active_barriers(struct i915_request *rq);
 
 void i915_active_print(struct i915_active *ref, struct drm_printer *m);
+void i915_active_unlock_wait(struct i915_active *ref);
 
 #endif /* _I915_ACTIVE_H_ */
diff --git a/drivers/gpu/drm/i915/i915_active_types.h b/drivers/gpu/drm/i915/i915_active_types.h
index 96aed0ee700a..6360c3e4b765 100644
--- a/drivers/gpu/drm/i915/i915_active_types.h
+++ b/drivers/gpu/drm/i915/i915_active_types.h
@@ -20,21 +20,6 @@
 struct i915_active_fence {
 	struct dma_fence __rcu *fence;
 	struct dma_fence_cb cb;
-#if IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM)
-	/*
-	 * Incorporeal!
-	 *
-	 * Updates to the i915_active_request must be serialised under a lock
-	 * to ensure that the timeline is ordered. Normally, this is the
-	 * timeline->mutex, but another mutex may be used so long as it is
-	 * done so consistently.
-	 *
-	 * For lockdep tracking of the above, we store the lock we intend
-	 * to always use for updates of this i915_active_request during
-	 * construction and assert that is held on every update.
-	 */
-	struct mutex *lock;
-#endif
 };
 
 struct active_node;
diff --git a/drivers/gpu/drm/i915/i915_cmd_parser.c b/drivers/gpu/drm/i915/i915_cmd_parser.c
index f24096e27bef..a0e437aa65b7 100644
--- a/drivers/gpu/drm/i915/i915_cmd_parser.c
+++ b/drivers/gpu/drm/i915/i915_cmd_parser.c
@@ -235,7 +235,7 @@ static const struct drm_i915_cmd_descriptor gen7_common_cmds[] = {
 	/*
 	 * MI_BATCH_BUFFER_START requires some special handling. It's not
 	 * really a 'skip' action but it doesn't seem like it's worth adding
-	 * a new action. See i915_parse_cmds().
+	 * a new action. See intel_engine_cmd_parser().
 	 */
 	CMD(  MI_BATCH_BUFFER_START,            SMI,   !F,  0xFF,   S  ),
 };
@@ -731,7 +731,7 @@ static u32 gen7_render_get_cmd_length_mask(u32 cmd_header)
 			return 0xFF;
 	}
 
-	DRM_DEBUG_DRIVER("CMD: Abnormal rcs cmd length! 0x%08X\n", cmd_header);
+	DRM_DEBUG("CMD: Abnormal rcs cmd length! 0x%08X\n", cmd_header);
 	return 0;
 }
 
@@ -754,7 +754,7 @@ static u32 gen7_bsd_get_cmd_length_mask(u32 cmd_header)
 			return 0xFF;
 	}
 
-	DRM_DEBUG_DRIVER("CMD: Abnormal bsd cmd length! 0x%08X\n", cmd_header);
+	DRM_DEBUG("CMD: Abnormal bsd cmd length! 0x%08X\n", cmd_header);
 	return 0;
 }
 
@@ -767,7 +767,7 @@ static u32 gen7_blt_get_cmd_length_mask(u32 cmd_header)
 	else if (client == INSTR_BC_CLIENT)
 		return 0xFF;
 
-	DRM_DEBUG_DRIVER("CMD: Abnormal blt cmd length! 0x%08X\n", cmd_header);
+	DRM_DEBUG("CMD: Abnormal blt cmd length! 0x%08X\n", cmd_header);
 	return 0;
 }
 
@@ -778,7 +778,7 @@ static u32 gen9_blt_get_cmd_length_mask(u32 cmd_header)
 	if (client == INSTR_MI_CLIENT || client == INSTR_BC_CLIENT)
 		return 0xFF;
 
-	DRM_DEBUG_DRIVER("CMD: Abnormal blt cmd length! 0x%08X\n", cmd_header);
+	DRM_DEBUG("CMD: Abnormal blt cmd length! 0x%08X\n", cmd_header);
 	return 0;
 }
 
@@ -1127,79 +1127,71 @@ find_reg(const struct intel_engine_cs *engine, u32 addr)
 /* Returns a vmap'd pointer to dst_obj, which the caller must unmap */
 static u32 *copy_batch(struct drm_i915_gem_object *dst_obj,
 		       struct drm_i915_gem_object *src_obj,
-		       u32 batch_start_offset,
-		       u32 batch_len,
-		       bool *needs_clflush_after)
+		       u32 offset, u32 length)
 {
-	unsigned int src_needs_clflush;
-	unsigned int dst_needs_clflush;
+	bool needs_clflush;
 	void *dst, *src;
 	int ret;
 
-	ret = i915_gem_object_prepare_write(dst_obj, &dst_needs_clflush);
-	if (ret)
-		return ERR_PTR(ret);
-
 	dst = i915_gem_object_pin_map(dst_obj, I915_MAP_FORCE_WB);
-	i915_gem_object_finish_access(dst_obj);
 	if (IS_ERR(dst))
 		return dst;
 
-	ret = i915_gem_object_prepare_read(src_obj, &src_needs_clflush);
+	ret = i915_gem_object_pin_pages(src_obj);
 	if (ret) {
 		i915_gem_object_unpin_map(dst_obj);
 		return ERR_PTR(ret);
 	}
 
+	needs_clflush =
+		!(src_obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ);
+
 	src = ERR_PTR(-ENODEV);
-	if (src_needs_clflush &&
-	    i915_can_memcpy_from_wc(NULL, batch_start_offset, 0)) {
+	if (needs_clflush && i915_has_memcpy_from_wc()) {
 		src = i915_gem_object_pin_map(src_obj, I915_MAP_WC);
 		if (!IS_ERR(src)) {
-			i915_memcpy_from_wc(dst,
-					    src + batch_start_offset,
-					    ALIGN(batch_len, 16));
+			i915_unaligned_memcpy_from_wc(dst,
+						      src + offset,
+						      length);
 			i915_gem_object_unpin_map(src_obj);
 		}
 	}
 	if (IS_ERR(src)) {
 		void *ptr;
-		int offset, n;
+		int x, n;
 
-		offset = offset_in_page(batch_start_offset);
-
-		/* We can avoid clflushing partial cachelines before the write
+		/*
+		 * We can avoid clflushing partial cachelines before the write
 		 * if we only every write full cache-lines. Since we know that
 		 * both the source and destination are in multiples of
 		 * PAGE_SIZE, we can simply round up to the next cacheline.
 		 * We don't care about copying too much here as we only
 		 * validate up to the end of the batch.
 		 */
-		if (dst_needs_clflush & CLFLUSH_BEFORE)
-			batch_len = roundup(batch_len,
-					    boot_cpu_data.x86_clflush_size);
+		if (!(dst_obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_READ))
+			length = round_up(length,
+					  boot_cpu_data.x86_clflush_size);
 
 		ptr = dst;
-		for (n = batch_start_offset >> PAGE_SHIFT; batch_len; n++) {
-			int len = min_t(int, batch_len, PAGE_SIZE - offset);
+		x = offset_in_page(offset);
+		for (n = offset >> PAGE_SHIFT; length; n++) {
+			int len = min_t(int, length, PAGE_SIZE - x);
 
 			src = kmap_atomic(i915_gem_object_get_page(src_obj, n));
-			if (src_needs_clflush)
-				drm_clflush_virt_range(src + offset, len);
-			memcpy(ptr, src + offset, len);
+			if (needs_clflush)
+				drm_clflush_virt_range(src + x, len);
+			memcpy(ptr, src + x, len);
 			kunmap_atomic(src);
 
 			ptr += len;
-			batch_len -= len;
-			offset = 0;
+			length -= len;
+			x = 0;
 		}
 	}
 
-	i915_gem_object_finish_access(src_obj);
+	i915_gem_object_unpin_pages(src_obj);
 
 	/* dst_obj is returned with vmap pinned */
-	*needs_clflush_after = dst_needs_clflush & CLFLUSH_AFTER;
-
 	return dst;
 }
 
@@ -1211,7 +1203,7 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 		return true;
 
 	if (desc->flags & CMD_DESC_REJECT) {
-		DRM_DEBUG_DRIVER("CMD: Rejected command: 0x%08X\n", *cmd);
+		DRM_DEBUG("CMD: Rejected command: 0x%08X\n", *cmd);
 		return false;
 	}
 
@@ -1231,8 +1223,8 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 				find_reg(engine, reg_addr);
 
 			if (!reg) {
-				DRM_DEBUG_DRIVER("CMD: Rejected register 0x%08X in command: 0x%08X (%s)\n",
-						 reg_addr, *cmd, engine->name);
+				DRM_DEBUG("CMD: Rejected register 0x%08X in command: 0x%08X (%s)\n",
+					  reg_addr, *cmd, engine->name);
 				return false;
 			}
 
@@ -1242,22 +1234,22 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 			 */
 			if (reg->mask) {
 				if (desc->cmd.value == MI_LOAD_REGISTER_MEM) {
-					DRM_DEBUG_DRIVER("CMD: Rejected LRM to masked register 0x%08X\n",
-							 reg_addr);
+					DRM_DEBUG("CMD: Rejected LRM to masked register 0x%08X\n",
+						  reg_addr);
 					return false;
 				}
 
 				if (desc->cmd.value == MI_LOAD_REGISTER_REG) {
-					DRM_DEBUG_DRIVER("CMD: Rejected LRR to masked register 0x%08X\n",
-							 reg_addr);
+					DRM_DEBUG("CMD: Rejected LRR to masked register 0x%08X\n",
+						  reg_addr);
 					return false;
 				}
 
 				if (desc->cmd.value == MI_LOAD_REGISTER_IMM(1) &&
 				    (offset + 2 > length ||
 				     (cmd[offset + 1] & reg->mask) != reg->value)) {
-					DRM_DEBUG_DRIVER("CMD: Rejected LRI to masked register 0x%08X\n",
-							 reg_addr);
+					DRM_DEBUG("CMD: Rejected LRI to masked register 0x%08X\n",
+						  reg_addr);
 					return false;
 				}
 			}
@@ -1284,8 +1276,8 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 			}
 
 			if (desc->bits[i].offset >= length) {
-				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X, too short to check bitmask (%s)\n",
-						 *cmd, engine->name);
+				DRM_DEBUG("CMD: Rejected command 0x%08X, too short to check bitmask (%s)\n",
+					  *cmd, engine->name);
 				return false;
 			}
 
@@ -1293,11 +1285,11 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 				desc->bits[i].mask;
 
 			if (dword != desc->bits[i].expected) {
-				DRM_DEBUG_DRIVER("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (%s)\n",
-						 *cmd,
-						 desc->bits[i].mask,
-						 desc->bits[i].expected,
-						 dword, engine->name);
+				DRM_DEBUG("CMD: Rejected command 0x%08X for bitmask 0x%08X (exp=0x%08X act=0x%08X) (%s)\n",
+					  *cmd,
+					  desc->bits[i].mask,
+					  desc->bits[i].expected,
+					  dword, engine->name);
 				return false;
 			}
 		}
@@ -1306,17 +1298,17 @@ static bool check_cmd(const struct intel_engine_cs *engine,
 	return true;
 }
 
-static int check_bbstart(const struct i915_gem_context *ctx,
-			 u32 *cmd, u32 offset, u32 length,
-			 u32 batch_len,
-			 u64 batch_start,
-			 u64 shadow_batch_start)
+static int check_bbstart(u32 *cmd, u32 offset, u32 length,
+			 u32 batch_length,
+			 u64 batch_addr,
+			 u64 shadow_addr,
+			 const unsigned long *jump_whitelist)
 {
 	u64 jump_offset, jump_target;
 	u32 target_cmd_offset, target_cmd_index;
 
 	/* For igt compatibility on older platforms */
-	if (CMDPARSER_USES_GGTT(ctx->i915)) {
+	if (!jump_whitelist) {
 		DRM_DEBUG("CMD: Rejecting BB_START for ggtt based submission\n");
 		return -EACCES;
 	}
@@ -1327,14 +1319,14 @@ static int check_bbstart(const struct i915_gem_context *ctx,
 		return -EINVAL;
 	}
 
-	jump_target = *(u64*)(cmd+1);
-	jump_offset = jump_target - batch_start;
+	jump_target = *(u64 *)(cmd + 1);
+	jump_offset = jump_target - batch_addr;
 
 	/*
 	 * Any underflow of jump_target is guaranteed to be outside the range
 	 * of a u32, so >= test catches both too large and too small
 	 */
-	if (jump_offset >= batch_len) {
+	if (jump_offset >= batch_length) {
 		DRM_DEBUG("CMD: BB_START to 0x%llx jumps out of BB\n",
 			  jump_target);
 		return -EINVAL;
@@ -1342,20 +1334,20 @@ static int check_bbstart(const struct i915_gem_context *ctx,
 
 	/*
 	 * This cannot overflow a u32 because we already checked jump_offset
-	 * is within the BB, and the batch_len is a u32
+	 * is within the BB, and the batch_length is a u32
 	 */
 	target_cmd_offset = lower_32_bits(jump_offset);
 	target_cmd_index = target_cmd_offset / sizeof(u32);
 
-	*(u64*)(cmd + 1) = shadow_batch_start + target_cmd_offset;
+	*(u64 *)(cmd + 1) = shadow_addr + target_cmd_offset;
 
 	if (target_cmd_index == offset)
 		return 0;
 
-	if (ctx->jump_whitelist_cmds <= target_cmd_index) {
-		DRM_DEBUG("CMD: Rejecting BB_START - truncated whitelist array\n");
-		return -EINVAL;
-	} else if (!test_bit(target_cmd_index, ctx->jump_whitelist)) {
+	if (IS_ERR(jump_whitelist))
+		return PTR_ERR(jump_whitelist);
+
+	if (!test_bit(target_cmd_index, jump_whitelist)) {
 		DRM_DEBUG("CMD: BB_START to 0x%llx not a previously executed cmd\n",
 			  jump_target);
 		return -EINVAL;
@@ -1364,54 +1356,40 @@ static int check_bbstart(const struct i915_gem_context *ctx,
 	return 0;
 }
 
-static void init_whitelist(struct i915_gem_context *ctx, u32 batch_len)
+static unsigned long *alloc_whitelist(u32 batch_length)
 {
-	const u32 batch_cmds = DIV_ROUND_UP(batch_len, sizeof(u32));
-	const u32 exact_size = BITS_TO_LONGS(batch_cmds);
-	u32 next_size = BITS_TO_LONGS(roundup_pow_of_two(batch_cmds));
-	unsigned long *next_whitelist;
+	unsigned long *jmp;
 
-	if (CMDPARSER_USES_GGTT(ctx->i915))
-		return;
-
-	if (batch_cmds <= ctx->jump_whitelist_cmds) {
-		bitmap_zero(ctx->jump_whitelist, batch_cmds);
-		return;
-	}
-
-again:
-	next_whitelist = kcalloc(next_size, sizeof(long), GFP_KERNEL);
-	if (next_whitelist) {
-		kfree(ctx->jump_whitelist);
-		ctx->jump_whitelist = next_whitelist;
-		ctx->jump_whitelist_cmds =
-			next_size * BITS_PER_BYTE * sizeof(long);
-		return;
-	}
-
-	if (next_size > exact_size) {
-		next_size = exact_size;
-		goto again;
-	}
+	/*
+	 * We expect batch_length to be less than 256KiB for known users,
+	 * i.e. we need at most an 8KiB bitmap allocation which should be
+	 * reasonably cheap due to kmalloc caches.
+	 */
 
-	DRM_DEBUG("CMD: Failed to extend whitelist. BB_START may be disallowed\n");
-	bitmap_zero(ctx->jump_whitelist, ctx->jump_whitelist_cmds);
+	/* Prefer to report transient allocation failure rather than hit oom */
+	jmp = bitmap_zalloc(DIV_ROUND_UP(batch_length, sizeof(u32)),
+			    GFP_KERNEL | __GFP_RETRY_MAYFAIL | __GFP_NOWARN);
+	if (!jmp)
+		return ERR_PTR(-ENOMEM);
 
-	return;
+	return jmp;
 }
 
 #define LENGTH_BIAS 2
 
+static bool shadow_needs_clflush(struct drm_i915_gem_object *obj)
+{
+	return !(obj->cache_coherent & I915_BO_CACHE_COHERENT_FOR_WRITE);
+}
+
 /**
- * i915_parse_cmds() - parse a submitted batch buffer for privilege violations
- * @ctx: the context in which the batch is to execute
+ * intel_engine_cmd_parser() - parse a batch buffer for privilege violations
  * @engine: the engine on which the batch is to execute
- * @batch_obj: the batch buffer in question
- * @batch_start: Canonical base address of batch
- * @batch_start_offset: byte offset in the batch at which execution starts
- * @batch_len: length of the commands in batch_obj
- * @shadow_batch_obj: copy of the batch buffer in question
- * @shadow_batch_start: Canonical base address of shadow_batch_obj
+ * @batch: the batch buffer in question
+ * @batch_offset: byte offset in the batch at which execution starts
+ * @batch_length: length of the commands in batch_obj
+ * @shadow: validated copy of the batch buffer in question
+ * @trampoline: whether to emit a conditional trampoline at the end of the batch
  *
  * Parses the specified batch buffer looking for privilege violations as
  * described in the overview.
@@ -1419,38 +1397,46 @@ again:
  * Return: non-zero if the parser finds violations or otherwise fails; -EACCES
  * if the batch appears legal but should use hardware parsing
  */
-
-int intel_engine_cmd_parser(struct i915_gem_context *ctx,
-			    struct intel_engine_cs *engine,
-			    struct drm_i915_gem_object *batch_obj,
-			    u64 batch_start,
-			    u32 batch_start_offset,
-			    u32 batch_len,
-			    struct drm_i915_gem_object *shadow_batch_obj,
-			    u64 shadow_batch_start)
+int intel_engine_cmd_parser(struct intel_engine_cs *engine,
+			    struct i915_vma *batch,
+			    u32 batch_offset,
+			    u32 batch_length,
+			    struct i915_vma *shadow,
+			    bool trampoline)
 {
 	u32 *cmd, *batch_end, offset = 0;
 	struct drm_i915_cmd_descriptor default_desc = noop_desc;
 	const struct drm_i915_cmd_descriptor *desc = &default_desc;
-	bool needs_clflush_after = false;
+	unsigned long *jump_whitelist;
+	u64 batch_addr, shadow_addr;
 	int ret = 0;
 
-	cmd = copy_batch(shadow_batch_obj, batch_obj,
-			 batch_start_offset, batch_len,
-			 &needs_clflush_after);
+	GEM_BUG_ON(!IS_ALIGNED(batch_offset, sizeof(*cmd)));
+	GEM_BUG_ON(!IS_ALIGNED(batch_length, sizeof(*cmd)));
+	GEM_BUG_ON(range_overflows_t(u64, batch_offset, batch_length,
+				     batch->size));
+	GEM_BUG_ON(!batch_length);
+
+	cmd = copy_batch(shadow->obj, batch->obj, batch_offset, batch_length);
 	if (IS_ERR(cmd)) {
-		DRM_DEBUG_DRIVER("CMD: Failed to copy batch\n");
+		DRM_DEBUG("CMD: Failed to copy batch\n");
 		return PTR_ERR(cmd);
 	}
 
-	init_whitelist(ctx, batch_len);
+	jump_whitelist = NULL;
+	if (!trampoline)
+		/* Defer failure until attempted use */
+		jump_whitelist = alloc_whitelist(batch_length);
+
+	shadow_addr = gen8_canonical_addr(shadow->node.start);
+	batch_addr = gen8_canonical_addr(batch->node.start + batch_offset);
 
 	/*
 	 * We use the batch length as size because the shadow object is as
 	 * large or larger and copy_batch() will write MI_NOPs to the extra
 	 * space. Parsing should be faster in some cases this way.
 	 */
-	batch_end = cmd + (batch_len / sizeof(*batch_end));
+	batch_end = cmd + batch_length / sizeof(*batch_end);
 	do {
 		u32 length;
 
@@ -1459,61 +1445,99 @@ int intel_engine_cmd_parser(struct i915_gem_context *ctx,
 
 		desc = find_cmd(engine, *cmd, desc, &default_desc);
 		if (!desc) {
-			DRM_DEBUG_DRIVER("CMD: Unrecognized command: 0x%08X\n",
-					 *cmd);
+			DRM_DEBUG("CMD: Unrecognized command: 0x%08X\n", *cmd);
 			ret = -EINVAL;
-			goto err;
+			break;
 		}
 
 		if (desc->flags & CMD_DESC_FIXED)
 			length = desc->length.fixed;
 		else
-			length = ((*cmd & desc->length.mask) + LENGTH_BIAS);
+			length = (*cmd & desc->length.mask) + LENGTH_BIAS;
 
 		if ((batch_end - cmd) < length) {
-			DRM_DEBUG_DRIVER("CMD: Command length exceeds batch length: 0x%08X length=%u batchlen=%td\n",
-					 *cmd,
-					 length,
-					 batch_end - cmd);
+			DRM_DEBUG("CMD: Command length exceeds batch length: 0x%08X length=%u batchlen=%td\n",
+				  *cmd,
+				  length,
+				  batch_end - cmd);
 			ret = -EINVAL;
-			goto err;
+			break;
 		}
 
 		if (!check_cmd(engine, desc, cmd, length)) {
 			ret = -EACCES;
-			goto err;
+			break;
 		}
 
 		if (desc->cmd.value == MI_BATCH_BUFFER_START) {
-			ret = check_bbstart(ctx, cmd, offset, length,
-					    batch_len, batch_start,
-					    shadow_batch_start);
-
-			if (ret)
-				goto err;
+			ret = check_bbstart(cmd, offset, length, batch_length,
+					    batch_addr, shadow_addr,
+					    jump_whitelist);
 			break;
 		}
 
-		if (ctx->jump_whitelist_cmds > offset)
-			set_bit(offset, ctx->jump_whitelist);
+		if (!IS_ERR_OR_NULL(jump_whitelist))
+			__set_bit(offset, jump_whitelist);
 
 		cmd += length;
 		offset += length;
 		if  (cmd >= batch_end) {
-			DRM_DEBUG_DRIVER("CMD: Got to the end of the buffer w/o a BBE cmd!\n");
+			DRM_DEBUG("CMD: Got to the end of the buffer w/o a BBE cmd!\n");
 			ret = -EINVAL;
-			goto err;
+			break;
 		}
 	} while (1);
 
-	if (needs_clflush_after) {
-		void *ptr = page_mask_bits(shadow_batch_obj->mm.mapping);
+	if (trampoline) {
+		/*
+		 * With the trampoline, the shadow is executed twice.
+		 *
+		 *   1 - starting at offset 0, in privileged mode
+		 *   2 - starting at offset batch_len, as non-privileged
+		 *
+		 * Only if the batch is valid and safe to execute, do we
+		 * allow the first privileged execution to proceed. If not,
+		 * we terminate the first batch and use the second batchbuffer
+		 * entry to chain to the original unsafe non-privileged batch,
+		 * leaving it to the HW to validate.
+		 */
+		*batch_end = MI_BATCH_BUFFER_END;
+
+		if (ret) {
+			/* Batch unsafe to execute with privileges, cancel! */
+			cmd = page_mask_bits(shadow->obj->mm.mapping);
+			*cmd = MI_BATCH_BUFFER_END;
+
+			/* If batch is unsafe but valid, jump to the original */
+			if (ret == -EACCES) {
+				unsigned int flags;
+
+				flags = MI_BATCH_NON_SECURE_I965;
+				if (IS_HASWELL(engine->i915))
+					flags = MI_BATCH_NON_SECURE_HSW;
+
+				GEM_BUG_ON(!IS_GEN_RANGE(engine->i915, 6, 7));
+				__gen6_emit_bb_start(batch_end,
+						     batch_addr,
+						     flags);
+
+				ret = 0; /* allow execution */
+			}
+		}
+
+		if (shadow_needs_clflush(shadow->obj))
+			drm_clflush_virt_range(batch_end, 8);
+	}
+
+	if (shadow_needs_clflush(shadow->obj)) {
+		void *ptr = page_mask_bits(shadow->obj->mm.mapping);
 
 		drm_clflush_virt_range(ptr, (void *)(cmd + 1) - ptr);
 	}
 
-err:
-	i915_gem_object_unpin_map(shadow_batch_obj);
+	if (!IS_ERR_OR_NULL(jump_whitelist))
+		kfree(jump_whitelist);
+	i915_gem_object_unpin_map(shadow->obj);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c
index 8016484ebcd3..d28468eaed57 100644
--- a/drivers/gpu/drm/i915/i915_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_debugfs.c
@@ -61,24 +61,14 @@ static inline struct drm_i915_private *node_to_i915(struct drm_info_node *node)
 
 static int i915_capabilities(struct seq_file *m, void *data)
 {
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	const struct intel_device_info *info = INTEL_INFO(dev_priv);
+	struct drm_i915_private *i915 = node_to_i915(m->private);
 	struct drm_printer p = drm_seq_file_printer(m);
-	const char *msg;
 
-	seq_printf(m, "gen: %d\n", INTEL_GEN(dev_priv));
-	seq_printf(m, "platform: %s\n", intel_platform_name(info->platform));
-	seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(dev_priv));
+	seq_printf(m, "pch: %d\n", INTEL_PCH_TYPE(i915));
 
-	msg = "n/a";
-#ifdef CONFIG_INTEL_IOMMU
-	msg = enableddisabled(intel_iommu_gfx_mapped);
-#endif
-	seq_printf(m, "iommu: %s\n", msg);
-
-	intel_device_info_dump_flags(info, &p);
-	intel_device_info_dump_runtime(RUNTIME_INFO(dev_priv), &p);
-	intel_driver_caps_print(&dev_priv->caps, &p);
+	intel_device_info_print_static(INTEL_INFO(i915), &p);
+	intel_device_info_print_runtime(RUNTIME_INFO(i915), &p);
+	intel_driver_caps_print(&i915->caps, &p);
 
 	kernel_param_lock(THIS_MODULE);
 	i915_params_dump(&i915_modparams, &p);
@@ -891,7 +881,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused)
 		rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK;
 		rpcurdown = I915_READ(GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK;
 		rpprevdown = I915_READ(GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK;
-		cagf = intel_gpu_freq(rps, intel_get_cagf(rps, rpstat));
+		cagf = intel_rps_read_actual_frequency(rps);
 
 		intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
 
@@ -1633,21 +1623,6 @@ static int i915_rps_boost_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
 	struct intel_rps *rps = &dev_priv->gt.rps;
-	u32 act_freq = rps->cur_freq;
-	intel_wakeref_t wakeref;
-
-	with_intel_runtime_pm_if_in_use(&dev_priv->runtime_pm, wakeref) {
-		if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
-			vlv_punit_get(dev_priv);
-			act_freq = vlv_punit_read(dev_priv,
-						  PUNIT_REG_GPU_FREQ_STS);
-			vlv_punit_put(dev_priv);
-			act_freq = (act_freq >> 8) & 0xff;
-		} else {
-			act_freq = intel_get_cagf(rps,
-						  I915_READ(GEN6_RPSTAT1));
-		}
-	}
 
 	seq_printf(m, "RPS enabled? %d\n", rps->enabled);
 	seq_printf(m, "GPU busy? %s\n", yesno(dev_priv->gt.awake));
@@ -1656,7 +1631,7 @@ static int i915_rps_boost_info(struct seq_file *m, void *data)
 	seq_printf(m, "Interactive? %d\n", READ_ONCE(rps->power.interactive));
 	seq_printf(m, "Frequency requested %d, actual %d\n",
 		   intel_gpu_freq(rps, rps->cur_freq),
-		   intel_gpu_freq(rps, act_freq));
+		   intel_rps_read_actual_frequency(rps));
 	seq_printf(m, "  min hard:%d, soft:%d; max soft:%d, hard:%d\n",
 		   intel_gpu_freq(rps, rps->min_freq),
 		   intel_gpu_freq(rps, rps->min_freq_softlimit),
@@ -1802,30 +1777,12 @@ static void i915_guc_log_info(struct seq_file *m,
 static int i915_guc_info(struct seq_file *m, void *data)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	const struct intel_guc *guc = &dev_priv->gt.uc.guc;
-	struct intel_guc_client *client = guc->execbuf_client;
 
 	if (!USES_GUC(dev_priv))
 		return -ENODEV;
 
 	i915_guc_log_info(m, dev_priv);
 
-	if (!USES_GUC_SUBMISSION(dev_priv))
-		return 0;
-
-	GEM_BUG_ON(!guc->execbuf_client);
-
-	seq_printf(m, "\nDoorbell map:\n");
-	seq_printf(m, "\t%*pb\n", GUC_NUM_DOORBELLS, guc->doorbell_bitmap);
-	seq_printf(m, "Doorbell next cacheline: 0x%x\n", guc->db_cacheline);
-
-	seq_printf(m, "\nGuC execbuf client @ %p:\n", client);
-	seq_printf(m, "\tPriority %d, GuC stage index: %u, PD offset 0x%x\n",
-		   client->priority,
-		   client->stage_id,
-		   client->proc_desc_offset);
-	seq_printf(m, "\tDoorbell id %d, offset: 0x%lx\n",
-		   client->doorbell_id, client->doorbell_offset);
 	/* Add more as required ... */
 
 	return 0;
@@ -2367,7 +2324,7 @@ out:
 }
 
 static void intel_seq_print_mode(struct seq_file *m, int tabs,
-				 struct drm_display_mode *mode)
+				 const struct drm_display_mode *mode)
 {
 	int i;
 
@@ -2378,59 +2335,35 @@ static void intel_seq_print_mode(struct seq_file *m, int tabs,
 }
 
 static void intel_encoder_info(struct seq_file *m,
-			       struct intel_crtc *intel_crtc,
-			       struct intel_encoder *intel_encoder)
+			       struct intel_crtc *crtc,
+			       struct intel_encoder *encoder)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
-	struct drm_crtc *crtc = &intel_crtc->base;
-	struct intel_connector *intel_connector;
-	struct drm_encoder *encoder;
-
-	encoder = &intel_encoder->base;
-	seq_printf(m, "\tencoder %d: type: %s, connectors:\n",
-		   encoder->base.id, encoder->name);
-	for_each_connector_on_encoder(dev, encoder, intel_connector) {
-		struct drm_connector *connector = &intel_connector->base;
-		seq_printf(m, "\t\tconnector %d: type: %s, status: %s",
-			   connector->base.id,
-			   connector->name,
-			   drm_get_connector_status_name(connector->status));
-		if (connector->status == connector_status_connected) {
-			struct drm_display_mode *mode = &crtc->mode;
-			seq_printf(m, ", mode:\n");
-			intel_seq_print_mode(m, 2, mode);
-		} else {
-			seq_putc(m, '\n');
-		}
-	}
-}
+	struct drm_connector_list_iter conn_iter;
+	struct drm_connector *connector;
 
-static void intel_crtc_info(struct seq_file *m, struct intel_crtc *intel_crtc)
-{
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
-	struct drm_crtc *crtc = &intel_crtc->base;
-	struct intel_encoder *intel_encoder;
-	struct drm_plane_state *plane_state = crtc->primary->state;
-	struct drm_framebuffer *fb = plane_state->fb;
+	seq_printf(m, "\t[ENCODER:%d:%s]: connectors:\n",
+		   encoder->base.base.id, encoder->base.name);
 
-	if (fb)
-		seq_printf(m, "\tfb: %d, pos: %dx%d, size: %dx%d\n",
-			   fb->base.id, plane_state->src_x >> 16,
-			   plane_state->src_y >> 16, fb->width, fb->height);
-	else
-		seq_puts(m, "\tprimary plane disabled\n");
-	for_each_encoder_on_crtc(dev, crtc, intel_encoder)
-		intel_encoder_info(m, intel_crtc, intel_encoder);
+	drm_connector_list_iter_begin(&dev_priv->drm, &conn_iter);
+	drm_for_each_connector_iter(connector, &conn_iter) {
+		const struct drm_connector_state *conn_state =
+			connector->state;
+
+		if (conn_state->best_encoder != &encoder->base)
+			continue;
+
+		seq_printf(m, "\t\t[CONNECTOR:%d:%s]\n",
+			   connector->base.id, connector->name);
+	}
+	drm_connector_list_iter_end(&conn_iter);
 }
 
 static void intel_panel_info(struct seq_file *m, struct intel_panel *panel)
 {
-	struct drm_display_mode *mode = panel->fixed_mode;
+	const struct drm_display_mode *mode = panel->fixed_mode;
 
-	seq_printf(m, "\tfixed mode:\n");
-	intel_seq_print_mode(m, 2, mode);
+	seq_printf(m, "\tfixed mode: " DRM_MODE_FMT "\n", DRM_MODE_ARG(mode));
 }
 
 static void intel_hdcp_info(struct seq_file *m,
@@ -2508,10 +2441,12 @@ static void intel_connector_info(struct seq_file *m,
 				 struct drm_connector *connector)
 {
 	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct intel_encoder *intel_encoder = intel_connector->encoder;
-	struct drm_display_mode *mode;
+	const struct drm_connector_state *conn_state = connector->state;
+	struct intel_encoder *encoder =
+		to_intel_encoder(conn_state->best_encoder);
+	const struct drm_display_mode *mode;
 
-	seq_printf(m, "connector %d: type %s, status: %s\n",
+	seq_printf(m, "[CONNECTOR:%d:%s]: status: %s\n",
 		   connector->base.id, connector->name,
 		   drm_get_connector_status_name(connector->status));
 
@@ -2525,24 +2460,24 @@ static void intel_connector_info(struct seq_file *m,
 		   drm_get_subpixel_order_name(connector->display_info.subpixel_order));
 	seq_printf(m, "\tCEA rev: %d\n", connector->display_info.cea_rev);
 
-	if (!intel_encoder)
+	if (!encoder)
 		return;
 
 	switch (connector->connector_type) {
 	case DRM_MODE_CONNECTOR_DisplayPort:
 	case DRM_MODE_CONNECTOR_eDP:
-		if (intel_encoder->type == INTEL_OUTPUT_DP_MST)
+		if (encoder->type == INTEL_OUTPUT_DP_MST)
 			intel_dp_mst_info(m, intel_connector);
 		else
 			intel_dp_info(m, intel_connector);
 		break;
 	case DRM_MODE_CONNECTOR_LVDS:
-		if (intel_encoder->type == INTEL_OUTPUT_LVDS)
+		if (encoder->type == INTEL_OUTPUT_LVDS)
 			intel_lvds_info(m, intel_connector);
 		break;
 	case DRM_MODE_CONNECTOR_HDMIA:
-		if (intel_encoder->type == INTEL_OUTPUT_HDMI ||
-		    intel_encoder->type == INTEL_OUTPUT_DDI)
+		if (encoder->type == INTEL_OUTPUT_HDMI ||
+		    encoder->type == INTEL_OUTPUT_DDI)
 			intel_hdmi_info(m, intel_connector);
 		break;
 	default:
@@ -2589,70 +2524,88 @@ static void plane_rotation(char *buf, size_t bufsize, unsigned int rotation)
 		 rotation);
 }
 
-static void intel_plane_info(struct seq_file *m, struct intel_crtc *intel_crtc)
+static void intel_plane_uapi_info(struct seq_file *m, struct intel_plane *plane)
 {
-	struct drm_i915_private *dev_priv = node_to_i915(m->private);
-	struct drm_device *dev = &dev_priv->drm;
-	struct intel_plane *intel_plane;
+	const struct intel_plane_state *plane_state =
+		to_intel_plane_state(plane->base.state);
+	const struct drm_framebuffer *fb = plane_state->uapi.fb;
+	struct drm_format_name_buf format_name;
+	struct drm_rect src, dst;
+	char rot_str[48];
 
-	for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane) {
-		struct drm_plane_state *state;
-		struct drm_plane *plane = &intel_plane->base;
-		struct drm_format_name_buf format_name;
-		char rot_str[48];
+	src = drm_plane_state_src(&plane_state->uapi);
+	dst = drm_plane_state_dest(&plane_state->uapi);
 
-		if (!plane->state) {
-			seq_puts(m, "plane->state is NULL!\n");
-			continue;
-		}
+	if (fb)
+		drm_get_format_name(fb->format->format, &format_name);
 
-		state = plane->state;
+	plane_rotation(rot_str, sizeof(rot_str),
+		       plane_state->uapi.rotation);
 
-		if (state->fb) {
-			drm_get_format_name(state->fb->format->format,
-					    &format_name);
-		} else {
-			sprintf(format_name.str, "N/A");
-		}
+	seq_printf(m, "\t\tuapi: fb=%d,%s,%dx%d, src=" DRM_RECT_FP_FMT ", dst=" DRM_RECT_FMT ", rotation=%s\n",
+		   fb ? fb->base.id : 0, fb ? format_name.str : "n/a",
+		   fb ? fb->width : 0, fb ? fb->height : 0,
+		   DRM_RECT_FP_ARG(&src),
+		   DRM_RECT_ARG(&dst),
+		   rot_str);
+}
+
+static void intel_plane_hw_info(struct seq_file *m, struct intel_plane *plane)
+{
+	const struct intel_plane_state *plane_state =
+		to_intel_plane_state(plane->base.state);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
+	struct drm_format_name_buf format_name;
+	char rot_str[48];
+
+	if (!fb)
+		return;
+
+	drm_get_format_name(fb->format->format, &format_name);
 
-		plane_rotation(rot_str, sizeof(rot_str), state->rotation);
-
-		seq_printf(m, "\t--Plane id %d: type=%s, crtc_pos=%4dx%4d, crtc_size=%4dx%4d, src_pos=%d.%04ux%d.%04u, src_size=%d.%04ux%d.%04u, format=%s, rotation=%s\n",
-			   plane->base.id,
-			   plane_type(intel_plane->base.type),
-			   state->crtc_x, state->crtc_y,
-			   state->crtc_w, state->crtc_h,
-			   (state->src_x >> 16),
-			   ((state->src_x & 0xffff) * 15625) >> 10,
-			   (state->src_y >> 16),
-			   ((state->src_y & 0xffff) * 15625) >> 10,
-			   (state->src_w >> 16),
-			   ((state->src_w & 0xffff) * 15625) >> 10,
-			   (state->src_h >> 16),
-			   ((state->src_h & 0xffff) * 15625) >> 10,
-			   format_name.str,
-			   rot_str);
+	plane_rotation(rot_str, sizeof(rot_str),
+		       plane_state->hw.rotation);
+
+	seq_printf(m, "\t\thw: fb=%d,%s,%dx%d, visible=%s, src=" DRM_RECT_FP_FMT ", dst=" DRM_RECT_FMT ", rotation=%s\n",
+		   fb->base.id, format_name.str,
+		   fb->width, fb->height,
+		   yesno(plane_state->uapi.visible),
+		   DRM_RECT_FP_ARG(&plane_state->uapi.src),
+		   DRM_RECT_ARG(&plane_state->uapi.dst),
+		   rot_str);
+}
+
+static void intel_plane_info(struct seq_file *m, struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = node_to_i915(m->private);
+	struct intel_plane *plane;
+
+	for_each_intel_plane_on_crtc(&dev_priv->drm, crtc, plane) {
+		seq_printf(m, "\t[PLANE:%d:%s]: type=%s\n",
+			   plane->base.base.id, plane->base.name,
+			   plane_type(plane->base.type));
+		intel_plane_uapi_info(m, plane);
+		intel_plane_hw_info(m, plane);
 	}
 }
 
-static void intel_scaler_info(struct seq_file *m, struct intel_crtc *intel_crtc)
+static void intel_scaler_info(struct seq_file *m, struct intel_crtc *crtc)
 {
-	struct intel_crtc_state *pipe_config;
-	int num_scalers = intel_crtc->num_scalers;
+	const struct intel_crtc_state *crtc_state =
+		to_intel_crtc_state(crtc->base.state);
+	int num_scalers = crtc->num_scalers;
 	int i;
 
-	pipe_config = to_intel_crtc_state(intel_crtc->base.state);
-
 	/* Not all platformas have a scaler */
 	if (num_scalers) {
 		seq_printf(m, "\tnum_scalers=%d, scaler_users=%x scaler_id=%d",
 			   num_scalers,
-			   pipe_config->scaler_state.scaler_users,
-			   pipe_config->scaler_state.scaler_id);
+			   crtc_state->scaler_state.scaler_users,
+			   crtc_state->scaler_state.scaler_id);
 
 		for (i = 0; i < num_scalers; i++) {
-			struct intel_scaler *sc =
-					&pipe_config->scaler_state.scalers[i];
+			const struct intel_scaler *sc =
+				&crtc_state->scaler_state.scalers[i];
 
 			seq_printf(m, ", scalers[%d]: use=%s, mode=%x",
 				   i, yesno(sc->in_use), sc->mode);
@@ -2663,6 +2616,44 @@ static void intel_scaler_info(struct seq_file *m, struct intel_crtc *intel_crtc)
 	}
 }
 
+static void intel_crtc_info(struct seq_file *m, struct intel_crtc *crtc)
+{
+	struct drm_i915_private *dev_priv = node_to_i915(m->private);
+	const struct intel_crtc_state *crtc_state =
+		to_intel_crtc_state(crtc->base.state);
+	struct intel_encoder *encoder;
+
+	seq_printf(m, "[CRTC:%d:%s]:\n",
+		   crtc->base.base.id, crtc->base.name);
+
+	seq_printf(m, "\tuapi: enable=%s, active=%s, mode=" DRM_MODE_FMT "\n",
+		   yesno(crtc_state->uapi.enable),
+		   yesno(crtc_state->uapi.active),
+		   DRM_MODE_ARG(&crtc_state->uapi.mode));
+
+	if (crtc_state->hw.enable) {
+		seq_printf(m, "\thw: active=%s, adjusted_mode=" DRM_MODE_FMT "\n",
+			   yesno(crtc_state->hw.active),
+			   DRM_MODE_ARG(&crtc_state->hw.adjusted_mode));
+
+		seq_printf(m, "\tpipe src size=%dx%d, dither=%s, bpp=%d\n",
+			   crtc_state->pipe_src_w, crtc_state->pipe_src_h,
+			   yesno(crtc_state->dither), crtc_state->pipe_bpp);
+
+		intel_scaler_info(m, crtc);
+	}
+
+	for_each_intel_encoder_mask(&dev_priv->drm, encoder,
+				    crtc_state->uapi.encoder_mask)
+		intel_encoder_info(m, crtc, encoder);
+
+	intel_plane_info(m, crtc);
+
+	seq_printf(m, "\tunderrun reporting: cpu=%s pch=%s\n",
+		   yesno(!crtc->cpu_fifo_underrun_disabled),
+		   yesno(!crtc->pch_fifo_underrun_disabled));
+}
+
 static int i915_display_info(struct seq_file *m, void *unused)
 {
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
@@ -2674,52 +2665,22 @@ static int i915_display_info(struct seq_file *m, void *unused)
 
 	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
 
+	drm_modeset_lock_all(dev);
+
 	seq_printf(m, "CRTC info\n");
 	seq_printf(m, "---------\n");
-	for_each_intel_crtc(dev, crtc) {
-		struct intel_crtc_state *pipe_config;
-
-		drm_modeset_lock(&crtc->base.mutex, NULL);
-		pipe_config = to_intel_crtc_state(crtc->base.state);
-
-		seq_printf(m, "CRTC %d: pipe: %c, active=%s, (size=%dx%d), dither=%s, bpp=%d\n",
-			   crtc->base.base.id, pipe_name(crtc->pipe),
-			   yesno(pipe_config->base.active),
-			   pipe_config->pipe_src_w, pipe_config->pipe_src_h,
-			   yesno(pipe_config->dither), pipe_config->pipe_bpp);
-
-		if (pipe_config->base.active) {
-			struct intel_plane *cursor =
-				to_intel_plane(crtc->base.cursor);
-
-			intel_crtc_info(m, crtc);
-
-			seq_printf(m, "\tcursor visible? %s, position (%d, %d), size %dx%d, addr 0x%08x\n",
-				   yesno(cursor->base.state->visible),
-				   cursor->base.state->crtc_x,
-				   cursor->base.state->crtc_y,
-				   cursor->base.state->crtc_w,
-				   cursor->base.state->crtc_h,
-				   cursor->cursor.base);
-			intel_scaler_info(m, crtc);
-			intel_plane_info(m, crtc);
-		}
-
-		seq_printf(m, "\tunderrun reporting: cpu=%s pch=%s \n",
-			   yesno(!crtc->cpu_fifo_underrun_disabled),
-			   yesno(!crtc->pch_fifo_underrun_disabled));
-		drm_modeset_unlock(&crtc->base.mutex);
-	}
+	for_each_intel_crtc(dev, crtc)
+		intel_crtc_info(m, crtc);
 
 	seq_printf(m, "\n");
 	seq_printf(m, "Connector info\n");
 	seq_printf(m, "--------------\n");
-	mutex_lock(&dev->mode_config.mutex);
 	drm_connector_list_iter_begin(dev, &conn_iter);
 	drm_for_each_connector_iter(connector, &conn_iter)
 		intel_connector_info(m, connector);
 	drm_connector_list_iter_end(&conn_iter);
-	mutex_unlock(&dev->mode_config.mutex);
+
+	drm_modeset_unlock_all(dev);
 
 	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
 
@@ -2755,7 +2716,7 @@ static int i915_rcs_topology(struct seq_file *m, void *unused)
 	struct drm_i915_private *dev_priv = node_to_i915(m->private);
 	struct drm_printer p = drm_seq_file_printer(m);
 
-	intel_device_info_dump_topology(&RUNTIME_INFO(dev_priv)->sseu, &p);
+	intel_device_info_print_topology(&RUNTIME_INFO(dev_priv)->sseu, &p);
 
 	return 0;
 }
@@ -4161,11 +4122,11 @@ static int i915_drrs_ctl_set(void *data, u64 val)
 
 		crtc_state = to_intel_crtc_state(crtc->base.state);
 
-		if (!crtc_state->base.active ||
+		if (!crtc_state->hw.active ||
 		    !crtc_state->has_drrs)
 			goto out;
 
-		commit = crtc_state->base.commit;
+		commit = crtc_state->uapi.commit;
 		if (commit) {
 			ret = wait_for_completion_interruptible(&commit->hw_done);
 			if (ret)
@@ -4177,7 +4138,7 @@ static int i915_drrs_ctl_set(void *data, u64 val)
 			struct intel_encoder *encoder;
 			struct intel_dp *intel_dp;
 
-			if (!(crtc_state->base.connector_mask &
+			if (!(crtc_state->uapi.connector_mask &
 			      drm_connector_mask(connector)))
 				continue;
 
@@ -4236,14 +4197,14 @@ i915_fifo_underrun_reset_write(struct file *filp,
 			return ret;
 
 		crtc_state = to_intel_crtc_state(intel_crtc->base.state);
-		commit = crtc_state->base.commit;
+		commit = crtc_state->uapi.commit;
 		if (commit) {
 			ret = wait_for_completion_interruptible(&commit->hw_done);
 			if (!ret)
 				ret = wait_for_completion_interruptible(&commit->flip_done);
 		}
 
-		if (!ret && crtc_state->base.active) {
+		if (!ret && crtc_state->hw.active) {
 			DRM_DEBUG_KMS("Re-arming FIFO underruns on pipe %c\n",
 				      pipe_name(intel_crtc->pipe));
 
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index fee7dfee1f91..59525094d0e3 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -61,6 +61,7 @@
 
 #include "gem/i915_gem_context.h"
 #include "gem/i915_gem_ioctls.h"
+#include "gem/i915_gem_mman.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
 #include "gt/intel_rc6.h"
@@ -618,7 +619,6 @@ err_bridge:
  */
 static void i915_driver_mmio_release(struct drm_i915_private *dev_priv)
 {
-	intel_engines_cleanup(&dev_priv->gt);
 	intel_teardown_mchbar(dev_priv);
 	intel_uncore_fini_mmio(&dev_priv->uncore);
 	pci_dev_put(dev_priv->bridge_dev);
@@ -1052,7 +1052,7 @@ intel_get_dram_info(struct drm_i915_private *dev_priv)
 	 */
 	dram_info->is_16gb_dimm = !IS_GEN9_LP(dev_priv);
 
-	if (INTEL_GEN(dev_priv) < 9)
+	if (INTEL_GEN(dev_priv) < 9 || !HAS_DISPLAY(dev_priv))
 		return;
 
 	if (IS_GEN9_LP(dev_priv))
@@ -1396,8 +1396,8 @@ static void i915_welcome_messages(struct drm_i915_private *dev_priv)
 					     INTEL_INFO(dev_priv)->platform),
 			   INTEL_GEN(dev_priv));
 
-		intel_device_info_dump_flags(INTEL_INFO(dev_priv), &p);
-		intel_device_info_dump_runtime(RUNTIME_INFO(dev_priv), &p);
+		intel_device_info_print_static(INTEL_INFO(dev_priv), &p);
+		intel_device_info_print_runtime(RUNTIME_INFO(dev_priv), &p);
 	}
 
 	if (IS_ENABLED(CONFIG_DRM_I915_DEBUG))
@@ -1817,8 +1817,6 @@ static int i915_drm_resume(struct drm_device *dev)
 
 	disable_rpm_wakeref_asserts(&dev_priv->runtime_pm);
 
-	intel_rc6_ctx_wa_resume(&dev_priv->gt.rc6);
-
 	intel_gt_sanitize(&dev_priv->gt, true);
 
 	ret = i915_ggtt_enable_hw(dev_priv);
@@ -2662,18 +2660,12 @@ const struct dev_pm_ops i915_pm_ops = {
 	.runtime_resume = intel_runtime_resume,
 };
 
-static const struct vm_operations_struct i915_gem_vm_ops = {
-	.fault = i915_gem_fault,
-	.open = drm_gem_vm_open,
-	.close = drm_gem_vm_close,
-};
-
 static const struct file_operations i915_driver_fops = {
 	.owner = THIS_MODULE,
 	.open = drm_open,
 	.release = drm_release,
 	.unlocked_ioctl = drm_ioctl,
-	.mmap = drm_gem_mmap,
+	.mmap = i915_gem_mmap,
 	.poll = drm_poll,
 	.read = drm_read,
 	.compat_ioctl = i915_compat_ioctl,
@@ -2720,7 +2712,7 @@ static const struct drm_ioctl_desc i915_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(I915_GEM_PREAD, i915_gem_pread_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_PWRITE, i915_gem_pwrite_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_MMAP, i915_gem_mmap_ioctl, DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF_DRV(I915_GEM_MMAP_GTT, i915_gem_mmap_gtt_ioctl, DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(I915_GEM_MMAP_OFFSET, i915_gem_mmap_offset_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_SET_DOMAIN, i915_gem_set_domain_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_SW_FINISH, i915_gem_sw_finish_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_SET_TILING, i915_gem_set_tiling_ioctl, DRM_RENDER_ALLOW),
@@ -2762,7 +2754,6 @@ static struct drm_driver driver = {
 
 	.gem_close_object = i915_gem_close_object,
 	.gem_free_object_unlocked = i915_gem_free_object,
-	.gem_vm_ops = &i915_gem_vm_ops,
 
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
@@ -2773,7 +2764,8 @@ static struct drm_driver driver = {
 	.get_scanout_position = i915_get_crtc_scanoutpos,
 
 	.dumb_create = i915_gem_dumb_create,
-	.dumb_map_offset = i915_gem_mmap_gtt,
+	.dumb_map_offset = i915_gem_dumb_mmap_offset,
+
 	.ioctls = i915_ioctls,
 	.num_ioctls = ARRAY_SIZE(i915_ioctls),
 	.fops = &i915_driver_fops,
@@ -2784,7 +2776,3 @@ static struct drm_driver driver = {
 	.minor = DRIVER_MINOR,
 	.patchlevel = DRIVER_PATCHLEVEL,
 };
-
-#if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
-#include "selftests/mock_drm.c"
-#endif
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index e29bc137e7ba..d05a968227f7 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -110,8 +110,8 @@
 
 #define DRIVER_NAME		"i915"
 #define DRIVER_DESC		"Intel Graphics"
-#define DRIVER_DATE		"20191101"
-#define DRIVER_TIMESTAMP	1572604873
+#define DRIVER_DATE		"20191223"
+#define DRIVER_TIMESTAMP	1577120893
 
 struct drm_i915_gem_object;
 
@@ -273,11 +273,11 @@ struct drm_i915_display_funcs {
 	int (*compute_pipe_wm)(struct intel_crtc_state *crtc_state);
 	int (*compute_intermediate_wm)(struct intel_crtc_state *crtc_state);
 	void (*initial_watermarks)(struct intel_atomic_state *state,
-				   struct intel_crtc_state *crtc_state);
+				   struct intel_crtc *crtc);
 	void (*atomic_update_watermarks)(struct intel_atomic_state *state,
-					 struct intel_crtc_state *crtc_state);
+					 struct intel_crtc *crtc);
 	void (*optimize_watermarks)(struct intel_atomic_state *state,
-				    struct intel_crtc_state *crtc_state);
+				    struct intel_crtc *crtc);
 	int (*compute_global_watermarks)(struct intel_atomic_state *state);
 	void (*update_wm)(struct intel_crtc *crtc);
 	int (*modeset_calc_cdclk)(struct intel_atomic_state *state);
@@ -290,10 +290,10 @@ struct drm_i915_display_funcs {
 					 struct intel_initial_plane_config *);
 	int (*crtc_compute_clock)(struct intel_crtc *crtc,
 				  struct intel_crtc_state *crtc_state);
-	void (*crtc_enable)(struct intel_crtc_state *pipe_config,
-			    struct intel_atomic_state *old_state);
-	void (*crtc_disable)(struct intel_crtc_state *old_crtc_state,
-			     struct intel_atomic_state *old_state);
+	void (*crtc_enable)(struct intel_atomic_state *state,
+			    struct intel_crtc *crtc);
+	void (*crtc_disable)(struct intel_atomic_state *state,
+			     struct intel_crtc *crtc);
 	void (*commit_modeset_enables)(struct intel_atomic_state *state);
 	void (*commit_modeset_disables)(struct intel_atomic_state *state);
 	void (*audio_codec_enable)(struct intel_encoder *encoder,
@@ -366,7 +366,6 @@ struct intel_fbc {
 	unsigned threshold;
 	unsigned int possible_framebuffer_bits;
 	unsigned int busy_bits;
-	unsigned int visible_pipes_mask;
 	struct intel_crtc *crtc;
 
 	struct drm_mm_node compressed_fb;
@@ -374,8 +373,8 @@ struct intel_fbc {
 
 	bool false_color;
 
-	bool enabled;
 	bool active;
+	bool activated;
 	bool flip_pending;
 
 	bool underrun_detected;
@@ -387,9 +386,6 @@ struct intel_fbc {
 	 * these problems.
 	 */
 	struct intel_fbc_state_cache {
-		struct i915_vma *vma;
-		unsigned long flags;
-
 		struct {
 			unsigned int mode_flags;
 			u32 hsw_bdw_pixel_rate;
@@ -418,6 +414,8 @@ struct intel_fbc {
 			const struct drm_format_info *format;
 			unsigned int stride;
 		} fb;
+		u16 gen9_wa_cfb_stride;
+		s8 fence_id;
 	} state_cache;
 
 	/*
@@ -428,9 +426,6 @@ struct intel_fbc {
 	 * are supposed to read from it in order to program the registers.
 	 */
 	struct intel_fbc_reg_params {
-		struct i915_vma *vma;
-		unsigned long flags;
-
 		struct {
 			enum pipe pipe;
 			enum i9xx_plane_id i9xx_plane;
@@ -443,7 +438,9 @@ struct intel_fbc {
 		} fb;
 
 		int cfb_size;
-		unsigned int gen9_wa_cfb_stride;
+		u16 gen9_wa_cfb_stride;
+		s8 fence_id;
+		bool plane_visible;
 	} params;
 
 	const char *no_fbc_reason;
@@ -621,19 +618,18 @@ struct i915_gem_mm {
 
 #define I915_ENGINE_WEDGED_TIMEOUT  (60 * HZ)  /* Reset but no recovery? */
 
+/* Amount of SAGV/QGV points, BSpec precisely defines this */
+#define I915_NUM_QGV_POINTS 8
+
 struct ddi_vbt_port_info {
 	/* Non-NULL if port present. */
 	const struct child_device_config *child;
 
 	int max_tmds_clock;
 
-	/*
-	 * This is an index in the HDMI/DVI DDI buffer translation table.
-	 * The special value HDMI_LEVEL_SHIFT_UNKNOWN means the VBT didn't
-	 * populate this field.
-	 */
-#define HDMI_LEVEL_SHIFT_UNKNOWN	0xff
+	/* This is an index in the HDMI/DVI DDI buffer translation table. */
 	u8 hdmi_level_shift;
+	u8 hdmi_level_shift_set:1;
 
 	u8 supports_dvi:1;
 	u8 supports_hdmi:1;
@@ -724,8 +720,7 @@ struct intel_vbt_data {
 
 	int crt_ddc_pin;
 
-	int child_dev_num;
-	struct child_device_config *child_dev;
+	struct list_head display_devices;
 
 	struct ddi_vbt_port_info ddi_port_info[I915_MAX_PORTS];
 	struct sdvo_device_mapping sdvo_mappings[2];
@@ -891,6 +886,10 @@ struct intel_cdclk_state {
 	u8 voltage_level;
 };
 
+struct i915_selftest_stash {
+	atomic_t counter;
+};
+
 struct drm_i915_private {
 	struct drm_device drm;
 
@@ -956,9 +955,6 @@ struct drm_i915_private {
 
 	struct pci_dev *bridge_dev;
 
-	/* Context used internally to idle the GPU and setup initial state */
-	struct i915_gem_context *kernel_context;
-
 	struct intel_engine_cs *engine[I915_NUM_ENGINES];
 	struct rb_root uabi_engines;
 
@@ -1233,7 +1229,8 @@ struct drm_i915_private {
 	} dram_info;
 
 	struct intel_bw_info {
-		unsigned int deratedbw[3]; /* for each QGV point */
+		/* for each QGV point */
+		unsigned int deratedbw[I915_NUM_QGV_POINTS];
 		u8 num_qgv_points;
 		u8 num_planes;
 	} max_bw[6];
@@ -1248,8 +1245,6 @@ struct drm_i915_private {
 	struct intel_gt gt;
 
 	struct {
-		struct notifier_block pm_notifier;
-
 		struct i915_gem_contexts {
 			spinlock_t lock; /* locks list */
 			struct list_head list;
@@ -1286,6 +1281,8 @@ struct drm_i915_private {
 	/* Mutex to protect the above hdcp component related values. */
 	struct mutex hdcp_comp_mutex;
 
+	I915_SELFTEST_DECLARE(struct i915_selftest_stash selftest;)
+
 	/*
 	 * NOTE: This is the dri1/ums dungeon, don't add stuff here. Your patch
 	 * will be rejected. Instead look for a better place.
@@ -1661,7 +1658,7 @@ IS_SUBPLATFORM(const struct drm_i915_private *i915,
 
 /* WaRsDisableCoarsePowerGating:skl,cnl */
 #define NEEDS_WaRsDisableCoarsePowerGating(dev_priv) \
-	(IS_CANNONLAKE(dev_priv) || IS_GEN(dev_priv, 9))
+	IS_GEN_RANGE(dev_priv, 9, 10)
 
 #define HAS_GMBUS_IRQ(dev_priv) (INTEL_GEN(dev_priv) >= 4)
 #define HAS_GMBUS_BURST_READ(dev_priv) (INTEL_GEN(dev_priv) >= 10 || \
@@ -1842,14 +1839,7 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
 int i915_gem_object_unbind(struct drm_i915_gem_object *obj,
 			   unsigned long flags);
 #define I915_GEM_OBJECT_UNBIND_ACTIVE BIT(0)
-
-struct i915_vma * __must_check
-i915_gem_object_pin(struct drm_i915_gem_object *obj,
-		    struct i915_address_space *vm,
-		    const struct i915_ggtt_view *view,
-		    u64 size,
-		    u64 alignment,
-		    u64 flags);
+#define I915_GEM_OBJECT_UNBIND_BARRIER BIT(1)
 
 void i915_gem_runtime_suspend(struct drm_i915_private *dev_priv);
 
@@ -1862,9 +1852,6 @@ i915_mutex_lock_interruptible(struct drm_device *dev)
 int i915_gem_dumb_create(struct drm_file *file_priv,
 			 struct drm_device *dev,
 			 struct drm_mode_create_dumb *args);
-int i915_gem_mmap_gtt(struct drm_file *file_priv, struct drm_device *dev,
-		      u32 handle, u64 *offset);
-int i915_gem_mmap_gtt_version(void);
 
 int __must_check i915_gem_set_global_seqno(struct drm_device *dev, u32 seqno);
 
@@ -1887,7 +1874,6 @@ void i915_gem_driver_release(struct drm_i915_private *dev_priv);
 void i915_gem_suspend(struct drm_i915_private *dev_priv);
 void i915_gem_suspend_late(struct drm_i915_private *dev_priv);
 void i915_gem_resume(struct drm_i915_private *dev_priv);
-vm_fault_t i915_gem_fault(struct vm_fault *vmf);
 
 int i915_gem_open(struct drm_i915_private *i915, struct drm_file *file);
 void i915_gem_release(struct drm_device *dev, struct drm_file *file);
@@ -1956,14 +1942,13 @@ const char *i915_cache_level_str(struct drm_i915_private *i915, int type);
 int i915_cmd_parser_get_version(struct drm_i915_private *dev_priv);
 void intel_engine_init_cmd_parser(struct intel_engine_cs *engine);
 void intel_engine_cleanup_cmd_parser(struct intel_engine_cs *engine);
-int intel_engine_cmd_parser(struct i915_gem_context *cxt,
-			    struct intel_engine_cs *engine,
-			    struct drm_i915_gem_object *batch_obj,
-			    u64 user_batch_start,
-			    u32 batch_start_offset,
-			    u32 batch_len,
-			    struct drm_i915_gem_object *shadow_batch_obj,
-			    u64 shadow_batch_start);
+int intel_engine_cmd_parser(struct intel_engine_cs *engine,
+			    struct i915_vma *batch,
+			    u32 batch_offset,
+			    u32 batch_length,
+			    struct i915_vma *shadow,
+			    bool trampoline);
+#define I915_CMD_PARSER_TRAMPOLINE_SIZE 8
 
 /* intel_device_info.c */
 static inline struct intel_device_info *
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index d034fa413164..9ddcf17230e6 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -44,20 +44,13 @@
 #include "gem/i915_gem_clflush.h"
 #include "gem/i915_gem_context.h"
 #include "gem/i915_gem_ioctls.h"
-#include "gem/i915_gem_pm.h"
-#include "gt/intel_context.h"
+#include "gem/i915_gem_mman.h"
 #include "gt/intel_engine_user.h"
 #include "gt/intel_gt.h"
 #include "gt/intel_gt_pm.h"
-#include "gt/intel_gt_requests.h"
-#include "gt/intel_mocs.h"
-#include "gt/intel_reset.h"
-#include "gt/intel_renderstate.h"
-#include "gt/intel_rps.h"
 #include "gt/intel_workarounds.h"
 
 #include "i915_drv.h"
-#include "i915_scatterlist.h"
 #include "i915_trace.h"
 #include "i915_vgpu.h"
 
@@ -119,33 +112,65 @@ i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data,
 int i915_gem_object_unbind(struct drm_i915_gem_object *obj,
 			   unsigned long flags)
 {
-	struct i915_vma *vma;
+	struct intel_runtime_pm *rpm = &to_i915(obj->base.dev)->runtime_pm;
 	LIST_HEAD(still_in_list);
-	int ret = 0;
+	intel_wakeref_t wakeref;
+	struct i915_vma *vma;
+	int ret;
+
+	if (!atomic_read(&obj->bind_count))
+		return 0;
+
+	/*
+	 * As some machines use ACPI to handle runtime-resume callbacks, and
+	 * ACPI is quite kmalloc happy, we cannot resume beneath the vm->mutex
+	 * as they are required by the shrinker. Ergo, we wake the device up
+	 * first just in case.
+	 */
+	wakeref = intel_runtime_pm_get(rpm);
 
+try_again:
+	ret = 0;
 	spin_lock(&obj->vma.lock);
 	while (!ret && (vma = list_first_entry_or_null(&obj->vma.list,
 						       struct i915_vma,
 						       obj_link))) {
 		struct i915_address_space *vm = vma->vm;
 
-		ret = -EBUSY;
+		list_move_tail(&vma->obj_link, &still_in_list);
+		if (!i915_vma_is_bound(vma, I915_VMA_BIND_MASK))
+			continue;
+
+		ret = -EAGAIN;
 		if (!i915_vm_tryopen(vm))
 			break;
 
-		list_move_tail(&vma->obj_link, &still_in_list);
+		/* Prevent vma being freed by i915_vma_parked as we unbind */
+		vma = __i915_vma_get(vma);
 		spin_unlock(&obj->vma.lock);
 
-		if (flags & I915_GEM_OBJECT_UNBIND_ACTIVE ||
-		    !i915_vma_is_active(vma))
-			ret = i915_vma_unbind(vma);
+		if (vma) {
+			ret = -EBUSY;
+			if (flags & I915_GEM_OBJECT_UNBIND_ACTIVE ||
+			    !i915_vma_is_active(vma))
+				ret = i915_vma_unbind(vma);
+
+			__i915_vma_put(vma);
+		}
 
 		i915_vm_close(vm);
 		spin_lock(&obj->vma.lock);
 	}
-	list_splice(&still_in_list, &obj->vma.list);
+	list_splice_init(&still_in_list, &obj->vma.list);
 	spin_unlock(&obj->vma.lock);
 
+	if (ret == -EAGAIN && flags & I915_GEM_OBJECT_UNBIND_BARRIER) {
+		rcu_barrier(); /* flush the i915_vm_release() */
+		goto try_again;
+	}
+
+	intel_runtime_pm_put(rpm, wakeref);
+
 	return ret;
 }
 
@@ -161,7 +186,7 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
 	 * We manually control the domain here and pretend that it
 	 * remains coherent i.e. in the GTT domain, like shmem_pwrite.
 	 */
-	intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU);
+	i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
 
 	if (copy_from_user(vaddr, user_data, args->size))
 		return -EFAULT;
@@ -169,7 +194,7 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj,
 	drm_clflush_virt_range(vaddr, args->size);
 	intel_gt_chipset_flush(&to_i915(obj->base.dev)->gt);
 
-	intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU);
+	i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
 	return 0;
 }
 
@@ -589,7 +614,7 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 		goto out_unpin;
 	}
 
-	intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU);
+	i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
 
 	user_data = u64_to_user_ptr(args->data_ptr);
 	offset = args->offset;
@@ -631,11 +656,12 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj,
 		user_data += page_length;
 		offset += page_length;
 	}
-	intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU);
+
+	intel_gt_flush_ggtt_writes(ggtt->vm.gt);
+	i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
 
 	i915_gem_object_unlock_fence(obj, fence);
 out_unpin:
-	intel_gt_flush_ggtt_writes(ggtt->vm.gt);
 	if (drm_mm_node_allocated(&node)) {
 		ggtt->vm.clear_range(&ggtt->vm, node.start, node.size);
 		remove_mappable_node(ggtt, &node);
@@ -721,7 +747,7 @@ i915_gem_shmem_pwrite(struct drm_i915_gem_object *obj,
 		offset = 0;
 	}
 
-	intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU);
+	i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
 	i915_gem_object_unlock_fence(obj, fence);
 
 	return ret;
@@ -855,7 +881,7 @@ void i915_gem_runtime_suspend(struct drm_i915_private *i915)
 
 	list_for_each_entry_safe(obj, on,
 				 &i915->ggtt.userfault_list, userfault_link)
-		__i915_gem_object_release_mmap(obj);
+		__i915_gem_object_release_mmap_gtt(obj);
 
 	/*
 	 * The fence will be lost when the device powers down. If any were
@@ -892,22 +918,8 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj,
 			 u64 alignment,
 			 u64 flags)
 {
-	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
-	struct i915_address_space *vm = &dev_priv->ggtt.vm;
-
-	return i915_gem_object_pin(obj, vm, view, size, alignment,
-				   flags | PIN_GLOBAL);
-}
-
-struct i915_vma *
-i915_gem_object_pin(struct drm_i915_gem_object *obj,
-		    struct i915_address_space *vm,
-		    const struct i915_ggtt_view *view,
-		    u64 size,
-		    u64 alignment,
-		    u64 flags)
-{
-	struct drm_i915_private *dev_priv = to_i915(obj->base.dev);
+	struct drm_i915_private *i915 = to_i915(obj->base.dev);
+	struct i915_ggtt *ggtt = &i915->ggtt;
 	struct i915_vma *vma;
 	int ret;
 
@@ -916,17 +928,19 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 
 	if (flags & PIN_MAPPABLE &&
 	    (!view || view->type == I915_GGTT_VIEW_NORMAL)) {
-		/* If the required space is larger than the available
+		/*
+		 * If the required space is larger than the available
 		 * aperture, we will not able to find a slot for the
 		 * object and unbinding the object now will be in
 		 * vain. Worse, doing so may cause us to ping-pong
 		 * the object in and out of the Global GTT and
 		 * waste a lot of cycles under the mutex.
 		 */
-		if (obj->base.size > dev_priv->ggtt.mappable_end)
+		if (obj->base.size > ggtt->mappable_end)
 			return ERR_PTR(-E2BIG);
 
-		/* If NONBLOCK is set the caller is optimistically
+		/*
+		 * If NONBLOCK is set the caller is optimistically
 		 * trying to cache the full object within the mappable
 		 * aperture, and *must* have a fallback in place for
 		 * situations where we cannot bind the object. We
@@ -942,11 +956,11 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 		 * we could try to minimise harm to others.
 		 */
 		if (flags & PIN_NONBLOCK &&
-		    obj->base.size > dev_priv->ggtt.mappable_end / 2)
+		    obj->base.size > ggtt->mappable_end / 2)
 			return ERR_PTR(-ENOSPC);
 	}
 
-	vma = i915_vma_instance(obj, vm, view);
+	vma = i915_vma_instance(obj, &ggtt->vm, view);
 	if (IS_ERR(vma))
 		return vma;
 
@@ -956,7 +970,7 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 				return ERR_PTR(-ENOSPC);
 
 			if (flags & PIN_MAPPABLE &&
-			    vma->fence_size > dev_priv->ggtt.mappable_end / 2)
+			    vma->fence_size > ggtt->mappable_end / 2)
 				return ERR_PTR(-ENOSPC);
 		}
 
@@ -966,14 +980,14 @@ i915_gem_object_pin(struct drm_i915_gem_object *obj,
 	}
 
 	if (vma->fence && !i915_gem_object_is_tiled(obj)) {
-		mutex_lock(&vma->vm->mutex);
+		mutex_lock(&ggtt->vm.mutex);
 		ret = i915_vma_revoke_fence(vma);
-		mutex_unlock(&vma->vm->mutex);
+		mutex_unlock(&ggtt->vm.mutex);
 		if (ret)
 			return ERR_PTR(ret);
 	}
 
-	ret = i915_vma_pin(vma, size, alignment, flags);
+	ret = i915_vma_pin(vma, size, alignment, flags | PIN_GLOBAL);
 	if (ret)
 		return ERR_PTR(ret);
 
@@ -1054,172 +1068,6 @@ out:
 	return err;
 }
 
-static int __intel_context_flush_retire(struct intel_context *ce)
-{
-	struct intel_timeline *tl;
-
-	tl = intel_context_timeline_lock(ce);
-	if (IS_ERR(tl))
-		return PTR_ERR(tl);
-
-	intel_context_timeline_unlock(tl);
-	return 0;
-}
-
-static int __intel_engines_record_defaults(struct intel_gt *gt)
-{
-	struct i915_request *requests[I915_NUM_ENGINES] = {};
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	int err = 0;
-
-	/*
-	 * As we reset the gpu during very early sanitisation, the current
-	 * register state on the GPU should reflect its defaults values.
-	 * We load a context onto the hw (with restore-inhibit), then switch
-	 * over to a second context to save that default register state. We
-	 * can then prime every new context with that state so they all start
-	 * from the same default HW values.
-	 */
-
-	for_each_engine(engine, gt, id) {
-		struct intel_context *ce;
-		struct i915_request *rq;
-
-		/* We must be able to switch to something! */
-		GEM_BUG_ON(!engine->kernel_context);
-		engine->serial++; /* force the kernel context switch */
-
-		ce = intel_context_create(engine->kernel_context->gem_context,
-					  engine);
-		if (IS_ERR(ce)) {
-			err = PTR_ERR(ce);
-			goto out;
-		}
-
-		rq = intel_context_create_request(ce);
-		if (IS_ERR(rq)) {
-			err = PTR_ERR(rq);
-			intel_context_put(ce);
-			goto out;
-		}
-
-		err = intel_engine_emit_ctx_wa(rq);
-		if (err)
-			goto err_rq;
-
-		err = intel_renderstate_emit(rq);
-		if (err)
-			goto err_rq;
-
-err_rq:
-		requests[id] = i915_request_get(rq);
-		i915_request_add(rq);
-		if (err)
-			goto out;
-	}
-
-	/* Flush the default context image to memory, and enable powersaving. */
-	if (intel_gt_wait_for_idle(gt, I915_GEM_IDLE_TIMEOUT) == -ETIME) {
-		err = -EIO;
-		goto out;
-	}
-
-	for (id = 0; id < ARRAY_SIZE(requests); id++) {
-		struct i915_request *rq;
-		struct i915_vma *state;
-		void *vaddr;
-
-		rq = requests[id];
-		if (!rq)
-			continue;
-
-		GEM_BUG_ON(!test_bit(CONTEXT_ALLOC_BIT,
-				     &rq->hw_context->flags));
-		state = rq->hw_context->state;
-		if (!state)
-			continue;
-
-		/* Serialise with retirement on another CPU */
-		err = __intel_context_flush_retire(rq->hw_context);
-		if (err)
-			goto out;
-
-		/* We want to be able to unbind the state from the GGTT */
-		GEM_BUG_ON(intel_context_is_pinned(rq->hw_context));
-
-		/*
-		 * As we will hold a reference to the logical state, it will
-		 * not be torn down with the context, and importantly the
-		 * object will hold onto its vma (making it possible for a
-		 * stray GTT write to corrupt our defaults). Unmap the vma
-		 * from the GTT to prevent such accidents and reclaim the
-		 * space.
-		 */
-		err = i915_vma_unbind(state);
-		if (err)
-			goto out;
-
-		i915_gem_object_lock(state->obj);
-		err = i915_gem_object_set_to_cpu_domain(state->obj, false);
-		i915_gem_object_unlock(state->obj);
-		if (err)
-			goto out;
-
-		i915_gem_object_set_cache_coherency(state->obj, I915_CACHE_LLC);
-
-		/* Check we can acquire the image of the context state */
-		vaddr = i915_gem_object_pin_map(state->obj, I915_MAP_FORCE_WB);
-		if (IS_ERR(vaddr)) {
-			err = PTR_ERR(vaddr);
-			goto out;
-		}
-
-		rq->engine->default_state = i915_gem_object_get(state->obj);
-		i915_gem_object_unpin_map(state->obj);
-	}
-
-out:
-	/*
-	 * If we have to abandon now, we expect the engines to be idle
-	 * and ready to be torn-down. The quickest way we can accomplish
-	 * this is by declaring ourselves wedged.
-	 */
-	if (err)
-		intel_gt_set_wedged(gt);
-
-	for (id = 0; id < ARRAY_SIZE(requests); id++) {
-		struct intel_context *ce;
-		struct i915_request *rq;
-
-		rq = requests[id];
-		if (!rq)
-			continue;
-
-		ce = rq->hw_context;
-		i915_request_put(rq);
-		intel_context_put(ce);
-	}
-	return err;
-}
-
-static int intel_engines_verify_workarounds(struct intel_gt *gt)
-{
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
-	int err = 0;
-
-	if (!IS_ENABLED(CONFIG_DRM_I915_DEBUG_GEM))
-		return 0;
-
-	for_each_engine(engine, gt, id) {
-		if (intel_engine_verify_workarounds(engine, "load"))
-			err = -EIO;
-	}
-
-	return err;
-}
-
 int i915_gem_init(struct drm_i915_private *dev_priv)
 {
 	int ret;
@@ -1229,8 +1077,6 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 		mkwrite_device_info(dev_priv)->page_sizes =
 			I915_GTT_PAGE_SIZE_4K;
 
-	intel_timelines_init(dev_priv);
-
 	ret = i915_gem_init_userptr(dev_priv);
 	if (ret)
 		return ret;
@@ -1238,51 +1084,12 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	intel_uc_fetch_firmwares(&dev_priv->gt.uc);
 	intel_wopcm_init(&dev_priv->wopcm);
 
-	/* This is just a security blanket to placate dragons.
-	 * On some systems, we very sporadically observe that the first TLBs
-	 * used by the CS may be stale, despite us poking the TLB reset. If
-	 * we hold the forcewake during initialisation these problems
-	 * just magically go away.
-	 */
-	intel_uncore_forcewake_get(&dev_priv->uncore, FORCEWAKE_ALL);
-
 	ret = i915_init_ggtt(dev_priv);
 	if (ret) {
 		GEM_BUG_ON(ret == -EIO);
 		goto err_unlock;
 	}
 
-	intel_gt_init(&dev_priv->gt);
-
-	ret = intel_engines_setup(&dev_priv->gt);
-	if (ret) {
-		GEM_BUG_ON(ret == -EIO);
-		goto err_unlock;
-	}
-
-	ret = i915_gem_init_contexts(dev_priv);
-	if (ret) {
-		GEM_BUG_ON(ret == -EIO);
-		goto err_scratch;
-	}
-
-	ret = intel_engines_init(&dev_priv->gt);
-	if (ret) {
-		GEM_BUG_ON(ret == -EIO);
-		goto err_context;
-	}
-
-	intel_uc_init(&dev_priv->gt.uc);
-
-	ret = intel_gt_init_hw(&dev_priv->gt);
-	if (ret)
-		goto err_uc_init;
-
-	/* Only when the HW is re-initialised, can we replay the requests */
-	ret = intel_gt_resume(&dev_priv->gt);
-	if (ret)
-		goto err_init_hw;
-
 	/*
 	 * Despite its name intel_init_clock_gating applies both display
 	 * clock gating workarounds; GT mmio workarounds and the occasional
@@ -1294,23 +1101,9 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	 */
 	intel_init_clock_gating(dev_priv);
 
-	ret = intel_engines_verify_workarounds(&dev_priv->gt);
-	if (ret)
-		goto err_gt;
-
-	ret = __intel_engines_record_defaults(&dev_priv->gt);
-	if (ret)
-		goto err_gt;
-
-	ret = i915_inject_probe_error(dev_priv, -ENODEV);
-	if (ret)
-		goto err_gt;
-
-	ret = i915_inject_probe_error(dev_priv, -EIO);
+	ret = intel_gt_init(&dev_priv->gt);
 	if (ret)
-		goto err_gt;
-
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
+		goto err_unlock;
 
 	return 0;
 
@@ -1320,31 +1113,12 @@ int i915_gem_init(struct drm_i915_private *dev_priv)
 	 * HW as irrevisibly wedged, but keep enough state around that the
 	 * driver doesn't explode during runtime.
 	 */
-err_gt:
-	intel_gt_set_wedged_on_init(&dev_priv->gt);
-	i915_gem_suspend(dev_priv);
-	i915_gem_suspend_late(dev_priv);
-
-	i915_gem_drain_workqueue(dev_priv);
-err_init_hw:
-	intel_uc_fini_hw(&dev_priv->gt.uc);
-err_uc_init:
-	if (ret != -EIO) {
-		intel_uc_fini(&dev_priv->gt.uc);
-		intel_engines_cleanup(&dev_priv->gt);
-	}
-err_context:
-	if (ret != -EIO)
-		i915_gem_driver_release__contexts(dev_priv);
-err_scratch:
-	intel_gt_driver_release(&dev_priv->gt);
 err_unlock:
-	intel_uncore_forcewake_put(&dev_priv->uncore, FORCEWAKE_ALL);
+	i915_gem_drain_workqueue(dev_priv);
 
 	if (ret != -EIO) {
 		intel_uc_cleanup_firmwares(&dev_priv->gt.uc);
 		i915_gem_cleanup_userptr(dev_priv);
-		intel_timelines_fini(dev_priv);
 	}
 
 	if (ret == -EIO) {
@@ -1388,27 +1162,24 @@ void i915_gem_driver_remove(struct drm_i915_private *dev_priv)
 
 	i915_gem_suspend_late(dev_priv);
 	intel_gt_driver_remove(&dev_priv->gt);
+	dev_priv->uabi_engines = RB_ROOT;
 
 	/* Flush any outstanding unpin_work. */
 	i915_gem_drain_workqueue(dev_priv);
 
-	intel_uc_fini_hw(&dev_priv->gt.uc);
-	intel_uc_fini(&dev_priv->gt.uc);
-
 	i915_gem_drain_freed_objects(dev_priv);
 }
 
 void i915_gem_driver_release(struct drm_i915_private *dev_priv)
 {
-	intel_engines_cleanup(&dev_priv->gt);
-	i915_gem_driver_release__contexts(dev_priv);
 	intel_gt_driver_release(&dev_priv->gt);
 
 	intel_wa_list_free(&dev_priv->gt_wa_list);
 
 	intel_uc_cleanup_firmwares(&dev_priv->gt.uc);
 	i915_gem_cleanup_userptr(dev_priv);
-	intel_timelines_fini(dev_priv);
+
+	i915_gem_driver_release__contexts(dev_priv);
 
 	i915_gem_drain_freed_objects(dev_priv);
 
@@ -1430,6 +1201,7 @@ static void i915_gem_init__mm(struct drm_i915_private *i915)
 void i915_gem_init_early(struct drm_i915_private *dev_priv)
 {
 	i915_gem_init__mm(dev_priv);
+	i915_gem_init__contexts(dev_priv);
 
 	spin_lock_init(&dev_priv->fb_tracking.lock);
 }
diff --git a/drivers/gpu/drm/i915/i915_gem.h b/drivers/gpu/drm/i915/i915_gem.h
index 5ff051764373..1753c84d6c0d 100644
--- a/drivers/gpu/drm/i915/i915_gem.h
+++ b/drivers/gpu/drm/i915/i915_gem.h
@@ -30,6 +30,8 @@
 
 #include <drm/drm_drv.h>
 
+#include "i915_utils.h"
+
 struct drm_i915_private;
 
 #ifdef CONFIG_DRM_I915_DEBUG_GEM
@@ -39,6 +41,7 @@ struct drm_i915_private;
 #define GEM_BUG_ON(condition) do { if (unlikely((condition))) {	\
 		GEM_TRACE_ERR("%s:%d GEM_BUG_ON(%s)\n", \
 			      __func__, __LINE__, __stringify(condition)); \
+		GEM_TRACE_DUMP(); \
 		BUG(); \
 		} \
 	} while(0)
@@ -68,9 +71,10 @@ struct drm_i915_private;
 	pr_err(__VA_ARGS__);						\
 	trace_printk(__VA_ARGS__);					\
 } while (0)
-#define GEM_TRACE_DUMP() ftrace_dump(DUMP_ALL)
+#define GEM_TRACE_DUMP() \
+	do { ftrace_dump(DUMP_ALL); add_taint_for_CI(TAINT_WARN); } while (0)
 #define GEM_TRACE_DUMP_ON(expr) \
-	do { if (expr) ftrace_dump(DUMP_ALL); } while (0)
+	do { if (expr) GEM_TRACE_DUMP(); } while (0)
 #else
 #define GEM_TRACE(...) do { } while (0)
 #define GEM_TRACE_ERR(...) do { } while (0)
diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c
index 7e62c310290f..0697bedebeef 100644
--- a/drivers/gpu/drm/i915/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/i915_gem_evict.c
@@ -359,9 +359,7 @@ int i915_gem_evict_for_node(struct i915_address_space *vm,
  */
 int i915_gem_evict_vm(struct i915_address_space *vm)
 {
-	struct list_head eviction_list;
-	struct i915_vma *vma, *next;
-	int ret;
+	int ret = 0;
 
 	lockdep_assert_held(&vm->mutex);
 	trace_i915_gem_evict_vm(vm);
@@ -377,21 +375,30 @@ int i915_gem_evict_vm(struct i915_address_space *vm)
 			return ret;
 	}
 
-	INIT_LIST_HEAD(&eviction_list);
-	list_for_each_entry(vma, &vm->bound_list, vm_link) {
-		if (i915_vma_is_pinned(vma))
-			continue;
+	do {
+		struct i915_vma *vma, *vn;
+		LIST_HEAD(eviction_list);
 
-		__i915_vma_pin(vma);
-		list_add(&vma->evict_link, &eviction_list);
-	}
+		list_for_each_entry(vma, &vm->bound_list, vm_link) {
+			if (i915_vma_is_pinned(vma))
+				continue;
+
+			__i915_vma_pin(vma);
+			list_add(&vma->evict_link, &eviction_list);
+		}
+		if (list_empty(&eviction_list))
+			break;
+
+		ret = 0;
+		list_for_each_entry_safe(vma, vn, &eviction_list, evict_link) {
+			__i915_vma_unpin(vma);
+			if (ret == 0)
+				ret = __i915_vma_unbind(vma);
+			if (ret != -EINTR) /* "Get me out of here!" */
+				ret = 0;
+		}
+	} while (ret == 0);
 
-	ret = 0;
-	list_for_each_entry_safe(vma, next, &eviction_list, evict_link) {
-		__i915_vma_unpin(vma);
-		if (ret == 0)
-			ret = __i915_vma_unbind(vma);
-	}
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 6239a9adbf14..1efe58ad0ce9 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -53,6 +53,8 @@
 #define DBG(...)
 #endif
 
+#define NALLOC 3 /* 1 normal, 1 for concurrent threads, 1 for preallocation */
+
 /**
  * DOC: Global GTT views
  *
@@ -123,6 +125,16 @@ static void gen6_ggtt_invalidate(struct i915_ggtt *ggtt)
 {
 	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
 
+	spin_lock_irq(&uncore->lock);
+	intel_uncore_write_fw(uncore, GFX_FLSH_CNTL_GEN6, GFX_FLSH_CNTL_EN);
+	intel_uncore_read_fw(uncore, GFX_FLSH_CNTL_GEN6);
+	spin_unlock_irq(&uncore->lock);
+}
+
+static void gen8_ggtt_invalidate(struct i915_ggtt *ggtt)
+{
+	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
+
 	/*
 	 * Note that as an uncached mmio write, this will flush the
 	 * WCB of the writes into the GGTT before it triggers the invalidate.
@@ -135,7 +147,7 @@ static void guc_ggtt_invalidate(struct i915_ggtt *ggtt)
 	struct intel_uncore *uncore = ggtt->vm.gt->uncore;
 	struct drm_i915_private *i915 = ggtt->vm.i915;
 
-	gen6_ggtt_invalidate(ggtt);
+	gen8_ggtt_invalidate(ggtt);
 
 	if (INTEL_GEN(i915) >= 12)
 		intel_uncore_write_fw(uncore, GEN12_GUC_TLB_INV_CR,
@@ -522,7 +534,7 @@ void __i915_vm_close(struct i915_address_space *vm)
 
 		atomic_and(~I915_VMA_PIN_MASK, &vma->flags);
 		WARN_ON(__i915_vma_unbind(vma));
-		i915_vma_destroy(vma);
+		__i915_vma_put(vma);
 
 		i915_gem_object_put(obj);
 	}
@@ -783,7 +795,7 @@ __set_pd_entry(struct i915_page_directory * const pd,
 	       u64 (*encode)(const dma_addr_t, const enum i915_cache_level))
 {
 	/* Each thread pre-pins the pd, and we may have a thread per pde. */
-	GEM_BUG_ON(atomic_read(px_used(pd)) > 2 * ARRAY_SIZE(pd->entry));
+	GEM_BUG_ON(atomic_read(px_used(pd)) > NALLOC * ARRAY_SIZE(pd->entry));
 
 	atomic_inc(px_used(pd));
 	pd->entry[idx] = to;
@@ -1118,7 +1130,7 @@ static int __gen8_ppgtt_alloc(struct i915_address_space * const vm,
 
 			atomic_add(count, &pt->used);
 			/* All other pdes may be simultaneously removed */
-			GEM_BUG_ON(atomic_read(&pt->used) > 2 * I915_PDES);
+			GEM_BUG_ON(atomic_read(&pt->used) > NALLOC * I915_PDES);
 			*start += count;
 		}
 	} while (idx++, --len);
@@ -1363,12 +1375,8 @@ static int gen8_init_scratch(struct i915_address_space *vm)
 	 * If everybody agrees to not to write into the scratch page,
 	 * we can reuse it for all vm, keeping contexts and processes separate.
 	 */
-	if (vm->has_read_only &&
-	    vm->i915->kernel_context &&
-	    vm->i915->kernel_context->vm) {
-		struct i915_address_space *clone =
-			rcu_dereference_protected(vm->i915->kernel_context->vm,
-						  true); /* static */
+	if (vm->has_read_only && vm->gt->vm && !i915_is_ggtt(vm->gt->vm)) {
+		struct i915_address_space *clone = vm->gt->vm;
 
 		GEM_BUG_ON(!clone->has_read_only);
 
@@ -1683,6 +1691,28 @@ static void gen6_ppgtt_insert_entries(struct i915_address_space *vm,
 	vma->page_sizes.gtt = I915_GTT_PAGE_SIZE;
 }
 
+static void gen6_flush_pd(struct gen6_ppgtt *ppgtt, u64 start, u64 end)
+{
+	struct i915_page_directory * const pd = ppgtt->base.pd;
+	struct i915_page_table *pt;
+	unsigned int pde;
+
+	start = round_down(start, SZ_64K);
+	end = round_up(end, SZ_64K) - start;
+
+	mutex_lock(&ppgtt->flush);
+
+	gen6_for_each_pde(pt, pd, start, end, pde)
+		gen6_write_pde(ppgtt, pde, pt);
+
+	mb();
+	ioread32(ppgtt->pd_addr + pde - 1);
+	gen6_ggtt_invalidate(ppgtt->base.vm.gt->ggtt);
+	mb();
+
+	mutex_unlock(&ppgtt->flush);
+}
+
 static int gen6_alloc_va_range(struct i915_address_space *vm,
 			       u64 start, u64 length)
 {
@@ -1692,7 +1722,6 @@ static int gen6_alloc_va_range(struct i915_address_space *vm,
 	intel_wakeref_t wakeref;
 	u64 from = start;
 	unsigned int pde;
-	bool flush = false;
 	int ret = 0;
 
 	wakeref = intel_runtime_pm_get(&vm->i915->runtime_pm);
@@ -1717,11 +1746,6 @@ static int gen6_alloc_va_range(struct i915_address_space *vm,
 			spin_lock(&pd->lock);
 			if (pd->entry[pde] == &vm->scratch[1]) {
 				pd->entry[pde] = pt;
-				if (i915_vma_is_bound(ppgtt->vma,
-						      I915_VMA_GLOBAL_BIND)) {
-					gen6_write_pde(ppgtt, pde, pt);
-					flush = true;
-				}
 			} else {
 				alloc = pt;
 				pt = pd->entry[pde];
@@ -1732,8 +1756,8 @@ static int gen6_alloc_va_range(struct i915_address_space *vm,
 	}
 	spin_unlock(&pd->lock);
 
-	if (flush)
-		gen6_ggtt_invalidate(vm->gt->ggtt);
+	if (i915_vma_is_bound(ppgtt->vma, I915_VMA_GLOBAL_BIND))
+		gen6_flush_pd(ppgtt, from, start);
 
 	goto out;
 
@@ -1788,11 +1812,12 @@ static void gen6_ppgtt_cleanup(struct i915_address_space *vm)
 {
 	struct gen6_ppgtt *ppgtt = to_gen6_ppgtt(i915_vm_to_ppgtt(vm));
 
-	i915_vma_destroy(ppgtt->vma);
+	__i915_vma_put(ppgtt->vma);
 
 	gen6_ppgtt_free_pd(ppgtt);
 	free_scratch(vm);
 
+	mutex_destroy(&ppgtt->flush);
 	mutex_destroy(&ppgtt->pin_mutex);
 	kfree(ppgtt->base.pd);
 }
@@ -1817,17 +1842,11 @@ static int pd_vma_bind(struct i915_vma *vma,
 	struct i915_ggtt *ggtt = i915_vm_to_ggtt(vma->vm);
 	struct gen6_ppgtt *ppgtt = vma->private;
 	u32 ggtt_offset = i915_ggtt_offset(vma) / I915_GTT_PAGE_SIZE;
-	struct i915_page_table *pt;
-	unsigned int pde;
 
 	px_base(ppgtt->base.pd)->ggtt_offset = ggtt_offset * sizeof(gen6_pte_t);
 	ppgtt->pd_addr = (gen6_pte_t __iomem *)ggtt->gsm + ggtt_offset;
 
-	gen6_for_all_pdes(pt, ppgtt->base.pd, pde)
-		gen6_write_pde(ppgtt, pde, pt);
-
-	gen6_ggtt_invalidate(ggtt);
-
+	gen6_flush_pd(ppgtt, 0, ppgtt->base.vm.total);
 	return 0;
 }
 
@@ -1876,6 +1895,7 @@ static struct i915_vma *pd_vma_create(struct gen6_ppgtt *ppgtt, int size)
 
 	i915_active_init(&vma->active, NULL, NULL);
 
+	kref_init(&vma->ref);
 	mutex_init(&vma->pages_mutex);
 	vma->vm = i915_vm_get(&ggtt->vm);
 	vma->ops = &pd_vma_ops;
@@ -1917,9 +1937,7 @@ int gen6_ppgtt_pin(struct i915_ppgtt *base)
 	 * size. We allocate at the top of the GTT to avoid fragmentation.
 	 */
 	if (!atomic_read(&ppgtt->pin_count)) {
-		err = i915_vma_pin(ppgtt->vma,
-				   0, GEN6_PD_ALIGN,
-				   PIN_GLOBAL | PIN_HIGH);
+		err = i915_ggtt_pin(ppgtt->vma, GEN6_PD_ALIGN, PIN_HIGH);
 	}
 	if (!err)
 		atomic_inc(&ppgtt->pin_count);
@@ -1958,6 +1976,7 @@ static struct i915_ppgtt *gen6_ppgtt_create(struct drm_i915_private *i915)
 	if (!ppgtt)
 		return ERR_PTR(-ENOMEM);
 
+	mutex_init(&ppgtt->flush);
 	mutex_init(&ppgtt->pin_mutex);
 
 	ppgtt_init(&ppgtt->base, &i915->gt);
@@ -1994,6 +2013,7 @@ err_scratch:
 err_pd:
 	kfree(ppgtt->base.pd);
 err_free:
+	mutex_destroy(&ppgtt->pin_mutex);
 	kfree(ppgtt);
 	return ERR_PTR(err);
 }
@@ -3041,7 +3061,7 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt)
 			ggtt->vm.clear_range = bxt_vtd_ggtt_clear_range__BKL;
 	}
 
-	ggtt->invalidate = gen6_ggtt_invalidate;
+	ggtt->invalidate = gen8_ggtt_invalidate;
 
 	ggtt->vm.vma_ops.bind_vma    = ggtt_bind_vma;
 	ggtt->vm.vma_ops.unbind_vma  = ggtt_unbind_vma;
@@ -3281,7 +3301,7 @@ int i915_ggtt_enable_hw(struct drm_i915_private *dev_priv)
 
 void i915_ggtt_enable_guc(struct i915_ggtt *ggtt)
 {
-	GEM_BUG_ON(ggtt->invalidate != gen6_ggtt_invalidate);
+	GEM_BUG_ON(ggtt->invalidate != gen8_ggtt_invalidate);
 
 	ggtt->invalidate = guc_ggtt_invalidate;
 
@@ -3291,13 +3311,13 @@ void i915_ggtt_enable_guc(struct i915_ggtt *ggtt)
 void i915_ggtt_disable_guc(struct i915_ggtt *ggtt)
 {
 	/* XXX Temporary pardon for error unload */
-	if (ggtt->invalidate == gen6_ggtt_invalidate)
+	if (ggtt->invalidate == gen8_ggtt_invalidate)
 		return;
 
 	/* We should only be called after i915_ggtt_enable_guc() */
 	GEM_BUG_ON(ggtt->invalidate != guc_ggtt_invalidate);
 
-	ggtt->invalidate = gen6_ggtt_invalidate;
+	ggtt->invalidate = gen8_ggtt_invalidate;
 
 	ggtt->invalidate(ggtt);
 }
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h
index 402283ce2864..31a4a96ddd0d 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.h
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.h
@@ -443,6 +443,7 @@ struct i915_ppgtt {
 struct gen6_ppgtt {
 	struct i915_ppgtt base;
 
+	struct mutex flush;
 	struct i915_vma *vma;
 	gen6_pte_t __iomem *pd_addr;
 
diff --git a/drivers/gpu/drm/i915/i915_getparam.c b/drivers/gpu/drm/i915/i915_getparam.c
index cf8a8c3ef047..54fce81d5724 100644
--- a/drivers/gpu/drm/i915/i915_getparam.c
+++ b/drivers/gpu/drm/i915/i915_getparam.c
@@ -2,6 +2,7 @@
  * SPDX-License-Identifier: MIT
  */
 
+#include "gem/i915_gem_mman.h"
 #include "gt/intel_engine_user.h"
 
 #include "i915_drv.h"
diff --git a/drivers/gpu/drm/i915/i915_globals.c b/drivers/gpu/drm/i915/i915_globals.c
index be127cd28931..3aa213684293 100644
--- a/drivers/gpu/drm/i915/i915_globals.c
+++ b/drivers/gpu/drm/i915/i915_globals.c
@@ -20,7 +20,10 @@ static LIST_HEAD(globals);
 static atomic_t active;
 static atomic_t epoch;
 static struct park_work {
-	struct rcu_work work;
+	struct delayed_work work;
+	struct rcu_head rcu;
+	unsigned long flags;
+#define PENDING 0
 	int epoch;
 } park;
 
@@ -37,11 +40,33 @@ static void i915_globals_shrink(void)
 		global->shrink();
 }
 
+static void __i915_globals_grace(struct rcu_head *rcu)
+{
+	/* Ratelimit parking as shrinking is quite slow */
+	schedule_delayed_work(&park.work, round_jiffies_up_relative(2 * HZ));
+}
+
+static void __i915_globals_queue_rcu(void)
+{
+	park.epoch = atomic_inc_return(&epoch);
+	if (!atomic_read(&active)) {
+		init_rcu_head(&park.rcu);
+		call_rcu(&park.rcu, __i915_globals_grace);
+	}
+}
+
 static void __i915_globals_park(struct work_struct *work)
 {
+	destroy_rcu_head(&park.rcu);
+
 	/* Confirm nothing woke up in the last grace period */
-	if (park.epoch == atomic_read(&epoch))
-		i915_globals_shrink();
+	if (park.epoch != atomic_read(&epoch)) {
+		__i915_globals_queue_rcu();
+		return;
+	}
+
+	clear_bit(PENDING, &park.flags);
+	i915_globals_shrink();
 }
 
 void __init i915_global_register(struct i915_global *global)
@@ -85,7 +110,7 @@ int __init i915_globals_init(void)
 		}
 	}
 
-	INIT_RCU_WORK(&park.work, __i915_globals_park);
+	INIT_DELAYED_WORK(&park.work, __i915_globals_park);
 	return 0;
 }
 
@@ -103,8 +128,9 @@ void i915_globals_park(void)
 	if (!atomic_dec_and_test(&active))
 		return;
 
-	park.epoch = atomic_inc_return(&epoch);
-	queue_rcu_work(system_wq, &park.work);
+	/* Queue cleanup after the next RCU grace period has freed slabs */
+	if (!test_and_set_bit(PENDING, &park.flags))
+		__i915_globals_queue_rcu();
 }
 
 void i915_globals_unpark(void)
@@ -113,12 +139,21 @@ void i915_globals_unpark(void)
 	atomic_inc(&active);
 }
 
+static void __exit __i915_globals_flush(void)
+{
+	atomic_inc(&active); /* skip shrinking */
+
+	rcu_barrier(); /* wait for the work to be queued */
+	flush_delayed_work(&park.work);
+
+	atomic_dec(&active);
+}
+
 void __exit i915_globals_exit(void)
 {
-	/* Flush any residual park_work */
-	atomic_inc(&epoch);
-	flush_rcu_work(&park.work);
+	GEM_BUG_ON(atomic_read(&active));
 
+	__i915_globals_flush();
 	__i915_globals_cleanup();
 
 	/* And ensure that our DESTROY_BY_RCU slabs are truly destroyed */
diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c
index 3c85cb0ee99f..fda0977d2059 100644
--- a/drivers/gpu/drm/i915/i915_gpu_error.c
+++ b/drivers/gpu/drm/i915/i915_gpu_error.c
@@ -599,9 +599,10 @@ static void err_print_capabilities(struct drm_i915_error_state_buf *m,
 {
 	struct drm_printer p = i915_error_printer(m);
 
-	intel_device_info_dump_flags(info, &p);
+	intel_device_info_print_static(info, &p);
+	intel_device_info_print_runtime(runtime, &p);
+	intel_device_info_print_topology(&runtime->sseu, &p);
 	intel_driver_caps_print(caps, &p);
-	intel_device_info_dump_topology(&runtime->sseu, &p);
 }
 
 static void err_print_params(struct drm_i915_error_state_buf *m,
@@ -1045,7 +1046,7 @@ i915_error_object_create(struct drm_i915_private *i915,
 
 			s = kmap(page);
 			ret = compress_page(compress, s, dst);
-			kunmap(s);
+			kunmap(page);
 
 			drm_clflush_pages(&page, 1);
 
@@ -1220,7 +1221,7 @@ static void error_record_engine_registers(struct i915_gpu_state *error,
 static void record_request(const struct i915_request *request,
 			   struct drm_i915_error_request *erq)
 {
-	const struct i915_gem_context *ctx = request->gem_context;
+	const struct i915_gem_context *ctx;
 
 	erq->flags = request->fence.flags;
 	erq->context = request->fence.context;
@@ -1231,8 +1232,11 @@ static void record_request(const struct i915_request *request,
 	erq->head = request->head;
 	erq->tail = request->tail;
 
+	erq->pid = 0;
 	rcu_read_lock();
-	erq->pid = ctx->pid ? pid_nr(ctx->pid) : 0;
+	ctx = rcu_dereference(request->context->gem_context);
+	if (ctx)
+		erq->pid = pid_nr(ctx->pid);
 	rcu_read_unlock();
 }
 
@@ -1300,25 +1304,34 @@ static void error_record_engine_execlists(const struct intel_engine_cs *engine,
 static bool record_context(struct drm_i915_error_context *e,
 			   const struct i915_request *rq)
 {
-	const struct i915_gem_context *ctx = rq->gem_context;
+	struct i915_gem_context *ctx;
+	struct task_struct *task;
+	bool capture;
 
-	if (ctx->pid) {
-		struct task_struct *task;
+	rcu_read_lock();
+	ctx = rcu_dereference(rq->context->gem_context);
+	if (ctx && !kref_get_unless_zero(&ctx->ref))
+		ctx = NULL;
+	rcu_read_unlock();
+	if (!ctx)
+		return false;
 
-		rcu_read_lock();
-		task = pid_task(ctx->pid, PIDTYPE_PID);
-		if (task) {
-			strcpy(e->comm, task->comm);
-			e->pid = task->pid;
-		}
-		rcu_read_unlock();
+	rcu_read_lock();
+	task = pid_task(ctx->pid, PIDTYPE_PID);
+	if (task) {
+		strcpy(e->comm, task->comm);
+		e->pid = task->pid;
 	}
+	rcu_read_unlock();
 
 	e->sched_attr = ctx->sched;
 	e->guilty = atomic_read(&ctx->guilty_count);
 	e->active = atomic_read(&ctx->active_count);
 
-	return i915_gem_context_no_error_capture(ctx);
+	capture = i915_gem_context_no_error_capture(ctx);
+
+	i915_gem_context_put(ctx);
+	return capture;
 }
 
 struct capture_vma {
@@ -1454,7 +1467,7 @@ gem_record_rings(struct i915_gpu_state *error, struct compress *compress)
 		capture = request_record_user_bo(request, ee, capture);
 
 		capture = capture_vma(capture,
-				      request->hw_context->state,
+				      request->context->state,
 				      &ee->ctx);
 
 		capture = capture_vma(capture,
diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c
index dae00f7dd7df..42b79f577500 100644
--- a/drivers/gpu/drm/i915/i915_irq.c
+++ b/drivers/gpu/drm/i915/i915_irq.c
@@ -2453,6 +2453,25 @@ static inline void gen11_master_intr_enable(void __iomem * const regs)
 	raw_reg_write(regs, GEN11_GFX_MSTR_IRQ, GEN11_MASTER_IRQ);
 }
 
+static void
+gen11_display_irq_handler(struct drm_i915_private *i915)
+{
+	void __iomem * const regs = i915->uncore.regs;
+	const u32 disp_ctl = raw_reg_read(regs, GEN11_DISPLAY_INT_CTL);
+
+	disable_rpm_wakeref_asserts(&i915->runtime_pm);
+	/*
+	 * GEN11_DISPLAY_INT_CTL has same format as GEN8_MASTER_IRQ
+	 * for the display related bits.
+	 */
+	raw_reg_write(regs, GEN11_DISPLAY_INT_CTL, 0x0);
+	gen8_de_irq_handler(i915, disp_ctl);
+	raw_reg_write(regs, GEN11_DISPLAY_INT_CTL,
+		      GEN11_DISPLAY_IRQ_ENABLE);
+
+	enable_rpm_wakeref_asserts(&i915->runtime_pm);
+}
+
 static __always_inline irqreturn_t
 __gen11_irq_handler(struct drm_i915_private * const i915,
 		    u32 (*intr_disable)(void __iomem * const regs),
@@ -2476,17 +2495,8 @@ __gen11_irq_handler(struct drm_i915_private * const i915,
 	gen11_gt_irq_handler(gt, master_ctl);
 
 	/* IRQs are synced during runtime_suspend, we don't require a wakeref */
-	if (master_ctl & GEN11_DISPLAY_IRQ) {
-		const u32 disp_ctl = raw_reg_read(regs, GEN11_DISPLAY_INT_CTL);
-
-		disable_rpm_wakeref_asserts(&i915->runtime_pm);
-		/*
-		 * GEN11_DISPLAY_INT_CTL has same format as GEN8_MASTER_IRQ
-		 * for the display related bits.
-		 */
-		gen8_de_irq_handler(i915, disp_ctl);
-		enable_rpm_wakeref_asserts(&i915->runtime_pm);
-	}
+	if (master_ctl & GEN11_DISPLAY_IRQ)
+		gen11_display_irq_handler(i915);
 
 	gu_misc_iir = gen11_gu_misc_irq_ack(gt, master_ctl);
 
@@ -2788,15 +2798,11 @@ static void gen8_irq_reset(struct drm_i915_private *dev_priv)
 		ibx_irq_reset(dev_priv);
 }
 
-static void gen11_irq_reset(struct drm_i915_private *dev_priv)
+static void gen11_display_irq_reset(struct drm_i915_private *dev_priv)
 {
 	struct intel_uncore *uncore = &dev_priv->uncore;
 	enum pipe pipe;
 
-	gen11_master_intr_disable(dev_priv->uncore.regs);
-
-	gen11_gt_irq_reset(&dev_priv->gt);
-
 	intel_uncore_write(uncore, GEN11_DISPLAY_INT_CTL, 0);
 
 	if (INTEL_GEN(dev_priv) >= 12) {
@@ -2825,13 +2831,24 @@ static void gen11_irq_reset(struct drm_i915_private *dev_priv)
 	GEN3_IRQ_RESET(uncore, GEN8_DE_PORT_);
 	GEN3_IRQ_RESET(uncore, GEN8_DE_MISC_);
 	GEN3_IRQ_RESET(uncore, GEN11_DE_HPD_);
-	GEN3_IRQ_RESET(uncore, GEN11_GU_MISC_);
-	GEN3_IRQ_RESET(uncore, GEN8_PCU_);
 
 	if (INTEL_PCH_TYPE(dev_priv) >= PCH_ICP)
 		GEN3_IRQ_RESET(uncore, SDE);
 }
 
+static void gen11_irq_reset(struct drm_i915_private *dev_priv)
+{
+	struct intel_uncore *uncore = &dev_priv->uncore;
+
+	gen11_master_intr_disable(dev_priv->uncore.regs);
+
+	gen11_gt_irq_reset(&dev_priv->gt);
+	gen11_display_irq_reset(dev_priv);
+
+	GEN3_IRQ_RESET(uncore, GEN11_GU_MISC_);
+	GEN3_IRQ_RESET(uncore, GEN8_PCU_);
+}
+
 void gen8_irq_power_well_post_enable(struct drm_i915_private *dev_priv,
 				     u8 pipe_mask)
 {
@@ -2976,6 +2993,8 @@ static void icp_hpd_irq_setup(struct drm_i915_private *dev_priv,
 	hotplug_irqs = sde_ddi_mask | sde_tc_mask;
 	enabled_irqs = intel_hpd_enabled_irqs(dev_priv, pins);
 
+	I915_WRITE(SHPD_FILTER_CNT, SHPD_FILTER_CNT_500_ADJ);
+
 	ibx_display_interrupt_update(dev_priv, hotplug_irqs, enabled_irqs);
 
 	icp_hpd_detection_setup(dev_priv, ddi_enable_mask, tc_enable_mask);
@@ -3081,6 +3100,9 @@ static void spt_hpd_irq_setup(struct drm_i915_private *dev_priv)
 {
 	u32 hotplug_irqs, enabled_irqs;
 
+	if (INTEL_PCH_TYPE(dev_priv) >= PCH_CNP)
+		I915_WRITE(SHPD_FILTER_CNT, SHPD_FILTER_CNT_500_ADJ);
+
 	hotplug_irqs = SDE_HOTPLUG_MASK_SPT;
 	enabled_irqs = intel_hpd_enabled_irqs(dev_priv, hpd_spt);
 
@@ -3597,7 +3619,7 @@ static irqreturn_t i8xx_irq_handler(int irq, void *arg)
 		intel_uncore_write16(&dev_priv->uncore, GEN2_IIR, iir);
 
 		if (iir & I915_USER_INTERRUPT)
-			intel_engine_breadcrumbs_irq(dev_priv->engine[RCS0]);
+			intel_engine_signal_breadcrumbs(dev_priv->engine[RCS0]);
 
 		if (iir & I915_MASTER_ERROR_INTERRUPT)
 			i8xx_error_irq_handler(dev_priv, eir, eir_stuck);
@@ -3702,7 +3724,7 @@ static irqreturn_t i915_irq_handler(int irq, void *arg)
 		I915_WRITE(GEN2_IIR, iir);
 
 		if (iir & I915_USER_INTERRUPT)
-			intel_engine_breadcrumbs_irq(dev_priv->engine[RCS0]);
+			intel_engine_signal_breadcrumbs(dev_priv->engine[RCS0]);
 
 		if (iir & I915_MASTER_ERROR_INTERRUPT)
 			i9xx_error_irq_handler(dev_priv, eir, eir_stuck);
@@ -3844,10 +3866,10 @@ static irqreturn_t i965_irq_handler(int irq, void *arg)
 		I915_WRITE(GEN2_IIR, iir);
 
 		if (iir & I915_USER_INTERRUPT)
-			intel_engine_breadcrumbs_irq(dev_priv->engine[RCS0]);
+			intel_engine_signal_breadcrumbs(dev_priv->engine[RCS0]);
 
 		if (iir & I915_BSD_USER_INTERRUPT)
-			intel_engine_breadcrumbs_irq(dev_priv->engine[VCS0]);
+			intel_engine_signal_breadcrumbs(dev_priv->engine[VCS0]);
 
 		if (iir & I915_MASTER_ERROR_INTERRUPT)
 			i9xx_error_irq_handler(dev_priv, eir, eir_stuck);
diff --git a/drivers/gpu/drm/i915/i915_memcpy.c b/drivers/gpu/drm/i915/i915_memcpy.c
index 07b04b0acb77..fdd550405fd3 100644
--- a/drivers/gpu/drm/i915/i915_memcpy.c
+++ b/drivers/gpu/drm/i915/i915_memcpy.c
@@ -27,6 +27,12 @@
 
 #include "i915_memcpy.h"
 
+#if IS_ENABLED(CONFIG_DRM_I915_DEBUG)
+#define CI_BUG_ON(expr) BUG_ON(expr)
+#else
+#define CI_BUG_ON(expr) BUILD_BUG_ON_INVALID(expr)
+#endif
+
 static DEFINE_STATIC_KEY_FALSE(has_movntdqa);
 
 #ifdef CONFIG_AS_MOVNTDQA
@@ -34,7 +40,6 @@ static void __memcpy_ntdqa(void *dst, const void *src, unsigned long len)
 {
 	kernel_fpu_begin();
 
-	len >>= 4;
 	while (len >= 4) {
 		asm("movntdqa   (%0), %%xmm0\n"
 		    "movntdqa 16(%0), %%xmm1\n"
@@ -59,6 +64,38 @@ static void __memcpy_ntdqa(void *dst, const void *src, unsigned long len)
 
 	kernel_fpu_end();
 }
+
+static void __memcpy_ntdqu(void *dst, const void *src, unsigned long len)
+{
+	kernel_fpu_begin();
+
+	while (len >= 4) {
+		asm("movntdqa   (%0), %%xmm0\n"
+		    "movntdqa 16(%0), %%xmm1\n"
+		    "movntdqa 32(%0), %%xmm2\n"
+		    "movntdqa 48(%0), %%xmm3\n"
+		    "movups %%xmm0,   (%1)\n"
+		    "movups %%xmm1, 16(%1)\n"
+		    "movups %%xmm2, 32(%1)\n"
+		    "movups %%xmm3, 48(%1)\n"
+		    :: "r" (src), "r" (dst) : "memory");
+		src += 64;
+		dst += 64;
+		len -= 4;
+	}
+	while (len--) {
+		asm("movntdqa (%0), %%xmm0\n"
+		    "movups %%xmm0, (%1)\n"
+		    :: "r" (src), "r" (dst) : "memory");
+		src += 16;
+		dst += 16;
+	}
+
+	kernel_fpu_end();
+}
+#else
+static void __memcpy_ntdqa(void *dst, const void *src, unsigned long len) {}
+static void __memcpy_ntdqu(void *dst, const void *src, unsigned long len) {}
 #endif
 
 /**
@@ -83,17 +120,47 @@ bool i915_memcpy_from_wc(void *dst, const void *src, unsigned long len)
 	if (unlikely(((unsigned long)dst | (unsigned long)src | len) & 15))
 		return false;
 
-#ifdef CONFIG_AS_MOVNTDQA
 	if (static_branch_likely(&has_movntdqa)) {
 		if (likely(len))
-			__memcpy_ntdqa(dst, src, len);
+			__memcpy_ntdqa(dst, src, len >> 4);
 		return true;
 	}
-#endif
 
 	return false;
 }
 
+/**
+ * i915_unaligned_memcpy_from_wc: perform a mostly accelerated read from WC
+ * @dst: destination pointer
+ * @src: source pointer
+ * @len: how many bytes to copy
+ *
+ * Like i915_memcpy_from_wc(), the unaligned variant copies @len bytes from
+ * @src to @dst using * non-temporal instructions where available, but
+ * accepts that its arguments may not be aligned, but are valid for the
+ * potential 16-byte read past the end.
+ */
+void i915_unaligned_memcpy_from_wc(void *dst, void *src, unsigned long len)
+{
+	unsigned long addr;
+
+	CI_BUG_ON(!i915_has_memcpy_from_wc());
+
+	addr = (unsigned long)src;
+	if (!IS_ALIGNED(addr, 16)) {
+		unsigned long x = min(ALIGN(addr, 16) - addr, len);
+
+		memcpy(dst, src, x);
+
+		len -= x;
+		dst += x;
+		src += x;
+	}
+
+	if (likely(len))
+		__memcpy_ntdqu(dst, src, DIV_ROUND_UP(len, 16));
+}
+
 void i915_memcpy_init_early(struct drm_i915_private *dev_priv)
 {
 	/*
diff --git a/drivers/gpu/drm/i915/i915_memcpy.h b/drivers/gpu/drm/i915/i915_memcpy.h
index 970d84b16987..e36d30edd987 100644
--- a/drivers/gpu/drm/i915/i915_memcpy.h
+++ b/drivers/gpu/drm/i915/i915_memcpy.h
@@ -11,7 +11,9 @@
 struct drm_i915_private;
 
 void i915_memcpy_init_early(struct drm_i915_private *i915);
+
 bool i915_memcpy_from_wc(void *dst, const void *src, unsigned long len);
+void i915_unaligned_memcpy_from_wc(void *dst, void *src, unsigned long len);
 
 /* The movntdqa instructions used for memcpy-from-wc require 16-byte alignment,
  * as well as SSE4.1 support. i915_memcpy_from_wc() will report if it cannot
diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c
index 1bb701d32a5d..9571611b4b16 100644
--- a/drivers/gpu/drm/i915/i915_pci.c
+++ b/drivers/gpu/drm/i915/i915_pci.c
@@ -30,6 +30,7 @@
 #include "display/intel_fbdev.h"
 
 #include "i915_drv.h"
+#include "i915_perf.h"
 #include "i915_globals.h"
 #include "i915_selftest.h"
 
@@ -436,7 +437,7 @@ static const struct intel_device_info intel_sandybridge_m_gt2_info = {
 	.has_rc6 = 1, \
 	.has_rc6p = 1, \
 	.has_rps = true, \
-	.ppgtt_type = INTEL_PPGTT_ALIASING, \
+	.ppgtt_type = INTEL_PPGTT_FULL, \
 	.ppgtt_size = 31, \
 	IVB_PIPE_OFFSETS, \
 	IVB_CURSOR_OFFSETS, \
@@ -493,7 +494,7 @@ static const struct intel_device_info intel_valleyview_info = {
 	.has_rps = true,
 	.display.has_gmch = 1,
 	.display.has_hotplug = 1,
-	.ppgtt_type = INTEL_PPGTT_ALIASING,
+	.ppgtt_type = INTEL_PPGTT_FULL,
 	.ppgtt_size = 31,
 	.has_snoop = true,
 	.has_coherent_ggtt = false,
@@ -897,6 +898,8 @@ static const struct pci_device_id pciidlist[] = {
 	INTEL_WHL_U_GT3_IDS(&intel_coffeelake_gt3_info),
 	INTEL_CML_GT1_IDS(&intel_coffeelake_gt1_info),
 	INTEL_CML_GT2_IDS(&intel_coffeelake_gt2_info),
+	INTEL_CML_U_GT1_IDS(&intel_coffeelake_gt1_info),
+	INTEL_CML_U_GT2_IDS(&intel_coffeelake_gt2_info),
 	INTEL_CNL_IDS(&intel_cannonlake_info),
 	INTEL_ICL_11_IDS(&intel_icelake_11_info),
 	INTEL_EHL_IDS(&intel_elkhartlake_info),
@@ -1003,6 +1006,12 @@ static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		return err > 0 ? -ENOTTY : err;
 	}
 
+	err = i915_perf_selftests(pdev);
+	if (err) {
+		i915_pci_remove(pdev);
+		return err > 0 ? -ENOTTY : err;
+	}
+
 	return 0;
 }
 
@@ -1045,7 +1054,12 @@ static int __init i915_init(void)
 		return 0;
 	}
 
-	return pci_register_driver(&i915_pci_driver);
+	err = pci_register_driver(&i915_pci_driver);
+	if (err)
+		return err;
+
+	i915_perf_sysctl_register();
+	return 0;
 }
 
 static void __exit i915_exit(void)
@@ -1053,6 +1067,7 @@ static void __exit i915_exit(void)
 	if (!i915_pci_driver.driver.owner)
 		return;
 
+	i915_perf_sysctl_unregister();
 	pci_unregister_driver(&i915_pci_driver);
 	i915_globals_exit();
 }
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 2ae14bc14931..84350c7bc711 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -387,6 +387,8 @@ struct i915_oa_config_bo {
 	struct i915_vma *vma;
 };
 
+static struct ctl_table_header *sysctl_header;
+
 static enum hrtimer_restart oa_poll_check_timer_cb(struct hrtimer *hrtimer);
 
 void i915_oa_config_release(struct kref *ref)
@@ -1777,6 +1779,8 @@ static int alloc_noa_wait(struct i915_perf_stream *stream)
 	*cs++ = MI_MATH_ADD;
 	*cs++ = MI_MATH_STOREINV(MI_MATH_REG(JUMP_PREDICATE), MI_MATH_REG_CF);
 
+	*cs++ = MI_ARB_CHECK;
+
 	/*
 	 * Transfer the result into the predicate register to be used for the
 	 * predicated jump.
@@ -1966,7 +1970,9 @@ static int emit_oa_config(struct i915_perf_stream *stream,
 	if (err)
 		goto err_vma_put;
 
+	intel_engine_pm_get(ce->engine);
 	rq = i915_request_create(ce);
+	intel_engine_pm_put(ce->engine);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto err_vma_unpin;
@@ -2155,7 +2161,7 @@ static int gen8_modify_context(struct intel_context *ce,
 
 	lockdep_assert_held(&ce->pin_mutex);
 
-	rq = i915_request_create(ce->engine->kernel_context);
+	rq = intel_engine_create_kernel_request(ce->engine);
 	if (IS_ERR(rq))
 		return PTR_ERR(rq);
 
@@ -2317,9 +2323,6 @@ static int oa_configure_all_contexts(struct i915_perf_stream *stream,
 	 */
 	spin_lock(&i915->gem.contexts.lock);
 	list_for_each_entry_safe(ctx, cn, &i915->gem.contexts.list, link) {
-		if (ctx == i915->kernel_context)
-			continue;
-
 		if (!kref_get_unless_zero(&ctx->ref))
 			continue;
 
@@ -2455,6 +2458,13 @@ static int gen8_enable_metric_set(struct i915_perf_stream *stream)
 	return emit_oa_config(stream, oa_config, oa_context(stream));
 }
 
+static u32 oag_report_ctx_switches(const struct i915_perf_stream *stream)
+{
+	return _MASKED_FIELD(GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS,
+			     (stream->sample_flags & SAMPLE_OA_REPORT) ?
+			     0 : GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS);
+}
+
 static int gen12_enable_metric_set(struct i915_perf_stream *stream)
 {
 	struct intel_uncore *uncore = stream->uncore;
@@ -2468,12 +2478,10 @@ static int gen12_enable_metric_set(struct i915_perf_stream *stream)
 			   _MASKED_BIT_ENABLE(GEN12_OAG_OA_DEBUG_DISABLE_CLK_RATIO_REPORTS |
 					      GEN12_OAG_OA_DEBUG_INCLUDE_CLK_RATIO) |
 			   /*
-			    * If the user didn't require OA reports, instruct the
-			    * hardware not to emit ctx switch reports.
+			    * If the user didn't require OA reports, instruct
+			    * the hardware not to emit ctx switch reports.
 			    */
-			   !(stream->sample_flags & SAMPLE_OA_REPORT) ?
-			   _MASKED_BIT_ENABLE(GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS) :
-			   _MASKED_BIT_DISABLE(GEN12_OAG_OA_DEBUG_DISABLE_CTX_SWITCH_REPORTS));
+			   oag_report_ctx_switches(stream));
 
 	intel_uncore_write(uncore, GEN12_OAG_OAGLBCTXCTRL, periodic ?
 			   (GEN12_OAG_OAGLBCTXCTRL_COUNTER_RESUME |
@@ -3101,7 +3109,7 @@ static void i915_perf_enable_locked(struct i915_perf_stream *stream)
 		stream->ops->enable(stream);
 
 	if (stream->hold_preemption)
-		i915_gem_context_set_nopreempt(stream->ctx);
+		intel_context_set_nopreempt(stream->pinned_ctx);
 }
 
 /**
@@ -3127,7 +3135,7 @@ static void i915_perf_disable_locked(struct i915_perf_stream *stream)
 	stream->enabled = false;
 
 	if (stream->hold_preemption)
-		i915_gem_context_clear_nopreempt(stream->ctx);
+		intel_context_clear_nopreempt(stream->pinned_ctx);
 
 	if (stream->ops->disable)
 		stream->ops->disable(stream);
@@ -3981,7 +3989,7 @@ int i915_perf_add_config_ioctl(struct drm_device *dev, void *data,
 	struct i915_perf *perf = &to_i915(dev)->perf;
 	struct drm_i915_perf_oa_config *args = data;
 	struct i915_oa_config *oa_config, *tmp;
-	static struct i915_oa_reg *regs;
+	struct i915_oa_reg *regs;
 	int err, id;
 
 	if (!perf->i915) {
@@ -4219,7 +4227,7 @@ static struct ctl_table dev_root[] = {
 };
 
 /**
- * i915_perf_init - initialize i915-perf state on module load
+ * i915_perf_init - initialize i915-perf state on module bind
  * @i915: i915 device instance
  *
  * Initializes i915-perf state without exposing anything to userspace.
@@ -4336,7 +4344,6 @@ void i915_perf_init(struct drm_i915_private *i915)
 
 		oa_sample_rate_hard_limit = 1000 *
 			(RUNTIME_INFO(i915)->cs_timestamp_frequency_khz / 2);
-		perf->sysctl_header = register_sysctl_table(dev_root);
 
 		mutex_init(&perf->metrics_lock);
 		idr_init(&perf->metrics_idr);
@@ -4372,6 +4379,16 @@ static int destroy_config(int id, void *p, void *data)
 	return 0;
 }
 
+void i915_perf_sysctl_register(void)
+{
+	sysctl_header = register_sysctl_table(dev_root);
+}
+
+void i915_perf_sysctl_unregister(void)
+{
+	unregister_sysctl_table(sysctl_header);
+}
+
 /**
  * i915_perf_fini - Counter part to i915_perf_init()
  * @i915: i915 device instance
@@ -4386,8 +4403,6 @@ void i915_perf_fini(struct drm_i915_private *i915)
 	idr_for_each(&perf->metrics_idr, destroy_config, perf);
 	idr_destroy(&perf->metrics_idr);
 
-	unregister_sysctl_table(perf->sysctl_header);
-
 	memset(&perf->ops, 0, sizeof(perf->ops));
 	perf->i915 = NULL;
 }
diff --git a/drivers/gpu/drm/i915/i915_perf.h b/drivers/gpu/drm/i915/i915_perf.h
index 4ceebce72060..882fdd0a7680 100644
--- a/drivers/gpu/drm/i915/i915_perf.h
+++ b/drivers/gpu/drm/i915/i915_perf.h
@@ -23,6 +23,8 @@ void i915_perf_fini(struct drm_i915_private *i915);
 void i915_perf_register(struct drm_i915_private *i915);
 void i915_perf_unregister(struct drm_i915_private *i915);
 int i915_perf_ioctl_version(void);
+void i915_perf_sysctl_register(void);
+void i915_perf_sysctl_unregister(void);
 
 int i915_perf_open_ioctl(struct drm_device *dev, void *data,
 			 struct drm_file *file);
diff --git a/drivers/gpu/drm/i915/i915_perf_types.h b/drivers/gpu/drm/i915/i915_perf_types.h
index 74ddc20a0d37..45e581455f5d 100644
--- a/drivers/gpu/drm/i915/i915_perf_types.h
+++ b/drivers/gpu/drm/i915/i915_perf_types.h
@@ -380,7 +380,6 @@ struct i915_perf {
 	struct drm_i915_private *i915;
 
 	struct kobject *metrics_kobj;
-	struct ctl_table_header *sysctl_header;
 
 	/*
 	 * Lock associated with adding/modifying/removing OA configs
diff --git a/drivers/gpu/drm/i915/i915_pmu.c b/drivers/gpu/drm/i915/i915_pmu.c
index 2814218c5ba1..f3ef6700a5f2 100644
--- a/drivers/gpu/drm/i915/i915_pmu.c
+++ b/drivers/gpu/drm/i915/i915_pmu.c
@@ -144,61 +144,40 @@ static inline s64 ktime_since(const ktime_t kt)
 	return ktime_to_ns(ktime_sub(ktime_get(), kt));
 }
 
-static u64 __pmu_estimate_rc6(struct i915_pmu *pmu)
-{
-	u64 val;
-
-	/*
-	 * We think we are runtime suspended.
-	 *
-	 * Report the delta from when the device was suspended to now,
-	 * on top of the last known real value, as the approximated RC6
-	 * counter value.
-	 */
-	val = ktime_since(pmu->sleep_last);
-	val += pmu->sample[__I915_SAMPLE_RC6].cur;
-
-	pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur = val;
-
-	return val;
-}
-
-static u64 __pmu_update_rc6(struct i915_pmu *pmu, u64 val)
-{
-	/*
-	 * If we are coming back from being runtime suspended we must
-	 * be careful not to report a larger value than returned
-	 * previously.
-	 */
-	if (val >= pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur) {
-		pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur = 0;
-		pmu->sample[__I915_SAMPLE_RC6].cur = val;
-	} else {
-		val = pmu->sample[__I915_SAMPLE_RC6_ESTIMATED].cur;
-	}
-
-	return val;
-}
-
 static u64 get_rc6(struct intel_gt *gt)
 {
 	struct drm_i915_private *i915 = gt->i915;
 	struct i915_pmu *pmu = &i915->pmu;
 	unsigned long flags;
+	bool awake = false;
 	u64 val;
 
-	val = 0;
 	if (intel_gt_pm_get_if_awake(gt)) {
 		val = __get_rc6(gt);
 		intel_gt_pm_put_async(gt);
+		awake = true;
 	}
 
 	spin_lock_irqsave(&pmu->lock, flags);
 
-	if (val)
-		val = __pmu_update_rc6(pmu, val);
+	if (awake) {
+		pmu->sample[__I915_SAMPLE_RC6].cur = val;
+	} else {
+		/*
+		 * We think we are runtime suspended.
+		 *
+		 * Report the delta from when the device was suspended to now,
+		 * on top of the last known real value, as the approximated RC6
+		 * counter value.
+		 */
+		val = ktime_since(pmu->sleep_last);
+		val += pmu->sample[__I915_SAMPLE_RC6].cur;
+	}
+
+	if (val < pmu->sample[__I915_SAMPLE_RC6_LAST_REPORTED].cur)
+		val = pmu->sample[__I915_SAMPLE_RC6_LAST_REPORTED].cur;
 	else
-		val = __pmu_estimate_rc6(pmu);
+		pmu->sample[__I915_SAMPLE_RC6_LAST_REPORTED].cur = val;
 
 	spin_unlock_irqrestore(&pmu->lock, flags);
 
@@ -210,20 +189,11 @@ static void park_rc6(struct drm_i915_private *i915)
 	struct i915_pmu *pmu = &i915->pmu;
 
 	if (pmu->enable & config_enabled_mask(I915_PMU_RC6_RESIDENCY))
-		__pmu_update_rc6(pmu, __get_rc6(&i915->gt));
+		pmu->sample[__I915_SAMPLE_RC6].cur = __get_rc6(&i915->gt);
 
 	pmu->sleep_last = ktime_get();
 }
 
-static void unpark_rc6(struct drm_i915_private *i915)
-{
-	struct i915_pmu *pmu = &i915->pmu;
-
-	/* Estimate how long we slept and accumulate that into rc6 counters */
-	if (pmu->enable & config_enabled_mask(I915_PMU_RC6_RESIDENCY))
-		__pmu_estimate_rc6(pmu);
-}
-
 #else
 
 static u64 get_rc6(struct intel_gt *gt)
@@ -232,7 +202,6 @@ static u64 get_rc6(struct intel_gt *gt)
 }
 
 static void park_rc6(struct drm_i915_private *i915) {}
-static void unpark_rc6(struct drm_i915_private *i915) {}
 
 #endif
 
@@ -281,8 +250,6 @@ void i915_pmu_gt_unparked(struct drm_i915_private *i915)
 	 */
 	__i915_pmu_maybe_start_timer(pmu);
 
-	unpark_rc6(i915);
-
 	spin_unlock_irq(&pmu->lock);
 }
 
@@ -292,6 +259,16 @@ add_sample(struct i915_pmu_sample *sample, u32 val)
 	sample->cur += val;
 }
 
+static bool exclusive_mmio_access(const struct drm_i915_private *i915)
+{
+	/*
+	 * We have to avoid concurrent mmio cache line access on gen7 or
+	 * risk a machine hang. For a fun history lesson dig out the old
+	 * userspace intel_gpu_top and run it on Ivybridge or Haswell!
+	 */
+	return IS_GEN(i915, 7);
+}
+
 static void
 engines_sample(struct intel_gt *gt, unsigned int period_ns)
 {
@@ -302,8 +279,12 @@ engines_sample(struct intel_gt *gt, unsigned int period_ns)
 	if ((i915->pmu.enable & ENGINE_SAMPLE_MASK) == 0)
 		return;
 
+	if (!intel_gt_pm_is_awake(gt))
+		return;
+
 	for_each_engine(engine, gt, id) {
 		struct intel_engine_pmu *pmu = &engine->pmu;
+		spinlock_t *mmio_lock;
 		unsigned long flags;
 		bool busy;
 		u32 val;
@@ -311,7 +292,12 @@ engines_sample(struct intel_gt *gt, unsigned int period_ns)
 		if (!intel_engine_pm_get_if_awake(engine))
 			continue;
 
-		spin_lock_irqsave(&engine->uncore->lock, flags);
+		mmio_lock = NULL;
+		if (exclusive_mmio_access(i915))
+			mmio_lock = &engine->uncore->lock;
+
+		if (unlikely(mmio_lock))
+			spin_lock_irqsave(mmio_lock, flags);
 
 		val = ENGINE_READ_FW(engine, RING_CTL);
 		if (val == 0) /* powerwell off => engine idle */
@@ -342,7 +328,8 @@ engines_sample(struct intel_gt *gt, unsigned int period_ns)
 			add_sample(&pmu->sample[I915_SAMPLE_BUSY], period_ns);
 
 skip:
-		spin_unlock_irqrestore(&engine->uncore->lock, flags);
+		if (unlikely(mmio_lock))
+			spin_unlock_irqrestore(mmio_lock, flags);
 		intel_engine_pm_put_async(engine);
 	}
 }
@@ -353,6 +340,13 @@ add_sample_mult(struct i915_pmu_sample *sample, u32 val, u32 mul)
 	sample->cur += mul_u32_u32(val, mul);
 }
 
+static bool frequency_sampling_enabled(struct i915_pmu *pmu)
+{
+	return pmu->enable &
+	       (config_enabled_mask(I915_PMU_ACTUAL_FREQUENCY) |
+		config_enabled_mask(I915_PMU_REQUESTED_FREQUENCY));
+}
+
 static void
 frequency_sample(struct intel_gt *gt, unsigned int period_ns)
 {
@@ -361,19 +355,33 @@ frequency_sample(struct intel_gt *gt, unsigned int period_ns)
 	struct i915_pmu *pmu = &i915->pmu;
 	struct intel_rps *rps = &gt->rps;
 
+	if (!frequency_sampling_enabled(pmu))
+		return;
+
+	/* Report 0/0 (actual/requested) frequency while parked. */
+	if (!intel_gt_pm_get_if_awake(gt))
+		return;
+
 	if (pmu->enable & config_enabled_mask(I915_PMU_ACTUAL_FREQUENCY)) {
 		u32 val;
 
-		val = rps->cur_freq;
-		if (intel_gt_pm_get_if_awake(gt)) {
-			val = intel_uncore_read_notrace(uncore, GEN6_RPSTAT1);
-			val = intel_get_cagf(rps, val);
-			intel_gt_pm_put_async(gt);
-		}
+		/*
+		 * We take a quick peek here without using forcewake
+		 * so that we don't perturb the system under observation
+		 * (forcewake => !rc6 => increased power use). We expect
+		 * that if the read fails because it is outside of the
+		 * mmio power well, then it will return 0 -- in which
+		 * case we assume the system is running at the intended
+		 * frequency. Fortunately, the read should rarely fail!
+		 */
+		val = intel_uncore_read_fw(uncore, GEN6_RPSTAT1);
+		if (val)
+			val = intel_rps_get_cagf(rps, val);
+		else
+			val = rps->cur_freq;
 
 		add_sample_mult(&pmu->sample[__I915_SAMPLE_FREQ_ACT],
-				intel_gpu_freq(rps, val),
-				period_ns / 1000);
+				intel_gpu_freq(rps, val), period_ns / 1000);
 	}
 
 	if (pmu->enable & config_enabled_mask(I915_PMU_REQUESTED_FREQUENCY)) {
@@ -381,6 +389,8 @@ frequency_sample(struct intel_gt *gt, unsigned int period_ns)
 				intel_gpu_freq(rps, rps->cur_freq),
 				period_ns / 1000);
 	}
+
+	intel_gt_pm_put_async(gt);
 }
 
 static enum hrtimer_restart i915_sample(struct hrtimer *hrtimer)
diff --git a/drivers/gpu/drm/i915/i915_pmu.h b/drivers/gpu/drm/i915/i915_pmu.h
index bf52e3983631..6c1647c5daf2 100644
--- a/drivers/gpu/drm/i915/i915_pmu.h
+++ b/drivers/gpu/drm/i915/i915_pmu.h
@@ -18,7 +18,7 @@ enum {
 	__I915_SAMPLE_FREQ_ACT = 0,
 	__I915_SAMPLE_FREQ_REQ,
 	__I915_SAMPLE_RC6,
-	__I915_SAMPLE_RC6_ESTIMATED,
+	__I915_SAMPLE_RC6_LAST_REPORTED,
 	__I915_NUM_PMU_SAMPLERS
 };
 
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 73079b503724..bbfedeb00b7f 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -5042,14 +5042,20 @@ enum {
 #define   BLM_PCH_POLARITY			(1 << 29)
 #define BLC_PWM_PCH_CTL2	_MMIO(0xc8254)
 
-#define UTIL_PIN_CTL		_MMIO(0x48400)
-#define   UTIL_PIN_ENABLE	(1 << 31)
-
-#define   UTIL_PIN_PIPE(x)     ((x) << 29)
-#define   UTIL_PIN_PIPE_MASK   (3 << 29)
-#define   UTIL_PIN_MODE_PWM    (1 << 24)
-#define   UTIL_PIN_MODE_MASK   (0xf << 24)
-#define   UTIL_PIN_POLARITY    (1 << 22)
+#define UTIL_PIN_CTL			_MMIO(0x48400)
+#define   UTIL_PIN_ENABLE		(1 << 31)
+#define   UTIL_PIN_PIPE_MASK		(3 << 29)
+#define   UTIL_PIN_PIPE(x)		((x) << 29)
+#define   UTIL_PIN_MODE_MASK		(0xf << 24)
+#define   UTIL_PIN_MODE_DATA		(0 << 24)
+#define   UTIL_PIN_MODE_PWM		(1 << 24)
+#define   UTIL_PIN_MODE_VBLANK		(4 << 24)
+#define   UTIL_PIN_MODE_VSYNC		(5 << 24)
+#define   UTIL_PIN_MODE_EYE_LEVEL	(8 << 24)
+#define   UTIL_PIN_OUTPUT_DATA		(1 << 23)
+#define   UTIL_PIN_POLARITY		(1 << 22)
+#define   UTIL_PIN_DIRECTION_INPUT	(1 << 19)
+#define   UTIL_PIN_INPUT_DATA		(1 << 16)
 
 /* BXT backlight register definition. */
 #define _BXT_BLC_PWM_CTL1			0xC8250
@@ -5751,7 +5757,8 @@ enum {
 #define   PIPECONF_DOUBLE_WIDE	(1 << 30)
 #define   I965_PIPECONF_ACTIVE	(1 << 30)
 #define   PIPECONF_DSI_PLL_LOCKED	(1 << 29) /* vlv & pipe A only */
-#define   PIPECONF_FRAME_START_DELAY_MASK (3 << 27)
+#define   PIPECONF_FRAME_START_DELAY_MASK	(3 << 27) /* pre-hsw */
+#define   PIPECONF_FRAME_START_DELAY(x)		((x) << 27) /* pre-hsw: 0-3 */
 #define   PIPECONF_SINGLE_WIDE	0
 #define   PIPECONF_PIPE_UNLOCKED 0
 #define   PIPECONF_PIPE_LOCKED	(1 << 25)
@@ -6359,6 +6366,7 @@ enum {
 #define   DISPPLANE_RGBX101010			(0x8 << 26)
 #define   DISPPLANE_RGBA101010			(0x9 << 26)
 #define   DISPPLANE_BGRX101010			(0xa << 26)
+#define   DISPPLANE_BGRA101010			(0xb << 26)
 #define   DISPPLANE_RGBX161616			(0xc << 26)
 #define   DISPPLANE_RGBX888			(0xe << 26)
 #define   DISPPLANE_RGBA888			(0xf << 26)
@@ -6633,12 +6641,15 @@ enum {
 #define   SP_ENABLE			(1 << 31)
 #define   SP_GAMMA_ENABLE		(1 << 30)
 #define   SP_PIXFORMAT_MASK		(0xf << 26)
-#define   SP_FORMAT_YUV422		(0 << 26)
-#define   SP_FORMAT_BGR565		(5 << 26)
-#define   SP_FORMAT_BGRX8888		(6 << 26)
-#define   SP_FORMAT_BGRA8888		(7 << 26)
-#define   SP_FORMAT_RGBX1010102		(8 << 26)
-#define   SP_FORMAT_RGBA1010102		(9 << 26)
+#define   SP_FORMAT_YUV422		(0x0 << 26)
+#define   SP_FORMAT_8BPP		(0x2 << 26)
+#define   SP_FORMAT_BGR565		(0x5 << 26)
+#define   SP_FORMAT_BGRX8888		(0x6 << 26)
+#define   SP_FORMAT_BGRA8888		(0x7 << 26)
+#define   SP_FORMAT_RGBX1010102		(0x8 << 26)
+#define   SP_FORMAT_RGBA1010102		(0x9 << 26)
+#define   SP_FORMAT_BGRX1010102		(0xa << 26) /* CHV pipe B */
+#define   SP_FORMAT_BGRA1010102		(0xb << 26) /* CHV pipe B */
 #define   SP_FORMAT_RGBX8888		(0xe << 26)
 #define   SP_FORMAT_RGBA8888		(0xf << 26)
 #define   SP_ALPHA_PREMULTIPLY		(1 << 23) /* CHV pipe B */
@@ -6789,6 +6800,7 @@ enum {
 #define   PLANE_CTL_YUV422_VYUY			(3 << 16)
 #define   PLANE_CTL_RENDER_DECOMPRESSION_ENABLE	(1 << 15)
 #define   PLANE_CTL_TRICKLE_FEED_DISABLE	(1 << 14)
+#define   PLANE_CTL_CLEAR_COLOR_DISABLE		(1 << 13) /* TGL+ */
 #define   PLANE_CTL_PLANE_GAMMA_DISABLE		(1 << 13) /* Pre-GLK */
 #define   PLANE_CTL_TILED_MASK			(0x7 << 10)
 #define   PLANE_CTL_TILED_LINEAR		(0 << 10)
@@ -7506,11 +7518,15 @@ enum {
 #define GEN8_DE_PORT_IMR _MMIO(0x44444)
 #define GEN8_DE_PORT_IIR _MMIO(0x44448)
 #define GEN8_DE_PORT_IER _MMIO(0x4444c)
+#define  DSI1_NON_TE			(1 << 31)
+#define  DSI0_NON_TE			(1 << 30)
 #define  ICL_AUX_CHANNEL_E		(1 << 29)
 #define  CNL_AUX_CHANNEL_F		(1 << 28)
 #define  GEN9_AUX_CHANNEL_D		(1 << 27)
 #define  GEN9_AUX_CHANNEL_C		(1 << 26)
 #define  GEN9_AUX_CHANNEL_B		(1 << 25)
+#define  DSI1_TE			(1 << 24)
+#define  DSI0_TE			(1 << 23)
 #define  BXT_DE_PORT_HP_DDIC		(1 << 5)
 #define  BXT_DE_PORT_HP_DDIB		(1 << 4)
 #define  BXT_DE_PORT_HP_DDIA		(1 << 3)
@@ -7725,6 +7741,8 @@ enum {
 					    [TRANSCODER_B] = _CHICKEN_TRANS_B, \
 					    [TRANSCODER_C] = _CHICKEN_TRANS_C, \
 					    [TRANSCODER_D] = _CHICKEN_TRANS_D))
+#define  HSW_FRAME_START_DELAY_MASK	(3 << 27)
+#define  HSW_FRAME_START_DELAY(x)	((x) << 27) /* 0-3 */
 #define  VSC_DATA_SEL_SOFTWARE_CONTROL	(1 << 25) /* GLK and CNL+ */
 #define  DDI_TRAINING_OVERRIDE_ENABLE	(1 << 19)
 #define  DDI_TRAINING_OVERRIDE_VALUE	(1 << 18)
@@ -7748,6 +7766,14 @@ enum {
 #define GEN7_MSG_CTL	_MMIO(0x45010)
 #define  WAIT_FOR_PCH_RESET_ACK		(1 << 1)
 #define  WAIT_FOR_PCH_FLR_ACK		(1 << 0)
+
+#define BW_BUDDY1_CTL			_MMIO(0x45140)
+#define BW_BUDDY2_CTL			_MMIO(0x45150)
+#define   BW_BUDDY_DISABLE		REG_BIT(31)
+
+#define BW_BUDDY1_PAGE_MASK		_MMIO(0x45144)
+#define BW_BUDDY2_PAGE_MASK		_MMIO(0x45154)
+
 #define HSW_NDE_RSTWRN_OPT	_MMIO(0x46408)
 #define  RESET_PCH_HANDSHAKE_ENABLE	(1 << 4)
 
@@ -7905,6 +7931,10 @@ enum {
 #define   PIXEL_ROUNDING_TRUNC_FB_PASSTHRU 	(1 << 15)
 #define   PER_PIXEL_ALPHA_BYPASS_EN		(1 << 7)
 
+#define FF_MODE2			_MMIO(0x6604)
+#define   FF_MODE2_TDS_TIMER_MASK	REG_GENMASK(23, 16)
+#define   FF_MODE2_TDS_TIMER_128	REG_FIELD_PREP(FF_MODE2_TDS_TIMER_MASK, 4)
+
 /* PCH */
 
 #define PCH_DISPLAY_BASE	0xc0000u
@@ -8093,6 +8123,10 @@ enum {
 
 #define SHOTPLUG_CTL_TC				_MMIO(0xc4034)
 #define   ICP_TC_HPD_ENABLE(tc_port)		(8 << (tc_port) * 4)
+
+#define SHPD_FILTER_CNT				_MMIO(0xc4038)
+#define   SHPD_FILTER_CNT_500_ADJ		0x001D9
+
 /* Icelake DSC Rate Control Range Parameter Registers */
 #define DSCA_RC_RANGE_PARAMETERS_0		_MMIO(0x6B240)
 #define DSCA_RC_RANGE_PARAMETERS_0_UDW		_MMIO(0x6B240 + 4)
@@ -8443,10 +8477,8 @@ enum {
 #define  TRANS_STATE_MASK       (1 << 30)
 #define  TRANS_STATE_DISABLE    (0 << 30)
 #define  TRANS_STATE_ENABLE     (1 << 30)
-#define  TRANS_FSYNC_DELAY_HB1  (0 << 27)
-#define  TRANS_FSYNC_DELAY_HB2  (1 << 27)
-#define  TRANS_FSYNC_DELAY_HB3  (2 << 27)
-#define  TRANS_FSYNC_DELAY_HB4  (3 << 27)
+#define  TRANS_FRAME_START_DELAY_MASK	(3 << 27) /* ibx */
+#define  TRANS_FRAME_START_DELAY(x)	((x) << 27) /* ibx: 0-3 */
 #define  TRANS_INTERLACE_MASK   (7 << 21)
 #define  TRANS_PROGRESSIVE      (0 << 21)
 #define  TRANS_INTERLACED       (3 << 21)
@@ -8467,6 +8499,7 @@ enum {
 #define  TRANS_CHICKEN2_TIMING_OVERRIDE			(1 << 31)
 #define  TRANS_CHICKEN2_FDI_POLARITY_REVERSED		(1 << 29)
 #define  TRANS_CHICKEN2_FRAME_START_DELAY_MASK		(3 << 27)
+#define  TRANS_CHICKEN2_FRAME_START_DELAY(x)		((x) << 27) /* 0-3 */
 #define  TRANS_CHICKEN2_DISABLE_DEEP_COLOR_COUNTER	(1 << 26)
 #define  TRANS_CHICKEN2_DISABLE_DEEP_COLOR_MODESWITCH	(1 << 25)
 
@@ -9405,11 +9438,9 @@ enum skl_power_gate {
 #define _ICL_AUX_REG_IDX(pw_idx)	((pw_idx) - ICL_PW_CTL_IDX_AUX_A)
 #define _ICL_AUX_ANAOVRD1_A		0x162398
 #define _ICL_AUX_ANAOVRD1_B		0x6C398
-#define _TGL_AUX_ANAOVRD1_C		0x160398
 #define ICL_AUX_ANAOVRD1(pw_idx)	_MMIO(_PICK(_ICL_AUX_REG_IDX(pw_idx), \
 						    _ICL_AUX_ANAOVRD1_A, \
-						    _ICL_AUX_ANAOVRD1_B, \
-						    _TGL_AUX_ANAOVRD1_C))
+						    _ICL_AUX_ANAOVRD1_B))
 #define   ICL_AUX_ANAOVRD1_LDO_BYPASS	(1 << 7)
 #define   ICL_AUX_ANAOVRD1_ENABLE	(1 << 0)
 
@@ -9671,6 +9702,7 @@ enum skl_power_gate {
 #define  TRANS_DDI_EDP_INPUT_A_ONOFF	(4 << 12)
 #define  TRANS_DDI_EDP_INPUT_B_ONOFF	(5 << 12)
 #define  TRANS_DDI_EDP_INPUT_C_ONOFF	(6 << 12)
+#define  TRANS_DDI_EDP_INPUT_D_ONOFF	(7 << 12)
 #define  TRANS_DDI_MST_TRANSPORT_SELECT_MASK	REG_GENMASK(11, 10)
 #define  TRANS_DDI_MST_TRANSPORT_SELECT(trans)	\
 	REG_FIELD_PREP(TRANS_DDI_MST_TRANSPORT_SELECT_MASK, trans)
@@ -10776,6 +10808,57 @@ enum skl_power_gate {
 #define  ICL_ESC_CLK_DIV_SHIFT			0
 #define DSI_MAX_ESC_CLK			20000		/* in KHz */
 
+#define _DSI_CMD_FRMCTL_0		0x6b034
+#define _DSI_CMD_FRMCTL_1		0x6b834
+#define DSI_CMD_FRMCTL(port)		_MMIO_PORT(port,	\
+						   _DSI_CMD_FRMCTL_0,\
+						   _DSI_CMD_FRMCTL_1)
+#define   DSI_FRAME_UPDATE_REQUEST		(1 << 31)
+#define   DSI_PERIODIC_FRAME_UPDATE_ENABLE	(1 << 29)
+#define   DSI_NULL_PACKET_ENABLE		(1 << 28)
+#define   DSI_FRAME_IN_PROGRESS			(1 << 0)
+
+#define _DSI_INTR_MASK_REG_0		0x6b070
+#define _DSI_INTR_MASK_REG_1		0x6b870
+#define DSI_INTR_MASK_REG(port)		_MMIO_PORT(port,	\
+						   _DSI_INTR_MASK_REG_0,\
+						   _DSI_INTR_MASK_REG_1)
+
+#define _DSI_INTR_IDENT_REG_0		0x6b074
+#define _DSI_INTR_IDENT_REG_1		0x6b874
+#define DSI_INTR_IDENT_REG(port)	_MMIO_PORT(port,	\
+						   _DSI_INTR_IDENT_REG_0,\
+						   _DSI_INTR_IDENT_REG_1)
+#define   DSI_TE_EVENT				(1 << 31)
+#define   DSI_RX_DATA_OR_BTA_TERMINATED		(1 << 30)
+#define   DSI_TX_DATA				(1 << 29)
+#define   DSI_ULPS_ENTRY_DONE			(1 << 28)
+#define   DSI_NON_TE_TRIGGER_RECEIVED		(1 << 27)
+#define   DSI_HOST_CHKSUM_ERROR			(1 << 26)
+#define   DSI_HOST_MULTI_ECC_ERROR		(1 << 25)
+#define   DSI_HOST_SINGL_ECC_ERROR		(1 << 24)
+#define   DSI_HOST_CONTENTION_DETECTED		(1 << 23)
+#define   DSI_HOST_FALSE_CONTROL_ERROR		(1 << 22)
+#define   DSI_HOST_TIMEOUT_ERROR		(1 << 21)
+#define   DSI_HOST_LOW_POWER_TX_SYNC_ERROR	(1 << 20)
+#define   DSI_HOST_ESCAPE_MODE_ENTRY_ERROR	(1 << 19)
+#define   DSI_FRAME_UPDATE_DONE			(1 << 16)
+#define   DSI_PROTOCOL_VIOLATION_REPORTED	(1 << 15)
+#define   DSI_INVALID_TX_LENGTH			(1 << 13)
+#define   DSI_INVALID_VC			(1 << 12)
+#define   DSI_INVALID_DATA_TYPE			(1 << 11)
+#define   DSI_PERIPHERAL_CHKSUM_ERROR		(1 << 10)
+#define   DSI_PERIPHERAL_MULTI_ECC_ERROR	(1 << 9)
+#define   DSI_PERIPHERAL_SINGLE_ECC_ERROR	(1 << 8)
+#define   DSI_PERIPHERAL_CONTENTION_DETECTED	(1 << 7)
+#define   DSI_PERIPHERAL_FALSE_CTRL_ERROR	(1 << 6)
+#define   DSI_PERIPHERAL_TIMEOUT_ERROR		(1 << 5)
+#define   DSI_PERIPHERAL_LP_TX_SYNC_ERROR	(1 << 4)
+#define   DSI_PERIPHERAL_ESC_MODE_ENTRY_CMD_ERR	(1 << 3)
+#define   DSI_EOT_SYNC_ERROR			(1 << 2)
+#define   DSI_SOT_SYNC_ERROR			(1 << 1)
+#define   DSI_SOT_ERROR				(1 << 0)
+
 /* Gen4+ Timestamp and Pipe Frame time stamp registers */
 #define GEN4_TIMESTAMP		_MMIO(0x2358)
 #define ILK_TIMESTAMP_HI	_MMIO(0x70070)
@@ -11380,6 +11463,7 @@ enum skl_power_gate {
 #define  CMD_MODE_TE_GATE		(0x1 << 28)
 #define  VIDEO_MODE_SYNC_EVENT		(0x2 << 28)
 #define  VIDEO_MODE_SYNC_PULSE		(0x3 << 28)
+#define  TE_SOURCE_GPIO			(1 << 27)
 #define  LINK_READY			(1 << 20)
 #define  PIX_FMT_MASK			(0x3 << 16)
 #define  PIX_FMT_SHIFT			16
@@ -11632,13 +11716,18 @@ enum skl_power_gate {
 /* MOCS (Memory Object Control State) registers */
 #define GEN9_LNCFCMOCS(i)	_MMIO(0xb020 + (i) * 4)	/* L3 Cache Control */
 
-#define GEN9_GFX_MOCS(i)	_MMIO(0xc800 + (i) * 4)	/* Graphics MOCS registers */
-#define GEN9_MFX0_MOCS(i)	_MMIO(0xc900 + (i) * 4)	/* Media 0 MOCS registers */
-#define GEN9_MFX1_MOCS(i)	_MMIO(0xca00 + (i) * 4)	/* Media 1 MOCS registers */
-#define GEN9_VEBOX_MOCS(i)	_MMIO(0xcb00 + (i) * 4)	/* Video MOCS registers */
-#define GEN9_BLT_MOCS(i)	_MMIO(0xcc00 + (i) * 4)	/* Blitter MOCS registers */
-/* Media decoder 2 MOCS registers */
-#define GEN11_MFX2_MOCS(i)	_MMIO(0x10000 + (i) * 4)
+#define __GEN9_RCS0_MOCS0	0xc800
+#define GEN9_GFX_MOCS(i)	_MMIO(__GEN9_RCS0_MOCS0 + (i) * 4)
+#define __GEN9_VCS0_MOCS0	0xc900
+#define GEN9_MFX0_MOCS(i)	_MMIO(__GEN9_VCS0_MOCS0 + (i) * 4)
+#define __GEN9_VCS1_MOCS0	0xca00
+#define GEN9_MFX1_MOCS(i)	_MMIO(__GEN9_VCS1_MOCS0 + (i) * 4)
+#define __GEN9_VECS0_MOCS0	0xcb00
+#define GEN9_VEBOX_MOCS(i)	_MMIO(__GEN9_VECS0_MOCS0 + (i) * 4)
+#define __GEN9_BCS0_MOCS0	0xcc00
+#define GEN9_BLT_MOCS(i)	_MMIO(__GEN9_BCS0_MOCS0 + (i) * 4)
+#define __GEN11_VCS2_MOCS0	0x10000
+#define GEN11_MFX2_MOCS(i)	_MMIO(__GEN11_VCS2_MOCS0 + (i) * 4)
 
 #define GEN10_SCRATCH_LNCF2		_MMIO(0xb0a0)
 #define   PMFLUSHDONE_LNICRSDROP	(1 << 20)
@@ -11994,7 +12083,7 @@ enum skl_power_gate {
 /* This register controls the Display State Buffer (DSB) engines. */
 #define _DSBSL_INSTANCE_BASE		0x70B00
 #define DSBSL_INSTANCE(pipe, id)	(_DSBSL_INSTANCE_BASE + \
-					 (pipe) * 0x1000 + (id) * 100)
+					 (pipe) * 0x1000 + (id) * 0x100)
 #define DSB_HEAD(pipe, id)		_MMIO(DSBSL_INSTANCE(pipe, id) + 0x0)
 #define DSB_TAIL(pipe, id)		_MMIO(DSBSL_INSTANCE(pipe, id) + 0x4)
 #define DSB_CTRL(pipe, id)		_MMIO(DSBSL_INSTANCE(pipe, id) + 0x8)
diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c
index bbd71af00a91..44a0d1a950c5 100644
--- a/drivers/gpu/drm/i915/i915_request.c
+++ b/drivers/gpu/drm/i915/i915_request.c
@@ -57,11 +57,13 @@ static struct i915_global_request {
 
 static const char *i915_fence_get_driver_name(struct dma_fence *fence)
 {
-	return "i915";
+	return dev_name(to_request(fence)->i915->drm.dev);
 }
 
 static const char *i915_fence_get_timeline_name(struct dma_fence *fence)
 {
+	const struct i915_gem_context *ctx;
+
 	/*
 	 * The timeline struct (as part of the ppgtt underneath a context)
 	 * may be freed when the request is no longer in use by the GPU.
@@ -74,7 +76,11 @@ static const char *i915_fence_get_timeline_name(struct dma_fence *fence)
 	if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
 		return "signaled";
 
-	return to_request(fence)->gem_context->name ?: "[i915]";
+	ctx = i915_request_gem_context(to_request(fence));
+	if (!ctx)
+		return "[" DRIVER_NAME "]";
+
+	return ctx->name;
 }
 
 static bool i915_fence_signaled(struct dma_fence *fence)
@@ -188,7 +194,7 @@ static void free_capture_list(struct i915_request *request)
 {
 	struct i915_capture_list *capture;
 
-	capture = request->capture_list;
+	capture = fetch_and_zero(&request->capture_list);
 	while (capture) {
 		struct i915_capture_list *next = capture->next;
 
@@ -214,7 +220,7 @@ static void remove_from_engine(struct i915_request *rq)
 		spin_lock(&engine->active.lock);
 		locked = engine;
 	}
-	list_del(&rq->sched.link);
+	list_del_init(&rq->sched.link);
 	spin_unlock_irq(&locked->active.lock);
 }
 
@@ -223,10 +229,7 @@ bool i915_request_retire(struct i915_request *rq)
 	if (!i915_request_completed(rq))
 		return false;
 
-	GEM_TRACE("%s fence %llx:%lld, current %d\n",
-		  rq->engine->name,
-		  rq->fence.context, rq->fence.seqno,
-		  hwsp_seqno(rq));
+	RQ_TRACE(rq, "\n");
 
 	GEM_BUG_ON(!i915_sw_fence_signaled(&rq->submit));
 	trace_i915_request_retire(rq);
@@ -272,8 +275,8 @@ bool i915_request_retire(struct i915_request *rq)
 	remove_from_client(rq);
 	list_del(&rq->link);
 
-	intel_context_exit(rq->hw_context);
-	intel_context_unpin(rq->hw_context);
+	intel_context_exit(rq->context);
+	intel_context_unpin(rq->context);
 
 	free_capture_list(rq);
 	i915_sched_node_fini(&rq->sched);
@@ -287,10 +290,7 @@ void i915_request_retire_upto(struct i915_request *rq)
 	struct intel_timeline * const tl = i915_request_timeline(rq);
 	struct i915_request *tmp;
 
-	GEM_TRACE("%s fence %llx:%lld, current %d\n",
-		  rq->engine->name,
-		  rq->fence.context, rq->fence.seqno,
-		  hwsp_seqno(rq));
+	RQ_TRACE(rq, "\n");
 
 	GEM_BUG_ON(!i915_request_completed(rq));
 
@@ -300,11 +300,11 @@ void i915_request_retire_upto(struct i915_request *rq)
 }
 
 static int
-__i915_request_await_execution(struct i915_request *rq,
-			       struct i915_request *signal,
-			       void (*hook)(struct i915_request *rq,
-					    struct dma_fence *signal),
-			       gfp_t gfp)
+__await_execution(struct i915_request *rq,
+		  struct i915_request *signal,
+		  void (*hook)(struct i915_request *rq,
+			       struct dma_fence *signal),
+		  gfp_t gfp)
 {
 	struct execute_cb *cb;
 
@@ -341,6 +341,8 @@ __i915_request_await_execution(struct i915_request *rq,
 	}
 	spin_unlock_irq(&signal->lock);
 
+	/* Copy across semaphore status as we need the same behaviour */
+	rq->sched.flags |= signal->sched.flags;
 	return 0;
 }
 
@@ -349,10 +351,7 @@ bool __i915_request_submit(struct i915_request *request)
 	struct intel_engine_cs *engine = request->engine;
 	bool result = false;
 
-	GEM_TRACE("%s fence %llx:%lld, current %d\n",
-		  engine->name,
-		  request->fence.context, request->fence.seqno,
-		  hwsp_seqno(request));
+	RQ_TRACE(request, "\n");
 
 	GEM_BUG_ON(!irqs_disabled());
 	lockdep_assert_held(&engine->active.lock);
@@ -376,7 +375,7 @@ bool __i915_request_submit(struct i915_request *request)
 	if (i915_request_completed(request))
 		goto xfer;
 
-	if (i915_gem_context_is_banned(request->gem_context))
+	if (intel_context_is_banned(request->context))
 		i915_request_skip(request, -EIO);
 
 	/*
@@ -415,7 +414,7 @@ xfer:	/* We may be recursing from the signal callback of another i915 fence */
 	if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &request->fence.flags) &&
 	    !test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &request->fence.flags) &&
 	    !i915_request_enable_breadcrumb(request))
-		intel_engine_queue_breadcrumbs(engine);
+		intel_engine_signal_breadcrumbs(engine);
 
 	__notify_execute_cb(request);
 
@@ -441,10 +440,7 @@ void __i915_request_unsubmit(struct i915_request *request)
 {
 	struct intel_engine_cs *engine = request->engine;
 
-	GEM_TRACE("%s fence %llx:%lld, current %d\n",
-		  engine->name,
-		  request->fence.context, request->fence.seqno,
-		  hwsp_seqno(request));
+	RQ_TRACE(request, "\n");
 
 	GEM_BUG_ON(!irqs_disabled());
 	lockdep_assert_held(&engine->active.lock);
@@ -586,6 +582,21 @@ out:
 	return kmem_cache_alloc(global.slab_requests, gfp);
 }
 
+static void __i915_request_ctor(void *arg)
+{
+	struct i915_request *rq = arg;
+
+	spin_lock_init(&rq->lock);
+	i915_sched_node_init(&rq->sched);
+	i915_sw_fence_init(&rq->submit, submit_notify);
+	i915_sw_fence_init(&rq->semaphore, semaphore_notify);
+
+	rq->file_priv = NULL;
+	rq->capture_list = NULL;
+
+	INIT_LIST_HEAD(&rq->execute_cb);
+}
+
 struct i915_request *
 __i915_request_create(struct intel_context *ce, gfp_t gfp)
 {
@@ -643,35 +654,32 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp)
 		goto err_free;
 
 	rq->i915 = ce->engine->i915;
-	rq->hw_context = ce;
-	rq->gem_context = ce->gem_context;
+	rq->context = ce;
 	rq->engine = ce->engine;
 	rq->ring = ce->ring;
 	rq->execution_mask = ce->engine->mask;
+	rq->flags = 0;
 
-	rcu_assign_pointer(rq->timeline, tl);
+	RCU_INIT_POINTER(rq->timeline, tl);
+	RCU_INIT_POINTER(rq->hwsp_cacheline, tl->hwsp_cacheline);
 	rq->hwsp_seqno = tl->hwsp_seqno;
-	rq->hwsp_cacheline = tl->hwsp_cacheline;
 
 	rq->rcustate = get_state_synchronize_rcu(); /* acts as smp_mb() */
 
-	spin_lock_init(&rq->lock);
 	dma_fence_init(&rq->fence, &i915_fence_ops, &rq->lock,
 		       tl->fence_context, seqno);
 
 	/* We bump the ref for the fence chain */
-	i915_sw_fence_init(&i915_request_get(rq)->submit, submit_notify);
-	i915_sw_fence_init(&i915_request_get(rq)->semaphore, semaphore_notify);
+	i915_sw_fence_reinit(&i915_request_get(rq)->submit);
+	i915_sw_fence_reinit(&i915_request_get(rq)->semaphore);
 
-	i915_sched_node_init(&rq->sched);
+	i915_sched_node_reinit(&rq->sched);
 
-	/* No zalloc, must clear what we need by hand */
-	rq->file_priv = NULL;
+	/* No zalloc, everything must be cleared after use */
 	rq->batch = NULL;
-	rq->capture_list = NULL;
-	rq->flags = 0;
-
-	INIT_LIST_HEAD(&rq->execute_cb);
+	GEM_BUG_ON(rq->file_priv);
+	GEM_BUG_ON(rq->capture_list);
+	GEM_BUG_ON(!list_empty(&rq->execute_cb));
 
 	/*
 	 * Reserve space in the ring buffer for all the commands required to
@@ -753,34 +761,37 @@ err_unlock:
 static int
 i915_request_await_start(struct i915_request *rq, struct i915_request *signal)
 {
-	struct intel_timeline *tl;
 	struct dma_fence *fence;
 	int err;
 
 	GEM_BUG_ON(i915_request_timeline(rq) ==
 		   rcu_access_pointer(signal->timeline));
 
+	fence = NULL;
 	rcu_read_lock();
-	tl = rcu_dereference(signal->timeline);
-	if (i915_request_started(signal) || !kref_get_unless_zero(&tl->kref))
-		tl = NULL;
-	rcu_read_unlock();
-	if (!tl) /* already started or maybe even completed */
-		return 0;
+	spin_lock_irq(&signal->lock);
+	if (!i915_request_started(signal) &&
+	    !list_is_first(&signal->link,
+			   &rcu_dereference(signal->timeline)->requests)) {
+		struct i915_request *prev = list_prev_entry(signal, link);
 
-	fence = ERR_PTR(-EBUSY);
-	if (mutex_trylock(&tl->mutex)) {
-		fence = NULL;
-		if (!i915_request_started(signal) &&
-		    !list_is_first(&signal->link, &tl->requests)) {
-			signal = list_prev_entry(signal, link);
-			fence = dma_fence_get(&signal->fence);
+		/*
+		 * Peek at the request before us in the timeline. That
+		 * request will only be valid before it is retired, so
+		 * after acquiring a reference to it, confirm that it is
+		 * still part of the signaler's timeline.
+		 */
+		if (i915_request_get_rcu(prev)) {
+			if (list_next_entry(prev, link) == signal)
+				fence = &prev->fence;
+			else
+				i915_request_put(prev);
 		}
-		mutex_unlock(&tl->mutex);
 	}
-	intel_timeline_put(tl);
-	if (IS_ERR_OR_NULL(fence))
-		return PTR_ERR_OR_ZERO(fence);
+	spin_unlock_irq(&signal->lock);
+	rcu_read_unlock();
+	if (!fence)
+		return 0;
 
 	err = 0;
 	if (intel_timeline_sync_is_later(i915_request_timeline(rq), fence))
@@ -811,31 +822,21 @@ already_busywaiting(struct i915_request *rq)
 }
 
 static int
-emit_semaphore_wait(struct i915_request *to,
-		    struct i915_request *from,
-		    gfp_t gfp)
+__emit_semaphore_wait(struct i915_request *to,
+		      struct i915_request *from,
+		      u32 seqno)
 {
 	const int has_token = INTEL_GEN(to->i915) >= 12;
 	u32 hwsp_offset;
-	int len;
+	int len, err;
 	u32 *cs;
 
 	GEM_BUG_ON(INTEL_GEN(to->i915) < 8);
 
-	/* Just emit the first semaphore we see as request space is limited. */
-	if (already_busywaiting(to) & from->engine->mask)
-		goto await_fence;
-
-	if (i915_request_await_start(to, from) < 0)
-		goto await_fence;
-
-	/* Only submit our spinner after the signaler is running! */
-	if (__i915_request_await_execution(to, from, NULL, gfp))
-		goto await_fence;
-
 	/* We need to pin the signaler's HWSP until we are finished reading. */
-	if (intel_timeline_read_hwsp(from, to, &hwsp_offset))
-		goto await_fence;
+	err = intel_timeline_read_hwsp(from, to, &hwsp_offset);
+	if (err)
+		return err;
 
 	len = 4;
 	if (has_token)
@@ -858,7 +859,7 @@ emit_semaphore_wait(struct i915_request *to,
 		 MI_SEMAPHORE_POLL |
 		 MI_SEMAPHORE_SAD_GTE_SDD) +
 		has_token;
-	*cs++ = from->fence.seqno;
+	*cs++ = seqno;
 	*cs++ = hwsp_offset;
 	*cs++ = 0;
 	if (has_token) {
@@ -867,6 +868,28 @@ emit_semaphore_wait(struct i915_request *to,
 	}
 
 	intel_ring_advance(to, cs);
+	return 0;
+}
+
+static int
+emit_semaphore_wait(struct i915_request *to,
+		    struct i915_request *from,
+		    gfp_t gfp)
+{
+	/* Just emit the first semaphore we see as request space is limited. */
+	if (already_busywaiting(to) & from->engine->mask)
+		goto await_fence;
+
+	if (i915_request_await_start(to, from) < 0)
+		goto await_fence;
+
+	/* Only submit our spinner after the signaler is running! */
+	if (__await_execution(to, from, NULL, gfp))
+		goto await_fence;
+
+	if (__emit_semaphore_wait(to, from, from->fence.seqno))
+		goto await_fence;
+
 	to->sched.semaphores |= from->engine->mask;
 	to->sched.flags |= I915_SCHED_HAS_SEMAPHORE_CHAIN;
 	return 0;
@@ -894,18 +917,16 @@ i915_request_await_request(struct i915_request *to, struct i915_request *from)
 			return ret;
 	}
 
-	if (to->engine == from->engine) {
+	if (to->engine == from->engine)
 		ret = i915_sw_fence_await_sw_fence_gfp(&to->submit,
 						       &from->submit,
 						       I915_FENCE_GFP);
-	} else if (intel_engine_has_semaphores(to->engine) &&
-		   to->gem_context->sched.priority >= I915_PRIORITY_NORMAL) {
+	else if (intel_context_use_semaphores(to->context))
 		ret = emit_semaphore_wait(to, from, I915_FENCE_GFP);
-	} else {
+	else
 		ret = i915_sw_fence_await_dma_fence(&to->submit,
 						    &from->fence, 0,
 						    I915_FENCE_GFP);
-	}
 	if (ret < 0)
 		return ret;
 
@@ -945,8 +966,10 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 
 	do {
 		fence = *child++;
-		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
+			i915_sw_fence_set_error_once(&rq->submit, fence->error);
 			continue;
+		}
 
 		/*
 		 * Requests on the same timeline are explicitly ordered, along
@@ -980,6 +1003,57 @@ i915_request_await_dma_fence(struct i915_request *rq, struct dma_fence *fence)
 	return 0;
 }
 
+static bool intel_timeline_sync_has_start(struct intel_timeline *tl,
+					  struct dma_fence *fence)
+{
+	return __intel_timeline_sync_is_later(tl,
+					      fence->context,
+					      fence->seqno - 1);
+}
+
+static int intel_timeline_sync_set_start(struct intel_timeline *tl,
+					 const struct dma_fence *fence)
+{
+	return __intel_timeline_sync_set(tl, fence->context, fence->seqno - 1);
+}
+
+static int
+__i915_request_await_execution(struct i915_request *to,
+			       struct i915_request *from,
+			       void (*hook)(struct i915_request *rq,
+					    struct dma_fence *signal))
+{
+	int err;
+
+	/* Submit both requests at the same time */
+	err = __await_execution(to, from, hook, I915_FENCE_GFP);
+	if (err)
+		return err;
+
+	/* Squash repeated depenendices to the same timelines */
+	if (intel_timeline_sync_has_start(i915_request_timeline(to),
+					  &from->fence))
+		return 0;
+
+	/* Ensure both start together [after all semaphores in signal] */
+	if (intel_engine_has_semaphores(to->engine))
+		err = __emit_semaphore_wait(to, from, from->fence.seqno - 1);
+	else
+		err = i915_request_await_start(to, from);
+	if (err < 0)
+		return err;
+
+	/* Couple the dependency tree for PI on this exposed to->fence */
+	if (to->engine->schedule) {
+		err = i915_sched_node_add_dependency(&to->sched, &from->sched);
+		if (err < 0)
+			return err;
+	}
+
+	return intel_timeline_sync_set_start(i915_request_timeline(to),
+					     &from->fence);
+}
+
 int
 i915_request_await_execution(struct i915_request *rq,
 			     struct dma_fence *fence,
@@ -1002,8 +1076,10 @@ i915_request_await_execution(struct i915_request *rq,
 
 	do {
 		fence = *child++;
-		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
+			i915_sw_fence_set_error_once(&rq->submit, fence->error);
 			continue;
+		}
 
 		/*
 		 * We don't squash repeated fence dependencies here as we
@@ -1013,8 +1089,7 @@ i915_request_await_execution(struct i915_request *rq,
 		if (dma_fence_is_i915(fence))
 			ret = __i915_request_await_execution(rq,
 							     to_request(fence),
-							     hook,
-							     I915_FENCE_GFP);
+							     hook);
 		else
 			ret = i915_sw_fence_await_dma_fence(&rq->submit, fence,
 							    I915_FENCE_TIMEOUT,
@@ -1180,8 +1255,7 @@ struct i915_request *__i915_request_commit(struct i915_request *rq)
 	struct intel_ring *ring = rq->ring;
 	u32 *cs;
 
-	GEM_TRACE("%s fence %llx:%lld\n",
-		  engine->name, rq->fence.context, rq->fence.seqno);
+	RQ_TRACE(rq, "\n");
 
 	/*
 	 * To ensure that this call will not fail, space for its emissions
@@ -1227,8 +1301,8 @@ void __i915_request_queue(struct i915_request *rq,
 
 void i915_request_add(struct i915_request *rq)
 {
-	struct i915_sched_attr attr = rq->gem_context->sched;
 	struct intel_timeline * const tl = i915_request_timeline(rq);
+	struct i915_sched_attr attr = {};
 	struct i915_request *prev;
 
 	lockdep_assert_held(&tl->mutex);
@@ -1238,6 +1312,9 @@ void i915_request_add(struct i915_request *rq)
 
 	prev = __i915_request_commit(rq);
 
+	if (rcu_access_pointer(rq->context->gem_context))
+		attr = i915_request_gem_context(rq)->sched;
+
 	/*
 	 * Boost actual workloads past semaphores!
 	 *
@@ -1533,10 +1610,14 @@ static struct i915_global_request global = { {
 
 int __init i915_global_request_init(void)
 {
-	global.slab_requests = KMEM_CACHE(i915_request,
-					  SLAB_HWCACHE_ALIGN |
-					  SLAB_RECLAIM_ACCOUNT |
-					  SLAB_TYPESAFE_BY_RCU);
+	global.slab_requests =
+		kmem_cache_create("i915_request",
+				  sizeof(struct i915_request),
+				  __alignof__(struct i915_request),
+				  SLAB_HWCACHE_ALIGN |
+				  SLAB_RECLAIM_ACCOUNT |
+				  SLAB_TYPESAFE_BY_RCU,
+				  __i915_request_ctor);
 	if (!global.slab_requests)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/i915/i915_request.h b/drivers/gpu/drm/i915/i915_request.h
index 96991d64759c..565322640378 100644
--- a/drivers/gpu/drm/i915/i915_request.h
+++ b/drivers/gpu/drm/i915/i915_request.h
@@ -28,8 +28,10 @@
 #include <linux/dma-fence.h>
 #include <linux/lockdep.h>
 
+#include "gem/i915_gem_context_types.h"
 #include "gt/intel_context_types.h"
 #include "gt/intel_engine_types.h"
+#include "gt/intel_timeline_types.h"
 
 #include "i915_gem.h"
 #include "i915_scheduler.h"
@@ -41,14 +43,19 @@
 struct drm_file;
 struct drm_i915_gem_object;
 struct i915_request;
-struct intel_timeline;
-struct intel_timeline_cacheline;
 
 struct i915_capture_list {
 	struct i915_capture_list *next;
 	struct i915_vma *vma;
 };
 
+#define RQ_TRACE(rq, fmt, ...) do {					\
+	const struct i915_request *rq__ = (rq);				\
+	ENGINE_TRACE(rq__->engine, "fence %llx:%lld, current %d" fmt,	\
+		     rq__->fence.context, rq__->fence.seqno,		\
+		     hwsp_seqno(rq__), ##__VA_ARGS__);			\
+} while (0)
+
 enum {
 	/*
 	 * I915_FENCE_FLAG_ACTIVE - this request is currently submitted to HW.
@@ -109,9 +116,8 @@ struct i915_request {
 	 * i915_request_free() will then decrement the refcount on the
 	 * context.
 	 */
-	struct i915_gem_context *gem_context;
 	struct intel_engine_cs *engine;
-	struct intel_context *hw_context;
+	struct intel_context *context;
 	struct intel_ring *ring;
 	struct intel_timeline __rcu *timeline;
 	struct list_head signal_link;
@@ -144,6 +150,10 @@ struct i915_request {
 	union {
 		wait_queue_entry_t submitq;
 		struct i915_sw_dma_fence_cb dmaq;
+		struct i915_request_duration_cb {
+			struct dma_fence_cb cb;
+			ktime_t emitted;
+		} duration;
 	};
 	struct list_head execute_cb;
 	struct i915_sw_fence semaphore;
@@ -176,7 +186,7 @@ struct i915_request {
 	 * inside the timeline's HWSP vma, but it is only valid while this
 	 * request has not completed and guarded by the timeline mutex.
 	 */
-	struct intel_timeline_cacheline *hwsp_cacheline;
+	struct intel_timeline_cacheline __rcu *hwsp_cacheline;
 
 	/** Position in the ring of the start of the request */
 	u32 head;
@@ -454,6 +464,13 @@ i915_request_timeline(struct i915_request *rq)
 					 lockdep_is_held(&rcu_access_pointer(rq->timeline)->mutex));
 }
 
+static inline struct i915_gem_context *
+i915_request_gem_context(struct i915_request *rq)
+{
+	/* Valid only while the request is being constructed (or retired). */
+	return rcu_dereference_protected(rq->context->gem_context, true);
+}
+
 static inline struct intel_timeline *
 i915_request_active_timeline(struct i915_request *rq)
 {
diff --git a/drivers/gpu/drm/i915/i915_scheduler.c b/drivers/gpu/drm/i915/i915_scheduler.c
index 010d67f48ad9..bf87c70bfdd9 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.c
+++ b/drivers/gpu/drm/i915/i915_scheduler.c
@@ -213,7 +213,7 @@ static void kick_submission(struct intel_engine_cs *engine,
 	 * If we are already the currently executing context, don't
 	 * bother evaluating if we should preempt ourselves.
 	 */
-	if (inflight->hw_context == rq->hw_context)
+	if (inflight->context == rq->context)
 		goto unlock;
 
 	engine->execlists.queue_priority_hint = prio;
@@ -387,9 +387,19 @@ void i915_sched_node_init(struct i915_sched_node *node)
 	INIT_LIST_HEAD(&node->signalers_list);
 	INIT_LIST_HEAD(&node->waiters_list);
 	INIT_LIST_HEAD(&node->link);
+
+	i915_sched_node_reinit(node);
+}
+
+void i915_sched_node_reinit(struct i915_sched_node *node)
+{
 	node->attr.priority = I915_PRIORITY_INVALID;
 	node->semaphores = 0;
 	node->flags = 0;
+
+	GEM_BUG_ON(!list_empty(&node->signalers_list));
+	GEM_BUG_ON(!list_empty(&node->waiters_list));
+	GEM_BUG_ON(!list_empty(&node->link));
 }
 
 static struct i915_dependency *
@@ -474,13 +484,13 @@ void i915_sched_node_fini(struct i915_sched_node *node)
 	 * so we may be called out-of-order.
 	 */
 	list_for_each_entry_safe(dep, tmp, &node->signalers_list, signal_link) {
-		GEM_BUG_ON(!node_signaled(dep->signaler));
 		GEM_BUG_ON(!list_empty(&dep->dfs_link));
 
 		list_del(&dep->wait_link);
 		if (dep->flags & I915_DEPENDENCY_ALLOC)
 			i915_dependency_free(dep);
 	}
+	INIT_LIST_HEAD(&node->signalers_list);
 
 	/* Remove ourselves from everyone who depends upon us */
 	list_for_each_entry_safe(dep, tmp, &node->waiters_list, wait_link) {
@@ -491,6 +501,7 @@ void i915_sched_node_fini(struct i915_sched_node *node)
 		if (dep->flags & I915_DEPENDENCY_ALLOC)
 			i915_dependency_free(dep);
 	}
+	INIT_LIST_HEAD(&node->waiters_list);
 
 	spin_unlock_irq(&schedule_lock);
 }
diff --git a/drivers/gpu/drm/i915/i915_scheduler.h b/drivers/gpu/drm/i915/i915_scheduler.h
index 07d243acf553..d1dc4efef77b 100644
--- a/drivers/gpu/drm/i915/i915_scheduler.h
+++ b/drivers/gpu/drm/i915/i915_scheduler.h
@@ -26,6 +26,7 @@
 					 sched.link)
 
 void i915_sched_node_init(struct i915_sched_node *node);
+void i915_sched_node_reinit(struct i915_sched_node *node);
 
 bool __i915_sched_node_add_dependency(struct i915_sched_node *node,
 				      struct i915_sched_node *signal,
diff --git a/drivers/gpu/drm/i915/i915_selftest.h b/drivers/gpu/drm/i915/i915_selftest.h
index 4d88205de51b..98bcb6fa0ab4 100644
--- a/drivers/gpu/drm/i915/i915_selftest.h
+++ b/drivers/gpu/drm/i915/i915_selftest.h
@@ -36,6 +36,7 @@ struct i915_selftest {
 	char *filter;
 	int mock;
 	int live;
+	int perf;
 };
 
 #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST)
@@ -45,6 +46,7 @@ extern struct i915_selftest i915_selftest;
 
 int i915_mock_selftests(void);
 int i915_live_selftests(struct pci_dev *pdev);
+int i915_perf_selftests(struct pci_dev *pdev);
 
 /* We extract the function declarations from i915_mock_selftests.h and
  * i915_live_selftests.h Add your unit test declarations there!
@@ -61,6 +63,7 @@ int i915_live_selftests(struct pci_dev *pdev);
 #undef selftest
 #define selftest(name, func) int func(struct drm_i915_private *i915);
 #include "selftests/i915_live_selftests.h"
+#include "selftests/i915_perf_selftests.h"
 #undef selftest
 
 struct i915_subtest {
@@ -109,6 +112,7 @@ int __i915_subtests(const char *caller,
 
 static inline int i915_mock_selftests(void) { return 0; }
 static inline int i915_live_selftests(struct pci_dev *pdev) { return 0; }
+static inline int i915_perf_selftests(struct pci_dev *pdev) { return 0; }
 
 #define I915_SELFTEST_DECLARE(x)
 #define I915_SELFTEST_ONLY(x) 0
diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c
index 6a88db291252..51ba97daf2a0 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence.c
+++ b/drivers/gpu/drm/i915/i915_sw_fence.c
@@ -12,6 +12,12 @@
 #include "i915_sw_fence.h"
 #include "i915_selftest.h"
 
+#if IS_ENABLED(CONFIG_DRM_I915_DEBUG)
+#define I915_SW_FENCE_BUG_ON(expr) BUG_ON(expr)
+#else
+#define I915_SW_FENCE_BUG_ON(expr) BUILD_BUG_ON_INVALID(expr)
+#endif
+
 #define I915_SW_FENCE_FLAG_ALLOC BIT(3) /* after WQ_FLAG_* for safety */
 
 static DEFINE_SPINLOCK(i915_sw_fence_lock);
@@ -218,13 +224,21 @@ void __i915_sw_fence_init(struct i915_sw_fence *fence,
 {
 	BUG_ON(!fn || (unsigned long)fn & ~I915_SW_FENCE_MASK);
 
+	__init_waitqueue_head(&fence->wait, name, key);
+	fence->flags = (unsigned long)fn;
+
+	i915_sw_fence_reinit(fence);
+}
+
+void i915_sw_fence_reinit(struct i915_sw_fence *fence)
+{
 	debug_fence_init(fence);
 
-	__init_waitqueue_head(&fence->wait, name, key);
 	atomic_set(&fence->pending, 1);
 	fence->error = 0;
 
-	fence->flags = (unsigned long)fn;
+	I915_SW_FENCE_BUG_ON(!fence->flags);
+	I915_SW_FENCE_BUG_ON(!list_empty(&fence->wait.head));
 }
 
 void i915_sw_fence_commit(struct i915_sw_fence *fence)
@@ -414,8 +428,10 @@ static void dma_i915_sw_fence_wake_timer(struct dma_fence *dma,
 	struct i915_sw_fence *fence;
 
 	fence = xchg(&cb->base.fence, NULL);
-	if (fence)
+	if (fence) {
+		i915_sw_fence_set_error_once(fence, dma->error);
 		i915_sw_fence_complete(fence);
+	}
 
 	irq_work_queue(&cb->work);
 }
@@ -443,8 +459,10 @@ int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,
 	debug_fence_assert(fence);
 	might_sleep_if(gfpflags_allow_blocking(gfp));
 
-	if (dma_fence_is_signaled(dma))
+	if (dma_fence_is_signaled(dma)) {
+		i915_sw_fence_set_error_once(fence, dma->error);
 		return 0;
+	}
 
 	cb = kmalloc(timeout ?
 		     sizeof(struct i915_sw_dma_fence_cb_timer) :
@@ -454,7 +472,12 @@ int i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,
 		if (!gfpflags_allow_blocking(gfp))
 			return -ENOMEM;
 
-		return dma_fence_wait(dma, false);
+		ret = dma_fence_wait(dma, false);
+		if (ret)
+			return ret;
+
+		i915_sw_fence_set_error_once(fence, dma->error);
+		return 0;
 	}
 
 	cb->fence = fence;
@@ -504,8 +527,10 @@ int __i915_sw_fence_await_dma_fence(struct i915_sw_fence *fence,
 
 	debug_fence_assert(fence);
 
-	if (dma_fence_is_signaled(dma))
+	if (dma_fence_is_signaled(dma)) {
+		i915_sw_fence_set_error_once(fence, dma->error);
 		return 0;
+	}
 
 	cb->fence = fence;
 	i915_sw_fence_await(fence);
@@ -539,8 +564,7 @@ int i915_sw_fence_await_reservation(struct i915_sw_fence *fence,
 		struct dma_fence **shared;
 		unsigned int count, i;
 
-		ret = dma_resv_get_fences_rcu(resv,
-							&excl, &count, &shared);
+		ret = dma_resv_get_fences_rcu(resv, &excl, &count, &shared);
 		if (ret)
 			return ret;
 
diff --git a/drivers/gpu/drm/i915/i915_sw_fence.h b/drivers/gpu/drm/i915/i915_sw_fence.h
index ab7d58bd0b9d..19e806ce43bc 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence.h
+++ b/drivers/gpu/drm/i915/i915_sw_fence.h
@@ -54,6 +54,8 @@ do {								\
 	__i915_sw_fence_init((fence), (fn), NULL, NULL)
 #endif
 
+void i915_sw_fence_reinit(struct i915_sw_fence *fence);
+
 #ifdef CONFIG_DRM_I915_SW_FENCE_DEBUG_OBJECTS
 void i915_sw_fence_fini(struct i915_sw_fence *fence);
 #else
@@ -110,7 +112,8 @@ static inline void i915_sw_fence_wait(struct i915_sw_fence *fence)
 static inline void
 i915_sw_fence_set_error_once(struct i915_sw_fence *fence, int error)
 {
-	cmpxchg(&fence->error, 0, error);
+	if (unlikely(error))
+		cmpxchg(&fence->error, 0, error);
 }
 
 #endif /* _I915_SW_FENCE_H_ */
diff --git a/drivers/gpu/drm/i915/i915_sw_fence_work.c b/drivers/gpu/drm/i915/i915_sw_fence_work.c
index 07552cd544f2..997b2998f1f2 100644
--- a/drivers/gpu/drm/i915/i915_sw_fence_work.c
+++ b/drivers/gpu/drm/i915/i915_sw_fence_work.c
@@ -6,6 +6,13 @@
 
 #include "i915_sw_fence_work.h"
 
+static void fence_complete(struct dma_fence_work *f)
+{
+	if (f->ops->release)
+		f->ops->release(f);
+	dma_fence_signal(&f->dma);
+}
+
 static void fence_work(struct work_struct *work)
 {
 	struct dma_fence_work *f = container_of(work, typeof(*f), work);
@@ -14,7 +21,8 @@ static void fence_work(struct work_struct *work)
 	err = f->ops->work(f);
 	if (err)
 		dma_fence_set_error(&f->dma, err);
-	dma_fence_signal(&f->dma);
+
+	fence_complete(f);
 	dma_fence_put(&f->dma);
 }
 
@@ -32,7 +40,7 @@ fence_notify(struct i915_sw_fence *fence, enum i915_sw_fence_notify state)
 			dma_fence_get(&f->dma);
 			queue_work(system_unbound_wq, &f->work);
 		} else {
-			dma_fence_signal(&f->dma);
+			fence_complete(f);
 		}
 		break;
 
@@ -60,9 +68,6 @@ static void fence_release(struct dma_fence *fence)
 {
 	struct dma_fence_work *f = container_of(fence, typeof(*f), dma);
 
-	if (f->ops->release)
-		f->ops->release(f);
-
 	i915_sw_fence_fini(&f->chain);
 
 	BUILD_BUG_ON(offsetof(typeof(*f), dma));
@@ -78,12 +83,11 @@ static const struct dma_fence_ops fence_ops = {
 void dma_fence_work_init(struct dma_fence_work *f,
 			 const struct dma_fence_work_ops *ops)
 {
+	f->ops = ops;
 	spin_lock_init(&f->lock);
 	dma_fence_init(&f->dma, &fence_ops, &f->lock, 0, 0);
 	i915_sw_fence_init(&f->chain, fence_notify);
 	INIT_WORK(&f->work, fence_work);
-
-	f->ops = ops;
 }
 
 int dma_fence_work_chain(struct dma_fence_work *f, struct dma_fence *signal)
diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c
index 65476909d1bf..ad2b1b833d7b 100644
--- a/drivers/gpu/drm/i915/i915_sysfs.c
+++ b/drivers/gpu/drm/i915/i915_sysfs.c
@@ -259,33 +259,18 @@ static const struct bin_attribute dpf_attrs_1 = {
 static ssize_t gt_act_freq_mhz_show(struct device *kdev,
 				    struct device_attribute *attr, char *buf)
 {
-	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt.rps;
-	intel_wakeref_t wakeref;
-	u32 freq;
-
-	wakeref = intel_runtime_pm_get(&dev_priv->runtime_pm);
-
-	if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
-		vlv_punit_get(dev_priv);
-		freq = vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS);
-		vlv_punit_put(dev_priv);
-
-		freq = (freq >> 8) & 0xff;
-	} else {
-		freq = intel_get_cagf(rps, I915_READ(GEN6_RPSTAT1));
-	}
-
-	intel_runtime_pm_put(&dev_priv->runtime_pm, wakeref);
+	struct drm_i915_private *i915 = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &i915->gt.rps;
 
-	return snprintf(buf, PAGE_SIZE, "%d\n", intel_gpu_freq(rps, freq));
+	return snprintf(buf, PAGE_SIZE, "%d\n",
+			intel_rps_read_actual_frequency(rps));
 }
 
 static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
 				    struct device_attribute *attr, char *buf)
 {
-	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt.rps;
+	struct drm_i915_private *i915 = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &i915->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
 			intel_gpu_freq(rps, rps->cur_freq));
@@ -293,8 +278,8 @@ static ssize_t gt_cur_freq_mhz_show(struct device *kdev,
 
 static ssize_t gt_boost_freq_mhz_show(struct device *kdev, struct device_attribute *attr, char *buf)
 {
-	struct drm_i915_private *dev_priv = kdev_minor_to_i915(kdev);
-	struct intel_rps *rps = &dev_priv->gt.rps;
+	struct drm_i915_private *i915 = kdev_minor_to_i915(kdev);
+	struct intel_rps *rps = &i915->gt.rps;
 
 	return snprintf(buf, PAGE_SIZE, "%d\n",
 			intel_gpu_freq(rps, rps->boost_freq));
diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h
index 7ef7a1e1664c..233a97a2c276 100644
--- a/drivers/gpu/drm/i915/i915_trace.h
+++ b/drivers/gpu/drm/i915/i915_trace.h
@@ -341,7 +341,7 @@ TRACE_EVENT(intel_disable_plane,
 
 /* pipe updates */
 
-TRACE_EVENT(i915_pipe_update_start,
+TRACE_EVENT(intel_pipe_update_start,
 	    TP_PROTO(struct intel_crtc *crtc),
 	    TP_ARGS(crtc),
 
@@ -366,7 +366,7 @@ TRACE_EVENT(i915_pipe_update_start,
 		       __entry->scanline, __entry->min, __entry->max)
 );
 
-TRACE_EVENT(i915_pipe_update_vblank_evaded,
+TRACE_EVENT(intel_pipe_update_vblank_evaded,
 	    TP_PROTO(struct intel_crtc *crtc),
 	    TP_ARGS(crtc),
 
@@ -391,7 +391,7 @@ TRACE_EVENT(i915_pipe_update_vblank_evaded,
 		       __entry->scanline, __entry->min, __entry->max)
 );
 
-TRACE_EVENT(i915_pipe_update_end,
+TRACE_EVENT(intel_pipe_update_end,
 	    TP_PROTO(struct intel_crtc *crtc, u32 frame, int scanline_end),
 	    TP_ARGS(crtc, frame, scanline_end),
 
diff --git a/drivers/gpu/drm/i915/i915_utils.h b/drivers/gpu/drm/i915/i915_utils.h
index 04139ba1191e..b0ade76bec90 100644
--- a/drivers/gpu/drm/i915/i915_utils.h
+++ b/drivers/gpu/drm/i915/i915_utils.h
@@ -69,7 +69,7 @@ bool i915_error_injected(void);
 
 #else
 
-#define i915_inject_probe_error(_i915, _err) 0
+#define i915_inject_probe_error(i915, e) ({ BUILD_BUG_ON_INVALID(i915); 0; })
 #define i915_error_injected() false
 
 #endif
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index e5512f26e20a..cbd783c31adb 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -28,7 +28,9 @@
 #include "display/intel_frontbuffer.h"
 
 #include "gt/intel_engine.h"
+#include "gt/intel_engine_heartbeat.h"
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_requests.h"
 
 #include "i915_drv.h"
 #include "i915_globals.h"
@@ -112,6 +114,7 @@ vma_create(struct drm_i915_gem_object *obj,
 	if (vma == NULL)
 		return ERR_PTR(-ENOMEM);
 
+	kref_init(&vma->ref);
 	mutex_init(&vma->pages_mutex);
 	vma->vm = i915_vm_get(vm);
 	vma->ops = &vm->vma_ops;
@@ -290,6 +293,7 @@ i915_vma_instance(struct drm_i915_gem_object *obj,
 struct i915_vma_work {
 	struct dma_fence_work base;
 	struct i915_vma *vma;
+	struct drm_i915_gem_object *pinned;
 	enum i915_cache_level cache_level;
 	unsigned int flags;
 };
@@ -304,15 +308,21 @@ static int __vma_bind(struct dma_fence_work *work)
 	if (err)
 		atomic_or(I915_VMA_ERROR, &vma->flags);
 
-	if (vma->obj)
-		__i915_gem_object_unpin_pages(vma->obj);
-
 	return err;
 }
 
+static void __vma_release(struct dma_fence_work *work)
+{
+	struct i915_vma_work *vw = container_of(work, typeof(*vw), base);
+
+	if (vw->pinned)
+		__i915_gem_object_unpin_pages(vw->pinned);
+}
+
 static const struct dma_fence_work_ops bind_ops = {
 	.name = "bind",
 	.work = __vma_bind,
+	.release = __vma_release,
 };
 
 struct i915_vma_work *i915_vma_work(void)
@@ -393,8 +403,10 @@ int i915_vma_bind(struct i915_vma *vma,
 		i915_active_set_exclusive(&vma->active, &work->base.dma);
 		work->base.dma.error = 0; /* enable the queue_work() */
 
-		if (vma->obj)
+		if (vma->obj) {
 			__i915_gem_object_pin_pages(vma->obj);
+			work->pinned = vma->obj;
+		}
 	} else {
 		GEM_BUG_ON((bind_flags & ~vma_flags) & vma->vm->bind_async_flags);
 		ret = vma->ops->bind_vma(vma, cache_level, bind_flags);
@@ -939,6 +951,38 @@ err_pages:
 	return err;
 }
 
+static void flush_idle_contexts(struct intel_gt *gt)
+{
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+
+	for_each_engine(engine, gt, id)
+		intel_engine_flush_barriers(engine);
+
+	intel_gt_wait_for_idle(gt, MAX_SCHEDULE_TIMEOUT);
+}
+
+int i915_ggtt_pin(struct i915_vma *vma, u32 align, unsigned int flags)
+{
+	struct i915_address_space *vm = vma->vm;
+	int err;
+
+	GEM_BUG_ON(!i915_vma_is_ggtt(vma));
+
+	do {
+		err = i915_vma_pin(vma, 0, align, flags | PIN_GLOBAL);
+		if (err != -ENOSPC)
+			return err;
+
+		/* Unlike i915_vma_pin, we don't take no for an answer! */
+		flush_idle_contexts(vm->gt);
+		if (mutex_lock_interruptible(&vm->mutex) == 0) {
+			i915_gem_evict_vm(vm);
+			mutex_unlock(&vm->mutex);
+		}
+	} while (1);
+}
+
 void i915_vma_close(struct i915_vma *vma)
 {
 	struct intel_gt *gt = vma->vm->gt;
@@ -978,8 +1022,10 @@ void i915_vma_reopen(struct i915_vma *vma)
 		__i915_vma_remove_closed(vma);
 }
 
-void i915_vma_destroy(struct i915_vma *vma)
+void i915_vma_release(struct kref *ref)
 {
+	struct i915_vma *vma = container_of(ref, typeof(*vma), ref);
+
 	if (drm_mm_node_allocated(&vma->node)) {
 		mutex_lock(&vma->vm->mutex);
 		atomic_and(~I915_VMA_PIN_MASK, &vma->flags);
@@ -1019,7 +1065,9 @@ void i915_vma_parked(struct intel_gt *gt)
 		if (!kref_get_unless_zero(&obj->base.refcount))
 			continue;
 
-		if (!i915_vm_tryopen(vm)) {
+		if (i915_vm_tryopen(vm)) {
+			list_del_init(&vma->closed_link);
+		} else {
 			i915_gem_object_put(obj);
 			obj = NULL;
 		}
@@ -1027,7 +1075,7 @@ void i915_vma_parked(struct intel_gt *gt)
 		spin_unlock_irq(&gt->closed_lock);
 
 		if (obj) {
-			i915_vma_destroy(vma);
+			__i915_vma_put(vma);
 			i915_gem_object_put(obj);
 		}
 
@@ -1054,17 +1102,16 @@ static void __i915_vma_iounmap(struct i915_vma *vma)
 
 void i915_vma_revoke_mmap(struct i915_vma *vma)
 {
-	struct drm_vma_offset_node *node = &vma->obj->base.vma_node;
+	struct drm_vma_offset_node *node;
 	u64 vma_offset;
 
-	lockdep_assert_held(&vma->vm->mutex);
-
 	if (!i915_vma_has_userfault(vma))
 		return;
 
 	GEM_BUG_ON(!i915_vma_is_map_and_fenceable(vma));
 	GEM_BUG_ON(!vma->obj->userfault_count);
 
+	node = &vma->mmo->vma_node;
 	vma_offset = vma->ggtt_view.partial.offset << PAGE_SHIFT;
 	unmap_mapping_range(vma->vm->i915->drm.anon_inode->i_mapping,
 			    drm_vma_node_offset_addr(node) + vma_offset,
@@ -1104,8 +1151,14 @@ int i915_vma_move_to_active(struct i915_vma *vma,
 		return err;
 
 	if (flags & EXEC_OBJECT_WRITE) {
-		if (intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CS))
-			i915_active_add_request(&obj->frontbuffer->write, rq);
+		struct intel_frontbuffer *front;
+
+		front = __intel_frontbuffer_get(obj);
+		if (unlikely(front)) {
+			if (intel_frontbuffer_invalidate(front, ORIGIN_CS))
+				i915_active_add_request(&front->write, rq);
+			intel_frontbuffer_put(front);
+		}
 
 		dma_resv_add_excl_fence(vma->resv, &rq->fence);
 		obj->write_domain = I915_GEM_DOMAIN_RENDER;
@@ -1146,7 +1199,7 @@ int __i915_vma_unbind(struct i915_vma *vma)
 	GEM_BUG_ON(i915_vma_is_active(vma));
 	if (i915_vma_is_pinned(vma)) {
 		vma_print_allocator(vma, "is pinned");
-		return -EBUSY;
+		return -EAGAIN;
 	}
 
 	GEM_BUG_ON(i915_vma_is_active(vma));
@@ -1186,7 +1239,7 @@ int __i915_vma_unbind(struct i915_vma *vma)
 	i915_vma_detach(vma);
 	vma_unbind_pages(vma);
 
-	drm_mm_remove_node(&vma->node); /* pairs with i915_vma_destroy() */
+	drm_mm_remove_node(&vma->node); /* pairs with i915_vma_release() */
 	return 0;
 }
 
@@ -1195,6 +1248,9 @@ int i915_vma_unbind(struct i915_vma *vma)
 	struct i915_address_space *vm = vma->vm;
 	int err;
 
+	if (!drm_mm_node_allocated(&vma->node))
+		return 0;
+
 	err = mutex_lock_interruptible(&vm->mutex);
 	if (err)
 		return err;
diff --git a/drivers/gpu/drm/i915/i915_vma.h b/drivers/gpu/drm/i915/i915_vma.h
index 465932813bc5..5fffa3c58908 100644
--- a/drivers/gpu/drm/i915/i915_vma.h
+++ b/drivers/gpu/drm/i915/i915_vma.h
@@ -51,18 +51,26 @@ enum i915_cache_level;
  */
 struct i915_vma {
 	struct drm_mm_node node;
-	struct drm_i915_gem_object *obj;
+
 	struct i915_address_space *vm;
 	const struct i915_vma_ops *ops;
-	struct i915_fence_reg *fence;
+
+	struct drm_i915_gem_object *obj;
 	struct dma_resv *resv; /** Alias of obj->resv */
+
 	struct sg_table *pages;
 	void __iomem *iomap;
 	void *private; /* owned by creator */
+
+	struct i915_fence_reg *fence;
+
 	u64 size;
 	u64 display_alignment;
 	struct i915_page_sizes page_sizes;
 
+	/* mmap-offset associated with fencing for this vma */
+	struct i915_mmap_offset	*mmo;
+
 	u32 fence_size;
 	u32 fence_alignment;
 
@@ -71,6 +79,7 @@ struct i915_vma {
 	 * handles (but same file) for execbuf, i.e. the number of aliases
 	 * that exist in the ctx->handle_vmas LUT for this vma.
 	 */
+	struct kref ref;
 	atomic_t open_count;
 	atomic_t flags;
 	/**
@@ -333,7 +342,20 @@ int __must_check i915_vma_unbind(struct i915_vma *vma);
 void i915_vma_unlink_ctx(struct i915_vma *vma);
 void i915_vma_close(struct i915_vma *vma);
 void i915_vma_reopen(struct i915_vma *vma);
-void i915_vma_destroy(struct i915_vma *vma);
+
+static inline struct i915_vma *__i915_vma_get(struct i915_vma *vma)
+{
+	if (kref_get_unless_zero(&vma->ref))
+		return vma;
+
+	return NULL;
+}
+
+void i915_vma_release(struct kref *ref);
+static inline void __i915_vma_put(struct i915_vma *vma)
+{
+	kref_put(&vma->ref, i915_vma_release);
+}
 
 #define assert_vma_held(vma) dma_resv_assert_held((vma)->resv)
 
@@ -349,6 +371,7 @@ static inline void i915_vma_unlock(struct i915_vma *vma)
 
 int __must_check
 i915_vma_pin(struct i915_vma *vma, u64 size, u64 alignment, u64 flags);
+int i915_ggtt_pin(struct i915_vma *vma, u32 align, unsigned int flags);
 
 static inline int i915_vma_pin_count(const struct i915_vma *vma)
 {
diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c
index a5b571364cf6..1acb5db77431 100644
--- a/drivers/gpu/drm/i915/intel_device_info.c
+++ b/drivers/gpu/drm/i915/intel_device_info.c
@@ -73,9 +73,30 @@ const char *intel_platform_name(enum intel_platform platform)
 	return platform_names[platform];
 }
 
-void intel_device_info_dump_flags(const struct intel_device_info *info,
-				  struct drm_printer *p)
+static const char *iommu_name(void)
 {
+	const char *msg = "n/a";
+
+#ifdef CONFIG_INTEL_IOMMU
+	msg = enableddisabled(intel_iommu_gfx_mapped);
+#endif
+
+	return msg;
+}
+
+void intel_device_info_print_static(const struct intel_device_info *info,
+				    struct drm_printer *p)
+{
+	drm_printf(p, "engines: %x\n", info->engine_mask);
+	drm_printf(p, "gen: %d\n", info->gen);
+	drm_printf(p, "gt: %d\n", info->gt);
+	drm_printf(p, "iommu: %s\n", iommu_name());
+	drm_printf(p, "memory-regions: %x\n", info->memory_regions);
+	drm_printf(p, "page-sizes: %x\n", info->page_sizes);
+	drm_printf(p, "platform: %s\n", intel_platform_name(info->platform));
+	drm_printf(p, "ppgtt-size: %d\n", info->ppgtt_size);
+	drm_printf(p, "ppgtt-type: %d\n", info->ppgtt_type);
+
 #define PRINT_FLAG(name) drm_printf(p, "%s: %s\n", #name, yesno(info->name));
 	DEV_INFO_FOR_EACH_FLAG(PRINT_FLAG);
 #undef PRINT_FLAG
@@ -106,8 +127,8 @@ static void sseu_dump(const struct sseu_dev_info *sseu, struct drm_printer *p)
 	drm_printf(p, "has EU power gating: %s\n", yesno(sseu->has_eu_pg));
 }
 
-void intel_device_info_dump_runtime(const struct intel_runtime_info *info,
-				    struct drm_printer *p)
+void intel_device_info_print_runtime(const struct intel_runtime_info *info,
+				     struct drm_printer *p)
 {
 	sseu_dump(&info->sseu, p);
 
@@ -148,8 +169,8 @@ static void sseu_set_eus(struct sseu_dev_info *sseu, int slice, int subslice,
 	}
 }
 
-void intel_device_info_dump_topology(const struct sseu_dev_info *sseu,
-				     struct drm_printer *p)
+void intel_device_info_print_topology(const struct sseu_dev_info *sseu,
+				      struct drm_printer *p)
 {
 	int s, ss;
 
@@ -808,6 +829,8 @@ static const u16 subplatform_ult_ids[] = {
 	INTEL_WHL_U_GT1_IDS(0),
 	INTEL_WHL_U_GT2_IDS(0),
 	INTEL_WHL_U_GT3_IDS(0),
+	INTEL_CML_U_GT1_IDS(0),
+	INTEL_CML_U_GT2_IDS(0),
 };
 
 static const u16 subplatform_ulx_ids[] = {
diff --git a/drivers/gpu/drm/i915/intel_device_info.h b/drivers/gpu/drm/i915/intel_device_info.h
index 4bdf8a6cfb47..2725cb7fc169 100644
--- a/drivers/gpu/drm/i915/intel_device_info.h
+++ b/drivers/gpu/drm/i915/intel_device_info.h
@@ -230,12 +230,13 @@ const char *intel_platform_name(enum intel_platform platform);
 
 void intel_device_info_subplatform_init(struct drm_i915_private *dev_priv);
 void intel_device_info_runtime_init(struct drm_i915_private *dev_priv);
-void intel_device_info_dump_flags(const struct intel_device_info *info,
-				  struct drm_printer *p);
-void intel_device_info_dump_runtime(const struct intel_runtime_info *info,
+
+void intel_device_info_print_static(const struct intel_device_info *info,
 				    struct drm_printer *p);
-void intel_device_info_dump_topology(const struct sseu_dev_info *sseu,
+void intel_device_info_print_runtime(const struct intel_runtime_info *info,
 				     struct drm_printer *p);
+void intel_device_info_print_topology(const struct sseu_dev_info *sseu,
+				      struct drm_printer *p);
 
 void intel_device_info_init_mmio(struct drm_i915_private *dev_priv);
 
diff --git a/drivers/gpu/drm/i915/intel_memory_region.c b/drivers/gpu/drm/i915/intel_memory_region.c
index baaeaecc64af..e24c280e5930 100644
--- a/drivers/gpu/drm/i915/intel_memory_region.c
+++ b/drivers/gpu/drm/i915/intel_memory_region.c
@@ -73,6 +73,9 @@ __intel_memory_region_get_pages_buddy(struct intel_memory_region *mem,
 		min_order = ilog2(size) - ilog2(mem->mm.chunk_size);
 	}
 
+	if (size > BIT(mem->mm.max_order) * mem->mm.chunk_size)
+		return -E2BIG;
+
 	n_pages = size >> ilog2(mem->mm.chunk_size);
 
 	mutex_lock(&mem->mm_lock);
diff --git a/drivers/gpu/drm/i915/intel_pch.c b/drivers/gpu/drm/i915/intel_pch.c
index 8fd92b9130a7..43b68b5fc562 100644
--- a/drivers/gpu/drm/i915/intel_pch.c
+++ b/drivers/gpu/drm/i915/intel_pch.c
@@ -89,6 +89,7 @@ intel_pch_type(const struct drm_i915_private *dev_priv, unsigned short id)
 		WARN_ON(!IS_ELKHARTLAKE(dev_priv));
 		return PCH_MCC;
 	case INTEL_PCH_TGP_DEVICE_ID_TYPE:
+	case INTEL_PCH_TGP2_DEVICE_ID_TYPE:
 		DRM_DEBUG_KMS("Found Tiger Lake LP PCH\n");
 		WARN_ON(!IS_TIGERLAKE(dev_priv));
 		return PCH_TGP;
diff --git a/drivers/gpu/drm/i915/intel_pch.h b/drivers/gpu/drm/i915/intel_pch.h
index d26c25dd8d54..3053d1ce398b 100644
--- a/drivers/gpu/drm/i915/intel_pch.h
+++ b/drivers/gpu/drm/i915/intel_pch.h
@@ -47,6 +47,7 @@ enum intel_pch {
 #define INTEL_PCH_ICP_DEVICE_ID_TYPE		0x3480
 #define INTEL_PCH_MCC_DEVICE_ID_TYPE		0x4B00
 #define INTEL_PCH_TGP_DEVICE_ID_TYPE		0xA080
+#define INTEL_PCH_TGP2_DEVICE_ID_TYPE		0x4380
 #define INTEL_PCH_JSP_DEVICE_ID_TYPE		0x4D80
 #define INTEL_PCH_JSP2_DEVICE_ID_TYPE		0x3880
 #define INTEL_PCH_P2X_DEVICE_ID_TYPE		0x7100
diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
index 29061685f764..31ec82337e4f 100644
--- a/drivers/gpu/drm/i915/intel_pm.c
+++ b/drivers/gpu/drm/i915/intel_pm.c
@@ -463,7 +463,7 @@ static const int pessimal_latency_ns = 5000;
 
 static void vlv_get_fifo_size(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct vlv_fifo_state *fifo_state = &crtc_state->wm.vlv.fifo_state;
 	enum pipe pipe = crtc->pipe;
@@ -794,10 +794,10 @@ static int intel_wm_num_levels(struct drm_i915_private *dev_priv)
 static bool intel_wm_plane_visible(const struct intel_crtc_state *crtc_state,
 				   const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 
 	/* FIXME check the 'enable' instead */
-	if (!crtc_state->base.active)
+	if (!crtc_state->hw.active)
 		return false;
 
 	/*
@@ -809,9 +809,28 @@ static bool intel_wm_plane_visible(const struct intel_crtc_state *crtc_state,
 	 * around this problem with the watermark code.
 	 */
 	if (plane->id == PLANE_CURSOR)
-		return plane_state->base.fb != NULL;
+		return plane_state->hw.fb != NULL;
 	else
-		return plane_state->base.visible;
+		return plane_state->uapi.visible;
+}
+
+static bool intel_crtc_active(struct intel_crtc *crtc)
+{
+	/* Be paranoid as we can arrive here with only partial
+	 * state retrieved from the hardware during setup.
+	 *
+	 * We can ditch the adjusted_mode.crtc_clock check as soon
+	 * as Haswell has gained clock readout/fastboot support.
+	 *
+	 * We can ditch the crtc->primary->state->fb check as soon as we can
+	 * properly reconstruct framebuffers.
+	 *
+	 * FIXME: The intel_crtc->active here should be switched to
+	 * crtc->state->active once we have proper CRTC states wired up
+	 * for atomic.
+	 */
+	return crtc->active && crtc->base.primary->state->fb &&
+		crtc->config->hw.adjusted_mode.crtc_clock;
 }
 
 static struct intel_crtc *single_enabled_crtc(struct drm_i915_private *dev_priv)
@@ -850,7 +869,7 @@ static void pineview_update_wm(struct intel_crtc *unused_crtc)
 	crtc = single_enabled_crtc(dev_priv);
 	if (crtc) {
 		const struct drm_display_mode *adjusted_mode =
-			&crtc->config->base.adjusted_mode;
+			&crtc->config->hw.adjusted_mode;
 		const struct drm_framebuffer *fb =
 			crtc->base.primary->state->fb;
 		int cpp = fb->format->cpp[0];
@@ -1083,10 +1102,10 @@ static u16 g4x_compute_wm(const struct intel_crtc_state *crtc_state,
 			  const struct intel_plane_state *plane_state,
 			  int level)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	unsigned int latency = dev_priv->wm.pri_latency[level] * 10;
 	unsigned int clock, htotal, cpp, width, wm;
 
@@ -1096,7 +1115,7 @@ static u16 g4x_compute_wm(const struct intel_crtc_state *crtc_state,
 	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
-	cpp = plane_state->base.fb->format->cpp[0];
+	cpp = plane_state->hw.fb->format->cpp[0];
 
 	/*
 	 * Not 100% sure which way ELK should go here as the
@@ -1116,7 +1135,7 @@ static u16 g4x_compute_wm(const struct intel_crtc_state *crtc_state,
 	clock = adjusted_mode->crtc_clock;
 	htotal = adjusted_mode->crtc_htotal;
 
-	width = drm_rect_width(&plane_state->base.dst);
+	width = drm_rect_width(&plane_state->uapi.dst);
 
 	if (plane->id == PLANE_CURSOR) {
 		wm = intel_wm_method2(clock, htotal, width, cpp, latency);
@@ -1143,7 +1162,7 @@ static u16 g4x_compute_wm(const struct intel_crtc_state *crtc_state,
 static bool g4x_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
 				 int level, enum plane_id plane_id, u16 value)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	bool dirty = false;
 
 	for (; level < intel_wm_num_levels(dev_priv); level++) {
@@ -1159,7 +1178,7 @@ static bool g4x_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
 static bool g4x_raw_fbc_wm_set(struct intel_crtc_state *crtc_state,
 			       int level, u16 value)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	bool dirty = false;
 
 	/* NORMAL level doesn't have an FBC watermark */
@@ -1182,7 +1201,7 @@ static u32 ilk_compute_fbc_wm(const struct intel_crtc_state *crtc_state,
 static bool g4x_raw_plane_wm_compute(struct intel_crtc_state *crtc_state,
 				     const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	int num_levels = intel_wm_num_levels(to_i915(plane->base.dev));
 	enum plane_id plane_id = plane->id;
 	bool dirty = false;
@@ -1261,7 +1280,7 @@ static bool g4x_raw_plane_wm_is_valid(const struct intel_crtc_state *crtc_state,
 static bool g4x_raw_crtc_wm_is_valid(const struct intel_crtc_state *crtc_state,
 				     int level)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 
 	if (level > dev_priv->wm.max_level)
 		return false;
@@ -1299,9 +1318,9 @@ static void g4x_invalidate_wms(struct intel_crtc *crtc,
 
 static int g4x_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(crtc_state->base.state);
+		to_intel_atomic_state(crtc_state->uapi.state);
 	struct g4x_wm_state *wm_state = &crtc_state->wm.g4x.optimal;
 	int num_active_planes = hweight8(crtc_state->active_planes &
 					 ~BIT(PLANE_CURSOR));
@@ -1316,8 +1335,8 @@ static int g4x_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	for_each_oldnew_intel_plane_in_state(state, plane,
 					     old_plane_state,
 					     new_plane_state, i) {
-		if (new_plane_state->base.crtc != &crtc->base &&
-		    old_plane_state->base.crtc != &crtc->base)
+		if (new_plane_state->hw.crtc != &crtc->base &&
+		    old_plane_state->hw.crtc != &crtc->base)
 			continue;
 
 		if (g4x_raw_plane_wm_compute(crtc_state, new_plane_state))
@@ -1388,17 +1407,17 @@ static int g4x_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 
 static int g4x_compute_intermediate_wm(struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct g4x_wm_state *intermediate = &new_crtc_state->wm.g4x.intermediate;
 	const struct g4x_wm_state *optimal = &new_crtc_state->wm.g4x.optimal;
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(intel_state, crtc);
 	const struct g4x_wm_state *active = &old_crtc_state->wm.g4x.optimal;
 	enum plane_id plane_id;
 
-	if (!new_crtc_state->base.active || drm_atomic_crtc_needs_modeset(&new_crtc_state->base)) {
+	if (!new_crtc_state->hw.active || drm_atomic_crtc_needs_modeset(&new_crtc_state->uapi)) {
 		*intermediate = *optimal;
 
 		intermediate->cxsr = false;
@@ -1528,10 +1547,11 @@ static void g4x_program_watermarks(struct drm_i915_private *dev_priv)
 }
 
 static void g4x_initial_watermarks(struct intel_atomic_state *state,
-				   struct intel_crtc_state *crtc_state)
+				   struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 
 	mutex_lock(&dev_priv->wm.wm_mutex);
 	crtc->wm.active.g4x = crtc_state->wm.g4x.intermediate;
@@ -1540,10 +1560,11 @@ static void g4x_initial_watermarks(struct intel_atomic_state *state,
 }
 
 static void g4x_optimize_watermarks(struct intel_atomic_state *state,
-				    struct intel_crtc_state *crtc_state)
+				    struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 
 	if (!crtc_state->wm.need_postvbl_update)
 		return;
@@ -1589,10 +1610,10 @@ static u16 vlv_compute_wm_level(const struct intel_crtc_state *crtc_state,
 				const struct intel_plane_state *plane_state,
 				int level)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	unsigned int clock, htotal, cpp, width, wm;
 
 	if (dev_priv->wm.pri_latency[level] == 0)
@@ -1601,7 +1622,7 @@ static u16 vlv_compute_wm_level(const struct intel_crtc_state *crtc_state,
 	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
-	cpp = plane_state->base.fb->format->cpp[0];
+	cpp = plane_state->hw.fb->format->cpp[0];
 	clock = adjusted_mode->crtc_clock;
 	htotal = adjusted_mode->crtc_htotal;
 	width = crtc_state->pipe_src_w;
@@ -1630,7 +1651,7 @@ static bool vlv_need_sprite0_fifo_workaround(unsigned int active_planes)
 
 static int vlv_compute_fifo(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	const struct g4x_pipe_wm *raw =
 		&crtc_state->wm.vlv.raw[VLV_WM_LEVEL_PM2];
 	struct vlv_fifo_state *fifo_state = &crtc_state->wm.vlv.fifo_state;
@@ -1742,7 +1763,7 @@ static u16 vlv_invert_wm_value(u16 wm, u16 fifo_size)
 static bool vlv_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
 				 int level, enum plane_id plane_id, u16 value)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	int num_levels = intel_wm_num_levels(dev_priv);
 	bool dirty = false;
 
@@ -1759,7 +1780,7 @@ static bool vlv_raw_plane_wm_set(struct intel_crtc_state *crtc_state,
 static bool vlv_raw_plane_wm_compute(struct intel_crtc_state *crtc_state,
 				     const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
 	enum plane_id plane_id = plane->id;
 	int num_levels = intel_wm_num_levels(to_i915(plane->base.dev));
 	int level;
@@ -1817,16 +1838,16 @@ static bool vlv_raw_crtc_wm_is_valid(const struct intel_crtc_state *crtc_state,
 
 static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_atomic_state *state =
-		to_intel_atomic_state(crtc_state->base.state);
+		to_intel_atomic_state(crtc_state->uapi.state);
 	struct vlv_wm_state *wm_state = &crtc_state->wm.vlv.optimal;
 	const struct vlv_fifo_state *fifo_state =
 		&crtc_state->wm.vlv.fifo_state;
 	int num_active_planes = hweight8(crtc_state->active_planes &
 					 ~BIT(PLANE_CURSOR));
-	bool needs_modeset = drm_atomic_crtc_needs_modeset(&crtc_state->base);
+	bool needs_modeset = drm_atomic_crtc_needs_modeset(&crtc_state->uapi);
 	const struct intel_plane_state *old_plane_state;
 	const struct intel_plane_state *new_plane_state;
 	struct intel_plane *plane;
@@ -1837,8 +1858,8 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	for_each_oldnew_intel_plane_in_state(state, plane,
 					     old_plane_state,
 					     new_plane_state, i) {
-		if (new_plane_state->base.crtc != &crtc->base &&
-		    old_plane_state->base.crtc != &crtc->base)
+		if (new_plane_state->hw.crtc != &crtc->base &&
+		    old_plane_state->hw.crtc != &crtc->base)
 			continue;
 
 		if (vlv_raw_plane_wm_compute(crtc_state, new_plane_state))
@@ -1923,11 +1944,12 @@ static int vlv_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 	(((value) << DSPARB_ ## plane ## _SHIFT_VLV) & DSPARB_ ## plane ## _MASK_VLV)
 
 static void vlv_atomic_update_fifo(struct intel_atomic_state *state,
-				   struct intel_crtc_state *crtc_state)
+				   struct intel_crtc *crtc)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_uncore *uncore = &dev_priv->uncore;
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 	const struct vlv_fifo_state *fifo_state =
 		&crtc_state->wm.vlv.fifo_state;
 	int sprite0_start, sprite1_start, fifo_size;
@@ -2021,17 +2043,17 @@ static void vlv_atomic_update_fifo(struct intel_atomic_state *state,
 
 static int vlv_compute_intermediate_wm(struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct vlv_wm_state *intermediate = &new_crtc_state->wm.vlv.intermediate;
 	const struct vlv_wm_state *optimal = &new_crtc_state->wm.vlv.optimal;
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(new_crtc_state->base.state);
+		to_intel_atomic_state(new_crtc_state->uapi.state);
 	const struct intel_crtc_state *old_crtc_state =
 		intel_atomic_get_old_crtc_state(intel_state, crtc);
 	const struct vlv_wm_state *active = &old_crtc_state->wm.vlv.optimal;
 	int level;
 
-	if (!new_crtc_state->base.active || drm_atomic_crtc_needs_modeset(&new_crtc_state->base)) {
+	if (!new_crtc_state->hw.active || drm_atomic_crtc_needs_modeset(&new_crtc_state->uapi)) {
 		*intermediate = *optimal;
 
 		intermediate->cxsr = false;
@@ -2147,10 +2169,11 @@ static void vlv_program_watermarks(struct drm_i915_private *dev_priv)
 }
 
 static void vlv_initial_watermarks(struct intel_atomic_state *state,
-				   struct intel_crtc_state *crtc_state)
+				   struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 
 	mutex_lock(&dev_priv->wm.wm_mutex);
 	crtc->wm.active.vlv = crtc_state->wm.vlv.intermediate;
@@ -2159,10 +2182,11 @@ static void vlv_initial_watermarks(struct intel_atomic_state *state,
 }
 
 static void vlv_optimize_watermarks(struct intel_atomic_state *state,
-				    struct intel_crtc_state *crtc_state)
+				    struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 
 	if (!crtc_state->wm.need_postvbl_update)
 		return;
@@ -2187,7 +2211,7 @@ static void i965_update_wm(struct intel_crtc *unused_crtc)
 		/* self-refresh has much higher latency */
 		static const int sr_latency_ns = 12000;
 		const struct drm_display_mode *adjusted_mode =
-			&crtc->config->base.adjusted_mode;
+			&crtc->config->hw.adjusted_mode;
 		const struct drm_framebuffer *fb =
 			crtc->base.primary->state->fb;
 		int clock = adjusted_mode->crtc_clock;
@@ -2268,7 +2292,7 @@ static void i9xx_update_wm(struct intel_crtc *unused_crtc)
 	crtc = intel_get_crtc_for_plane(dev_priv, PLANE_A);
 	if (intel_crtc_active(crtc)) {
 		const struct drm_display_mode *adjusted_mode =
-			&crtc->config->base.adjusted_mode;
+			&crtc->config->hw.adjusted_mode;
 		const struct drm_framebuffer *fb =
 			crtc->base.primary->state->fb;
 		int cpp;
@@ -2295,7 +2319,7 @@ static void i9xx_update_wm(struct intel_crtc *unused_crtc)
 	crtc = intel_get_crtc_for_plane(dev_priv, PLANE_B);
 	if (intel_crtc_active(crtc)) {
 		const struct drm_display_mode *adjusted_mode =
-			&crtc->config->base.adjusted_mode;
+			&crtc->config->hw.adjusted_mode;
 		const struct drm_framebuffer *fb =
 			crtc->base.primary->state->fb;
 		int cpp;
@@ -2343,7 +2367,7 @@ static void i9xx_update_wm(struct intel_crtc *unused_crtc)
 		/* self-refresh has much higher latency */
 		static const int sr_latency_ns = 6000;
 		const struct drm_display_mode *adjusted_mode =
-			&enabled->config->base.adjusted_mode;
+			&enabled->config->hw.adjusted_mode;
 		const struct drm_framebuffer *fb =
 			enabled->base.primary->state->fb;
 		int clock = adjusted_mode->crtc_clock;
@@ -2401,7 +2425,7 @@ static void i845_update_wm(struct intel_crtc *unused_crtc)
 	if (crtc == NULL)
 		return;
 
-	adjusted_mode = &crtc->config->base.adjusted_mode;
+	adjusted_mode = &crtc->config->hw.adjusted_mode;
 	planea_wm = intel_calculate_wm(adjusted_mode->crtc_clock,
 				       &i845_wm_info,
 				       dev_priv->display.get_fifo_size(dev_priv, PLANE_A),
@@ -2483,7 +2507,7 @@ static u32 ilk_compute_pri_wm(const struct intel_crtc_state *crtc_state,
 	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
-	cpp = plane_state->base.fb->format->cpp[0];
+	cpp = plane_state->hw.fb->format->cpp[0];
 
 	method1 = ilk_wm_method1(crtc_state->pixel_rate, cpp, mem_value);
 
@@ -2491,8 +2515,8 @@ static u32 ilk_compute_pri_wm(const struct intel_crtc_state *crtc_state,
 		return method1;
 
 	method2 = ilk_wm_method2(crtc_state->pixel_rate,
-				 crtc_state->base.adjusted_mode.crtc_htotal,
-				 drm_rect_width(&plane_state->base.dst),
+				 crtc_state->hw.adjusted_mode.crtc_htotal,
+				 drm_rect_width(&plane_state->uapi.dst),
 				 cpp, mem_value);
 
 	return min(method1, method2);
@@ -2515,12 +2539,12 @@ static u32 ilk_compute_spr_wm(const struct intel_crtc_state *crtc_state,
 	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
-	cpp = plane_state->base.fb->format->cpp[0];
+	cpp = plane_state->hw.fb->format->cpp[0];
 
 	method1 = ilk_wm_method1(crtc_state->pixel_rate, cpp, mem_value);
 	method2 = ilk_wm_method2(crtc_state->pixel_rate,
-				 crtc_state->base.adjusted_mode.crtc_htotal,
-				 drm_rect_width(&plane_state->base.dst),
+				 crtc_state->hw.adjusted_mode.crtc_htotal,
+				 drm_rect_width(&plane_state->uapi.dst),
 				 cpp, mem_value);
 	return min(method1, method2);
 }
@@ -2541,11 +2565,11 @@ static u32 ilk_compute_cur_wm(const struct intel_crtc_state *crtc_state,
 	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
-	cpp = plane_state->base.fb->format->cpp[0];
+	cpp = plane_state->hw.fb->format->cpp[0];
 
 	return ilk_wm_method2(crtc_state->pixel_rate,
-			      crtc_state->base.adjusted_mode.crtc_htotal,
-			      drm_rect_width(&plane_state->base.dst),
+			      crtc_state->hw.adjusted_mode.crtc_htotal,
+			      drm_rect_width(&plane_state->uapi.dst),
 			      cpp, mem_value);
 }
 
@@ -2559,9 +2583,10 @@ static u32 ilk_compute_fbc_wm(const struct intel_crtc_state *crtc_state,
 	if (!intel_wm_plane_visible(crtc_state, plane_state))
 		return 0;
 
-	cpp = plane_state->base.fb->format->cpp[0];
+	cpp = plane_state->hw.fb->format->cpp[0];
 
-	return ilk_wm_fbc(pri_val, drm_rect_width(&plane_state->base.dst), cpp);
+	return ilk_wm_fbc(pri_val, drm_rect_width(&plane_state->uapi.dst),
+			  cpp);
 }
 
 static unsigned int
@@ -2766,12 +2791,12 @@ static u32
 hsw_compute_linetime_wm(const struct intel_crtc_state *crtc_state)
 {
 	const struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(crtc_state->base.state);
+		to_intel_atomic_state(crtc_state->uapi.state);
 	const struct drm_display_mode *adjusted_mode =
-		&crtc_state->base.adjusted_mode;
+		&crtc_state->hw.adjusted_mode;
 	u32 linetime, ips_linetime;
 
-	if (!crtc_state->base.active)
+	if (!crtc_state->hw.active)
 		return 0;
 	if (WARN_ON(adjusted_mode->crtc_clock == 0))
 		return 0;
@@ -3081,11 +3106,9 @@ static bool ilk_validate_pipe_wm(const struct drm_i915_private *dev_priv,
 /* Compute new watermarks for the pipe */
 static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 {
-	struct drm_atomic_state *state = crtc_state->base.state;
-	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
+	struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct intel_pipe_wm *pipe_wm;
-	struct drm_device *dev = state->dev;
-	const struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_plane *plane;
 	const struct intel_plane_state *plane_state;
 	const struct intel_plane_state *pristate = NULL;
@@ -3105,12 +3128,12 @@ static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
 			curstate = plane_state;
 	}
 
-	pipe_wm->pipe_enabled = crtc_state->base.active;
+	pipe_wm->pipe_enabled = crtc_state->hw.active;
 	if (sprstate) {
-		pipe_wm->sprites_enabled = sprstate->base.visible;
-		pipe_wm->sprites_scaled = sprstate->base.visible &&
-			(drm_rect_width(&sprstate->base.dst) != drm_rect_width(&sprstate->base.src) >> 16 ||
-			 drm_rect_height(&sprstate->base.dst) != drm_rect_height(&sprstate->base.src) >> 16);
+		pipe_wm->sprites_enabled = sprstate->uapi.visible;
+		pipe_wm->sprites_scaled = sprstate->uapi.visible &&
+			(drm_rect_width(&sprstate->uapi.dst) != drm_rect_width(&sprstate->uapi.src) >> 16 ||
+			 drm_rect_height(&sprstate->uapi.dst) != drm_rect_height(&sprstate->uapi.src) >> 16);
 	}
 
 	usable_level = max_level;
@@ -3162,11 +3185,11 @@ static int ilk_compute_pipe_wm(struct intel_crtc_state *crtc_state)
  */
 static int ilk_compute_intermediate_wm(struct intel_crtc_state *newstate)
 {
-	struct intel_crtc *intel_crtc = to_intel_crtc(newstate->base.crtc);
+	struct intel_crtc *intel_crtc = to_intel_crtc(newstate->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev);
 	struct intel_pipe_wm *a = &newstate->wm.ilk.intermediate;
 	struct intel_atomic_state *intel_state =
-		to_intel_atomic_state(newstate->base.state);
+		to_intel_atomic_state(newstate->uapi.state);
 	const struct intel_crtc_state *oldstate =
 		intel_atomic_get_old_crtc_state(intel_state, intel_crtc);
 	const struct intel_pipe_wm *b = &oldstate->wm.ilk.optimal;
@@ -3178,7 +3201,7 @@ static int ilk_compute_intermediate_wm(struct intel_crtc_state *newstate)
 	 * and after the vblank.
 	 */
 	*a = newstate->wm.ilk.optimal;
-	if (!newstate->base.active || drm_atomic_crtc_needs_modeset(&newstate->base) ||
+	if (!newstate->hw.active || drm_atomic_crtc_needs_modeset(&newstate->uapi) ||
 	    intel_state->skip_intermediate_wm)
 		return 0;
 
@@ -3588,10 +3611,8 @@ static void ilk_write_wm_values(struct drm_i915_private *dev_priv,
 	dev_priv->wm.hw = *results;
 }
 
-bool ilk_disable_lp_wm(struct drm_device *dev)
+bool ilk_disable_lp_wm(struct drm_i915_private *dev_priv)
 {
-	struct drm_i915_private *dev_priv = to_i915(dev);
-
 	return _ilk_disable_lp_wm(dev_priv, WM_DIRTY_LP_ALL);
 }
 
@@ -3780,7 +3801,7 @@ bool intel_can_enable_sagv(struct intel_atomic_state *state)
 	crtc = intel_get_crtc_for_pipe(dev_priv, pipe);
 	crtc_state = to_intel_crtc_state(crtc->base.state);
 
-	if (crtc->base.state->adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
+	if (crtc_state->hw.adjusted_mode.flags & DRM_MODE_FLAG_INTERLACE)
 		return false;
 
 	for_each_intel_plane_on_crtc(dev, crtc, plane) {
@@ -3830,7 +3851,7 @@ static u16 intel_get_ddb_size(struct drm_i915_private *dev_priv,
 	if (INTEL_GEN(dev_priv) < 11)
 		return ddb_size - 4; /* 4 blocks for bypass path allocation */
 
-	adjusted_mode = &crtc_state->base.adjusted_mode;
+	adjusted_mode = &crtc_state->hw.adjusted_mode;
 	total_data_bw = total_data_rate * drm_mode_vrefresh(adjusted_mode);
 
 	/*
@@ -3859,16 +3880,16 @@ skl_ddb_get_pipe_allocation_limits(struct drm_i915_private *dev_priv,
 				   struct skl_ddb_entry *alloc, /* out */
 				   int *num_active /* out */)
 {
-	struct drm_atomic_state *state = crtc_state->base.state;
+	struct drm_atomic_state *state = crtc_state->uapi.state;
 	struct intel_atomic_state *intel_state = to_intel_atomic_state(state);
-	struct drm_crtc *for_crtc = crtc_state->base.crtc;
+	struct drm_crtc *for_crtc = crtc_state->uapi.crtc;
 	const struct intel_crtc *crtc;
 	u32 pipe_width = 0, total_width = 0, width_before_pipe = 0;
 	enum pipe for_pipe = to_intel_crtc(for_crtc)->pipe;
 	u16 ddb_size;
 	u32 i;
 
-	if (WARN_ON(!state) || !crtc_state->base.active) {
+	if (WARN_ON(!state) || !crtc_state->hw.active) {
 		alloc->start = 0;
 		alloc->end = 0;
 		*num_active = hweight8(dev_priv->active_pipes);
@@ -3907,11 +3928,11 @@ skl_ddb_get_pipe_allocation_limits(struct drm_i915_private *dev_priv,
 	 */
 	for_each_new_intel_crtc_in_state(intel_state, crtc, crtc_state, i) {
 		const struct drm_display_mode *adjusted_mode =
-			&crtc_state->base.adjusted_mode;
+			&crtc_state->hw.adjusted_mode;
 		enum pipe pipe = crtc->pipe;
 		int hdisplay, vdisplay;
 
-		if (!crtc_state->base.enable)
+		if (!crtc_state->hw.enable)
 			continue;
 
 		drm_mode_get_hv_timing(adjusted_mode, &hdisplay, &vdisplay);
@@ -3942,7 +3963,7 @@ static unsigned int
 skl_cursor_allocation(const struct intel_crtc_state *crtc_state,
 		      int num_active)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	int level, max_level = ilk_wm_max_level(dev_priv);
 	struct skl_wm_level wm = {};
 	int ret, min_ddb_alloc = 0;
@@ -4082,10 +4103,10 @@ skl_plane_downscale_amount(const struct intel_crtc_state *crtc_state,
 	 *
 	 * n.b., src is 16.16 fixed point, dst is whole integer.
 	 */
-	src_w = drm_rect_width(&plane_state->base.src) >> 16;
-	src_h = drm_rect_height(&plane_state->base.src) >> 16;
-	dst_w = drm_rect_width(&plane_state->base.dst);
-	dst_h = drm_rect_height(&plane_state->base.dst);
+	src_w = drm_rect_width(&plane_state->uapi.src) >> 16;
+	src_h = drm_rect_height(&plane_state->uapi.src) >> 16;
+	dst_w = drm_rect_width(&plane_state->uapi.dst);
+	dst_h = drm_rect_height(&plane_state->uapi.dst);
 
 	fp_w_ratio = div_fixed16(src_w, dst_w);
 	fp_h_ratio = div_fixed16(src_h, dst_h);
@@ -4100,21 +4121,21 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *crtc_state,
 			     const struct intel_plane_state *plane_state,
 			     int color_plane)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	u32 data_rate;
 	u32 width = 0, height = 0;
 	uint_fixed_16_16_t down_scale_amount;
 	u64 rate;
 
-	if (!plane_state->base.visible)
+	if (!plane_state->uapi.visible)
 		return 0;
 
 	if (plane->id == PLANE_CURSOR)
 		return 0;
 
 	if (color_plane == 1 &&
-	    !drm_format_info_is_yuv_semiplanar(fb->format))
+	    !intel_format_info_is_yuv_semiplanar(fb->format, fb->modifier))
 		return 0;
 
 	/*
@@ -4122,8 +4143,8 @@ skl_plane_relative_data_rate(const struct intel_crtc_state *crtc_state,
 	 * the 90/270 degree plane rotation cases (to match the
 	 * GTT mapping), hence no need to account for rotation here.
 	 */
-	width = drm_rect_width(&plane_state->base.src) >> 16;
-	height = drm_rect_height(&plane_state->base.src) >> 16;
+	width = drm_rect_width(&plane_state->uapi.src) >> 16;
+	height = drm_rect_height(&plane_state->uapi.src) >> 16;
 
 	/* UV plane does 1/2 pixel sub-sampling */
 	if (color_plane == 1) {
@@ -4146,7 +4167,7 @@ skl_get_total_relative_data_rate(struct intel_crtc_state *crtc_state,
 				 u64 *plane_data_rate,
 				 u64 *uv_plane_data_rate)
 {
-	struct drm_atomic_state *state = crtc_state->base.state;
+	struct drm_atomic_state *state = crtc_state->uapi.state;
 	struct intel_plane *plane;
 	const struct intel_plane_state *plane_state;
 	u64 total_data_rate = 0;
@@ -4181,7 +4202,7 @@ icl_get_total_relative_data_rate(struct intel_crtc_state *crtc_state,
 	const struct intel_plane_state *plane_state;
 	u64 total_data_rate = 0;
 
-	if (WARN_ON(!crtc_state->base.state))
+	if (WARN_ON(!crtc_state->uapi.state))
 		return 0;
 
 	/* Calculate and cache data rate for each plane */
@@ -4225,8 +4246,8 @@ static int
 skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state,
 		      struct skl_ddb_allocation *ddb /* out */)
 {
-	struct drm_atomic_state *state = crtc_state->base.state;
-	struct drm_crtc *crtc = crtc_state->base.crtc;
+	struct drm_atomic_state *state = crtc_state->uapi.state;
+	struct drm_crtc *crtc = crtc_state->uapi.crtc;
 	struct drm_i915_private *dev_priv = to_i915(crtc->dev);
 	struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
 	struct skl_ddb_entry *alloc = &crtc_state->wm.skl.ddb;
@@ -4248,7 +4269,7 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state,
 	if (WARN_ON(!state))
 		return 0;
 
-	if (!crtc_state->base.active) {
+	if (!crtc_state->hw.active) {
 		alloc->start = alloc->end = 0;
 		return 0;
 	}
@@ -4291,8 +4312,8 @@ skl_allocate_pipe_ddb(struct intel_crtc_state *crtc_state,
 				&crtc_state->wm.skl.optimal.planes[plane_id];
 
 			if (plane_id == PLANE_CURSOR) {
-				if (WARN_ON(wm->wm[level].min_ddb_alloc >
-					    total[PLANE_CURSOR])) {
+				if (wm->wm[level].min_ddb_alloc > total[PLANE_CURSOR]) {
+					WARN_ON(wm->wm[level].min_ddb_alloc != U16_MAX);
 					blocks = U32_MAX;
 					break;
 				}
@@ -4490,7 +4511,7 @@ intel_get_linetime_us(const struct intel_crtc_state *crtc_state)
 	u32 crtc_htotal;
 	uint_fixed_16_16_t linetime_us;
 
-	if (!crtc_state->base.active)
+	if (!crtc_state->hw.active)
 		return u32_to_fixed16(0);
 
 	pixel_rate = crtc_state->pixel_rate;
@@ -4498,7 +4519,7 @@ intel_get_linetime_us(const struct intel_crtc_state *crtc_state)
 	if (WARN_ON(pixel_rate == 0))
 		return u32_to_fixed16(0);
 
-	crtc_htotal = crtc_state->base.adjusted_mode.crtc_htotal;
+	crtc_htotal = crtc_state->hw.adjusted_mode.crtc_htotal;
 	linetime_us = div_fixed16(crtc_htotal * 1000, pixel_rate);
 
 	return linetime_us;
@@ -4533,12 +4554,13 @@ skl_compute_wm_params(const struct intel_crtc_state *crtc_state,
 		      u32 plane_pixel_rate, struct skl_wm_params *wp,
 		      int color_plane)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct intel_crtc *crtc = to_intel_crtc(crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	u32 interm_pbpl;
 
 	/* only planar format has two planes */
-	if (color_plane == 1 && !drm_format_info_is_yuv_semiplanar(format)) {
+	if (color_plane == 1 &&
+	    !intel_format_info_is_yuv_semiplanar(format, modifier)) {
 		DRM_DEBUG_KMS("Non planar format have single plane\n");
 		return -EINVAL;
 	}
@@ -4550,7 +4572,7 @@ skl_compute_wm_params(const struct intel_crtc_state *crtc_state,
 	wp->x_tiled = modifier == I915_FORMAT_MOD_X_TILED;
 	wp->rc_surface = modifier == I915_FORMAT_MOD_Y_TILED_CCS ||
 			 modifier == I915_FORMAT_MOD_Yf_TILED_CCS;
-	wp->is_planar = drm_format_info_is_yuv_semiplanar(format);
+	wp->is_planar = intel_format_info_is_yuv_semiplanar(format, modifier);
 
 	wp->width = width;
 	if (color_plane == 1 && wp->is_planar)
@@ -4622,7 +4644,7 @@ skl_compute_plane_wm_params(const struct intel_crtc_state *crtc_state,
 			    const struct intel_plane_state *plane_state,
 			    struct skl_wm_params *wp, int color_plane)
 {
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	int width;
 
 	/*
@@ -4630,11 +4652,11 @@ skl_compute_plane_wm_params(const struct intel_crtc_state *crtc_state,
 	 * the 90/270 degree plane rotation cases (to match the
 	 * GTT mapping), hence no need to account for rotation here.
 	 */
-	width = drm_rect_width(&plane_state->base.src) >> 16;
+	width = drm_rect_width(&plane_state->uapi.src) >> 16;
 
 	return skl_compute_wm_params(crtc_state, width,
 				     fb->format, fb->modifier,
-				     plane_state->base.rotation,
+				     plane_state->hw.rotation,
 				     skl_adjusted_plane_pixel_rate(crtc_state, plane_state),
 				     wp, color_plane);
 }
@@ -4654,7 +4676,7 @@ static void skl_compute_plane_wm(const struct intel_crtc_state *crtc_state,
 				 const struct skl_wm_level *result_prev,
 				 struct skl_wm_level *result /* out */)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	u32 latency = dev_priv->wm.skl_latency[level];
 	uint_fixed_16_16_t method1, method2;
 	uint_fixed_16_16_t selected_result;
@@ -4680,14 +4702,14 @@ static void skl_compute_plane_wm(const struct intel_crtc_state *crtc_state,
 	method1 = skl_wm_method1(dev_priv, wp->plane_pixel_rate,
 				 wp->cpp, latency, wp->dbuf_block_size);
 	method2 = skl_wm_method2(wp->plane_pixel_rate,
-				 crtc_state->base.adjusted_mode.crtc_htotal,
+				 crtc_state->hw.adjusted_mode.crtc_htotal,
 				 latency,
 				 wp->plane_blocks_per_line);
 
 	if (wp->y_tiled) {
 		selected_result = max_fixed16(method2, wp->y_tile_minimum);
 	} else {
-		if ((wp->cpp * crtc_state->base.adjusted_mode.crtc_htotal /
+		if ((wp->cpp * crtc_state->hw.adjusted_mode.crtc_htotal /
 		     wp->dbuf_block_size < 1) &&
 		     (wp->plane_bytes_per_line / wp->dbuf_block_size < 1)) {
 			selected_result = method2;
@@ -4778,7 +4800,7 @@ skl_compute_wm_levels(const struct intel_crtc_state *crtc_state,
 		      const struct skl_wm_params *wm_params,
 		      struct skl_wm_level *levels)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	int level, max_level = ilk_wm_max_level(dev_priv);
 	struct skl_wm_level *result_prev = &levels[0];
 
@@ -4795,7 +4817,7 @@ skl_compute_wm_levels(const struct intel_crtc_state *crtc_state,
 static u32
 skl_compute_linetime_wm(const struct intel_crtc_state *crtc_state)
 {
-	struct drm_atomic_state *state = crtc_state->base.state;
+	struct drm_atomic_state *state = crtc_state->uapi.state;
 	struct drm_i915_private *dev_priv = to_i915(state->dev);
 	uint_fixed_16_16_t linetime_us;
 	u32 linetime_wm;
@@ -4814,7 +4836,7 @@ static void skl_compute_transition_wm(const struct intel_crtc_state *crtc_state,
 				      const struct skl_wm_params *wp,
 				      struct skl_plane_wm *wm)
 {
-	struct drm_device *dev = crtc_state->base.crtc->dev;
+	struct drm_device *dev = crtc_state->uapi.crtc->dev;
 	const struct drm_i915_private *dev_priv = to_i915(dev);
 	u16 trans_min, trans_y_tile_min;
 	const u16 trans_amount = 10; /* This is configurable amount */
@@ -4912,8 +4934,8 @@ static int skl_build_plane_wm_uv(struct intel_crtc_state *crtc_state,
 static int skl_build_plane_wm(struct intel_crtc_state *crtc_state,
 			      const struct intel_plane_state *plane_state)
 {
-	struct intel_plane *plane = to_intel_plane(plane_state->base.plane);
-	const struct drm_framebuffer *fb = plane_state->base.fb;
+	struct intel_plane *plane = to_intel_plane(plane_state->uapi.plane);
+	const struct drm_framebuffer *fb = plane_state->hw.fb;
 	enum plane_id plane_id = plane->id;
 	int ret;
 
@@ -4938,7 +4960,7 @@ static int skl_build_plane_wm(struct intel_crtc_state *crtc_state,
 static int icl_build_plane_wm(struct intel_crtc_state *crtc_state,
 			      const struct intel_plane_state *plane_state)
 {
-	enum plane_id plane_id = to_intel_plane(plane_state->base.plane)->id;
+	enum plane_id plane_id = to_intel_plane(plane_state->uapi.plane)->id;
 	int ret;
 
 	/* Watermarks calculated in master */
@@ -4946,7 +4968,7 @@ static int icl_build_plane_wm(struct intel_crtc_state *crtc_state,
 		return 0;
 
 	if (plane_state->planar_linked_plane) {
-		const struct drm_framebuffer *fb = plane_state->base.fb;
+		const struct drm_framebuffer *fb = plane_state->hw.fb;
 		enum plane_id y_plane_id = plane_state->planar_linked_plane->id;
 
 		WARN_ON(!intel_wm_plane_visible(crtc_state, plane_state));
@@ -4974,7 +4996,7 @@ static int icl_build_plane_wm(struct intel_crtc_state *crtc_state,
 
 static int skl_build_pipe_wm(struct intel_crtc_state *crtc_state)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
+	struct drm_i915_private *dev_priv = to_i915(crtc_state->uapi.crtc->dev);
 	struct skl_pipe_wm *pipe_wm = &crtc_state->wm.skl.optimal;
 	struct intel_plane *plane;
 	const struct intel_plane_state *plane_state;
@@ -5151,8 +5173,8 @@ static int
 skl_ddb_add_affected_planes(const struct intel_crtc_state *old_crtc_state,
 			    struct intel_crtc_state *new_crtc_state)
 {
-	struct intel_atomic_state *state = to_intel_atomic_state(new_crtc_state->base.state);
-	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->base.crtc);
+	struct intel_atomic_state *state = to_intel_atomic_state(new_crtc_state->uapi.state);
+	struct intel_crtc *crtc = to_intel_crtc(new_crtc_state->uapi.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_plane *plane;
 
@@ -5356,7 +5378,7 @@ skl_ddb_add_affected_pipes(struct intel_atomic_state *state)
 		if (ret)
 			return ret;
 
-		state->active_pipe_changes = ~0;
+		state->active_pipe_changes = INTEL_INFO(dev_priv)->pipe_mask;
 
 		/*
 		 * We usually only initialize state->active_pipes if we
@@ -5382,7 +5404,7 @@ skl_ddb_add_affected_pipes(struct intel_atomic_state *state)
 	 * to grab the lock on *all* CRTC's.
 	 */
 	if (state->active_pipe_changes || state->modeset) {
-		state->wm_results.dirty_pipes = ~0;
+		state->wm_results.dirty_pipes = INTEL_INFO(dev_priv)->pipe_mask;
 
 		ret = intel_add_all_pipes(state);
 		if (ret)
@@ -5436,7 +5458,7 @@ static int skl_wm_add_affected_planes(struct intel_atomic_state *state,
 		 * power well the hardware state will go out of sync
 		 * with the software state.
 		 */
-		if (!drm_atomic_crtc_needs_modeset(&new_crtc_state->base) &&
+		if (!drm_atomic_crtc_needs_modeset(&new_crtc_state->uapi) &&
 		    skl_plane_wm_equals(dev_priv,
 					&old_crtc_state->wm.skl.optimal.planes[plane_id],
 					&new_crtc_state->wm.skl.optimal.planes[plane_id]))
@@ -5500,11 +5522,12 @@ skl_compute_wm(struct intel_atomic_state *state)
 }
 
 static void skl_atomic_update_crtc_wm(struct intel_atomic_state *state,
-				      struct intel_crtc_state *crtc_state)
+				      struct intel_crtc *crtc)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
-	struct drm_i915_private *dev_priv = to_i915(state->base.dev);
-	struct skl_pipe_wm *pipe_wm = &crtc_state->wm.skl.optimal;
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
+	const struct skl_pipe_wm *pipe_wm = &crtc_state->wm.skl.optimal;
 	enum pipe pipe = crtc->pipe;
 
 	if ((state->wm_results.dirty_pipes & BIT(crtc->pipe)) == 0)
@@ -5514,10 +5537,11 @@ static void skl_atomic_update_crtc_wm(struct intel_atomic_state *state,
 }
 
 static void skl_initial_wm(struct intel_atomic_state *state,
-			   struct intel_crtc_state *crtc_state)
+			   struct intel_crtc *crtc)
 {
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 	struct skl_ddb_values *results = &state->wm_results;
 
 	if ((results->dirty_pipes & BIT(crtc->pipe)) == 0)
@@ -5525,8 +5549,8 @@ static void skl_initial_wm(struct intel_atomic_state *state,
 
 	mutex_lock(&dev_priv->wm.wm_mutex);
 
-	if (crtc_state->base.active_changed)
-		skl_atomic_update_crtc_wm(state, crtc_state);
+	if (crtc_state->uapi.active_changed)
+		skl_atomic_update_crtc_wm(state, crtc);
 
 	mutex_unlock(&dev_priv->wm.wm_mutex);
 }
@@ -5582,10 +5606,11 @@ static void ilk_program_watermarks(struct drm_i915_private *dev_priv)
 }
 
 static void ilk_initial_watermarks(struct intel_atomic_state *state,
-				   struct intel_crtc_state *crtc_state)
+				   struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 
 	mutex_lock(&dev_priv->wm.wm_mutex);
 	crtc->wm.active.ilk = crtc_state->wm.ilk.intermediate;
@@ -5594,10 +5619,11 @@ static void ilk_initial_watermarks(struct intel_atomic_state *state,
 }
 
 static void ilk_optimize_watermarks(struct intel_atomic_state *state,
-				    struct intel_crtc_state *crtc_state)
+				    struct intel_crtc *crtc)
 {
-	struct drm_i915_private *dev_priv = to_i915(crtc_state->base.crtc->dev);
-	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
+	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
+	const struct intel_crtc_state *crtc_state =
+		intel_atomic_get_new_crtc_state(state, crtc);
 
 	if (!crtc_state->wm.need_postvbl_update)
 		return;
@@ -5938,7 +5964,7 @@ void g4x_wm_sanitize(struct drm_i915_private *dev_priv)
 		enum plane_id plane_id = plane->id;
 		int level;
 
-		if (plane_state->base.visible)
+		if (plane_state->uapi.visible)
 			continue;
 
 		for (level = 0; level < 3; level++) {
@@ -6093,7 +6119,7 @@ void vlv_wm_sanitize(struct drm_i915_private *dev_priv)
 		enum plane_id plane_id = plane->id;
 		int level;
 
-		if (plane_state->base.visible)
+		if (plane_state->uapi.visible)
 			continue;
 
 		for (level = 0; level < wm_state->num_levels; level++) {
@@ -6369,7 +6395,6 @@ static void cpt_init_clock_gating(struct drm_i915_private *dev_priv)
 		val &= ~TRANS_CHICKEN2_FDI_POLARITY_REVERSED;
 		if (dev_priv->vbt.fdi_rx_polarity_inverted)
 			val |= TRANS_CHICKEN2_FDI_POLARITY_REVERSED;
-		val &= ~TRANS_CHICKEN2_FRAME_START_DELAY_MASK;
 		val &= ~TRANS_CHICKEN2_DISABLE_DEEP_COLOR_COUNTER;
 		val &= ~TRANS_CHICKEN2_DISABLE_DEEP_COLOR_MODESWITCH;
 		I915_WRITE(TRANS_CHICKEN2(pipe), val);
diff --git a/drivers/gpu/drm/i915/intel_pm.h b/drivers/gpu/drm/i915/intel_pm.h
index b579c724b915..c06c6a846d9a 100644
--- a/drivers/gpu/drm/i915/intel_pm.h
+++ b/drivers/gpu/drm/i915/intel_pm.h
@@ -54,7 +54,7 @@ void skl_write_plane_wm(struct intel_plane *plane,
 			const struct intel_crtc_state *crtc_state);
 void skl_write_cursor_wm(struct intel_plane *plane,
 			 const struct intel_crtc_state *crtc_state);
-bool ilk_disable_lp_wm(struct drm_device *dev);
+bool ilk_disable_lp_wm(struct drm_i915_private *dev_priv);
 void intel_init_ipc(struct drm_i915_private *dev_priv);
 void intel_enable_ipc(struct drm_i915_private *dev_priv);
 
diff --git a/drivers/gpu/drm/i915/intel_region_lmem.c b/drivers/gpu/drm/i915/intel_region_lmem.c
index 583118095635..eddb392917aa 100644
--- a/drivers/gpu/drm/i915/intel_region_lmem.c
+++ b/drivers/gpu/drm/i915/intel_region_lmem.c
@@ -51,8 +51,10 @@ static int init_fake_lmem_bar(struct intel_memory_region *mem)
 
 static void release_fake_lmem_bar(struct intel_memory_region *mem)
 {
-	if (drm_mm_node_allocated(&mem->fake_mappable))
-		drm_mm_remove_node(&mem->fake_mappable);
+	if (!drm_mm_node_allocated(&mem->fake_mappable))
+		return;
+
+	drm_mm_remove_node(&mem->fake_mappable);
 
 	dma_unmap_resource(&mem->i915->drm.pdev->dev,
 			   mem->remap_addr,
diff --git a/drivers/gpu/drm/i915/intel_wakeref.h b/drivers/gpu/drm/i915/intel_wakeref.h
index da6e8fd506e6..8d945db94b7a 100644
--- a/drivers/gpu/drm/i915/intel_wakeref.h
+++ b/drivers/gpu/drm/i915/intel_wakeref.h
@@ -59,9 +59,7 @@ void __intel_wakeref_put_last(struct intel_wakeref *wf, unsigned long flags);
 
 /**
  * intel_wakeref_get: Acquire the wakeref
- * @i915: the drm_i915_private device
  * @wf: the wakeref
- * @fn: callback for acquired the wakeref, called only on first acquire.
  *
  * Acquire a hold on the wakeref. The first user to do so, will acquire
  * the runtime pm wakeref and then call the @fn underneath the wakeref
@@ -76,6 +74,7 @@ void __intel_wakeref_put_last(struct intel_wakeref *wf, unsigned long flags);
 static inline int
 intel_wakeref_get(struct intel_wakeref *wf)
 {
+	might_sleep();
 	if (unlikely(!atomic_inc_not_zero(&wf->count)))
 		return __intel_wakeref_get_first(wf);
 
@@ -83,6 +82,22 @@ intel_wakeref_get(struct intel_wakeref *wf)
 }
 
 /**
+ * __intel_wakeref_get: Acquire the wakeref, again
+ * @wf: the wakeref
+ *
+ * Increment the wakeref counter, only valid if it is already held by
+ * the caller.
+ *
+ * See intel_wakeref_get().
+ */
+static inline void
+__intel_wakeref_get(struct intel_wakeref *wf)
+{
+	INTEL_WAKEREF_BUG_ON(atomic_read(&wf->count) <= 0);
+	atomic_inc(&wf->count);
+}
+
+/**
  * intel_wakeref_get_if_in_use: Acquire the wakeref
  * @wf: the wakeref
  *
diff --git a/drivers/gpu/drm/i915/selftests/i915_active.c b/drivers/gpu/drm/i915/selftests/i915_active.c
index 260b0ee5d1e3..ef572a0c2566 100644
--- a/drivers/gpu/drm/i915/selftests/i915_active.c
+++ b/drivers/gpu/drm/i915/selftests/i915_active.c
@@ -99,7 +99,7 @@ __live_active_setup(struct drm_i915_private *i915)
 	for_each_uabi_engine(engine, i915) {
 		struct i915_request *rq;
 
-		rq = i915_request_create(engine->kernel_context);
+		rq = intel_engine_create_kernel_request(engine);
 		if (IS_ERR(rq)) {
 			err = PTR_ERR(rq);
 			break;
@@ -155,7 +155,11 @@ static int live_active_wait(void *arg)
 
 	i915_active_wait(&active->base);
 	if (!READ_ONCE(active->retired)) {
+		struct drm_printer p = drm_err_printer(__func__);
+
 		pr_err("i915_active not retired after waiting!\n");
+		i915_active_print(&active->base, &p);
+
 		err = -EINVAL;
 	}
 
@@ -184,7 +188,11 @@ static int live_active_retire(void *arg)
 		err = -EIO;
 
 	if (!READ_ONCE(active->retired)) {
+		struct drm_printer p = drm_err_printer(__func__);
+
 		pr_err("i915_active not retired after flushing!\n");
+		i915_active_print(&active->base, &p);
+
 		err = -EINVAL;
 	}
 
@@ -250,3 +258,36 @@ void i915_active_print(struct i915_active *ref, struct drm_printer *m)
 		i915_active_release(ref);
 	}
 }
+
+static void spin_unlock_wait(spinlock_t *lock)
+{
+	spin_lock_irq(lock);
+	spin_unlock_irq(lock);
+}
+
+void i915_active_unlock_wait(struct i915_active *ref)
+{
+	if (i915_active_acquire_if_busy(ref)) {
+		struct active_node *it, *n;
+
+		rcu_read_lock();
+		rbtree_postorder_for_each_entry_safe(it, n, &ref->tree, node) {
+			struct dma_fence *f;
+
+			/* Wait for all active callbacks */
+			f = rcu_dereference(it->base.fence);
+			if (f)
+				spin_unlock_wait(f->lock);
+		}
+		rcu_read_unlock();
+
+		i915_active_release(ref);
+	}
+
+	/* And wait for the retire callback */
+	spin_lock_irq(&ref->tree_lock);
+	spin_unlock_irq(&ref->tree_lock);
+
+	/* ... which may have been on a thread instead */
+	flush_work(&ref->work);
+}
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem.c b/drivers/gpu/drm/i915/selftests/i915_gem.c
index d83f6bf6d9d4..b37fc53973cc 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem.c
@@ -9,6 +9,7 @@
 #include "gem/selftests/igt_gem_utils.h"
 #include "gem/selftests/mock_context.h"
 #include "gt/intel_gt.h"
+#include "gt/intel_gt_pm.h"
 
 #include "i915_selftest.h"
 
@@ -136,7 +137,7 @@ static int igt_gem_suspend(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct i915_gem_context *ctx;
-	struct drm_file *file;
+	struct file *file;
 	int err;
 
 	file = mock_file(i915);
@@ -163,7 +164,7 @@ static int igt_gem_suspend(void *arg)
 
 	err = switch_to_context(ctx);
 out:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
@@ -171,7 +172,7 @@ static int igt_gem_hibernate(void *arg)
 {
 	struct drm_i915_private *i915 = arg;
 	struct i915_gem_context *ctx;
-	struct drm_file *file;
+	struct file *file;
 	int err;
 
 	file = mock_file(i915);
@@ -198,7 +199,7 @@ static int igt_gem_hibernate(void *arg)
 
 	err = switch_to_context(ctx);
 out:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_evict.c b/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
index 42e948144f1b..06ef88510209 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_evict.c
@@ -198,8 +198,8 @@ static int igt_overcommit(void *arg)
 	quirk_add(obj, &objects);
 
 	vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, 0);
-	if (!IS_ERR(vma) || PTR_ERR(vma) != -ENOSPC) {
-		pr_err("Failed to evict+insert, i915_gem_object_ggtt_pin returned err=%d\n", (int)PTR_ERR(vma));
+	if (vma != ERR_PTR(-ENOSPC)) {
+		pr_err("Failed to evict+insert, i915_gem_object_ggtt_pin returned err=%d\n", (int)PTR_ERR_OR_ZERO(vma));
 		err = -EINVAL;
 		goto cleanup;
 	}
@@ -466,7 +466,7 @@ static int igt_evict_contexts(void *arg)
 	/* Overfill the GGTT with context objects and so try to evict one. */
 	for_each_engine(engine, gt, id) {
 		struct i915_sw_fence fence;
-		struct drm_file *file;
+		struct file *file;
 
 		file = mock_file(i915);
 		if (IS_ERR(file)) {
@@ -515,7 +515,7 @@ static int igt_evict_contexts(void *arg)
 		pr_info("Submitted %lu contexts/requests on %s\n",
 			count, engine->name);
 
-		mock_file_free(i915, file);
+		fput(file);
 		if (err)
 			break;
 	}
diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
index 3f7e80fb3bbd..80cde5bda922 100644
--- a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c
@@ -212,10 +212,12 @@ static int lowlevel_hole(struct drm_i915_private *i915,
 			 unsigned long end_time)
 {
 	I915_RND_STATE(seed_prng);
+	struct i915_vma *mock_vma;
 	unsigned int size;
-	struct i915_vma mock_vma;
 
-	memset(&mock_vma, 0, sizeof(struct i915_vma));
+	mock_vma = kzalloc(sizeof(*mock_vma), GFP_KERNEL);
+	if (!mock_vma)
+		return -ENOMEM;
 
 	/* Keep creating larger objects until one cannot fit into the hole */
 	for (size = 12; (hole_end - hole_start) >> size; size++) {
@@ -239,8 +241,10 @@ static int lowlevel_hole(struct drm_i915_private *i915,
 			if (order)
 				break;
 		} while (count >>= 1);
-		if (!count)
+		if (!count) {
+			kfree(mock_vma);
 			return -ENOMEM;
+		}
 		GEM_BUG_ON(!order);
 
 		GEM_BUG_ON(count * BIT_ULL(size) > vm->total);
@@ -283,12 +287,12 @@ static int lowlevel_hole(struct drm_i915_private *i915,
 			    vm->allocate_va_range(vm, addr, BIT_ULL(size)))
 				break;
 
-			mock_vma.pages = obj->mm.pages;
-			mock_vma.node.size = BIT_ULL(size);
-			mock_vma.node.start = addr;
+			mock_vma->pages = obj->mm.pages;
+			mock_vma->node.size = BIT_ULL(size);
+			mock_vma->node.start = addr;
 
 			with_intel_runtime_pm(&i915->runtime_pm, wakeref)
-				vm->insert_entries(vm, &mock_vma,
+				vm->insert_entries(vm, mock_vma,
 						   I915_CACHE_NONE, 0);
 		}
 		count = n;
@@ -311,6 +315,7 @@ static int lowlevel_hole(struct drm_i915_private *i915,
 		cleanup_freed_objects(i915);
 	}
 
+	kfree(mock_vma);
 	return 0;
 }
 
@@ -1001,9 +1006,9 @@ static int exercise_ppgtt(struct drm_i915_private *dev_priv,
 				      u64 hole_start, u64 hole_end,
 				      unsigned long end_time))
 {
-	struct drm_file *file;
 	struct i915_ppgtt *ppgtt;
 	IGT_TIMEOUT(end_time);
+	struct file *file;
 	int err;
 
 	if (!HAS_FULL_PPGTT(dev_priv))
@@ -1026,7 +1031,7 @@ static int exercise_ppgtt(struct drm_i915_private *dev_priv,
 	i915_vm_put(&ppgtt->vm);
 
 out_free:
-	mock_file_free(dev_priv, file);
+	fput(file);
 	return err;
 }
 
@@ -1161,11 +1166,13 @@ static int igt_ggtt_page(void *arg)
 		goto out_free;
 
 	memset(&tmp, 0, sizeof(tmp));
+	mutex_lock(&ggtt->vm.mutex);
 	err = drm_mm_insert_node_in_range(&ggtt->vm.mm, &tmp,
 					  count * PAGE_SIZE, 0,
 					  I915_COLOR_UNEVICTABLE,
 					  0, ggtt->mappable_end,
 					  DRM_MM_INSERT_LOW);
+	mutex_unlock(&ggtt->vm.mutex);
 	if (err)
 		goto out_unpin;
 
@@ -1217,7 +1224,9 @@ static int igt_ggtt_page(void *arg)
 out_remove:
 	ggtt->vm.clear_range(&ggtt->vm, tmp.start, tmp.size);
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
+	mutex_lock(&ggtt->vm.mutex);
 	drm_mm_remove_node(&tmp);
+	mutex_unlock(&ggtt->vm.mutex);
 out_unpin:
 	i915_gem_object_unpin_pages(obj);
 out_free:
@@ -1782,9 +1791,9 @@ static int igt_cs_tlb(void *arg)
 	struct i915_address_space *vm;
 	struct i915_gem_context *ctx;
 	struct intel_context *ce;
-	struct drm_file *file;
 	struct i915_vma *vma;
 	I915_RND_STATE(prng);
+	struct file *file;
 	unsigned int i;
 	u32 *result;
 	u32 *batch;
@@ -2022,7 +2031,7 @@ out_put_bbe:
 out_vm:
 	i915_vm_put(vm);
 out_unlock:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
diff --git a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
index 4b3cac73e291..476fba2ed8bb 100644
--- a/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
+++ b/drivers/gpu/drm/i915/selftests/i915_live_selftests.h
@@ -16,6 +16,7 @@ selftest(gt_engines, intel_engine_live_selftests)
 selftest(gt_timelines, intel_timeline_live_selftests)
 selftest(gt_contexts, intel_context_live_selftests)
 selftest(gt_lrc, intel_lrc_live_selftests)
+selftest(gt_mocs, intel_mocs_live_selftests)
 selftest(gt_pm, intel_gt_pm_live_selftests)
 selftest(gt_heartbeat, intel_heartbeat_live_selftests)
 selftest(requests, i915_request_live_selftests)
@@ -36,5 +37,6 @@ selftest(reset, intel_reset_live_selftests)
 selftest(memory_region, intel_memory_region_live_selftests)
 selftest(hangcheck, intel_hangcheck_live_selftests)
 selftest(execlists, intel_execlists_live_selftests)
-selftest(guc, intel_guc_live_selftest)
 selftest(perf, i915_perf_live_selftests)
+/* Here be dragons: keep last to run last! */
+selftest(late_gt_pm, intel_gt_pm_late_selftests)
diff --git a/drivers/gpu/drm/i915/selftests/i915_perf.c b/drivers/gpu/drm/i915/selftests/i915_perf.c
index aabd07f67e49..d1a1568c47ba 100644
--- a/drivers/gpu/drm/i915/selftests/i915_perf.c
+++ b/drivers/gpu/drm/i915/selftests/i915_perf.c
@@ -132,7 +132,7 @@ static int live_noa_delay(void *arg)
 	for (i = 0; i < 4; i++)
 		intel_write_status_page(stream->engine, 0x100 + i, 0);
 
-	rq = i915_request_create(stream->engine->kernel_context);
+	rq = intel_engine_create_kernel_request(stream->engine);
 	if (IS_ERR(rq)) {
 		err = PTR_ERR(rq);
 		goto out;
diff --git a/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h b/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h
new file mode 100644
index 000000000000..f7129a243daa
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* List each unit test as selftest(name, function)
+ *
+ * The name is used as both an enum and expanded as subtest__name to create
+ * a module parameter. It must be unique and legal for a C identifier.
+ *
+ * The function should be of type int function(void). It may be conditionally
+ * compiled using #if IS_ENABLED(DRM_I915_SELFTEST).
+ *
+ * Tests are executed in order by igt/i915_selftest
+ */
+selftest(engine_cs, intel_engine_cs_perf_selftests)
+selftest(blt, i915_gem_object_blt_perf_selftests)
diff --git a/drivers/gpu/drm/i915/selftests/i915_request.c b/drivers/gpu/drm/i915/selftests/i915_request.c
index 8618a4dc0701..f89d9c42f1fa 100644
--- a/drivers/gpu/drm/i915/selftests/i915_request.c
+++ b/drivers/gpu/drm/i915/selftests/i915_request.c
@@ -27,11 +27,13 @@
 #include "gem/i915_gem_pm.h"
 #include "gem/selftests/mock_context.h"
 
+#include "gt/intel_engine_pm.h"
 #include "gt/intel_gt.h"
 
 #include "i915_random.h"
 #include "i915_selftest.h"
 #include "igt_live_test.h"
+#include "igt_spinner.h"
 #include "lib_sw_fence.h"
 
 #include "mock_drm.h"
@@ -540,6 +542,7 @@ static int live_nop_request(void *arg)
 		if (err)
 			return err;
 
+		intel_engine_pm_get(engine);
 		for_each_prime_number_from(prime, 1, 8192) {
 			struct i915_request *request = NULL;
 
@@ -578,6 +581,7 @@ static int live_nop_request(void *arg)
 			if (__igt_timeout(end_time, NULL))
 				break;
 		}
+		intel_engine_pm_put(engine);
 
 		err = igt_live_test_end(&t);
 		if (err)
@@ -692,10 +696,13 @@ static int live_empty_request(void *arg)
 		if (err)
 			goto out_batch;
 
+		intel_engine_pm_get(engine);
+
 		/* Warmup / preload */
 		request = empty_request(engine, batch);
 		if (IS_ERR(request)) {
 			err = PTR_ERR(request);
+			intel_engine_pm_put(engine);
 			goto out_batch;
 		}
 		i915_request_wait(request, 0, MAX_SCHEDULE_TIMEOUT);
@@ -708,6 +715,7 @@ static int live_empty_request(void *arg)
 				request = empty_request(engine, batch);
 				if (IS_ERR(request)) {
 					err = PTR_ERR(request);
+					intel_engine_pm_put(engine);
 					goto out_batch;
 				}
 			}
@@ -721,6 +729,7 @@ static int live_empty_request(void *arg)
 				break;
 		}
 		i915_request_put(request);
+		intel_engine_pm_put(engine);
 
 		err = igt_live_test_end(&t);
 		if (err)
@@ -740,10 +749,8 @@ out_batch:
 
 static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 {
-	struct i915_gem_context *ctx = i915->kernel_context;
 	struct drm_i915_gem_object *obj;
 	const int gen = INTEL_GEN(i915);
-	struct i915_address_space *vm;
 	struct i915_vma *vma;
 	u32 *cmd;
 	int err;
@@ -752,9 +759,7 @@ static struct i915_vma *recursive_batch(struct drm_i915_private *i915)
 	if (IS_ERR(obj))
 		return ERR_CAST(obj);
 
-	vm = i915_gem_context_get_vm_rcu(ctx);
-	vma = i915_vma_instance(obj, vm, NULL);
-	i915_vm_put(vm);
+	vma = i915_vma_instance(obj, i915->gt.vm, NULL);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		goto err;
@@ -845,7 +850,7 @@ static int live_all_engines(void *arg)
 
 	idx = 0;
 	for_each_uabi_engine(engine, i915) {
-		request[idx] = i915_request_create(engine->kernel_context);
+		request[idx] = intel_engine_create_kernel_request(engine);
 		if (IS_ERR(request[idx])) {
 			err = PTR_ERR(request[idx]);
 			pr_err("%s: Request allocation failed with err=%d\n",
@@ -962,7 +967,7 @@ static int live_sequential_engines(void *arg)
 			goto out_free;
 		}
 
-		request[idx] = i915_request_create(engine->kernel_context);
+		request[idx] = intel_engine_create_kernel_request(engine);
 		if (IS_ERR(request[idx])) {
 			err = PTR_ERR(request[idx]);
 			pr_err("%s: Request allocation failed for %s with err=%d\n",
@@ -1067,15 +1072,18 @@ static int __live_parallel_engine1(void *arg)
 	struct intel_engine_cs *engine = arg;
 	IGT_TIMEOUT(end_time);
 	unsigned long count;
+	int err = 0;
 
 	count = 0;
+	intel_engine_pm_get(engine);
 	do {
 		struct i915_request *rq;
-		int err;
 
 		rq = i915_request_create(engine->kernel_context);
-		if (IS_ERR(rq))
-			return PTR_ERR(rq);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			break;
+		}
 
 		i915_request_get(rq);
 		i915_request_add(rq);
@@ -1085,13 +1093,14 @@ static int __live_parallel_engine1(void *arg)
 			err = -ETIME;
 		i915_request_put(rq);
 		if (err)
-			return err;
+			break;
 
 		count++;
 	} while (!__igt_timeout(end_time, NULL));
+	intel_engine_pm_put(engine);
 
 	pr_info("%s: %lu request + sync\n", engine->name, count);
-	return 0;
+	return err;
 }
 
 static int __live_parallel_engineN(void *arg)
@@ -1099,21 +1108,100 @@ static int __live_parallel_engineN(void *arg)
 	struct intel_engine_cs *engine = arg;
 	IGT_TIMEOUT(end_time);
 	unsigned long count;
+	int err = 0;
 
 	count = 0;
+	intel_engine_pm_get(engine);
 	do {
 		struct i915_request *rq;
 
 		rq = i915_request_create(engine->kernel_context);
-		if (IS_ERR(rq))
-			return PTR_ERR(rq);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			break;
+		}
 
 		i915_request_add(rq);
 		count++;
 	} while (!__igt_timeout(end_time, NULL));
+	intel_engine_pm_put(engine);
 
 	pr_info("%s: %lu requests\n", engine->name, count);
-	return 0;
+	return err;
+}
+
+static bool wake_all(struct drm_i915_private *i915)
+{
+	if (atomic_dec_and_test(&i915->selftest.counter)) {
+		wake_up_var(&i915->selftest.counter);
+		return true;
+	}
+
+	return false;
+}
+
+static int wait_for_all(struct drm_i915_private *i915)
+{
+	if (wake_all(i915))
+		return 0;
+
+	if (wait_var_event_timeout(&i915->selftest.counter,
+				   !atomic_read(&i915->selftest.counter),
+				   i915_selftest.timeout_jiffies))
+		return 0;
+
+	return -ETIME;
+}
+
+static int __live_parallel_spin(void *arg)
+{
+	struct intel_engine_cs *engine = arg;
+	struct igt_spinner spin;
+	struct i915_request *rq;
+	int err = 0;
+
+	/*
+	 * Create a spinner running for eternity on each engine. If a second
+	 * spinner is incorrectly placed on the same engine, it will not be
+	 * able to start in time.
+	 */
+
+	if (igt_spinner_init(&spin, engine->gt)) {
+		wake_all(engine->i915);
+		return -ENOMEM;
+	}
+
+	intel_engine_pm_get(engine);
+	rq = igt_spinner_create_request(&spin,
+					engine->kernel_context,
+					MI_NOOP); /* no preemption */
+	intel_engine_pm_put(engine);
+	if (IS_ERR(rq)) {
+		err = PTR_ERR(rq);
+		if (err == -ENODEV)
+			err = 0;
+		wake_all(engine->i915);
+		goto out_spin;
+	}
+
+	i915_request_get(rq);
+	i915_request_add(rq);
+	if (igt_wait_for_spinner(&spin, rq)) {
+		/* Occupy this engine for the whole test */
+		err = wait_for_all(engine->i915);
+	} else {
+		pr_err("Failed to start spinner on %s\n", engine->name);
+		err = -EINVAL;
+	}
+	igt_spinner_end(&spin);
+
+	if (err == 0 && i915_request_wait(rq, 0, HZ / 5) < 0)
+		err = -EIO;
+	i915_request_put(rq);
+
+out_spin:
+	igt_spinner_fini(&spin);
+	return err;
 }
 
 static int live_parallel_engines(void *arg)
@@ -1122,6 +1210,7 @@ static int live_parallel_engines(void *arg)
 	static int (* const func[])(void *arg) = {
 		__live_parallel_engine1,
 		__live_parallel_engineN,
+		__live_parallel_spin,
 		NULL,
 	};
 	const unsigned int nengines = num_uabi_engines(i915);
@@ -1140,13 +1229,17 @@ static int live_parallel_engines(void *arg)
 		return -ENOMEM;
 
 	for (fn = func; !err && *fn; fn++) {
+		char name[KSYM_NAME_LEN];
 		struct igt_live_test t;
 		unsigned int idx;
 
-		err = igt_live_test_begin(&t, i915, __func__, "");
+		snprintf(name, sizeof(name), "%pS", fn);
+		err = igt_live_test_begin(&t, i915, __func__, name);
 		if (err)
 			break;
 
+		atomic_set(&i915->selftest.counter, nengines);
+
 		idx = 0;
 		for_each_uabi_engine(engine, i915) {
 			tsk[idx] = kthread_run(*fn, engine,
@@ -1230,9 +1323,9 @@ static int live_breadcrumbs_smoketest(void *arg)
 	struct task_struct **threads;
 	struct igt_live_test live;
 	intel_wakeref_t wakeref;
-	struct drm_file *file;
 	struct smoketest *smoke;
 	unsigned int n, idx;
+	struct file *file;
 	int ret = 0;
 
 	/*
@@ -1354,7 +1447,7 @@ out_threads:
 out_smoke:
 	kfree(smoke);
 out_file:
-	mock_file_free(i915, file);
+	fput(file);
 out_rpm:
 	intel_runtime_pm_put(&i915->runtime_pm, wakeref);
 
diff --git a/drivers/gpu/drm/i915/selftests/i915_selftest.c b/drivers/gpu/drm/i915/selftests/i915_selftest.c
index a6cca4ad96f6..d3bf9eefb682 100644
--- a/drivers/gpu/drm/i915/selftests/i915_selftest.c
+++ b/drivers/gpu/drm/i915/selftests/i915_selftest.c
@@ -57,6 +57,12 @@ enum {
 #undef selftest
 };
 
+enum {
+#define selftest(name, func) perf_##name,
+#include "i915_perf_selftests.h"
+#undef selftest
+};
+
 struct selftest {
 	bool enabled;
 	const char *name;
@@ -78,6 +84,12 @@ static struct selftest live_selftests[] = {
 };
 #undef selftest
 
+#define selftest(n, f) [perf_##n] = { .name = #n, { .live = f } },
+static struct selftest perf_selftests[] = {
+#include "i915_perf_selftests.h"
+};
+#undef selftest
+
 /* Embed the line number into the parameter name so that we can order tests */
 #define selftest(n, func) selftest_0(n, func, param(n))
 #define param(n) __PASTE(igt__, __PASTE(__LINE__, __mock_##n))
@@ -93,6 +105,13 @@ module_param_named(id, live_selftests[live_##n].enabled, bool, 0400);
 #include "i915_live_selftests.h"
 #undef selftest_0
 #undef param
+
+#define param(n) __PASTE(igt__, __PASTE(__LINE__, __perf_##n))
+#define selftest_0(n, func, id) \
+module_param_named(id, perf_selftests[perf_##n].enabled, bool, 0400);
+#include "i915_perf_selftests.h"
+#undef selftest_0
+#undef param
 #undef selftest
 
 static void set_default_test_all(struct selftest *st, unsigned int count)
@@ -200,6 +219,27 @@ int i915_live_selftests(struct pci_dev *pdev)
 	return 0;
 }
 
+int i915_perf_selftests(struct pci_dev *pdev)
+{
+	int err;
+
+	if (!i915_selftest.perf)
+		return 0;
+
+	err = run_selftests(perf, pdev_to_i915(pdev));
+	if (err) {
+		i915_selftest.perf = err;
+		return err;
+	}
+
+	if (i915_selftest.perf < 0) {
+		i915_selftest.perf = -ENOTTY;
+		return 1;
+	}
+
+	return 0;
+}
+
 static bool apply_subtest_filter(const char *caller, const char *name)
 {
 	char *filter, *sep, *tok;
@@ -365,3 +405,6 @@ MODULE_PARM_DESC(mock_selftests, "Run selftests before loading, using mock hardw
 
 module_param_named_unsafe(live_selftests, i915_selftest.live, int, 0400);
 MODULE_PARM_DESC(live_selftests, "Run selftests after driver initialisation on the live system (0:disabled [default], 1:run tests then continue, -1:run tests then exit module)");
+
+module_param_named_unsafe(perf_selftests, i915_selftest.perf, int, 0400);
+MODULE_PARM_DESC(perf_selftests, "Run performance orientated selftests after driver initialisation on the live system (0:disabled [default], 1:run tests then continue, -1:run tests then exit module)");
diff --git a/drivers/gpu/drm/i915/selftests/igt_mmap.c b/drivers/gpu/drm/i915/selftests/igt_mmap.c
new file mode 100644
index 000000000000..583a4ff8b8c9
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/igt_mmap.c
@@ -0,0 +1,39 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#include <drm/drm_file.h>
+
+#include "i915_drv.h"
+#include "igt_mmap.h"
+
+unsigned long igt_mmap_node(struct drm_i915_private *i915,
+			    struct drm_vma_offset_node *node,
+			    unsigned long addr,
+			    unsigned long prot,
+			    unsigned long flags)
+{
+	struct file *file;
+	int err;
+
+	/* Pretend to open("/dev/dri/card0") */
+	file = mock_drm_getfile(i915->drm.primary, O_RDWR);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	err = drm_vma_node_allow(node, file->private_data);
+	if (err) {
+		addr = err;
+		goto out_file;
+	}
+
+	addr = vm_mmap(file, addr, drm_vma_node_size(node) << PAGE_SHIFT,
+		       prot, flags, drm_vma_node_offset_addr(node));
+
+	drm_vma_node_revoke(node, file->private_data);
+out_file:
+	fput(file);
+	return addr;
+}
diff --git a/drivers/gpu/drm/i915/selftests/igt_mmap.h b/drivers/gpu/drm/i915/selftests/igt_mmap.h
new file mode 100644
index 000000000000..6e716cb59d7e
--- /dev/null
+++ b/drivers/gpu/drm/i915/selftests/igt_mmap.h
@@ -0,0 +1,19 @@
+/*
+ * SPDX-License-Identifier: MIT
+ *
+ * Copyright © 2019 Intel Corporation
+ */
+
+#ifndef IGT_MMAP_H
+#define IGT_MMAP_H
+
+struct drm_i915_private;
+struct drm_vma_offset_node;
+
+unsigned long igt_mmap_node(struct drm_i915_private *i915,
+			    struct drm_vma_offset_node *node,
+			    unsigned long addr,
+			    unsigned long prot,
+			    unsigned long flags);
+
+#endif /* IGT_MMAP_H */
diff --git a/drivers/gpu/drm/i915/selftests/igt_spinner.c b/drivers/gpu/drm/i915/selftests/igt_spinner.c
index ee8450b871da..e8a58fe49c39 100644
--- a/drivers/gpu/drm/i915/selftests/igt_spinner.c
+++ b/drivers/gpu/drm/i915/selftests/igt_spinner.c
@@ -15,8 +15,6 @@ int igt_spinner_init(struct igt_spinner *spin, struct intel_gt *gt)
 	void *vaddr;
 	int err;
 
-	GEM_BUG_ON(INTEL_GEN(gt->i915) < 8);
-
 	memset(spin, 0, sizeof(*spin));
 	spin->gt = gt;
 
@@ -95,11 +93,15 @@ igt_spinner_create_request(struct igt_spinner *spin,
 	struct intel_engine_cs *engine = ce->engine;
 	struct i915_request *rq = NULL;
 	struct i915_vma *hws, *vma;
+	unsigned int flags;
 	u32 *batch;
 	int err;
 
 	GEM_BUG_ON(spin->gt != ce->vm->gt);
 
+	if (!intel_engine_can_store_dword(ce->engine))
+		return ERR_PTR(-ENODEV);
+
 	vma = i915_vma_instance(spin->obj, ce->vm, NULL);
 	if (IS_ERR(vma))
 		return ERR_CAST(vma);
@@ -132,16 +134,37 @@ igt_spinner_create_request(struct igt_spinner *spin,
 
 	batch = spin->batch;
 
-	*batch++ = MI_STORE_DWORD_IMM_GEN4;
-	*batch++ = lower_32_bits(hws_address(hws, rq));
-	*batch++ = upper_32_bits(hws_address(hws, rq));
+	if (INTEL_GEN(rq->i915) >= 8) {
+		*batch++ = MI_STORE_DWORD_IMM_GEN4;
+		*batch++ = lower_32_bits(hws_address(hws, rq));
+		*batch++ = upper_32_bits(hws_address(hws, rq));
+	} else if (INTEL_GEN(rq->i915) >= 6) {
+		*batch++ = MI_STORE_DWORD_IMM_GEN4;
+		*batch++ = 0;
+		*batch++ = hws_address(hws, rq);
+	} else if (INTEL_GEN(rq->i915) >= 4) {
+		*batch++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT;
+		*batch++ = 0;
+		*batch++ = hws_address(hws, rq);
+	} else {
+		*batch++ = MI_STORE_DWORD_IMM | MI_MEM_VIRTUAL;
+		*batch++ = hws_address(hws, rq);
+	}
 	*batch++ = rq->fence.seqno;
 
 	*batch++ = arbitration_command;
 
-	*batch++ = MI_BATCH_BUFFER_START | 1 << 8 | 1;
+	if (INTEL_GEN(rq->i915) >= 8)
+		*batch++ = MI_BATCH_BUFFER_START | BIT(8) | 1;
+	else if (IS_HASWELL(rq->i915))
+		*batch++ = MI_BATCH_BUFFER_START | MI_BATCH_PPGTT_HSW;
+	else if (INTEL_GEN(rq->i915) >= 6)
+		*batch++ = MI_BATCH_BUFFER_START;
+	else
+		*batch++ = MI_BATCH_BUFFER_START | MI_BATCH_GTT;
 	*batch++ = lower_32_bits(vma->node.start);
 	*batch++ = upper_32_bits(vma->node.start);
+
 	*batch++ = MI_BATCH_BUFFER_END; /* not reached */
 
 	intel_gt_chipset_flush(engine->gt);
@@ -153,7 +176,10 @@ igt_spinner_create_request(struct igt_spinner *spin,
 			goto cancel_rq;
 	}
 
-	err = engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, 0);
+	flags = 0;
+	if (INTEL_GEN(rq->i915) <= 5)
+		flags |= I915_DISPATCH_SECURE;
+	err = engine->emit_bb_start(rq, vma->node.start, PAGE_SIZE, flags);
 
 cancel_rq:
 	if (err) {
diff --git a/drivers/gpu/drm/i915/selftests/intel_memory_region.c b/drivers/gpu/drm/i915/selftests/intel_memory_region.c
index 19e1cca8f143..04d0aa7b349e 100644
--- a/drivers/gpu/drm/i915/selftests/intel_memory_region.c
+++ b/drivers/gpu/drm/i915/selftests/intel_memory_region.c
@@ -32,7 +32,7 @@ static void close_objects(struct intel_memory_region *mem,
 		if (i915_gem_object_has_pinned_pages(obj))
 			i915_gem_object_unpin_pages(obj);
 		/* No polluting the memory region between tests */
-		__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+		__i915_gem_object_put_pages(obj);
 		list_del(&obj->st_link);
 		i915_gem_object_put(obj);
 	}
@@ -122,7 +122,7 @@ put:
 static void igt_object_release(struct drm_i915_gem_object *obj)
 {
 	i915_gem_object_unpin_pages(obj);
-	__i915_gem_object_put_pages(obj, I915_MM_NORMAL);
+	__i915_gem_object_put_pages(obj);
 	list_del(&obj->st_link);
 	i915_gem_object_put(obj);
 }
@@ -404,7 +404,7 @@ static int igt_lmem_write_gpu(void *arg)
 	struct drm_i915_private *i915 = arg;
 	struct drm_i915_gem_object *obj;
 	struct i915_gem_context *ctx;
-	struct drm_file *file;
+	struct file *file;
 	I915_RND_STATE(prng);
 	u32 sz;
 	int err;
@@ -439,7 +439,7 @@ static int igt_lmem_write_gpu(void *arg)
 out_put:
 	i915_gem_object_put(obj);
 out_file:
-	mock_file_free(i915, file);
+	fput(file);
 	return err;
 }
 
@@ -506,7 +506,9 @@ static int igt_lmem_write_cpu(void *arg)
 	}
 
 	/* Put the pages into a known state -- from the gpu for added fun */
+	intel_engine_pm_get(engine);
 	err = i915_gem_object_fill_blt(obj, engine->kernel_context, 0xdeadbeaf);
+	intel_engine_pm_put(engine);
 	if (err)
 		goto out_unpin;
 
diff --git a/drivers/gpu/drm/i915/selftests/mock_drm.c b/drivers/gpu/drm/i915/selftests/mock_drm.c
deleted file mode 100644
index 09c704153456..000000000000
--- a/drivers/gpu/drm/i915/selftests/mock_drm.c
+++ /dev/null
@@ -1,73 +0,0 @@
-/*
- * Copyright © 2017 Intel Corporation
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice (including the next
- * paragraph) shall be included in all copies or substantial portions of the
- * Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
- * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
- * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
- * IN THE SOFTWARE.
- *
- */
-
-#include "mock_drm.h"
-
-struct drm_file *mock_file(struct drm_i915_private *i915)
-{
-	struct file *filp;
-	struct inode *inode;
-	struct drm_file *file;
-	int err;
-
-	inode = kzalloc(sizeof(*inode), GFP_KERNEL);
-	if (!inode) {
-		err = -ENOMEM;
-		goto err;
-	}
-
-	inode->i_rdev = i915->drm.primary->index;
-
-	filp = kzalloc(sizeof(*filp), GFP_KERNEL);
-	if (!filp) {
-		err = -ENOMEM;
-		goto err_inode;
-	}
-
-	err = drm_open(inode, filp);
-	if (err)
-		goto err_filp;
-
-	file = filp->private_data;
-	memset(&file->filp, POISON_INUSE, sizeof(file->filp));
-	file->authenticated = true;
-
-	kfree(filp);
-	kfree(inode);
-	return file;
-
-err_filp:
-	kfree(filp);
-err_inode:
-	kfree(inode);
-err:
-	return ERR_PTR(err);
-}
-
-void mock_file_free(struct drm_i915_private *i915, struct drm_file *file)
-{
-	struct file filp = { .private_data = file };
-
-	drm_release(NULL, &filp);
-}
diff --git a/drivers/gpu/drm/i915/selftests/mock_drm.h b/drivers/gpu/drm/i915/selftests/mock_drm.h
index b39beee9f8f6..9916b6f95526 100644
--- a/drivers/gpu/drm/i915/selftests/mock_drm.h
+++ b/drivers/gpu/drm/i915/selftests/mock_drm.h
@@ -25,7 +25,21 @@
 #ifndef __MOCK_DRM_H
 #define __MOCK_DRM_H
 
-struct drm_file *mock_file(struct drm_i915_private *i915);
-void mock_file_free(struct drm_i915_private *i915, struct drm_file *file);
+#include <drm/drm_file.h>
+
+#include "i915_drv.h"
+
+struct drm_file;
+struct file;
+
+static inline struct file *mock_file(struct drm_i915_private *i915)
+{
+	return mock_drm_getfile(i915->drm.primary, O_RDWR);
+}
+
+static inline struct drm_file *to_drm_file(struct file *f)
+{
+	return f->private_data;
+}
 
 #endif /* !__MOCK_DRM_H */
diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
index 27ed3cee6a9b..ac641f5360e1 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c
+++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c
@@ -54,25 +54,17 @@ void mock_device_flush(struct drm_i915_private *i915)
 static void mock_device_release(struct drm_device *dev)
 {
 	struct drm_i915_private *i915 = to_i915(dev);
-	struct intel_engine_cs *engine;
-	enum intel_engine_id id;
 
 	mock_device_flush(i915);
+	intel_gt_driver_remove(&i915->gt);
 
 	i915_gem_drain_workqueue(i915);
-
-	for_each_engine(engine, &i915->gt, id)
-		mock_engine_free(engine);
-	i915_gem_driver_release__contexts(i915);
-
-	intel_timelines_fini(i915);
-
-	drain_workqueue(i915->wq);
 	i915_gem_drain_freed_objects(i915);
 
 	mock_fini_ggtt(&i915->ggtt);
 	destroy_workqueue(i915->wq);
 
+	intel_gt_driver_late_release(&i915->gt);
 	intel_memory_regions_driver_release(i915);
 
 	drm_mode_config_cleanup(&i915->drm);
@@ -180,9 +172,8 @@ struct drm_i915_private *mock_gem_device(void)
 
 	mock_init_contexts(i915);
 
-	intel_timelines_init(i915);
-
 	mock_init_ggtt(i915, &i915->ggtt);
+	i915->gt.vm = i915_vm_get(&i915->ggtt.vm);
 
 	mkwrite_device_info(i915)->engine_mask = BIT(0);
 
@@ -190,10 +181,6 @@ struct drm_i915_private *mock_gem_device(void)
 	if (!i915->engine[RCS0])
 		goto err_unlock;
 
-	i915->kernel_context = mock_context(i915, NULL);
-	if (!i915->kernel_context)
-		goto err_engine;
-
 	if (mock_engine_init(i915->engine[RCS0]))
 		goto err_context;
 
@@ -202,13 +189,11 @@ struct drm_i915_private *mock_gem_device(void)
 	return i915;
 
 err_context:
-	i915_gem_driver_release__contexts(i915);
-err_engine:
-	mock_engine_free(i915->engine[RCS0]);
+	intel_gt_driver_remove(&i915->gt);
 err_unlock:
-	intel_timelines_fini(i915);
 	destroy_workqueue(i915->wq);
 err_drv:
+	intel_gt_driver_late_release(&i915->gt);
 	intel_memory_regions_driver_release(i915);
 	drm_mode_config_cleanup(&i915->drm);
 	drm_dev_fini(&i915->drm);
diff --git a/drivers/gpu/drm/i915/selftests/mock_gtt.h b/drivers/gpu/drm/i915/selftests/mock_gtt.h
index 3387393286de..e3f224f43beb 100644
--- a/drivers/gpu/drm/i915/selftests/mock_gtt.h
+++ b/drivers/gpu/drm/i915/selftests/mock_gtt.h
@@ -25,6 +25,9 @@
 #ifndef __MOCK_GTT_H
 #define __MOCK_GTT_H
 
+struct drm_i915_private;
+struct i915_ggtt;
+
 void mock_init_ggtt(struct drm_i915_private *i915, struct i915_ggtt *ggtt);
 void mock_fini_ggtt(struct i915_ggtt *ggtt);
 
diff --git a/drivers/gpu/drm/i915/selftests/mock_region.h b/drivers/gpu/drm/i915/selftests/mock_region.h
index 24608089d833..329bf74dfaca 100644
--- a/drivers/gpu/drm/i915/selftests/mock_region.h
+++ b/drivers/gpu/drm/i915/selftests/mock_region.h
@@ -6,6 +6,11 @@
 #ifndef __MOCK_REGION_H
 #define __MOCK_REGION_H
 
+#include <linux/types.h>
+
+struct drm_i915_private;
+struct intel_memory_region;
+
 struct intel_memory_region *
 mock_region_create(struct drm_i915_private *i915,
 		   resource_size_t start,
diff --git a/drivers/gpu/drm/i915/selftests/mock_uncore.h b/drivers/gpu/drm/i915/selftests/mock_uncore.h
index 8a2cc553f466..7acf1ef4d488 100644
--- a/drivers/gpu/drm/i915/selftests/mock_uncore.h
+++ b/drivers/gpu/drm/i915/selftests/mock_uncore.h
@@ -25,6 +25,9 @@
 #ifndef __MOCK_UNCORE_H
 #define __MOCK_UNCORE_H
 
+struct drm_i915_private;
+struct intel_uncore;
+
 void mock_uncore_init(struct intel_uncore *uncore,
 		      struct drm_i915_private *i915);
 
diff --git a/drivers/mfd/intel_soc_pmic_core.c b/drivers/mfd/intel_soc_pmic_core.c
index c9f35378d391..47188df3080d 100644
--- a/drivers/mfd/intel_soc_pmic_core.c
+++ b/drivers/mfd/intel_soc_pmic_core.c
@@ -38,7 +38,7 @@ static struct gpiod_lookup_table panel_gpio_table = {
 
 /* PWM consumed by the Intel GFX */
 static struct pwm_lookup crc_pwm_lookup[] = {
-	PWM_LOOKUP("crystal_cove_pwm", 0, "0000:00:02.0", "pwm_backlight", 0, PWM_POLARITY_NORMAL),
+	PWM_LOOKUP("crystal_cove_pwm", 0, "0000:00:02.0", "pwm_pmic_backlight", 0, PWM_POLARITY_NORMAL),
 };
 
 static int intel_soc_pmic_i2c_probe(struct i2c_client *i2c,